studies on negative effects of homework

Search form

• Find Stories
• For Journalists

Stanford research shows pitfalls of homework

A Stanford researcher found that students in high-achieving communities who spend too much time on homework experience more stress, physical health problems, a lack of balance and even alienation from society. More than two hours of homework a night may be counterproductive, according to the study.

Education scholar Denise Pope has found that too much homework has negative effects on student well-being and behavioral engagement. (Image credit: L.A. Cicero)

A Stanford researcher found that too much homework can negatively affect kids, especially their lives away from school, where family, friends and activities matter.

“Our findings on the effects of homework challenge the traditional assumption that homework is inherently good,” wrote Denise Pope , a senior lecturer at the Stanford Graduate School of Education and a co-author of a study published in the Journal of Experimental Education .

The researchers used survey data to examine perceptions about homework, student well-being and behavioral engagement in a sample of 4,317 students from 10 high-performing high schools in upper-middle-class California communities. Along with the survey data, Pope and her colleagues used open-ended answers to explore the students’ views on homework.

Median household income exceeded $90,000 in these communities, and 93 percent of the students went on to college, either two-year or four-year.

Students in these schools average about 3.1 hours of homework each night.

“The findings address how current homework practices in privileged, high-performing schools sustain students’ advantage in competitive climates yet hinder learning, full engagement and well-being,” Pope wrote.

Pope and her colleagues found that too much homework can diminish its effectiveness and even be counterproductive. They cite prior research indicating that homework benefits plateau at about two hours per night, and that 90 minutes to two and a half hours is optimal for high school.

Their study found that too much homework is associated with:

• Greater stress: 56 percent of the students considered homework a primary source of stress, according to the survey data. Forty-three percent viewed tests as a primary stressor, while 33 percent put the pressure to get good grades in that category. Less than 1 percent of the students said homework was not a stressor.

• Reductions in health: In their open-ended answers, many students said their homework load led to sleep deprivation and other health problems. The researchers asked students whether they experienced health issues such as headaches, exhaustion, sleep deprivation, weight loss and stomach problems.

• Less time for friends, family and extracurricular pursuits: Both the survey data and student responses indicate that spending too much time on homework meant that students were “not meeting their developmental needs or cultivating other critical life skills,” according to the researchers. Students were more likely to drop activities, not see friends or family, and not pursue hobbies they enjoy.

A balancing act

The results offer empirical evidence that many students struggle to find balance between homework, extracurricular activities and social time, the researchers said. Many students felt forced or obligated to choose homework over developing other talents or skills.

Also, there was no relationship between the time spent on homework and how much the student enjoyed it. The research quoted students as saying they often do homework they see as “pointless” or “mindless” in order to keep their grades up.

“This kind of busy work, by its very nature, discourages learning and instead promotes doing homework simply to get points,” Pope said.

She said the research calls into question the value of assigning large amounts of homework in high-performing schools. Homework should not be simply assigned as a routine practice, she said.

“Rather, any homework assigned should have a purpose and benefit, and it should be designed to cultivate learning and development,” wrote Pope.

High-performing paradox

In places where students attend high-performing schools, too much homework can reduce their time to foster skills in the area of personal responsibility, the researchers concluded. “Young people are spending more time alone,” they wrote, “which means less time for family and fewer opportunities to engage in their communities.”

Student perspectives

The researchers say that while their open-ended or “self-reporting” methodology to gauge student concerns about homework may have limitations – some might regard it as an opportunity for “typical adolescent complaining” – it was important to learn firsthand what the students believe.

The paper was co-authored by Mollie Galloway from Lewis and Clark College and Jerusha Conner from Villanova University.

share this!

August 16, 2021

Is it time to get rid of homework? Mental health experts weigh in

by Sara M Moniuszko

It's no secret that kids hate homework. And as students grapple with an ongoing pandemic that has had a wide-range of mental health impacts, is it time schools start listening to their pleas over workloads?

Some teachers are turning to social media to take a stand against homework .

Tiktok user @misguided.teacher says he doesn't assign it because the "whole premise of homework is flawed."

For starters, he says he can't grade work on "even playing fields" when students' home environments can be vastly different.

"Even students who go home to a peaceful house, do they really want to spend their time on busy work? Because typically that's what a lot of homework is, it's busy work," he says in the video that has garnered 1.6 million likes. "You only get one year to be 7, you only got one year to be 10, you only get one year to be 16, 18."

Mental health experts agree heavy work loads have the potential do more harm than good for students, especially when taking into account the impacts of the pandemic. But they also say the answer may not be to eliminate homework altogether.

Emmy Kang, mental health counselor at Humantold, says studies have shown heavy workloads can be "detrimental" for students and cause a "big impact on their mental, physical and emotional health."

"More than half of students say that homework is their primary source of stress, and we know what stress can do on our bodies," she says, adding that staying up late to finish assignments also leads to disrupted sleep and exhaustion.

Cynthia Catchings, a licensed clinical social worker and therapist at Talkspace, says heavy workloads can also cause serious mental health problems in the long run, like anxiety and depression.

And for all the distress homework causes, it's not as useful as many may think, says Dr. Nicholas Kardaras, a psychologist and CEO of Omega Recovery treatment center.

"The research shows that there's really limited benefit of homework for elementary age students, that really the school work should be contained in the classroom," he says.

For older students, Kang says homework benefits plateau at about two hours per night.

"Most students, especially at these high-achieving schools, they're doing a minimum of three hours, and it's taking away time from their friends from their families, their extracurricular activities. And these are all very important things for a person's mental and emotional health."

Catchings, who also taught third to 12th graders for 12 years, says she's seen the positive effects of a no homework policy while working with students abroad.

"Not having homework was something that I always admired from the French students (and) the French schools, because that was helping the students to really have the time off and really disconnect from school ," she says.

The answer may not be to eliminate homework completely, but to be more mindful of the type of work students go home with, suggests Kang, who was a high-school teacher for 10 years.

"I don't think (we) should scrap homework, I think we should scrap meaningless, purposeless busy work-type homework. That's something that needs to be scrapped entirely," she says, encouraging teachers to be thoughtful and consider the amount of time it would take for students to complete assignments.

The pandemic made the conversation around homework more crucial

Mindfulness surrounding homework is especially important in the context of the last two years. Many students will be struggling with mental health issues that were brought on or worsened by the pandemic, making heavy workloads even harder to balance.

"COVID was just a disaster in terms of the lack of structure. Everything just deteriorated," Kardaras says, pointing to an increase in cognitive issues and decrease in attention spans among students. "School acts as an anchor for a lot of children, as a stabilizing force, and that disappeared."

But even if students transition back to the structure of in-person classes, Kardaras suspects students may still struggle after two school years of shifted schedules and disrupted sleeping habits.

"We've seen adults struggling to go back to in-person work environments from remote work environments. That effect is amplified with children because children have less resources to be able to cope with those transitions than adults do," he explains.

'Get organized' ahead of back-to-school

In order to make the transition back to in-person school easier, Kang encourages students to "get good sleep, exercise regularly (and) eat a healthy diet."

To help manage workloads, she suggests students "get organized."

"There's so much mental clutter up there when you're disorganized... sitting down and planning out their study schedules can really help manage their time," she says.

Breaking assignments up can also make things easier to tackle.

"I know that heavy workloads can be stressful, but if you sit down and you break down that studying into smaller chunks, they're much more manageable."

If workloads are still too much, Kang encourages students to advocate for themselves.

"They should tell their teachers when a homework assignment just took too much time or if it was too difficult for them to do on their own," she says. "It's good to speak up and ask those questions. Respectfully, of course, because these are your teachers. But still, I think sometimes teachers themselves need this feedback from their students."

©2021 USA Today Distributed by Tribune Content Agency, LLC.

Explore further

Feedback to editors

Using CO₂ and biomass, researchers find path to more environmentally friendly recyclable plastics

10 hours ago

Precision agriculture research identifies gene that controls production of flowers and fruits in pea plants

Long-term forest study shows tornado's effects linger 25 years later

A new coating method in mRNA engineering points the way to advanced therapies

ATLAS provides first measurement of the W-boson width at the LHC

12 hours ago

Smart vest turns fish into underwater spies, providing a glimpse into aquatic life like never before

Ants in Colorado are on the move due to climate change

Caterpillar 'noses' are surprisingly sophisticated, researchers find

Building footprints could help identify neighborhood sociodemographic traits

13 hours ago

Fossilized dinosaur eggshells can preserve amino acids, the building blocks of proteins, over millions of years

14 hours ago

Relevant PhysicsForums posts

Motivating high school physics students with popcorn physics.

Apr 3, 2024

How is Physics taught without Calculus?

Mar 29, 2024

Why are Physicists so informal with mathematics?

Mar 24, 2024

The changing physics curriculum in 1961

Suggestions for using math puzzles to stimulate my math students.

Mar 21, 2024

The New California Math Framework: Another Step Backwards?

Mar 14, 2024

More from STEM Educators and Teaching

Related Stories

Smartphones are lowering student's grades, study finds

Aug 18, 2020

Doing homework is associated with change in students' personality

Oct 6, 2017

Scholar suggests ways to craft more effective homework assignments

Oct 1, 2015

Should parents help their kids with homework?

Aug 29, 2019

How much math, science homework is too much?

Mar 23, 2015

Anxiety, depression, burnout rising as college students prepare to return to campus

Jul 26, 2021

Recommended for you

Earth, the sun and a bike wheel: Why your high-school textbook was wrong about the shape of Earth's orbit

Apr 8, 2024

Touchibo, a robot that fosters inclusion in education through touch

Apr 5, 2024

More than money, family and community bonds prep teens for college success: Study

Research reveals significant effects of onscreen instructors during video classes in aiding student learning

Mar 25, 2024

Prestigious journals make it hard for scientists who don't speak English to get published, study finds

Mar 23, 2024

Using Twitter/X to promote research findings found to have little impact on number of citations

Mar 22, 2024

Let us know if there is a problem with our content

Use this form if you have come across a typo, inaccuracy or would like to send an edit request for the content on this page. For general inquiries, please use our contact form . For general feedback, use the public comments section below (please adhere to guidelines ).

Please select the most appropriate category to facilitate processing of your request

Thank you for taking time to provide your feedback to the editors.

Your feedback is important to us. However, we do not guarantee individual replies due to the high volume of messages.

E-mail the story

Your email address is used only to let the recipient know who sent the email. Neither your address nor the recipient's address will be used for any other purpose. The information you enter will appear in your e-mail message and is not retained by Phys.org in any form.

Newsletter sign up

Get weekly and/or daily updates delivered to your inbox. You can unsubscribe at any time and we'll never share your details to third parties.

More information Privacy policy

Donate and enjoy an ad-free experience

We keep our content available to everyone. Consider supporting Science X's mission by getting a premium account.

E-mail newsletter

• Second Opinion
• Research & Innovation
• Patients & Families
• Health Professionals
• Recently Visited
• Segunda opinión
• Refer a patient
• MyChart Login

Healthier, Happy Lives Blog

Sort articles by..., sort by category.

• Celebrating Volunteers
• Community Outreach
• Construction Updates
• Family-Centered Care
• Healthy Eating
• Heart Center
• Interesting Things
• Mental Health
• Patient Stories
• Research and Innovation
• Safety Tips
• Sustainability
• World-Class Care

About Our Blog

• Back-to-School
• Pediatric Technology

Latest Posts

• NICU Sims Set Stage for Lifesaving Care
• How a Social Media Post Led a Teen to Find a ‘Kidney Buddy’ for Life
• Understanding Autism Spectrum Disorder in Young Children
• New Liver Gives a Toddler a Renewed Chance at Life
• Family Turns Newborn’s Rare Diagnosis Into Something Beautiful

Health Hazards of Homework

March 18, 2014 | Julie Greicius Pediatrics .

A new study by the Stanford Graduate School of Education and colleagues found that students in high-performing schools who did excessive hours of homework “experienced greater behavioral engagement in school but also more academic stress, physical health problems, and lack of balance in their lives.”

Those health problems ranged from stress, headaches, exhaustion, sleep deprivation, weight loss and stomach problems, to psycho-social effects like dropping activities, not seeing friends or family, and not pursuing hobbies they enjoy.

In the Stanford Report story about the research, Denise Pope , a senior lecturer at the Stanford Graduate School of Education and a co-author of the  study published in the  Journal of Experimental Education , says, “Our findings on the effects of homework challenge the traditional assumption that homework is inherently good.”

The study was based on survey data from a sample of 4,317 students from 10 high-performing high schools in California communities in which median household income exceeded $90,000. Of the students surveyed, homework volume averaged about 3.1 hours each night.

“It is time to re-evaluate how the school environment is preparing our high school student for today’s workplace,” says Neville Golden, MD , chief of adolescent medicine at Stanford Medicine Children’s Health and a professor at the School of Medicine. “This landmark study shows that excessive homework is counterproductive, leading to sleep deprivation, school stress and other health problems. Parents can best support their children in these demanding academic environments by advocating for them through direct communication with teachers and school administrators about homework load.”

Related Posts

Top-ranked group group in Los Gatos, Calif., is now a part of one of the…

The Stanford Medicine Children’s Health network continues to grow with our newest addition, Town and…

• Julie Greicius
• more by this author...

Connect with us:

Download our App:

ABOUT STANFORD MEDICINE CHILDREN'S HEALTH

• Leadership Team
• Vision, Mission & Values
• The Stanford Advantage
• Government and Community Relations

LUCILE PACKARD FOUNDATION FOR CHILDREN'S HEALTH

• Get Involved
• Volunteering Services
• Auxiliaries & Affiliates
• Our Hospital
• Send a Greeting Card
• New Hospital
• Refer a Patient
• Pay Your Bill

Also Find Us on:

• Notice of Nondiscrimination
• Terms of Use
• Privacy Policy
• Code of Conduct
• Price Transparency
• Stanford Medicine
• Stanford University
• Stanford Health Care

Request More Info

Fill out the form below and a member of our team will reach out right away!

" * " indicates required fields

Is Homework Necessary? Education Inequity and Its Impact on Students

The Problem with Homework: It Highlights Inequalities

How much homework is too much homework, when does homework actually help, negative effects of homework for students, how teachers can help.

Schools are getting rid of homework from Essex, Mass., to Los Angeles, Calif. Although the no-homework trend may sound alarming, especially to parents dreaming of their child’s acceptance to Harvard, Stanford or Yale, there is mounting evidence that eliminating homework in grade school may actually have great benefits , especially with regard to educational equity.

In fact, while the push to eliminate homework may come as a surprise to many adults, the debate is not new . Parents and educators have been talking about this subject for the last century, so that the educational pendulum continues to swing back and forth between the need for homework and the need to eliminate homework.

One of the most pressing talking points around homework is how it disproportionately affects students from less affluent families. The American Psychological Association (APA) explained:

“Kids from wealthier homes are more likely to have resources such as computers, internet connections, dedicated areas to do schoolwork and parents who tend to be more educated and more available to help them with tricky assignments. Kids from disadvantaged homes are more likely to work at afterschool jobs, or to be home without supervision in the evenings while their parents work multiple jobs.”

[RELATED] How to Advance Your Career: A Guide for Educators >> 

While students growing up in more affluent areas are likely playing sports, participating in other recreational activities after school, or receiving additional tutoring, children in disadvantaged areas are more likely headed to work after school, taking care of siblings while their parents work or dealing with an unstable home life. Adding homework into the mix is one more thing to deal with — and if the student is struggling, the task of completing homework can be too much to consider at the end of an already long school day.

While all students may groan at the mention of homework, it may be more than just a nuisance for poor and disadvantaged children, instead becoming another burden to carry and contend with.

Beyond the logistical issues, homework can negatively impact physical health and stress — and once again this may be a more significant problem among economically disadvantaged youth who typically already have a higher stress level than peers from more financially stable families .

Yet, today, it is not just the disadvantaged who suffer from the stressors that homework inflicts. A 2014 CNN article, “Is Homework Making Your Child Sick?” , covered the issue of extreme pressure placed on children of the affluent. The article looked at the results of a study surveying more than 4,300 students from 10 high-performing public and private high schools in upper-middle-class California communities.

“Their findings were troubling: Research showed that excessive homework is associated with high stress levels, physical health problems and lack of balance in children’s lives; 56% of the students in the study cited homework as a primary stressor in their lives,” according to the CNN story. “That children growing up in poverty are at-risk for a number of ailments is both intuitive and well-supported by research. More difficult to believe is the growing consensus that children on the other end of the spectrum, children raised in affluence, may also be at risk.”

When it comes to health and stress it is clear that excessive homework, for children at both ends of the spectrum, can be damaging. Which begs the question, how much homework is too much?

The National Education Association and the National Parent Teacher Association recommend that students spend 10 minutes per grade level per night on homework . That means that first graders should spend 10 minutes on homework, second graders 20 minutes and so on. But a study published by The American Journal of Family Therapy found that students are getting much more than that.

While 10 minutes per day doesn’t sound like much, that quickly adds up to an hour per night by sixth grade. The National Center for Education Statistics found that high school students get an average of 6.8 hours of homework per week, a figure that is much too high according to the Organization for Economic Cooperation and Development (OECD). It is also to be noted that this figure does not take into consideration the needs of underprivileged student populations.

In a study conducted by the OECD it was found that “after around four hours of homework per week, the additional time invested in homework has a negligible impact on performance .” That means that by asking our children to put in an hour or more per day of dedicated homework time, we are not only not helping them, but — according to the aforementioned studies — we are hurting them, both physically and emotionally.

What’s more is that homework is, as the name implies, to be completed at home, after a full day of learning that is typically six to seven hours long with breaks and lunch included. However, a study by the APA on how people develop expertise found that elite musicians, scientists and athletes do their most productive work for about only four hours per day. Similarly, companies like Tower Paddle Boards are experimenting with a five-hour workday, under the assumption that people are not able to be truly productive for much longer than that. CEO Stephan Aarstol told CNBC that he believes most Americans only get about two to three hours of work done in an eight-hour day.

In the scope of world history, homework is a fairly new construct in the U.S. Students of all ages have been receiving work to complete at home for centuries, but it was educational reformer Horace Mann who first brought the concept to America from Prussia. 

Since then, homework’s popularity has ebbed and flowed in the court of public opinion. In the 1930s, it was considered child labor (as, ironically, it compromised children’s ability to do chores at home). Then, in the 1950s, implementing mandatory homework was hailed as a way to ensure America’s youth were always one step ahead of Soviet children during the Cold War. Homework was formally mandated as a tool for boosting educational quality in 1986 by the U.S. Department of Education, and has remained in common practice ever since.  

School work assigned and completed outside of school hours is not without its benefits. Numerous studies have shown that regular homework has a hand in improving student performance and connecting students to their learning. When reviewing these studies, take them with a grain of salt; there are strong arguments for both sides, and only you will know which solution is best for your students or school. 

Homework improves student achievement.

• Source: The High School Journal, “ When is Homework Worth the Time?: Evaluating the Association between Homework and Achievement in High School Science and Math ,” 2012. 
• Source: IZA.org, “ Does High School Homework Increase Academic Achievement? ,” 2014. **Note: Study sample comprised only high school boys. 

Homework helps reinforce classroom learning.

• Source: “ Debunk This: People Remember 10 Percent of What They Read ,” 2015.

Homework helps students develop good study habits and life skills.

• Sources: The Repository @ St. Cloud State, “ Types of Homework and Their Effect on Student Achievement ,” 2017; Journal of Advanced Academics, “ Developing Self-Regulation Skills: The Important Role of Homework ,” 2011.
• Source: Journal of Advanced Academics, “ Developing Self-Regulation Skills: The Important Role of Homework ,” 2011.

Homework allows parents to be involved with their children’s learning.

• Parents can see what their children are learning and working on in school every day. 
• Parents can participate in their children’s learning by guiding them through homework assignments and reinforcing positive study and research habits.
• Homework observation and participation can help parents understand their children’s academic strengths and weaknesses, and even identify possible learning difficulties.
• Source: Phys.org, “ Sociologist Upends Notions about Parental Help with Homework ,” 2018.

While some amount of homework may help students connect to their learning and enhance their in-class performance, too much homework can have damaging effects. 

Students with too much homework have elevated stress levels. 

• Source: USA Today, “ Is It Time to Get Rid of Homework? Mental Health Experts Weigh In ,” 2021.
• Source: Stanford University, “ Stanford Research Shows Pitfalls of Homework ,” 2014.

Students with too much homework may be tempted to cheat. 

• Source: The Chronicle of Higher Education, “ High-Tech Cheating Abounds, and Professors Bear Some Blame ,” 2010.
• Source: The American Journal of Family Therapy, “ Homework and Family Stress: With Consideration of Parents’ Self Confidence, Educational Level, and Cultural Background ,” 2015.

Homework highlights digital inequity. 

• Sources: NEAToday.org, “ The Homework Gap: The ‘Cruelest Part of the Digital Divide’ ,” 2016; CNET.com, “ The Digital Divide Has Left Millions of School Kids Behind ,” 2021.
• Source: Investopedia, “ Digital Divide ,” 2022; International Journal of Education and Social Science, “ Getting the Homework Done: Social Class and Parents’ Relationship to Homework ,” 2015.
• Source: World Economic Forum, “ COVID-19 exposed the digital divide. Here’s how we can close it ,” 2021.

Homework does not help younger students.

• Source: Review of Educational Research, “ Does Homework Improve Academic Achievement? A Synthesis of Researcher, 1987-2003 ,” 2006.

To help students find the right balance and succeed, teachers and educators must start the homework conversation, both internally at their school and with parents. But in order to successfully advocate on behalf of students, teachers must be well educated on the subject, fully understanding the research and the outcomes that can be achieved by eliminating or reducing the homework burden. There is a plethora of research and writing on the subject for those interested in self-study.

For teachers looking for a more in-depth approach or for educators with a keen interest in educational equity, formal education may be the best route. If this latter option sounds appealing, there are now many reputable schools offering online master of education degree programs to help educators balance the demands of work and family life while furthering their education in the quest to help others.

YOU’RE INVITED! Watch Free Webinar on USD’s Online MEd Program >>

Be Sure To Share This Article

• Share on Twitter
• Share on Facebook
• Share on LinkedIn

Top 11 Reasons to get Your Master of Education Degree

Free 22-page Book

• Master of Education

Related Posts

Is Homework Good for Kids? Here’s What the Research Says

A s kids return to school, debate is heating up once again over how they should spend their time after they leave the classroom for the day.

The no-homework policy of a second-grade teacher in Texas went viral last week , earning praise from parents across the country who lament the heavy workload often assigned to young students. Brandy Young told parents she would not formally assign any homework this year, asking students instead to eat dinner with their families, play outside and go to bed early.

But the question of how much work children should be doing outside of school remains controversial, and plenty of parents take issue with no-homework policies, worried their kids are losing a potential academic advantage. Here’s what you need to know:

For decades, the homework standard has been a “10-minute rule,” which recommends a daily maximum of 10 minutes of homework per grade level. Second graders, for example, should do about 20 minutes of homework each night. High school seniors should complete about two hours of homework each night. The National PTA and the National Education Association both support that guideline.

But some schools have begun to give their youngest students a break. A Massachusetts elementary school has announced a no-homework pilot program for the coming school year, lengthening the school day by two hours to provide more in-class instruction. “We really want kids to go home at 4 o’clock, tired. We want their brain to be tired,” Kelly Elementary School Principal Jackie Glasheen said in an interview with a local TV station . “We want them to enjoy their families. We want them to go to soccer practice or football practice, and we want them to go to bed. And that’s it.”

A New York City public elementary school implemented a similar policy last year, eliminating traditional homework assignments in favor of family time. The change was quickly met with outrage from some parents, though it earned support from other education leaders.

New solutions and approaches to homework differ by community, and these local debates are complicated by the fact that even education experts disagree about what’s best for kids.

The research

The most comprehensive research on homework to date comes from a 2006 meta-analysis by Duke University psychology professor Harris Cooper, who found evidence of a positive correlation between homework and student achievement, meaning students who did homework performed better in school. The correlation was stronger for older students—in seventh through 12th grade—than for those in younger grades, for whom there was a weak relationship between homework and performance.

Cooper’s analysis focused on how homework impacts academic achievement—test scores, for example. His report noted that homework is also thought to improve study habits, attitudes toward school, self-discipline, inquisitiveness and independent problem solving skills. On the other hand, some studies he examined showed that homework can cause physical and emotional fatigue, fuel negative attitudes about learning and limit leisure time for children. At the end of his analysis, Cooper recommended further study of such potential effects of homework.

Despite the weak correlation between homework and performance for young children, Cooper argues that a small amount of homework is useful for all students. Second-graders should not be doing two hours of homework each night, he said, but they also shouldn’t be doing no homework.

Not all education experts agree entirely with Cooper’s assessment.

Cathy Vatterott, an education professor at the University of Missouri-St. Louis, supports the “10-minute rule” as a maximum, but she thinks there is not sufficient proof that homework is helpful for students in elementary school.

“Correlation is not causation,” she said. “Does homework cause achievement, or do high achievers do more homework?”

Vatterott, the author of Rethinking Homework: Best Practices That Support Diverse Needs , thinks there should be more emphasis on improving the quality of homework tasks, and she supports efforts to eliminate homework for younger kids.

“I have no concerns about students not starting homework until fourth grade or fifth grade,” she said, noting that while the debate over homework will undoubtedly continue, she has noticed a trend toward limiting, if not eliminating, homework in elementary school.

The issue has been debated for decades. A TIME cover in 1999 read: “Too much homework! How it’s hurting our kids, and what parents should do about it.” The accompanying story noted that the launch of Sputnik in 1957 led to a push for better math and science education in the U.S. The ensuing pressure to be competitive on a global scale, plus the increasingly demanding college admissions process, fueled the practice of assigning homework.

“The complaints are cyclical, and we’re in the part of the cycle now where the concern is for too much,” Cooper said. “You can go back to the 1970s, when you’ll find there were concerns that there was too little, when we were concerned about our global competitiveness.”

Cooper acknowledged that some students really are bringing home too much homework, and their parents are right to be concerned.

“A good way to think about homework is the way you think about medications or dietary supplements,” he said. “If you take too little, they’ll have no effect. If you take too much, they can kill you. If you take the right amount, you’ll get better.”

More Must-Reads From TIME

• Exclusive: Google Workers Revolt Over $1.2 Billion Contract With Israel
• Jane Fonda Champions Climate Action for Every Generation
• Stop Looking for Your Forever Home
• The Sympathizer Counters 50 Years of Hollywood Vietnam War Narratives
• The Bliss of Seeing the Eclipse From Cleveland
• Hormonal Birth Control Doesn’t Deserve Its Bad Reputation
• The Best TV Shows to Watch on Peacock
• Want Weekly Recs on What to Watch, Read, and More? Sign Up for Worth Your Time

Write to Katie Reilly at [email protected]

You May Also Like

Advertisement

Academic dishonesty when doing homework: How digital technologies are put to bad use in secondary schools

• Open access
• Published: 23 July 2022
• Volume 28 , pages 1251–1271, ( 2023 )

Cite this article

You have full access to this open access article

• Juliette C. Désiron   ORCID: orcid.org/0000-0002-3074-9018 1 &
• Dominik Petko   ORCID: orcid.org/0000-0003-1569-1302 1  

5763 Accesses

4 Citations

4 Altmetric

Explore all metrics

The growth in digital technologies in recent decades has offered many opportunities to support students’ learning and homework completion. However, it has also contributed to expanding the field of possibilities concerning homework avoidance. Although studies have investigated the factors of academic dishonesty, the focus has often been on college students and formal assessments. The present study aimed to determine what predicts homework avoidance using digital resources and whether engaging in these practices is another predictor of test performance. To address these questions, we analyzed data from the Program for International Student Assessment 2018 survey, which contained additional questionnaires addressing this issue, for the Swiss students. The results showed that about half of the students engaged in one kind or another of digitally-supported practices for homework avoidance at least once or twice a week. Students who were more likely to use digital resources to engage in dishonest practices were males who did not put much effort into their homework and were enrolled in non-higher education-oriented school programs. Further, we found that digitally-supported homework avoidance was a significant negative predictor of test performance when considering information and communication technology predictors. Thus, the present study not only expands the knowledge regarding the predictors of academic dishonesty with digital resources, but also confirms the negative impact of such practices on learning.

Similar content being viewed by others

The impact of smartphone use on learning effectiveness: A case study of primary school students

Jen Chun Wang, Chia-Yen Hsieh & Shih-Hao Kung

Adoption of online mathematics learning in Ugandan government universities during the COVID-19 pandemic: pre-service teachers’ behavioural intention and challenges

Geofrey Kansiime & Marjorie Sarah Kabuye Batiibwe

Exploring the role of social media in collaborative learning the new domain of learning

Jamal Abdul Nasir Ansari & Nawab Ali Khan

Avoid common mistakes on your manuscript.

1 Introduction

Academic dishonesty is a widespread and perpetual issue for teachers made even more easier to perpetrate with the rise of digital technologies (Blau & Eshet-Alkalai, 2017 ; Ma et al., 2008 ). Definitions vary but overall an academically dishonest practices correspond to learners engaging in unauthorized practice such as cheating and plagiarism. Differences in engaging in those two types of practices mainly resides in students’ perception that plagiarism is worse than cheating (Evering & Moorman, 2012 ; McCabe, 2005 ). Plagiarism is usually defined as the unethical act of copying part or all of someone else’s work, with or without editing it, while cheating is more about sharing practices (Krou et al., 2021 ). As a result, most students do report cheating in an exam or for homework (Ma et al., 2008 ). To note, other research follow a different distinction for those practices and consider that plagiarism is a specific – and common – type of cheating (Waltzer & Dahl, 2022 ). Digital technologies have contributed to opening possibilities of homework avoidance and technology-related distraction (Ma et al., 2008 ; Xu, 2015 ).

The question of whether the use of digital resources hinders or enhances homework has often been investigated in large-scale studies, such as the Program for International Student Assessment (PISA), the Trends in International Mathematics and Science Study (TIMSS), and the Progress in International Reading Literacy Study (PIRLS). While most of the early large-scale studies showed positive overall correlations between the use of digital technologies for learning at home and test scores in language, mathematics, and science (e.g., OECD, 2015 ; Petko et al., 2017 ; Skryabin et al., 2015 ), there have been more recent studies reporting negative associations as well (Agasisti et al., 2020 ; Odell et al., 2020 ). One reason for these inconclusive findings is certainly the complex interplay of related factors, which include diverse ways of measuring homework, gender, socioeconomic status, personality traits, learning goals, academic abilities, learning strategies, motivation, and effort, as well as support from teachers and parents. Despite this complexity, it needs to be acknowledged that doing homework digitally does not automatically lead to productive learning activities, and it might even be associated with counter-productive practices such as digital distraction or academic dishonesty. Digitally enhanced academic dishonesty has mostly been investigated regarding formal assessment-related examinations (Evering & Moorman, 2012 ; Ma et al., 2008 ); however, it might be equally important to investigate its effects regarding learning-related assignments such as homework. Although a large body of research exists on digital academic dishonesty regarding assignments in higher education, relatively few studies have investigated this topic on K12 homework. To investigate this issue, we integrated questionnaire items on homework engagement and digital homework avoidance in a national add-on to PISA 2018 in Switzerland. Data from the Swiss sample can serve as a case study for further research with a wider cultural background. This study provides an overview of the descriptive results and tries to identify predictors of the use of digital technology for academic dishonesty when completing homework.

1.1 Prevalence and factors of digital academic dishonesty in schools

According to Pavela’s ( 1997 ) framework, four different types of academic dishonesty can be distinguished: cheating by using unauthorized materials, plagiarism by copying the work of others, fabrication of invented evidence, and facilitation by helping others in their attempts at academic dishonesty. Academic dishonesty can happen in assessment situations, as well as in learning situations. In formal assessments, academic dishonesty usually serves the purpose of passing a test or getting a better grade despite lacking the proper abilities or knowledge. In learning-related situations such as homework, where assignments are mandatory, cheating practices equally qualify as academic dishonesty. For perpetrators, these practices can be seen as shortcuts in which the willingness to invest the proper time and effort into learning is missing (Chow, 2021; Waltzer & Dahl,  2022 ). The interviews by Waltzer & Dahl ( 2022 ) reveal that students do perceive cheating as being wrong but this does not prevent them from engaging in at least one type of dishonest practice. While academic dishonesty is not a new phenomenon, it has been changing together with the development of new digital technologies (Anderman & Koenka, 2017 ; Ercegovac & Richardson, 2004 ). With the rapid growth in technologies, new forms of homework avoidance, such as copying and plagiarism, are developing (Evering & Moorman, 2012 ; Ma et al., 2008 ) summarized the findings of the 2006 U.S. surveys of the Josephson Institute of Ethics with the conclusion that the internet has led to a deterioration of ethics among students. In 2006, one-third of high school students had copied an internet document in the past 12 months, and 60% had cheated on a test. In 2012, these numbers were updated to 32% and 51%, respectively (Josephson Institute of Ethics, 2012 ). Further, 75% reported having copied another’s homework. Surprisingly, only a few studies have provided more recent evidence on the prevalence of academic dishonesty in middle and high schools. The results from colleges and universities are hardly comparable, and until now, this topic has not been addressed in international large-scale studies on schooling and school performance.

Despite the lack of representative studies, research has identified many factors in smaller and non-representative samples that might explain why some students engage in dishonest practices and others do not. These include male gender (Whitley et al., 1999 ), the “dark triad” of personality traits in contrast to conscientiousness and agreeableness (e.g., Cuadrado et al., 2021 ; Giluk & Postlethwaite, 2015 ), extrinsic motivation and performance/avoidance goals in contrast to intrinsic motivation and mastery goals (e.g., Anderman & Koenka,  2017 ; Krou et al., 2021 ), self-efficacy and achievement scores (e.g., Nora & Zhang,  2010 ; Yaniv et al., 2017 ), unethical attitudes, and low fear of being caught (e.g., Cheng et al., 2021 ; Kam et al., 2018 ), influenced by the moral norms of peers and the conditions of the educational context (e.g., Isakov & Tripathy,  2017 ; Kapoor & Kaufman, 2021 ). Similar factors have been reported regarding research on the causes of plagiarism (Husain et al., 2017 ; Moss et al., 2018 ). Further, the systematic review from Chiang et al. ( 2022 ) focused on factors of academic dishonesty in online learning environments. The analyses, based on the six-components behavior engineering, showed that the most prominent factors were environmental (effect of incentives) and individual (effect of motivation). Despite these intensive research efforts, there is still no overarching model that can comprehensively explain the interplay of these factors.

1.2 Effects of homework engagement and digital dishonesty on school performance

In meta-analyses of schools, small but significant positive effects of homework have been found regarding learning and achievement (e.g., Baş et al., 2017 ; Chen & Chen, 2014 ; Fan et al., 2017 ). In their review, Fan et al. ( 2017 ) found lower effect sizes for studies focusing on the time or frequency of homework than for studies investigating homework completion, homework grades, or homework effort. In large surveys, such as PISA, homework measurement by estimating after-school working hours has been customary practice. However, this measure could hide some other variables, such as whether teachers even give homework, whether there are school or state policies regarding homework, where the homework is done, whether it is done alone, etc. (e.g., Fernández-Alonso et al., 2015 , 2017 ). Trautwein ( 2007 ) and Trautwein et al. ( 2009 ) repeatedly showed that homework effort rather than the frequency or the time spent on homework can be considered a better predictor for academic achievement Effort and engagement can be seen as closely interrelated. Martin et al. ( 2017 ) defined engagement as the expressed behavior corresponding to students’ motivation. This has been more recently expanded by the notion of the quality of homework completion (Rosário et al., 2018 ; Xu et al., 2021 ). Therefore, it is a plausible assumption that academic dishonesty when doing homework is closely related to low homework effort and a low quality of homework completion, which in turn affects academic achievement. However, almost no studies exist on the effects of homework avoidance or academic dishonesty on academic achievement. Studies investigating the relationship between academic dishonesty and academic achievement typically use academic achievement as a predictor of academic dishonesty, not the other way around (e.g., Cuadrado et al., 2019 ; McCabe et al., 2001 ). The results of these studies show that low-performing students tend to engage in dishonest practices more often. However, high-performing students also seem to be prone to cheating in highly competitive situations (Yaniv et al., 2017 ).

1.3 Present study and hypotheses

The present study serves three combined purposes.

First, based on the additional questionnaires integrated into the Program for International Student Assessment 2018 (PISA 2018) data collection in Switzerland, we provide descriptive figures on the frequency of homework effort and the various forms of digitally-supported homework avoidance practices.

Second, the data were used to identify possible factors that explain higher levels of digitally-supported homework avoidance practices. Based on our review of the literature presented in Section 1.1 , we hypothesized (Hypothesis 1 – H1) that these factors include homework effort, age, gender, socio-economic status, and study program.

Finally, we tested whether digitally-supported homework avoidance practices were a significant predictor of test score performance. We expected (Hypothesis 2 – H2) that technology-related factors influencing test scores include not only those reported by Petko et al. ( 2017 ) but also self-reported engagement in digital dishonesty practices. .

2.1 Participants

Our analyses were based on data collected for PISA 2018 in Switzerland, made available in June 2021 (Erzinger et al., 2021 ). The target sample of PISA was 15-year-old students, with a two-phase sampling: schools and then students (Erzinger et al., 2019 , p.7–8, OECD, 2019a ). A total of 228 schools were selected for Switzerland, with an original sample of 5822 students. Based on the PISA 2018 technical report (OECD, 2019a ), only participants with a minimum of three valid responses to each scale used in the statistical analyses were included (see Section 2.2 ). A final sample of 4771 responses (48% female) was used for statistical analyses. The mean age was 15 years and 9 months ( SD  = 3 months). As Switzerland is a multilingual country, 60% of the respondents completed the questionnaires in German, 23% in French, and 17% in Italian.

2.2 Measures

2.2.1 digital dishonesty in homework scale.

This six-item digital dishonesty for homework scale assesses the use of digital technology for homework avoidance and copying (IC801 C01 to C06), is intended to work as a single overall scale for digital homework dishonesty practice constructed to include items corresponding to two types of dishonest practices from Pavela ( 1997 ), namely cheating and plagiarism (see Table  1 ). Three items target individual digital practices to avoid homework, which can be referred to as plagiarism (items 1, 2 and 5). Two focus more on social digital practices, for which students are cheating together with peers (items 4 and 6). One item target cheating as peer authorized plagiarism. Response options are based on questions on the productive use of digital technologies for homework in the common PISA survey (IC010), with an additional distinction for the lowest frequency option (6-point Likert scale). The scale was not tested prior to its integration into the PISA questionnaire, as it was newly developed for the purposes of this study.

2.2.2 Homework engagement scale

The scale, originally developed by Trautwein et al. (Trautwein, 2007 ; Trautwein et al., 2006 ), measures homework engagement (IC800 C01 to C06) and can be subdivided into two sub-scales: homework compliance and homework effort. The reliability of the scale was tested and established in different variants, both in Germany (Trautwein et al., 2006 ; Trautwein & Köller, 2003 ) and in Switzerland (Schnyder et al., 2008 ; Schynder Godel, 2015 ). In the adaptation used in the PISA 2018 survey, four items were positively poled (items 1, 2, 4, and 6), and two items were negatively poled (items 3 and 5) and presented with a 4-point Likert scale ranging from “Does not apply at all” to “Applies absolutely.” This adaptation showed acceptable reliability in previous studies in Switzerland (α = 0.73 and α = 0.78). The present study focused on homework effort, and thus only data from the corresponding sub-scale was analyzed (items 2 [I always try to do all of my homework], 4 [When it comes to homework, I do my best], and 6 [On the whole, I think I do my homework more conscientiously than my classmates]).

2.2.3 Demographics

Previous studies showed that demographic characteristics, such as age, gender, and socioeconomic status, could impact learning outcomes (Jacobs et al., 2002 ) and intention to use digital tools for learning (Tarhini et al., 2014 ). Gender is a dummy variable (ST004), with 1 for female and 2 for male. Socioeconomic status was analyzed based on the PISA 2018 index of economic, social, and cultural status (ESCS). It is computed from three other indices (OECD, 2019b , Annex A1): parents’ highest level of education (PARED), parents’ highest occupational status (HISEI), and home possessions (HOMEPOS). The final ESCS score is transformed so that 0 corresponds to an average OECD student. More details can be found in Annex A1 from PISA 2018 Results Volume 3 (OECD, 2019b ).

2.2.4 Study program

Although large-scale studies on schools have accounted for the differences between schools, the study program can also be a factor that directly affects digital homework dishonesty practices. In Switzerland, 15-year-old students from the PISA sampling pool can be part of at least six main study programs, which greatly differ in terms of learning content. In this study, study programs distinguished both level and type of study: lower secondary education (gymnasial – n  = 798, basic requirements – n  = 897, advanced requirements – n  = 1235), vocational education (classic – n  = 571, with baccalaureate – n  = 275), and university entrance preparation ( n  = 745). An “other” category was also included ( n  = 250). This 6-level ordinal variable was dummy coded based on the available CNTSCHID variable.

2.2.5 Technologies and schools

The PISA 2015 ICT (Information and Communication Technology) familiarity questionnaire included most of the technology-related variables tested by Petko et al. ( 2017 ): ENTUSE (frequency of computer use at home for entertainment purposes), HOMESCH (frequency of computer use for school-related purposes at home), and USESCH (frequency of computer use at school). However, the measure of student’s attitudes toward ICT in the 2015 survey was different from that of the 2012 dataset. Based on previous studies (Arpacı et al., 2021 ; Kunina-Habenicht & Goldhammer, 2020 ), we thus included INICT (Student’s ICT interest), COMPICT (Students’ perceived ICT competence), AUTICT (Students’ perceived autonomy related to ICT use), and SOIACICT (Students’ ICT as a topic in social interaction) instead of the variable ICTATTPOS of the 2012 survey.

2.2.6 Test scores

The PISA science, mathematics, and reading test scores were used as dependent variables to test our second hypothesis. Following Aparicio et al. ( 2021 ), the mean scores from plausible values were computed for each test score and used in the test score analysis.

2.3 Data analyses

Our hypotheses aim to assess the factors explaining student digital homework dishonesty practices (H1) and test score performance (H2). At the student level, we used multilevel regression analyses to decompose the variance and estimate associations. As we used data for Switzerland, in which differences between school systems exist at the level of provinces (within and between), we also considered differences across schools (based on the variable CNTSCHID).

Data were downloaded from the main PISA repository, and additional data for Switzerland were available on forscenter.ch (Erzinger et al., 2021 ). Analyses were computed with Jamovi (v.1.8 for Microsoft Windows) statistics and R packages (GAMLj, lavaan).

3.1 Additional scales for Switzerland

3.1.1 digital dishonesty in homework practices.

The digital homework dishonesty scale (6 items), computed with the six items IC801, was found to be of very good reliability overall (α = 0.91, ω = 0.91). After checking for reliability, a mean score was computed for the overall scale. The confirmatory factor analysis for the one-dimensional model reached an adequate fit, with three modifications using residual covariances between single items χ 2 (6) = 220, p  < 0.001, TLI = 0.969, CFI = 0.988, RMSEA (Root Mean Square Error of Approximation) = 0.086, SRMR = 0.016).

On the one hand, the practice that was the least reported was copying something from the internet and presenting it as their own (51% never did). On the other hand, students were more likely to partially copy content from the internet and modify it to present as their own (47% did it at least once a month). Copying answers shared by friends was rather common, with 62% of the students reporting that they engaged in such practices at least once a month.

When all surveyed practices were taken together, 7.6% of the students reported that they had never engaged in digitally dishonest practices for homework, while 30.6% reported cheating once or twice a week, 12.1% almost every day, and 6.9% every day (Table  1 ).

3.1.2 Homework effort

The overall homework engagement scale consisted of six items (IC800), and it was found to be acceptably reliable (α = 0.76, ω = 0.79). Items 3 and 5 were reversed for this analysis. The homework compliance sub-scale had a low reliability (α = 0.58, ω = 0.64), whereas the homework effort sub-scale had an acceptable reliability (α = 0.78, ω = 0.79). Based on our rationale, the following statistical analyses used only the homework effort sub-scale. Furthermore, this focus is justified by the fact that the homework compliance scale might be statistically confounded with the digital dishonesty in homework scale.

Descriptive weighted statistics per item (Table  2 ) showed that while most students (80%) tried to complete all of their homework, only half of the students reported doing those diligently (53.3%). Most students also reported that they believed they put more effort into their homework than their peers (77.7%). The overall mean score of the composite scale was 2.81 ( SD  = 0.69).

3.2 Multilevel regression analysis: Predictors of digital dishonesty in homework (H1)

Mixed multilevel modeling was used to analyze predictors of digital homework avoidance while considering the effect of school (random component). Based on our first hypothesis, we compared several models by progressively including the following fixed effects: homework effort and personal traits (age, gender) (Model 2), then socio-economic status (Model 3), and finally, study program (Model 4). The results are presented in Table  3 . Except for the digital homework dishonesty and homework efforts scales, all other scales were based upon the scores computed according to the PISA technical report (OECD, 2019a ).

We first compared variance components. Variance was decomposed into student and school levels. Model 1 provides estimates of the variance component without any covariates. The intraclass coefficient (ICC) indicated that about 6.6% of the total variance was associated with schools. The parameter (b  = 2.56, SE b  = 0.025 ) falls within the 95% confidence interval. Further, CI is above 0 and thus we can reject the null hypothesis. Comparing the empty model to models with covariates, we found that Models 2, 3 and 4 showed an increase in total explained variance to 10%. Variance explained by the covariates was about 3% in Models 2 and 3, and about 4% in Model 4. Interestingly, in our models, student socio-economic status, measured by the PISA index, never accounted for variance in digitally-supported dishonest practices to complete homework.

Summary of the two-steps Model 4 (estimates - β, with standard errors and significance levels, *** p < 0.001)

Further, model comparison based on AIC indicates that Model 4, including homework effort, personal traits, socio-economic status, and study program, was the better fit for the data. In Model 4 (Table  3 ; Fig.  1 ), we observed that homework effort and gender were negatively associated with digital dishonesty. Male students who invested less effort in their homework were more prone to engage in digital dishonesty. The study program was positively but weakly associated with digital dishonesty. Students in programs that target higher education were less likely to engage in digital dishonesty when completing homework.

3.3 Multilevel regression analysis: Cheating and test scores (H2)

Our first hypothesis aimed to provide insights into characteristics of students reporting that they regularly use digital resources dishonestly when completing homework. Our second hypothesis focused on whether digitally-supported homework avoidance practices was linked to results of test scores. Mixed multilevel modeling was used to analyze predictors of test scores while considering the effect of school (random component). Based on the study by Petko et al. ( 2017 ), we compared several models by progressively including the following fixed effects ICT use (three measures) (Model 2), then attitude toward ICT (four measures) (Model 3), and finally, digital dishonesty in homework (single measure) (Model 4). The results are presented in Table  4 for science, Table  5 for mathematics, and Table  6 for reading.

Variance components were decomposed into student and school level. ICC for Model 1 indicated that 37.9% of the variance component without covariates was associated with schools.

Taking Model 1 as a reference, we observed an increase in total explained variance to 40.5% with factors related to ICT use (Model 2), to 40.8% with factors related to attitude toward ICT (Model 3), and to 41.1% with the single digital dishonesty factor. It is interesting to note that we obtained different results from those reported by Petko et al. ( 2017 ). In their study, they found significant effects on the explained variances of ENTUSE, USESCH, and ICTATTPOS but not of HOMESCH for Switzerland. In the present study (Model 3), HOMESCH and USESCH were significant predictors but not ENTUSE, and for attitude toward ICT, all but INTICT were significant predictors of the variance. However, factors corresponding to ICT use were negatively associated with test performance, as in the study by Petko et al. ( 2017 ). Similarly, all components of attitude toward ICT positively affected science test scores, except for students’ ICT as a topic in social interaction.

Based on the AIC values, Model 4, including ICT use, attitude toward ICT, and digital dishonesty, was the better fit for the data. The parameter ( b  = 498.00, SE b  = 3.550) shows that our sample falls within the 95% confidence interval and that we can reject the null hypothesis. In this model, all factors except the use of ICT outside of school for leisure were significant predictors of explained variance in science test scores. These results are consistent with those reported by Petko et al. ( 2017 ), in which more frequent use of ICT negatively affected science test scores, with an overall positive effect of positive attitude toward ICT. Further, we observed that homework avoidance with digital resources strongly negatively affected performance, with lower performance associated with students reporting a higher frequency of engagement in digital dishonesty practices.

For mathematics test scores, results from Models 2 and 3 showed a similar pattern than those for science, and Model 4 also explained the highest variance (41.2%). The results from Model 4 contrast with those found by Petko et al. ( 2017 ), as in this study, HOMESCH was the only significant variable of ICT use. Regarding attitudes toward ICT, only two measures (COMPICT and AUTICT) were significant positive factors in Model 4. As for science test scores, digital dishonesty practices were a significantly strong negative predictor. Students who reported cheating more frequently were more likely to perform poorly on mathematics tests.

The analyses of PISA test scores for reading in Model 2 was similar to that of science and mathematics, with ENTUSE being a non-significant predictor when we included only measures of ICT use as predictors. In Model 3, contrary to the science and mathematics test scores models, in which INICT was non-significant, all measures of attitude toward ICT were positively significant predictors. Nevertheless, as for science and mathematics, Model 4, which included digital dishonesty, explained the greater variance in reading test scores (42.2%). We observed that for reading, all predictors were significant in Model 4, with an overall negative effect of ICT use, a positive effect of attitude toward ICT—except for SOIAICT, and a negative effect of digital dishonesty on test scores. Interestingly, the detrimental effect of using digital resources to engage in dishonest homework completion was the strongest in reading test scores.

4 Discussion

In this study, we were able to provide descriptive statistics on the prevalence of digital dishonesty among secondary students in the Swiss sample of PISA 2018. Students from this country were selected because they received additional questions targeting both homework effort and the frequency with which they engaged in digital dishonesty when doing homework. Descriptive statistics indicated that fairly high numbers of students engage in dishonest homework practices, with 49.6% reporting digital dishonesty at least once or twice a week. The most frequently reported practice was copying answers from friends, which was undertaken at least once a month by more than two-thirds of respondents. Interestingly, the most infamous form of digital dishonesty, that is plagiarism by copy-pasting something from the internet (Evering & Moorman, 2012 ), was admitted to by close to half of the students (49%). These results for homework avoidance are close to those obtained by previous research on digital academic plagiarism (e.g., McCabe et al., 2001 ).

We then investigated what makes a cheater, based on students’ demographics and effort put in doing their homework (H1), before looking at digital dishonesty as an additional ICT predictor of PISA test scores (mathematics, reading, and science) (H2).

The goal of our first research hypothesis was to determine student-related factors that may predict digital homework avoidance practices. Here, we focused on factors linked to students’ personal characteristics and study programs. Our multilevel model explained about 10% of the variance overall. Our analysis of which students are more likely to digital resources to avoid homework revealed an increased probability for male students who did not put much effort into doing their homework and who were studying in a program that was not oriented toward higher education. Thus, our findings tend to support results from previous research that stresses the importance of gender and motivational factors for academic dishonesty (e.g., Anderman & Koenka,  2017 ; Krou et al., 2021 ). Yet, as our model only explained little variance and more research is needed to provide an accurate representation of the factors that lead to digital dishonesty. Future research could include more aspects that are linked to learning, such as peer-related or teaching-related factors. Possibly, how closely homework is embedded in the teaching and learning culture may play a key role in digital dishonesty. Additional factors might be linked to the overall availability and use of digital tools. For example, the report combining factors from the PISA 2018 school and student questionnaires showed that the higher the computer–student ratio, the lower students scored in the general tests (OECD, 2020b ). A positive association with reading disappeared when socio-economic background was considered. This is even more interesting when considering previous research indicating that while internet access is not a source of divide among youths, the quality of use is still different based on gender or socioeconomic status (Livingstone & Helsper, 2007 ). Thus, investigating the usage-related “digital divide” as a potential source of digital dishonesty is an interesting avenue for future research (Dolan, 2016 ).

Our second hypothesis considered that digital dishonesty in homework completion can be regarded as an additional ICT-related trait and thus could be included in models targeting the influence of traditional ICT on PISA test scores, such as Petko et al. ( 2017 ) study. Overall, our results on the influence of ICT use and attitudes toward ICT on test scores are in line with those reported by Petko et al. ( 2017 ). Digital dishonesty was found to negatively influence test scores, with a higher frequency of cheating leading to lower performance in all major PISA test domains, and particularly so for reading. For each subject, the combined models explained about 40% of the total variance.

4.1 Conclusions and recommendations

Our results have several practical implications. First, the amount of cheating on homework observed calls for new strategies for raising homework engagement, as this was found to be a clear predictor of digital dishonesty. This can be achieved by better explaining the goals and benefits of homework, the adverse effects of cheating on homework, and by providing adequate feedback on homework that was done properly. Second, teachers might consider new forms of homework that are less prone to cheating, such as doing homework in non-digital formats that are less easy to copy digitally or in proctored digital formats that allow for the monitoring of the process of homework completion, or by using plagiarism software to check homework. Sometimes, it might even be possible to give homework and explicitly encourage strategies that might be considered cheating, for example, by working together or using internet sources. As collaboration is one of the 21st century skills that students are expected to develop (Bray et al., 2020 ), this can be used to turn cheating into positive practice. There is already research showing the beneficial impact of computer-supported collaborative learning (e.g., Janssen et al., 2012 ). Zhang et al. ( 2011 ) compared three homework assignment (creation of a homepage) conditions: individually, in groups with specific instructions, and in groups with general instructions. Their results showed that computer supported collaborative homework led to better performance than individual settings, only when the instructions were general. Thus, promoting digital collaborative homework could support the development of students’ digital and collaborative skills.

Further, digital dishonesty in homework needs to be considered different from cheating in assessments. In research on assessment-related dishonesty, cheating is perceived as a reprehensible practice because grades obtained are a misrepresentation of student knowledge, and cheating “implies that efficient cheaters are good students, since they get good grades” (Bouville, 2010 , p. 69). However, regarding homework, this view is too restrictive. Indeed, not all homework is graded, and we cannot know for sure whether students answered this questionnaire while considering homework as a whole or only graded homework (assessments). Our study did not include questions about whether students displayed the same attitudes and practices toward assessments (graded) and practice exercises (non-graded), nor did it include questions on how assessments and homework were related. By cheating on ungraded practice exercises, students will primarily hamper their own learning process. Future research could investigate in more depth the kinds of homework students cheat on and why.

Finally, the question of how to foster engaging homework with digital tools becomes even more important in pandemic situations. Numerous studies following the switch to home schooling at the beginning of the 2020 COVID-19 pandemic have investigated the difficulties for parents in supporting their children (Bol, 2020 ; Parczewska, 2021 ); however, the question of digital homework has not been specifically addressed. It is unknown whether the increase in digital schooling paired with discrepancies in access to digital tools has led to an increase in digital dishonesty practices. Data from the PISA 2018 student questionnaires (OECD, 2020a ) indicated that about 90% of students have a computer for schoolwork (OECD average), but the availability per student remains unknown. Digital homework can be perceived as yet another factor of social differences (see for example Auxier & Anderson,  2020 ; Thorn & Vincent-Lancrin, 2022 ).

4.2 Limitations and directions

The limitations of the study include the format of the data collected, with the accuracy of self-reports to mirror actual practices restricted, as these measures are particularly likely to trigger response bias, such as social desirability. More objective data on digital dishonesty in homework-related purposes could, for example, be obtained by analyzing students’ homework with plagiarism software. Further, additional measures that provide a more complete landscape of contributing factors are necessary. For example, in considering digital homework as an alternative to traditional homework, parents’ involvement in homework and their attitudes toward ICT are factors that have not been considered in this study (Amzalag, 2021 ). Although our results are in line with studies on academic digital dishonesty, their scope is limited to the Swiss context. Moreover, our analyses focused on secondary students. Results might be different with a sample of younger students. As an example, Kiss and Teller ( 2022 ) measured primary students cheating practices and found that individual characteristics were not a stable predictor of cheating between age groups. Further, our models included school as a random component, yet other group variables, such as class and peer groups, may well affect digital homework avoidance strategies.

The findings of this study suggest that academic dishonesty when doing homework needs to be addressed in schools. One way, as suggested by Chow et al. ( 2021 ) and Djokovic et al. ( 2022 ), is to build on students’ practices to explain which need to be considered cheating. This recommendation for institutions to take preventive actions and explicit to students the punishment faced in case of digital academic behavior was also raised by Chiang et al. ( 2022 ). Another is that teachers may consider developing homework formats that discourage cheating and shortcuts (e.g., creating multimedia documents instead of text-based documents, using platforms where answers cannot be copied and pasted, or using advanced forms of online proctoring). It may also be possible to change homework formats toward more open formats, where today’s cheating practices are allowed when they are made transparent (open-book homework, collaborative homework). Further, experiences from the COVID-19 pandemic have stressed the importance of understanding the factors related to the successful integration of digital homework and the need to minimize the digital “homework gap” (Auxier & Anderson, 2020 ; Donnelly & Patrinos, 2021 ). Given that homework engagement is a core predictor of academic dishonesty, students should receive meaningful homework in preparation for upcoming lessons or for practicing what was learned in past lessons. Raising student’s awareness of the meaning and significance of homework might be an important piece of the puzzle to honesty in learning.

Data availability

The data that support the findings of this study are openly available in SISS base at https://doi.org/10.23662/FORS-DS-1285-1 , reference number 1285.

Agasisti, T., Gil-Izquierdo, M., & Han, S. W. (2020). ICT Use at home for school-related tasks: What is the effect on a student’s achievement? Empirical evidence from OECD PISA data. Education Economics, 28 (6), 601–620. https://doi.org/10.1080/09645292.2020.1822787

Article   Google Scholar  

Amzalag, M. (2021). Parent attitudes towards the integration of digital learning games as an alternative to traditional homework. International Journal of Information and Communication Technology Education, 17 (3), 151–167. https://doi.org/10.4018/IJICTE.20210701.oa10

Anderman, E. M., & Koenka, A. C. (2017). The relation between academic motivation and cheating. Theory into Practice, 56 (2), 95–102. https://doi.org/10.1080/00405841.2017.1308172

Aparicio, J., Cordero, J. M., & Ortiz, L. (2021). Efficiency analysis with educational data: How to deal with plausible values from international large-scale assessments. Mathematics, 9 (13), 1–16. https://doi.org/10.3390/math9131579

Arpacı, S., Mercan, F., & Arıkan, S. (2021). The differential relationships between PISA 2015 science performance and, ICT availability, ICT use and attitudes toward ICT across regions: evidence from 35 countries. Education and Information Technologies, 26 (5), 6299–6318. https://doi.org/10.1007/s10639-021-10576-2

Auxier, B., & Anderson, M. (2020, March 16). As schools close due to the coronavirus, some U.S. students face a digital “homework gap”. Pew Research Center, 1–8.  http://www.pewresearch.org/fact-tank/2018/10/19/5-charts-on-global-views-of-china/ . Retrieved November 29th, 2021

Baş, G., Şentürk, C., & Ciğerci, F. M. (2017). Homework and academic achievement: A meta-analytic review of research. Issues in Educational Research, 27 (1), 31–50.

Google Scholar  

Blau, I., & Eshet-Alkalai, Y. (2017). The ethical dissonance in digital and non-digital learning environments: Does technology promotes cheating among middle school students? Computers in Human Behavior, 73, 629–637. https://doi.org/10.1016/j.chb.2017.03.074

Bol, T. (2020). Inequality in homeschooling during the Corona crisis in the Netherlands. First results from the LISS Panel. https://doi.org/10.31235/osf.io/hf32q

Bouville, M. (2010). Why is cheating wrong? Studies in Philosophy and Education, 29 (1), 67–76. https://doi.org/10.1007/s11217-009-9148-0

Bray, A., Byrne, P., & O’Kelly, M. (2020). A short instrument for measuring students’ confidence with ‘key skills’ (SICKS): Development, validation and initial results. Thinking Skills and Creativity, 37 (June), 100700. https://doi.org/10.1016/j.tsc.2020.100700

Chen, C. M., & Chen, F. Y. (2014). Enhancing digital reading performance with a collaborative reading annotation system. Computers and Education, 77, 67–81. https://doi.org/10.1016/j.compedu.2014.04.010

Cheng, Y. C., Hung, F. C., & Hsu, H. M. (2021). The relationship between academic dishonesty, ethical attitude and ethical climate: The evidence from Taiwan. Sustainability (Switzerland), 13 (21), 1–16. https://doi.org/10.3390/su132111615

Chiang, F. K., Zhu, D., & Yu, W. (2022). A systematic review of academic dishonesty in online learning environments. Journal of Computer Assisted Learning , 907–928. https://doi.org/10.1111/jcal.12656

Chow, H. P. H., Jurdi-Hage, R., & Hage, H. S. (2021). Justifying academic dishonesty: A survey of Canadian university students. International Journal of Academic Research in Education , December. https://doi.org/10.17985/ijare.951714

Cuadrado, D., Salgado, J. F., & Moscoso, S. (2019). Prevalence and correlates of academic dishonesty: Towards a sustainable university. Sustainability (Switzerland) , 11 (21). https://doi.org/10.3390/su11216062

Cuadrado, D., Salgado, J. F., & Moscoso, S. (2021). Personality, intelligence, and counterproductive academic behaviors: A meta-analysis. Journal of Personality and Social Psychology, 120 (2), 504–537. https://doi.org/10.1037/pspp0000285

Djokovic, R., Janinovic, J., Pekovic, S., Vuckovic, D., & Blecic, M. (2022). Relying on technology for countering academic dishonesty: the impact of online tutorial on students’ perception of academic misconduct. Sustainability (Switzerland) , 14 (3). https://doi.org/10.3390/su14031756

Dolan, J. E. (2016). Splicing the divide: A review of research on the evolving digital divide among K–12 students. Journal of Research on Technology in Education, 48 (1), 16–37.

Donnelly, R., & Patrinos, H. A. (2021). Learning loss during Covid-19: An early systematic review. Prospects , 0123456789 . https://doi.org/10.1007/s11125-021-09582-6

Ercegovac, Z., & Richardson, J. V. (2004). Academic dishonesty, plagiarism included, in the digital age: A literature review. College & Research Libraries, 65 (4), 301–318. https://doi.org/10.5860/crl.65.4.301

Erzinger, A. B., Verner, M., König, N., Petrucci, F., Nidegger, C., Roos, E., & Salvisberg, M. (2019). PISA 2018: Les élèves de Suisse en comparaison internationale . SEFRI/CDIP et Consortium PISA.ch.

Erzinger, A. B., Verner, M., Salvisberg, M., Nidegger, C., & Seiler, S. (2021). PISA 2018 in Switzerland, add-on to the international dataset: Swiss specific variables [Dataset] . FORS. https://doi.org/10.23662/FORS-DS-1285-1

Evering, L. C., & Moorman, G. (2012). Rethinking plagiarism in the digital age. Journal of Adolescent & Adult Literacy, 56 (1), 35–44.

Fan, H., Xu, J., Cai, Z., He, J., & Fan, X. (2017). Homework and students’ achievement in math and science: A 30-year meta-analysis, 1986–2015. Educational Research Review, 20, 35–54. https://doi.org/10.1016/j.edurev.2016.11.003

Fernández-Alonso, R., álvarez-Díaz, M., Suárez-álvarez, J., & Muñiz, J. (2017). Students’ achievement and homework assignment strategies. Frontiers in Psychology, 8 (MAR), 1–11. https://doi.org/10.3389/fpsyg.2017.00286

Fernández-Alonso, R., Suárez-Álvarez, J., & Muñiz, J. (2015). Adolescents’ homework performance in mathematics and science: Personal factors and teaching practices. Journal of Educational Psychology, 107 (4), 1075–1085. https://doi.org/10.1037/edu0000032

Giluk, T. L., & Postlethwaite, B. E. (2015). Big Five personality and academic dishonesty: A meta-analytic review. Personality and Individual Differences, 72, 59–67. https://doi.org/10.1016/j.paid.2014.08.027

Husain, F. M., Al-Shaibani, G. K. S., & Mahfoodh, O. H. A. (2017). Perceptions of and attitudes toward plagiarism and factors contributing to plagiarism: A review of studies. Journal of Academic Ethics, 15 (2), 167–195. https://doi.org/10.1007/s10805-017-9274-1

Isakov, M., & Tripathy, A. (2017). Behavioral correlates of cheating: Environmental specificity and reward expectation. PLoS One1, 12 (10), 6–11. https://doi.org/10.1371/journal.pone.0186054

Jacobs, J. E., Lanza, S., Osgood, D. W., Eccles, J. S., & Wigfield, A. (2002). Changes in children’s self-competence and values: Gender and domain differences across grades one through twelve. Child Development, 73 (2), 509–527. https://doi.org/10.1111/1467-8624.00421

Janssen, J., Erkens, G., Kirschner, P., & Kanselaar, G. (2012). Task-related and social regulation during online collaborative learning. Metacognition and Learning, 7 (1), 25–43. https://doi.org/10.1007/s11409-010-9061-5

Josephson Institute of Ethics (2012). 2012 Report card on the ethics of American youth .  https://charactercounts.org/wp-content/uploads/2014/02/ReportCard-2012-DataTables.pdf . Retrieved January 24th, 2022

Kam, C. C. S., Hue, M. T., & Cheung, H. Y. (2018). Academic dishonesty among Hong Kong secondary school students: Application of theory of planned behavior. Educational Psychology, 38 (7), 945–963. https://doi.org/10.1080/01443410.2018.1454588

Kapoor, H., & Kaufman, J. C. (2021). Are cheaters common or creative?: Person-situation interactions of resistance in learning contexts. Journal of Academic Ethics, 19 (2), 157–174. https://doi.org/10.1007/s10805-020-09379-w

Kiss, H. J., & Keller, T. J. (2022). Individual characteristics do (not) matter in cheating. Available at SSRN 4001278. https://doi.org/10.2139/ssrn.4001278

Krou, M. R., Fong, C. J., & Hoff, M. A. (2021). Achievement motivation and academic dishonesty: A meta-analytic investigation. Educational Psychology Review, 33 (2), 427–458. https://doi.org/10.1007/s10648-020-09557-7

Kunina-Habenicht, O., & Goldhammer, F. (2020). ICT engagement: A new construct and its assessment in PISA 2015. Large-Scale Assessments in Education , 8 (1). https://doi.org/10.1186/s40536-020-00084-z

Livingstone, S., & Helsper, E. (2007). Gradations in digital inclusion: Children, young people and the digital divide. New Media and Society, 9 (4), 671–696. https://doi.org/10.1177/1461444807080335

Ma, H. J., Wan, G., & Lu, E. Y. (2008). Digital cheating and plagiarism in schools. Theory into Practice, 47 (3), 197–203. https://doi.org/10.1080/00405840802153809

Martin, A. J., Ginns, P., & Papworth, B. (2017). Motivation and engagement: Same or different? Does it matter? Learning and Individual Differences, 55, 150–162. https://doi.org/10.1016/j.lindif.2017.03.013

McCabe, D. L. (2005). It takes a village: Academic dishonesty & educational opportunity. Liberal Education, 91 (3), 26–31.

McCabe, D. L., Treviño, L. K., & Butterfield, K. D. (2001). Cheating in academic institutions: A decade of research. Ethics and Behavior, 11 (3), 219–232. https://doi.org/10.1207/S15327019EB1103_2

Moss, S. A., White, B., & Lee, J. (2018). A systematic review into the psychological causes and correlates of plagiarism. Ethics and Behavior, 28 (4), 261–283. https://doi.org/10.1080/10508422.2017.1341837

Nora, W. L. Y., & Zhang, K. C. (2010). Motives of cheating among secondary students: The role of self-efficacy and peer influence. Asia Pacific Education Review, 11 (4), 573–584. https://doi.org/10.1007/s12564-010-9104-2

Odell, B., Cutumisu, M., & Gierl, M. (2020). A scoping review of the relationship between students’ ICT and performance in mathematics and science in the PISA data. Social Psychology of Education , 23 (6). https://doi.org/10.1007/s11218-020-09591-x

OECD, & Publishing, O. E. C. D. (2015). Students, computers and learning: Making the connection . PISA. https://doi.org/10.1787/factbook-2015-68-en

OECD (2019a). Chapter 16. Scaling procedures and construct validation of context questionnaire data. In PISA 2018 Technical Report . OECD.

OECD (2019b). PISA 2018 Results - What school life means for students’ life (Vol. III). OECD Publishing.  https://www.oecd.org/pisa/publications/PISA2018_CN_IDN.pdf . Retrieved October 20th, 2021

OECD (2020a). Learning remotely when schools close . 1–13.  https://read.oecd-ilibrary.org/view/?ref=127_127063-iiwm328658&title=Learning-remotely-when-schools-close . Retrieved November 29th, 2021

OECD (2020b). PISA 2018 Results: Effective policies, successful schools (Vol. V). PISA, OECD Publishing. https://doi.org/10.1787/ca768d40-en

Parczewska, T. (2021). Difficult situations and ways of coping with them in the experiences of parents homeschooling their children during the COVID-19 pandemic in Poland. Education 3–13 , 49 (7), 889–900. https://doi.org/10.1080/03004279.2020.1812689

Pavela, G. (1997). Applying the power of association on campus: A model code of academic integrity. Law and Policy, 24 (1), 1–22.

Petko, D., Cantieni, A., & Prasse, D. (2017). Perceived quality of educational technology matters: A secondary analysis of students ICT use, ICTRelated attitudes, and PISA 2012 test scores. Journal of Educational Computing Research, 54 (8), 1070–1091. https://doi.org/10.1177/0735633116649373

Rosário, P., Carlos Núñez, J., Vallejo, G., Nunes, T., Cunha, J., Fuentes, S., & Valle, A. (2018). Homework purposes, homework behaviors, and academic achievement. Examining the mediating role of students’ perceived homework quality. Contemporary Educational Psychology, 53 (April), 168–180. https://doi.org/10.1016/j.cedpsych.2018.04.001

Schnyder, I., Niggli, A., & Trautwein, U. (2008). Hausaufgabenqualität im Französischunterricht aus der Sicht von Schülern, Lehrkräften und Experten und die Entwicklung von Leistung, Hausaufgabensorgfalt und Bewertung der Hausaufgaben. Zeitschrift Fur Padagogische Psychologie, 22 (3–4), 233–246. https://doi.org/10.1024/1010-0652.22.34.233

Schynder Godel, I. (2015). Die Hausaufgaben unter der Lupe. Eine empirische Untersuchung im Fach Französisch als Fremdsprache.

Skryabin, M., Zhang, J., Liu, L., & Zhang, D. (2015). How the ICT development level and usage influence student achievement in reading, mathematics, and science. Computers and Education, 85, 49–58. https://doi.org/10.1016/j.compedu.2015.02.004

Tarhini, A., Hone, K., & Liu, X. (2014). Measuring the moderating effect of gender and age on e-learning acceptance in England: A structural equation modeling approach for an extended technology acceptance model. Journal of Educational Computing Research, 51 (2), 163–184. https://doi.org/10.2190/EC.51.2.b

Thorn, W., & Vincent-Lancrin, S. (2022). Education in the time of COVID-19 in France, Ireland, the Unites Kingdom and the United States: The nature and impact of remote learning. In F. M. Reimers (Ed.), Primary and secondary education during Covid-19 (pp. 383–420). Springer. https://doi.org/10.1007/978-981-13-2632-5_2

Trautwein, U. (2007). The homework-achievement relation reconsidered: Differentiating homework time, homework frequency, and homework effort. Learning and Instruction, 17 (3), 372–388. https://doi.org/10.1016/j.learninstruc.2007.02.009

Trautwein, U., & Köller, O. (2003). Was lange währt, wird nicht immer gut: Zur Rolle selbstregulativer Strategien bei der Hausaufgabenerledigung. Zeitschrift Für Pädagogische Psychologie German Journal of Educational Psychology, 17 (3–4), 199–209.

Trautwein, U., Lüdtke, O., Schnyder, I., & Niggli, A. (2006). Predicting homework effort: Support for a domain-specific, multilevel homework model. Journal of Educational Psychology, 98 (2), 438–456. https://doi.org/10.1037/0022-0663.98.2.438

Trautwein, U., Schnyder, I., Niggli, A., Neumann, M., & Lüdtke, O. (2009). Chameleon effects in homework research: The homework-achievement association depends on the measures used and the level of analysis chosen. Contemporary Educational Psychology, 34 (1), 77–88. https://doi.org/10.1016/j.cedpsych.2008.09.001

Waltzer, T., & Dahl, A. (2022). Why do students cheat? Perceptions, evaluations, and motivations. Ethics and Behavior , 1–21. https://doi.org/10.1080/10508422.2022.2026775

Whitley, B. E., Nelson, A. B., & Jones, C. J. (1999). Gender differences in cheating attitudes and classroom cheating behavior: A meta-analysis. Sex Roles, 41 (9–10), 657–680. https://doi.org/10.1023/A:1018863909149

Xu, J. (2015). Investigating factors that influence conventional distraction and tech-related distraction in math homework. Computers and Education, 81, 304–314. https://doi.org/10.1016/j.compedu.2014.10.024

Xu, J., Du, J., Cunha, J., & Rosário, P. (2021). Student perceptions of homework quality, autonomy support, effort, and math achievement: Testing models of reciprocal effects. Teaching and Teacher Education , 108 . https://doi.org/10.1016/j.tate.2021.103508

Yaniv, G., Siniver, E., & Tobol, Y. (2017). Do higher achievers cheat less? An experiment of self-revealing individual cheating. Journal of Behavioral and Experimental Economics, 68, 91–96. https://doi.org/10.1016/j.socec.2017.04.005

Zhang, L., Ayres, P., & Chan, K. (2011). Examining different types of collaborative learning in a complex computer-based environment: A cognitive load approach. Computers in Human Behavior, 27 (1), 94–98. https://doi.org/10.1016/j.chb.2010.03.038

Download references

Open access funding provided by University of Zurich

Author information

Authors and affiliations.

Institute of Education, University of Zürich, Zürich, Switzerland

Juliette C. Désiron & Dominik Petko

You can also search for this author in PubMed   Google Scholar

Contributions

Juliette C. Désiron: Formal analysis, Writing (Original, Review and Editing), Dominik Petko: Conceptualization, Writing (Original, Review and Editing), Supervision.

Corresponding author

Correspondence to Juliette C. Désiron .

Ethics declarations

Competing of interests, additional information, publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

List of abbreviations related to PISA datasets

students’ perceived autonomy related to ICT use

students’ perceived ICT competence

frequency of computer use at home for entertainment purposes

index of economic, social, and cultural status (computed from PARED, HISEI and HOMEPOS)

parents’ highest occupational status

home possessions

frequency of computer use for school-related purposes at home

digital cheating for homework items for Switzerland

homework engagement items for Switzerland

positive attitude towards ICT as a learning tool

student’s ICT interest

parents’ highest level of education

students’ ICT as a topic in social interaction

frequency of computer use at school

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Désiron, J.C., Petko, D. Academic dishonesty when doing homework: How digital technologies are put to bad use in secondary schools. Educ Inf Technol 28 , 1251–1271 (2023). https://doi.org/10.1007/s10639-022-11225-y

Download citation

Received : 11 May 2022

Accepted : 05 July 2022

Published : 23 July 2022

Issue Date : February 2023

DOI : https://doi.org/10.1007/s10639-022-11225-y

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Academic dishonesty
• Digitally-supported cheating
• Secondary education
• Find a journal
• Publish with us
• Track your research

Recent Findings Suggest Math Homework May Negatively Affect Students and Families

R ecent research suggests that math homework, particularly when complex, might have detrimental effects on students and their families. The study, which involved a collaboration between the University of South Australia and St Francis Xavier University in Canada, engaged eight Canadian households in interviews exploring their experiences with mathematics homework and its influence on familial relations.

Children in grade 3, generally ages 8 or 9, were the focus, as this is the age at which children in the study area first encounter standardized math exams. The research found that math is perceived negatively by many and often seen as demanding extra effort.

“Homework is traditionally seen as conducive to reinforcing learning and aiding academic achievement,” remarks Lisa O’Keeffe, a mathematics education specialist from the University of South Australia.

“However, when homework goes beyond students’ capabilities, even with parental aid, questions arise regarding the purpose of assigning such homework,” she continues.

Families noted the difficulties presented by the homework, which often resulted in delayed bedtimes, infringed on family time, and elicited feelings of frustration and insufficiency.

Changing approaches to teaching mathematics can also create friction, particularly between parents who learned differently than their children are being taught today.

Mathematics instruction has indeed progressed, states O’Keeffe. The disparity between current and past teaching methods can create additional stress as parents struggle to accommodate new instructional approaches.

According to the study, this can result in intergenerational negativity. The responsibility for helping with homework largely fell to mothers, and when they found the tasks challenging, this potentially bolstered negative math stereotypes , especially the notion that females don’t excel in mathematics as “naturally” as males.

Such stereotypes can critically affect academic performance and future career paths .

The study, albeit with a limited sample size, reflects broader educational conversations. The authors advocate for better-tailored math homework assignments to prevent young students from being discouraged from math at an early age.

“Our objective isn’t to dissuade girls from nurturing a strong mathematical identity,” asserts education lecturer Sarah McDonald from the University of South Australia. Hence, there’s a need for a deeper understanding of homework policies and expectations.

While homework is often believed to teach non-academic skills, like independence, time management, and self-regulation, the study’s findings question this assumption.

The complete study is available in the British Journal of Sociology of Education .

FAQs about Math Homework Impact

The study implies that math homework that is too complex can cause stress, frustration, and interfere with family life and bedtimes.

Parents taught using different methodologies can struggle to help their children with homework, leading to frustration and the reinforcement of negative stereotypes.

Negative stereotypes, especially concerning gender, can influence a child’s academic performance and interest in pursuing math-related careers.

The study calls for better understanding and reform of homework policies to ensure that math homework is beneficial and not discouraging.

This recent study underscores the complexity surrounding math homework and its role in educational development. While homework has traditionally been viewed as a tool for reinforcing classroom learning, the findings from the University of South Australia and St Francis Xavier University highlight the potential negatives, including how it can exacerbate stress within families, create generational friction due to evolving teaching methods, and even reinforce negative gender stereotypes. There is a clear call-to-action for educators to reassess the assignment of math homework, ensuring that it aligns with the capability of students and supports positive learning outcomes, particularly for those at a formative stage in their education. Mindful consideration of these findings could help shape future homework policies that foster both academic growth and a healthy home environment.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Front Psychol

Effects of homework creativity on academic achievement and creativity disposition: Evidence from comparisons with homework time and completion based on two independent Chinese samples

Huiyong fan.

1 College of Educational Science, Bohai University, Jinzhou, China

2 Research Center of Brain and Cognitive Neuroscience, Liaoning Normal University, Dalian, China

Jianzhong Xu

3 Department of Counseling, Educational Psychology, and Foundations, College of Education, Mississippi State University, MS, United States

Shengli Guo

Associated data.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

During the past several decades, the previous studies have been focusing on the related theoretical issues and measuring tool of homework behaviors (mainly including homework time, completion, and homework creativity). However, the effects of these homework behaviors on general creativity remain unknown. Employing a number of questionnaires, this study investigated two samples from middle schools of Mainland China. The results showed that (1) the eight-item version of Homework Creativity Behaviors Scale had acceptable validity and reliability; (2) compared with homework completion and homework time, homework creativity explained less variety of academic achievement (3.7% for homework creativity; 5.4% for completion and time); (3) homework creativity explained more variance of general creativity than that of homework completion and homework time accounted (7.0% for homework creativity; 1.3% for completion and time); and (4) homework creativity was negatively associated with grade level. Contrary to the popular beliefs, homework completion and homework creativity have positive effects on the students’ general creativity. Several issues that need further studies were also discussed.

Introduction

Homework is an important part of the learning and instruction process. Each week, students around the world spend 3–14 hours on homework, with an average of 5 hours a week ( Dettmers et al., 2009 ; OECD, 2014 ). The results of the previous studies and meta-analysis showed that the homework time is correlated significantly with students’ gains on the academic tests ( Cooper et al., 2012 ; Fan et al., 2017 ; Fernández-Alonso et al., 2019 ).

Homework is a multi-faceted process which has many attributes – each attribute can be identified, defined, and measured independently ( Guo and Fan, 2018 ). Some attributes, such as homework time ( Núñez et al., 2013 ; Kalenkoski and Pabilonia, 2017 ), homework frequency ( Fernández-Alonso et al., 2015 ), homework completion ( Rosário et al., 2015 ), homework effort ( Trautwein and Lüdtke, 2007 ; Fernández-Alonso et al., 2015 ), homework purpose ( Trautwein and Lüdtke, 2009 ; Xu, 2010 , 2021 ), homework performance and problems ( Power et al., 2007 ), homework management behavior ( Xu, 2008 ), homework expectation ( Xu, 2017 ), and self-regulation of homework behavior ( Yang and Tu, 2020 ), have been well recorded in the literature, and operationally defined and measured.

Recently, a research community has noticed the “creativity” in homework (in short form, “homework creativity”) who have raised some speculations about its effects on students’ academic achievement and general creativity disposition ( Kaiipob, 1951 ; Beghetto and Kaufman, 2007 ; Kaufman and Beghetto, 2009 ; Guo, 2018 ; Guo and Fan, 2018 ; Chang, 2019 ). However, the scientific measurement of homework creativity has not been examined systematically. The relationship between homework creativity, academic achievement, and general creativity disposition, as well as the grade difference in homework creativity, are still in the state of conjectures consequently.

As a scientific probe to homework creativity, this study included three main sections. In the “Literature Review” section, the conceptualization and relevant measurement of homework creativity were summarized; the relationship between homework behaviors and academic achievements, general creativity, and the grade difference in homework behaviors and general creativity were also evaluated. These four main results related to the four research questions were also presented in the body of this article. They are reliability and validity of homework creativity behavior scale (HCBS), the relationships between the scores of HCBS and those of general creativity and academic achievement, and the grade effects of scores of HCBS. In the “Discussion” section, the scientific contributions and interpretations of the findings of this study were elaborated.

Homework creativity

Conceptual background of homework creativity.

As an attribute of homework process, homework creativity refers to the novelty and uniqueness of homework ( Guo and Fan, 2018 ). Specifically, the ways relating to homework creativity with extant theoretical literature are presented below.

First, creativity is a natural part of homework process which serves as a sub-process of learning. Guilford (1950) is the first psychologist who linked creativity with learning, pointing out that the acquisition of creativity is a typical quality of human learning, and that a complete learning theory must take creativity into account.

Second, according to the Four-C Model of Creativity (e.g., Kaufman and Beghetto, 2009 ), the homework creativity can be divided mainly into the category of “Transformative Learning” (Mini-C creativity), which is different from the “Everyday Innovation” (Little-C creativity), “Professional Expertise” (Pro-C creativity), or “Eminent Accomplishments” (Big-C creativity, Beghetto and Kaufman, 2007 ; Kaufman and Beghetto, 2009 ; Kozbelt et al., 2011 ).

The Mini-C is defined as a type of intrapersonal creativity which has personal meaning, not solid contribution or breakthrough in a field ( Beghetto and Kaufman, 2007 , p. 76, Table 1 ). The most important point which distinguishes Mini-C from other types of creativity is the level of novelty of product. The Mini-C creativity involves the personal insight or interpretation which is new to a particular individual, but may be ordinary to others. The Little-C creativity refers to any small, but solid innovation in daily life. The Pro-C creativity is represented in the form of professional contribution which is still not a breakthrough. The Big-C creativity generates a real breakthrough appears in some field which is considered as something new to all human beings. The other difference is related with the subjects of sub-types of creativity. The Mini-C creativity mainly happens in all kinds of students. The Little-C creativity can be widely found in normal people. The Pro-C creativity’s masters are those who are proficient in some field. The Big-C creativity is related frequently with those giants who has made eminent contribution to human being.

Basic information of samples 1 and 2 included.

The Mini-C creativity frequently happens in learning process. When the contribution of the Mini-C creativity grows big enough, it can move into the category of the Little-C creativity, or the Big-C creativity. Most homework creativity is of Mini-C creativity, and of which a small part may grow as the Little-C and Big-C creativities. For example, when students independently find a unique solution to a problem in homework which has scientific meaning, a Little-C or Big-C occurs.

Third, the education researchers have observed homework creativity for many years and been manipulating them in educational practice. Kaiipob (1951) described that homework is a semi-guide learning process in which homework such as composition, report, public speech, difficult and complex exercises, experiments, and making tools and models consumes a lot of time and accelerate the development of students’ creativity disposition (p. 153).

In the recent years, creativity has become a curriculum or instruction goal in many countries (the case of United Kingdom, see Smith and Smith, 2010 ; Chinese case, see Pang and Plucker, 2012 ). Homework is the most important way that accomplish this goal. Considering Chinese in primary and secondary schools in China as an example, the curriculum standards have clearly required homework to cultivate students’ creative spirit, creative thinking, and ability to imagination since the year 2000. The results of Qian’s (2006) investigation revealed that the percent of these creative homework items in each unit fluctuates between 29 and 45%.

Previous instruments of homework behaviors

Those existent instruments measuring homework behavior can be divided into the following two categories: The single-indicator instruments and the multi-dimension instruments ( Guo and Fan, 2018 ). The single-indicator instruments employ only one item to measure homework attributes, such as homework time (e.g., Trautwein and Lüdtke, 2007 ), homework frequency (e.g., De Jong et al., 2000 ), homework completion (e.g., Xu et al., 2019 ), and effort (e.g., Liu et al., 2013 ).

The typical multi-dimension instruments include Homework Process Inventory ( Cooper et al., 1998 ), Homework Purpose Scale ( Xu, 2010 ), Homework Performance Questionnaire ( Pendergast et al., 2014 ), Homework Management Scale (HMS; Xu and Corno, 2003 ), Homework Evaluating Scale ( Fernández-Alonso et al., 2015 ), Homework Problem Checklist ( Anesko et al., 1987 ), Science Homework Scale ( Tas et al., 2016 ), Homework Expectancy Value Scale ( Yang and Xu, 2017 ), and Online Homework Distraction Scale ( Xu et al., 2020 ).

Although the previous tools measured some dimensions of homework ( Guo and Fan, 2018 ), there is hardly any tool that can be employed to gauge the homework creativity. Guo and Fan (2018) extracted several attributes (i.e., time, completion, quality, purpose, effort, creativity, sociality, liking) represented in the existent instruments of homework behaviors, and put forth a multi-faceted model of homework behaviors which intuitionally predicts the existence of homework creativity.

Under the guideline of the multi-faceted model ( Guo and Fan, 2018 ), Guo (2018) developed a multi-dimensional homework behavior instrument, which detected the homework creativity as a dimension in the homework behavior of middle school students. A typical item of homework creativity in Guo (2018) is “The way I do my homework is different from others.” The subscale homework creativity reported by Guo (2018) needs to be improved because it has a small number of items with lower reliability.

Following Guo’s (2018) work, Chang (2019) conducted a new investigation focusing on homework creativity behavior. Using an open-ended questionnaire, a total of 30 students from primary, middle, and high schools were invited to answer this question, that is, “What characteristics can be considered as creative in the process of completing the homework?” Here, “creativity” refers to novelty, uniqueness, and high quality. A group of 23 specific behaviors were reported, among which the top 10 are as follows: Learning by analogy, open minded, one question with multiple solutions, unique solution, summarizing the cause of errors, constructing a personal understanding, analyzing knowledge points clearly, classifying homework contents, making more applications, having rich imagination, and a neat handwriting (see Chang, 2019 , Table 4 , p. 14). Based on these results of open-ended questionnaire, Chang (2019) invented a nine-item scale (see Table 1 and Supplementary Table S3 for details) called as the HCBS which has a good reliability coefficient (α = 0.87).

Regression analyses of homework creative behavior on academic achievement and general creativity.

AA, academic achievement; WCAPt, total score of WCAP; TWk, time spent on homework in week days; TWw, time spent on homework in weekend; HCp, homework completion; HCb, homework creativity behavior.

Previous studies on the relationship between homework behaviors and academic achievement

In the literature, homework behaviors is one cluster of variables typically including homework time, homework completion, effort, purpose, frequency, etc. Academic achievement is an outcome of homework which is operationally measured using the scores on the standardized tests, or non-standardized tests (including final examinations, or teachers’ grades, or estimations by participants themselves, those forms were used widely in the literature, see Fan et al., 2017 ). Academic achievement may be affected by a lot of factors inherited in the process of learning (see Hattie, 2009 for an overview of its correlates). The relationship between homework behaviors and academic achievement is one of the most important questions in homework field, because it is related to the effectiveness of homework ( Cooper et al., 2006 , 2012 ; Fan et al., 2017 ).

Most of the previous studies focused on the relationship between homework time and academic achievement. Cooper et al. (2006) synthesized the primary studies published from 1989 to 2003, and found that the correlation between homework time of America students and their academic achievement was about 0.15. Fan et al. (2017) reviewed those individual studies published before June 2015, and reported that the averaged correlation between homework time of international students and their science, technology, engineering, and mathematics (STEM) academic achievement was about 0.20. Fernández-Alonso et al. (2017) investigated a representative sample of Spanish students (more than 26,000), and the results of multi-level analysis indicated that the correlation between homework time and academic achievement was negative at student level, but positive at school level ( r = 0.16). Fernández-Alonso et al. (2019) took a survey on a big sample from 16 countries from Latin America, and reported that the relationship between homework time and academic achievement was very weak. Valle et al. (2019) analyzed the homework time, time management, and achievement of 968 Spain students finding that homework time management was positively related to academic achievement. Taken all these together, we will find that the homework has some small significant correlations with academic achievement, the average r = 0.15.

The correlation between homework completion and academic achievement has also been investigated for decades. Based on a review of 11 primary studies, Fan et al. (2017) reported a high correlation of 0.59 between them. Rosário et al. (2015) investigated 638 students, and demonstrated a correlation of 0.22 between amount of homework completed and math test scores. Xu et al. (2019) took a survey using a sample of 1,450 Chinese eighth graders, and found that the correlations between homework completion and the gains in math test scores ranged from 0.25 to 0.28. Dolean and Lervag (2022) employed the Randomized Controlled Trial design, and demonstrated that amount of homework completed has immediate effect on writing competency in which the effect of moderate amount of homework can last for 4 months. Integrating the aforementioned results, we can find that the averaged correlation between homework completion and academic achievement was higher than that between homework time with academic achievement.

Homework effort was also found to be correlated with academic achievement. Fan et al. (2017) reviewed four primary studies and returned that a medium correlation ( r = 0.31) between homework effort and academic achievement. Two recent investigations showed that this relationship is positively and reciprocally related ( r = 0.41–0.42) ( Xu, 2020 ; Xu et al., 2021 ).

The effect of homework purpose was also correlated with the academic achievement. Fan et al. (2017) summarized four existent primary studies and reported an averaged correlation of 0.11 between them. Later, Rosário et al. (2015) found a similar correlation coefficient of these two variables on a sample of 638 students. Xu’s (2018) investigation revealed that the correlation between purpose and academic achievement was about 0.40. Sun et al. (2021) investigated a larger sample ( N = 1,365), and found that the subscales of homework purpose had different correlation patterns with academic achievement (academic purpose is 0.40, self-regulatory purpose is 0.20, and approval-seeking purpose is 0.10).

Considering the case of homework creativity, there is only one study preliminarily investigated its relationship with academic achievement. Guo (2018) investigated a sample of 1,808 middle school students, and reported a significant correlation between homework creativity and academic achievement ( r = 0.34, p < 0.05).

Previous studies on the relationship between homework behaviors and general creativity

General creativity refers to the psychological attributes which can generate novel and valuable products ( Kaufman and Glăveanu, 2019 ; Sternberg and Karami, 2022 ). These psychological attributes typically included attitude (e.g., willing to take appropriate risk), motivations (e.g., intrinsic motivation, curiosity), abilities (e.g., divergent thinking), and personality (e.g., independence) ( Kaufman and Glăveanu, 2019 ; Long et al., 2022 ). These attributes can be assessed independently, or in the form of grouping ( Plucker et al., 2019 ; Sternberg, 2019 ). For instance, the divergent thinking was measured independently ( Kaufman et al., 2008 ). Also, the willing to take appropriate risk was measured in tools contain other variables ( Williams, 1979 ). There are many studies examined the relationship between learning and general creativity in the past several decades indicating that the correlation between them was around 0.22 (e.g., Gajda et al., 2017 ; Karwowski et al., 2020 ).

Regarding the relationship between homework behaviors and general creativity, there are few studies which presented some contradictory viewpoints. Kaiipob (1951) posited that homework could accelerate development of students’ general creativity disposition, because the tasks in homework provide opportunities to exercise creativity. Cooper et al. (2012) argued that homework can diminish creativity. Furthermore, Zheng (2013) insisted that homework will reduce curiosity and the ability to challenging – the two core components of creativity. The preliminary results of Chang (2019) indicated that the score of HCBS is significantly correlated with scores of a test of general creativity, Williams’ creativity packet ( r = 0.25–0.33, p < 0.05).

Previous studies on the relationship between homework behaviors and homework creativity

In Guo and Fan’s (2018) theoretical work, homework creativity was combined from two independent words, homework and creativity, which was defined as a new attribute of homework process and was considered as a new member of homework behaviors. Up till now, there are two works providing preliminary probe to the relationship between homework behaviors and homework creativity. Guo (2018) investigated a sample of 1808 middle school students, and found that homework creativity was correlated significantly with liking ( r = 0.33), correctness ( r = 0.47), completion ( r = 0.57), and purpose ( r = 0.53). Based on another sample of Chinese students (elementary school students, N = 300; middle school students, N = 518; high school students, N = 386), Chang (2019) showed that the score of homework creativity was correlated significantly with homework time ( r = 0.11), completion ( r = 0.39), correctness ( r = 0.63), effort ( r = 0.73), social interaction ( r = 0.35), quality ( r = 0.69), interpersonal relation purpose ( r = 0.17), and purpose of personal development ( r = 0.41).

Previous studies on grade differences of homework behaviors and general creativity

Grade differences of homework behaviors.

As a useful indicator, homework time was recorded frequently (e.g., Cooper et al., 2006 ; Fan et al., 2017 ). A recent meta-analysis included 172 primary studies (total N = 144,416) published from 2003 to 2019, and demonstrated that time Chinese K-12 students spent on homework increased significantly along with increasing of grades ( Zhai and Fan, 2021 , October).

Regarding homework managing time, some studies reported the grade difference was insignificant. Xu (2006) surveyed 426 middle school students and found that there was no difference between middle school students and high school students. Xu and Corno (2003) reported that urban junior school students ( N = 86) had no grade difference in homework Managing time. Yang and Tu (2020) surveyed 305 Chinese students in grades 7–9, and found that in managing time behavior, the grade differences were insignificant. The rest studies showed that the grade effect is significant. A survey by Xu et al. (2014) based on 1799 Chinese students in grades 10 and 11 showed that the higher level the grade, the lower level of time management.

Grade differences of general creativity

The findings from the previous studies suggested that the scores of general creativity deceases as the grade increases except for some dimensions. Kim (2011) reviewed the Torrance Tests of Creative thinking (TTCT) scores change using five datasets from 1974 to 2008, and reported that three dimensions of creative thinking (i.e., “Fluency,” “Originality,” and “Elaboration”) significantly decreased along with grades increase, while the rest dimension (i.e., “Abstractness of titles”) significantly increased when grades increase. Nie and Zheng (2005) investigated a sample of 3,729 participants from grades 3–12 using the Williams’ Creativity Assessment Packet (WCAP), and reported that the creativity scores decreased from grades 9–12. Said-Metwaly et al. (2021) synthesized 41 primary studies published in the past 60 years, and concluded that the ability of divergent thinking had a whole increase tendency from grades 1 to 12 with a decrease tendency from grades 8 to 11 at the same time.

The purpose and questions of this study

What we have known about homework creativity hitherto is nothing except for its notation and a preliminary version of measurement. To get deeper understanding of homework creativity, this study made an endeavor to examine its relationships with relevant variables based on a confirmation of the reliability and validity of HCBS. Specifically, there are four interrelated research questions, as the following paragraphs (and their corresponding hypotheses) described.

(i) What is the reliability and validity of the HCBS?

Because the earlier version of the HCBS showed a good Cronbach α coefficient of 0.87, and a set of well-fitting indices ( Chang, 2019 ), this study expected that the reliability and validity will also behave well in the current conditions as before. Then, we present the first set of hypotheses as follows:

H1a: The reliability coefficient will equal or greater than 0.80.

H1b: The one-factor model will also fit the current data well; and all indices will reach or over the criteria as the expertise suggested.

(ii) What degree is the score of the HCBS related with academic achievement?

As suggested by the review section, the correlations between homework behaviors and academic achievement ranged from 0.15 and 0.59 (e.g., Fan et al., 2017 ), then we expected that the relationship between homework creativity and academic achievement will fall into this range, because homework creativity is a member of homework behaviors.

The results of the previous studies also demonstrated that the correlation between general creativity and academic achievement changed in a range of 0.19–0.24 with a mean of 0.19 ( Gajda et al., 2017 ). Because it can be treated as a sub-category of general creativity, we predicted that homework creativity will have a similar behavior under the current condition.

Taken aforementioned information together, Hypothesis H2 is presented as follows:

H2: There will be a significant correlation between homework creativity and academic achievement which might fall into the interval of 0.15–0.59.

(iii) What degree is the relationship between HCBS and general creativity?

As discussed in the previous section, there are no inconsistent findings about the relationship between the score of HCBS and general creativity. Some studies postulated that these two variables be positive correlated (e.g., Kaiipob, 1951 ; Chang, 2019 ); other studies argued that this relationship be negative (e.g., Cooper et al., 2012 ; Zheng, 2013 ). Because homework creativity is a sub-category of general creativity, we expected that this relationship would be positive and its value might be equal or less than 0.33. Based on those reasoning, we presented our third hypothesis as follows:

H3: The correlation between homework creativity and general creativity would be equal or less than 0.33.

(iv) What effect does grade have on the HCBS score?

Concerning the grade effect of homework behaviors, the previous findings were contradictory ( Xu et al., 2014 ; Zhai and Fan, 2021 , October). However, the general creativity decreased as the level of grade increases from grade 8 to grade 11 ( Kim, 2011 ; Said-Metwaly et al., 2021 ). Taken these previous findings and the fact that repetitive exercises increase when grades go up ( Zheng, 2013 ), we were inclined to expect that the level of homework creativity is negative correlated with the level of grade. Thus, we presented our fourth hypothesis as follows:

H4: The score of HCBS might decrease as the level of grades goes up.

Materials and methods

Participants.

To get more robust result, this study investigated two convenient samples from six public schools in a medium-sized city in China. Among them, two schools were of high schools (including a key school and a non-key school), and the rest four schools were middle schools (one is key school, and the rest is non-key school). All these schools included here did not have free lunch system and written homework policy. Considering the students were mainly prepared for entrance examination of higher stage, the grades 9 and 12 were excluded in this survey. Consequently, students of grades 7, 8, 10, and 11 were included in our survey. After getting permission of the education bureau of the city investigated, the headmasters administrated the questions in October 2018 (sample 1) and November 2019 (sample 2).

A total of 850 questionnaires were released and the valid number of questionnaires returned is 639 with a valid return rate of 75.18%. Therefore, there were 639 valid participants in sample 1. Among them, there were 273 boys and 366 girls (57.2%); 149 participants from grade 7 (23.31%), 118 from grade 8 (18.47%), 183 from grade 10 (28.64%), and 189 from grade 11 (29.58%); the average age was 15.25 years, with a standard deviation (SD) of 1.73 years. See Table 1 for the information about each grade.

Those participants included received homework assignments every day (see Table 1 for the distribution of homework frequency). During the working days, the averaged homework time was 128.29 minutes with SD = 6.65 minutes. In the weekend, the average homework time was 3.75 hours, with SD = 0.22 hours. The percentage distribution here is similar with that of a national representative sample ( Sun et al., 2020 ), because the values of Chi-squared (χ 2 ) were 7.46 (father) and 8.46 (mother), all p -values were above 0.12 (see Supplementary Table S1 for details).

Another package of 850 questionnaires were released. The valid number of questionnaires returned is 710 with a valid return rate of 83.53%. Among them, there were 366 girls (51.50%); 171 participants from grade 7 (24.23%), 211 from grade 8 (26.06%), 190 from the grade 10 (22.96%), and 216 from grade 11 (26.76%); the average age was 15.06 years, with SD = 1.47 years.

Those participants included received homework assignments almost each day (see Table 1 for details for the distribution of homework frequency). During the working days, the averaged homework time was 123.02 minutes with SD = 6.13 minutes. In weekend, the average homework time was 3.47 hours, with SD = 0.21 hours.

The percentage distribution here is insignificantly different from that of a national representative sample ( Sun et al., 2020 ), because the values of χ 2 were 5.20 (father) and 6.05 (mother), p -values were above 0.30 (see Supplementary Table S1 for details).

Instruments

The homework creativity behavior scale.

The HCBS contains nine items representing students’ creativity behaviors in the process of completing homework (for example, “I do my homework in an innovative way”) ( Chang, 2019 , see Supplementary Table S3 for details). The HCBS employs a 5-point rating scale, where 1 means “completely disagree” and 5 means “completely agree.” The higher the score, the stronger the homework creative behavior students have. The reliability and validity of the HCBS can be found in Section “Reliability and validity of the homework creativity behavior scale” (see Table 2 and Figures 1 , ​ ,2 2 for details).

Results of item discrimination analysis and exploratory factor analysis.

**p < 0.01, two side-tailed. The same for below.

a Correlations for sample 1; b Correlations for sample 2. c Seventh item should be removed away according to the results of CFA (see section “Reliability and validity of the HCBS” for details).

Parallel analysis scree plots of the HCBS data.

The standardized solution for HCBS eight-item model. hcb, homework creativity behavior; it 1∼9, item1 ∼6, 8∼9.

Homework management scale

The HMS contains 22 items describing specific behaviors related to self-management in homework (for example, “I will choose a quiet place to do my homework” or “Tell myself to calm down when encountering difficulties”) ( Xu and Corno, 2003 ; Xu, 2008 ). The HMS employs a 5-point Likert scale, ranging from 1 (completely disagree) to 5 (completely agree). All items can be divided into five dimensions, i.e., arranging environment, managing time, focusing attention, monitoring motivation, and monitoring and controlling emotion. Among them, the monitoring and controlling emotion dimension adopts a method of reverse scoring.

Except for the internal consistency of arranging environment in sample 1, which is 0.63, the internal consistency coefficients of the five dimensions based two samples in this study are all greater than 0.7, ranging from 0.70 to 0.79. The Cronbach’s coefficients of the overall HMS-based two samples are 0.88 and 0.87, respectively. The ω coefficients of the dimensions of HMS ranged from 0.64 to 0.80. The ω coefficients of the HMS total scores were 0.88 and 0.87 for samples 1 and 2, respectively. Those reliability coefficients were acceptable for research purpose ( Clark and Watson, 1995 ; Peterson and Kim, 2013 ).

Williams’ creativity assessment packet

The WCAP including a total of 40 items is a revised version to measure general disposition of creativity (for example, “I like to ask some questions out of other’s expectation” or “I like to imagine something novel, even if it looks useless”) ( Williams, 1979 ; Wang and Lin, 1986 ; Liu et al., 2016 ). The WCAP uses a 3-point Likert scales, in which 1 = disagree, 2 = uncertain, and 3 = agree. The higher WCAP score, the higher is the general creativity level. All items of WCAP can be scattered into four dimensions: adventure, curiosity, imagination, and challenge ( Williams, 1979 ; Wang and Lin, 1986 ; Liu et al., 2016 ). In this study, the Cronbach’s α coefficients of adventure, curiosity, imagination, challenge, and total scale are 0.62, 0.71, 0.78, 0.64, and 0.90, respectively. The ω coefficients were in sequence 0.61, 0.70, 0.77, 0.63, and 0.90 for adventure, curiosity, imagination, challenge, and the total score of WCAP. The correlations between the four dimensions of WCAP are between 0.47 and 0.65. The patterns of reliability coefficients and correlations between dimensions are similar to those results reported by the previous studies ( Williams, 1979 ; Wang and Lin, 1986 ; Liu et al., 2016 ) which stand acceptable reliability and validity ( Clark and Watson, 1995 ; Peterson and Kim, 2013 ).

Homework indicators

Homework time.

The participants were asked to report the time spent on homework in the past week. This technique has been employed widely in many international survey programs, such as PISA from OECD (e.g., Trautwein and Lüdtke, 2007 ). The items are as follows: (1) “Every day, from Monday to Friday, in last week, how many minutes you spent on homework?” The options are as follows: (A) 0–30 min; (B) 31–60 min (C) 61–90 min (D) 91–120 min; (E) 121–180 min; (F) 181 min or more. (2) “In last weekend, how many hours you spent on homework?” The options are as follows: (A) 0–1 h; (B) 1.1–3 h; (C) 3.1–5 h; (D) 5.1–7 h; (E) 7.1 h or more.

Homework completion

The homework completion is a useful indicator demonstrated in the previous studies ( Welch et al., 1986 ; Austin, 1988 ; Swank, 1999 ; Pelletier, 2005 ; Wilson, 2010 ), and had large correlation with achievement, as a meta-analytic results suggested ( Fan et al., 2017 ). In the survey of this study, the participants were also asked to estimate a percent of the completion of homework in the past week and fill in the given blank space. It includes three items which are as follows: “What is the percentage of Chinese/Maths/English homework assignment you completed in the last week?” “Please estimate and write a number from 0 to 100 in the blank space.”

Academic achievement

To record the academic achievement, an item required participants to make a choice based on their real scores of tests, not estimate their tests scores. The item is, “In the last examination, what is the rank of your score in your grade?” (A) The first 2%; (B) The first 3–13%; (C) The first 14–50%; (D) The first 51–84%; (E) The last 16%. The options here correspond to the percentage in the normal distribution, it is convenient to compute a Z -score for each student.

The method employed here is effective to retrieve participants’ test scores. First, the self-report method is more effective than other method under the condition of anonymous investigation. To our knowledge, participants do not have the will to provide their real information in the real name format. Second, this method transforms test scores from different sources into the same space of norm distribution which benefits the comparisons. Third, the validity of this method has been supported by empirical data. Using another sample ( N = 234), we got the academic achievement they reported and real test scores their teacher recorded. The correlation between ranks self-reported and the real scores from Chinese test were r = 0.81, p < 0.001; and the correlation coefficient for mathematics was also large, i.e., r = 0.79, p < 0.001.

Data collection procedure

There are three phases in data collection. The first one is the design stage. At this stage, the corresponding author of this study designed the study content, prepared the survey tools, and got the ethical approve of this project authorized from research ethic committee of school the corresponding author belongs to.

The second stage is to releasing questionnaire prepared. The questionnaire was distributed and retrieved by the head master of those classes involved. Neither the teachers nor the students knew the purpose of this research. During this stage, students can stop answering at any time, or simply withdraw from the survey. None of the teachers and students in this study received payment.

The third stage is the data entry stage. At this stage, the corresponding author of this study recruited five volunteers majored in psychology and education, and explained to them the coding rules, missing value processing methods, identification of invalid questionnaires, and illustrated how to deal with these issues. The volunteers used the same data template for data entry. The corresponding author of this study controlled the data entry quality by selective check randomly.

Data analysis strategies

R packages employed.

The “psych” package in R environment ( R Core Team, 2019 ) was employed to do descriptive statistics, correlation analysis, mean difference comparisons, exploratory factor analysis (EFA), reliability Analysis ( Revelle, 2022 ); and the “lavaan” package was used in confirmatory factor analysis (CFA) and measurement invariance test ( Rosseel, 2012 ); and the “semPlot” package was employed to draw the picture of CFA’s outputs ( Epskamp et al., 2022 ).

Analysis strategies of exploratory factor analysis and reliability

Sample 1 was used for item analysis, EFA, reliability analysis. In EFA, factors were extracted using maximum likelihood, and the promax method served as the rotation method. The number of factors were determined according to the combination of the results from screen plot, and the rule of Eigenvalues exceeding 1.0, and parallel analysis ( Luo et al., 2019 ).

The Cronbach’s α and MacDonald’s ω test were employed to test the reliability of the scale. The rigorous criteria that α ≥ 0.70 ( Nunnally and Bernstein, 1994 ) and ω ≥ 0.7 ( Green and Yang, 2015 ) were taken as acceptable level of the reliability of HCBS.

Analysis strategies of confirmatory factor analysis

As suggested by Hu and Bentler (1999) , two absolute goodness-of-fit indices, namely, the root mean square error of approximation (RMSEA) and the standardized root mean square residual (SRMR), and two relative goodness-of-fit indices, namely, comparative fit index (CFI) and Tucker–Lewis Index (TLI) were recruited as fitting indicators. The absolute goodness-of-fit indices are less than 0.08, and the relative goodness-of-fit indices greater than 0.90 are considered as a good fit. The CFA was conducted using the second sample.

Strategies for measurement invariance

Measurement invariance testing included four models, they are Configural invariance (Model 1), which is to test whether the composition of latent variables between different groups is the same; Weak invariance (Factor loading invariance, Model 2), which is to test whether the factor loading is equal among the groups; Intercept invariance (Model 3), that is, whether the intercepts of the observed variables are equal; Strict equivalent (Residual Variance invariance, Model 4), that is, to test whether the error variances between different groups are equal ( Chen, 2007 ; Putnick and Bornstein, 2016 ).

Since the χ 2 test will be affected easily by the sample size, even small differences will result in significant differences as the sample size will increase. Therefore, this study used the changes of model fitting index CFI, RMSEA, and SRMR (ΔCFI, ΔRMSEA, and ΔSRMR) to evaluate the invariance of the measurement. When ΔCFI ≤ 0.010, ΔRMSEA ≤ 0.015, and ΔSRMR ≤ 0.030 (for metric invariance) or 0.015 (for scalar or residual invariance), the invariance model is considered acceptable ( Cheung and Rensvold, 2002 ; Chen, 2007 ; Putnick and Bornstein, 2016 ).

Strategies of controlling common methods biases

The strategy of controlling common methods biases is mainly hided in the directions. Each part of the printed questionnaire had a sub-direction which invites participants answer the printed questions honestly. The answer formats between any two neighboring parts were different from each other which requested participants change their mind in time. For example, on some part, the answering continuum varied from “1 = totally disagreed” to “5 = total agreed,” while the answering continuum on the neighboring part is the from “5 = totally disagreed” to “1 = total agreed.” Additionally, according to the suggestion of the previous studies, the one factor CFA model and the bi-factor model can be used to detect the common methods biases (e.g., Podsakoff et al., 2012 ).

Detection of common method biases

The fitting results of the one-common-factor model using CFA technique were as follows: χ 2 = 15,073, df = 3320, p < 0.001; χ 2 / df = 4.54, CFI = 0.323, TLI = 0.306, RMSEA = 0.071, 90% CI: 0.070–0.072, and SRMR = 0.101. The results of the bi-factor model under CFA framework were presented as follows: χ 2 = 2,225.826, df = 117, p < 0.001; χ 2 / df = 19.024, CFI = 0.650, TLI = 0.543, RMSEA = 0.159, 90% CI: 0.154–0.164, and SRMR = 0.127. These poor indices of the two models suggested that the one-common-factor model failed to fit the data well and that the biases of common method be ignored ( Podsakoff et al., 2012 ).

Reliability and validity of the homework creativity behavior scale

Item analysis.

Based on the sample 1, the correlation coefficients between the items of the HCBS were between 0.34 and 0.64, p -values were below 0.01. The correlations between the items and the total score of HCBS vary from 0.54 to 0.75 ( p -values are below 0.01). On the condition of sample 2, the correlations between the items fluctuate between 0.31 and 0.58, the correlation coefficients between the items and the total score of the HCBS change from 0.63 to 0.75 ( p -values were below 0.01). All correlation coefficients between items and total score are larger than those between items and reached the criterion suggested ( Ferketich, 1991 ; see Table 2 for details).

Results of exploratory factor analysis

The EFA results (based on sample 1) showed that the KMO was 0.89, and the χ 2 of Bartlett’s test = 1,666.07, p < 0.01. The rules combining eigenvalue larger than 1 and the results of parallel analysis (see Figure 1 for details) suggested that one factor should be extracted. The eigenvalue of the factor extracted was 3.63. The average variance extracted was 0.40. This factor accounts 40% variance with factor loadings fluctuating from 0.40 to 0.76 (see Table 2 ).

Results of confirmatory factor analysis

In the CFA situation (based on sample 2) the fitting indices of the nine-item model of the HCBS are acceptable marginally, they are χ 2 = 266.141; df = 27; χ 2 / df = 9.857; CFI = 0.904; TLI = 0.872; RMSEA = 0.112; 90% CI: 0.100–0.124; SRMR = 0.053.

The modification indices of item 7 were too big (MI value = 74.339, p < 0.01), so it is necessary to consider to delete item 7. Considering its content of “I designed a neat, clean and clear homework format by myself,” item 7 is an indicator of strictness which is weakly linked with creativity. Therefore, the item 7 should be deleted.

After removing item 7, the fitting results were, χ 2 = 106.111; df = 20; χ 2 / df = 5.306; CFI = 0.957; TLI = 0.939; RMSEA = 0.078; 90% CI: 0.064–0.093; SRMR = 0.038). The changes of the fitting indices of the two nested models (eight-item vs. nine-item models) are presented as follows: Δχ 2 = 160.03, Δ df = 7, χ 2 (α = 0.01, df = 7) = 18.48, p < 0.05. After deleting item 7, both CFI and TLI indices increased to above 0.93, and RMSEAs decreased below 0.08 which suggested that the factor model on which eight items loaded fitted the data well. The average variance extracted was 0.50 which is adequate according to the criteria suggested by Fornell and Larcker (1981) . The standardized solution for the eight-item model of the HCBS was shown in Figure 2 .

Correlations between the homework creativity behavior scale and similar concepts

The results showed that the score of the HCBS was significantly correlated with the total score and four dimensions of WCAP and their correlation coefficients ranged from 0.20 to 0.29, p -values were below 0.01. Similarly, the correlations between the score of the HCBS and the scores of arranging environment, managing time, motivation management, and controlling emotion, and total score of the HMS ranged from 0.08 to 0.22, p -values were 0.01; at the meanwhile, the correlation between the score of HCBS and the distraction dimension of the HMS was r = –0.14, p -values were 0.01. The HCBS score was also significantly related to homework completion ( r = 0.18, p < 0.01), but insignificantly related to homework time (see Table 3 for details).

Correlation matrix between variables included and the corresponding descriptive statistics.

About correlation between variables, the results of sample 1 and sample 2 were presented in the lower, upper triangle, respectively.

a In analyses, grades 7, 8, 10, and 11 were valued 1, 2, 3, and 4, respectively.

b TWk, the time spent on homework in the weekend; TWw, the time spent on homework from Monday to Friday; HCp, homework completion; HMSt, total score of homework management scale; AE, arrange environment; MT, manage time; MM, monitor motivation; CE, control emotion; FA, focus attention; WCAPt, WCAP total score; AD, adventure; CU, curiosity; IM, imagination; CH, challenging; HCb, homework creativity behavior; AA, academic achievement.

c Since sample 1 did not answer the WCAP, so the corresponding cells in the lower triangle are blank. *p < 0.05, two side-tailed, the same for below.

d Since there is only one item from variable 1 to 4, the α and ω coefficients cannot be computed.

Correlations between the homework creativity behavior scale and distinct concepts

The correlation analysis results demonstrated that both the correlation coefficients between the score of HCBS and the time spent on homework in week days, and time spent on in weekend days were insignificant ( r -values = 0.02, p -values were above 0.05), which indicated a non-overlap between two distinct constructs of homework creativity and time spent on homework.

Reliability analyses

The results revealed that both the Cronbach’s α coefficients of sample 1 and sample 2 were 0.86, which were greater than a 0.70 criteria the previous studies suggest ( Nunnally and Bernstein, 1994 ; Green and Yang, 2015 ).

Effect of homework creativity on academic achievement

The results (see Table 4 ) of hierarchical regression analyses demonstrated that (1) gender and grade explained 0.8% variation of the score of academic achievement. This number means closing to zero because the regression equation failed to pass the significance test; (2) homework time and completion explained 5.4% variation of academic achievement; considering the β coefficients of the time spent on homework is insignificant, this contribution should be attributed to homework completion totally, and (3) the score of the HCBS explained 3.7% variation of the academic achievement independently.

Effect of homework creativity on general creativity

The results showed the following (see Table 4 for details):

(1) Gender and grade explained 1.3% variation of the total score of general creativity (i.e., the total score of WACP); homework time and completion explained 1.3% variation of the total score of general creativity disposition; and the score of the HCBS independently explained 7.0% variation of the total score of general creativity.

(2) Gender and grade explained 1.7% variation of the adventure score, and homework time and completion explained 1.6% variation of the adventure score, and the score of the HCBS independently explained 6.4% variation of the adventure score.

(3) Gender and grade explained 2.4% variation of the curiosity score, and homework time and completion explained 1.1% variation of the curiosity score, and the score of the HCBS independently explained 5.1% variation of the curiosity score.

(4) Gender and grade explained 0.3% variation of the imagination score, homework time completion explained 0.3% variation of the imagination score. The real values of the two “0.3%” are zeros because both the regression equations and coefficients failed to pass the significance tests. Then the score of the HCBS independently explained 4.4% variation of the imagination score.

(5) Gender and grade explained 0.3% variation of the score of the challenge dimension, homework time and completion explained 2.3% variation of the challenge score, and the score of the HCBS independently explained 4.9% variation of the challenge score.

Grade differences of the homework creativity behavior scale

Test of measurement invariance.

The results of measurement invariance test across four grades indicated the following:

(1) The fitting states of the four models (Configural invariance, Factor loading invariance, Intercept invariance, and Residual variance invariance) were marginally acceptable, because values of CFIs (ranged from 0.89 to 0.93), TLIs (varied from 0.91 to 0.93), RMSEAs (fluctuated from 0.084 to 0.095), and SRMRs (changed from 0.043 to 0.074) located the cutoff intervals suggested by methodologists ( Cheung and Rensvold, 2002 ; Chen, 2007 ; Putnick and Bornstein, 2016 ; see Table 5 for fitting indices, and refer to Supplementary Table S2 for the estimation of parameters).

Fitting results of invariance tests across grades.

(2) When setting factor loadings equal across four grades (i.e., grades 7, 8, 10, and 11), the ΔCFA was –0.006, ΔRMSEA was –0.007, and ΔSRMR was 0.016 which indicated that it passed the test of factor loading invariance. After adding the limit of intercepts equal across four groups, the ΔCFA was –0.008, ΔRMSEA was –0.004, and the ΔSRMR was 0.005 which supported that it passed the test of intercept invariance. At the last step, the error variances were also added as equal, the ΔCFA was –0.027, ΔRMSEA was 0.005, and the ΔSRMR was 0.019 which failed to pass the test of residual variance invariance (see Table 5 for changes of fitting indices). Taking into these fitting indices into account, the subsequent comparisons between the means of factors can be conducted because the residuals are not part of the latent factor ( Cheung and Rensvold, 2002 ; Chen, 2007 ; Putnick and Bornstein, 2016 ).

Grade differences in homework creativity and general creativity

The results of ANOVA showed that there were significant differences in the HCBS among the four grades [ F (3,1345) = 27.49, p < 0.001, η 2 = 0.058, see Table 6 for details]. Further post-test tests returned that the scores of middle school students were significantly higher than those of high school students (Cohen’s d values ranged from 0.46 to 0.54; the averaged Cohen’s d = 0.494), and no significant difference occurs between grades 7 and 8, or between grades 10 and 11. See Figure 3 for details.

Grade differences in HCBS.

***p < 0.001.

The mean differences of the HCBS between the groups of grades.

To address the gap in the previous research on homework creativity, this study examined the psychometric proprieties of the HCBS and its relationship with academic achievement and general creativity. The main findings were (1) Hypotheses H1a and H1b were supported that the reliability and validity of the HCBS were acceptable; (2) Hypothesis H2 was supported that the correlation between the score of the HCBS and academic achievement was significant ( r -values = 0.23–0.26 for two samples); (3) Hypothesis H3 received support that the correlation between the scores of HCBS and WCAP was significant ( r -values = 0.20–0.29 for two samples); and (4) the H4 was supported from the current data that the score of high school students’ was lower than that of the middle school students’ (Cohen’s d = 0.49).

The positive correlations among homework creativity, homework completion, and general creativity

The first key finding should be noted is that the positive correlations with between pairs of homework creativity, homework completion, and general creativity. This result is inconsistent with prediction of an argument that homework diminishes creativity ( Cooper et al., 2012 ; Zheng, 2013 ). Specifically, the correlation between homework completion and curiosity was insignificant ( r = 0.08, p > 0.05) which did not support the argument that homework hurts curiosity of creativity ( Zheng, 2013 ). The possible reason may be homework can provide opportunities to foster some components of creativity by independently finding and developing new ways of understanding what students have learned in class, as Kaiipob (1951) argued. It may be the homework creativity that served as the way to practice the components of general creativity. In fact, the content of items of the HCBS are highly related with creative thinking (refer to Table 2 for details).

Possible reasons of the grade effect of the score of the homework creativity behavior scale

The second key finding should be noted is that the score of the HCBS decreased as the level of grades increased from 7 to 11. This is consistent with the basic trend recorded in the previous meta-analyses ( Kim, 2011 ; Said-Metwaly et al., 2021 ). There are three possible explanations leading to this grade effect. The first one is the repetitive exercises in homework. As Zheng (2013) observed, to get higher scores in the highly competitive entrance examination of high school and college, those Chinese students chose to practice a lot of repetitive exercises. The results of some behavior experiments suggested that repetitive activity could reduce the diverse thinking of subjects’ (e.g., Main et al., 2020 ). Furthermore, the repetitive exercises would lead to fast habituation (can be observed by skin conductance records) which hurts the creative thinking of participants ( Martindale et al., 1996 ). The second explanation is that the stress level in Chinese high schools is higher than in middle school because of the college entrance examination. The previous studies (e.g., Beversdorf, 2018 ) indicated that the high level of stress will trigger the increase activity of the noradrenergic system and the hypothalamic–pituitary–adrenal (HPA) axis which could debase the individual’s performance of creativity. Another likely explanation is the degree of the certainty of the college entrance examination. The level of certainty highly increases (success or failure) when time comes closer to the deadline of the entrance examination. The increase of degree of certainty will lead to the decrease of activity of the brain areas related to curiosity (e.g., Jepma et al., 2012 ).

The theoretical implications

From the theoretical perspective, there are two points deserving to be emphasized. First, the findings of this study extended the previous work ( Beghetto and Kaufman, 2007 ; Kaufman and Beghetto, 2009 ). This study revealed that homework creativity had two typical characteristics, including the personal meaning of students (as represented by the content of items of the HCBS) and the small size of “creativity” and limited in the scope of exercises (small correlations with general creativity). These characteristics are in line with what Mini-C described by the previous studies ( Beghetto and Kaufman, 2007 ; Kaufman and Beghetto, 2009 ). Second, this study deepened our understanding of the relationship between learning (homework is a part of learning) and creativity which has been discussed more than half a century. One of the main viewpoints is learning and creativity share some fundamental similarities, but no one explained what is the content of these “fundamental similarities” (e.g., Gajda et al., 2017 ). This study identified one similarity between learning and creativity in the context of homework, that is homework creativity. Homework creativity has the characteristics of homework and creativity at the same time which served as an inner factor in which homework promote creativity.

The practical implications

The findings in this study also have several potential practical implications. First, homework creativity should be a valuable goal of learning, because homework creativity may make contributions to academic achievement and general creativity simultaneously. They accounted for a total of 10.7% variance of academic achievement and general creativity which are the main goals of learning. Therefore, it is valuable to imbed homework creativity as a goal of learning, especially in the Chinese society ( Zheng, 2013 ).

Second, the items of the HCBS can be used as a vehicle to help students how to develop about homework creativity. Some studies indicated that the creative performance of students will improve just only under the simple requirement of “to be creative please” ( Niu and Sternberg, 2003 ). Similarly, some simple requirements, like “to do your homework in an innovative way,” “don’t stick to what you learned in class,” “to use a simpler method to do your homework,” “to use your imagination when you do homework,” “to design new problems on the basis what learnt,” “to find your own unique insights into your homework,” and “to find multiple solutions to the problem,” which rewritten from the items of the HCBS, can be used in the process of directing homework of students. In fact, these directions are typical behaviors of creative teaching (e.g., Soh, 2000 ); therefore, they are highly possible to be effective.

Third, the HCBS can be used to measure the degree of homework creativity in ordinary teaching or experimental situations. As demonstrated in the previous sections, the reliability and validity of the HCBS were good enough to play such a role. Based on this tool, the educators can collect the data of homework creativity, and make scientific decisions to improve the performance of people’s teaching or learning.

Strengths, limitations, and issues for further investigation

The main contribution is that this study accumulated some empirical knowledge about the relationship among homework creativity, homework completion, academic achievement, and general creativity, as well as the psychometric quality of the HCBS. However, the findings of this study should be treated with cautions because of the following limitations. First, our study did not collect the test–retest reliability of the HCBS. This makes it difficult for us to judge the HCBS’s stability over time. Second, the academic achievement data in our study were recorded by self-reported methods, and the objectivity may be more accurate. Third, the lower reliability coefficients existed in two dimensions employed, i.e., the arrange environment of the HMS (the α coefficient was 0.63), and the adventure of the WCAP (the α coefficient was 0.61). Fourth, the samples included here was not representative enough if we plan to generalize the finding to the population of middle and high school students in main land of China.

In addition to those questions listed as laminations, there are a number of issues deserve further examinations. (1) Can these findings from this study be generalized into other samples, especially into those from other cultures? For instances, can the reliability and validity of the HCBS be supported by the data from other samples? Or can the grade effect of the score of the HCBS be observed in other societies? Or can the correlation pattern among homework creativity, homework completion, and academic achievement be reproduced in other samples? (2) What is the role of homework creativity in the development of general creativity? Through longitudinal study, we can systematically observe the effect of homework creativity on individual’s general creativity, including creative skills, knowledge, and motivation. The micro-generating method ( Kupers et al., 2018 ) may be used to reveal how the homework creativity occurs in the learning process. (3) What factors affect homework creativity? Specifically, what effects do the individual factors (e.g., gender) and environmental factors (such as teaching styles of teachers) play in the development of homework creativity? (4) What training programs can be designed to improve homework creativity? What should these programs content? How about their effect on the development of homework creativity? What should the teachers do, if they want to promote creativity in their work situation? All those questions call for further explorations.

Homework is a complex thing which might have many aspects. Among them, homework creativity was the latest one being named ( Guo and Fan, 2018 ). Based on the testing of its reliability and validity, this study explored the relationships between homework creativity and academic achievement and general creativity, and its variation among different grade levels. The main findings of this study were (1) the eight-item version of the HCBS has good validity and reliability which can be employed in the further studies; (2) homework creativity had positive correlations with academic achievement and general creativity; (3) compared with homework completion, homework creativity made greater contribution to general creativity, but less to academic achievement; and (4) the score of homework creativity of high school students was lower than that of middle school students. Given that this is the first investigation, to our knowledge, that has systematically tapped into homework creativity, there is a critical need to pursue this line of investigation further.

Data availability statement

Ethics statement.

The studies involving human participants were reviewed and approved by the research ethic committee, School of Educational Science, Bohai University. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin.

Author contributions

HF designed the research, collected the data, and interpreted the results. YM and SG analyzed the data and wrote the manuscript. HF, JX, and YM revised the manuscript. YC and HF prepared the HCBS. All authors read and approved the final manuscript.

Acknowledgments

We thank Dr. Liwei Zhang for his supports in collecting data, and Lu Qiao, Dounan Lu, Xiao Zhang for their helps in the process of inputting data.

This work was supported by the LiaoNing Revitalization Talents Program (grant no. XLYC2007134) and the Funding for Teaching Leader of Bohai University.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg.2022.923882/full#supplementary-material

• Anesko K. M., Schoiock G., Ramirez R., Levine F. M. (1987). The homework problem checklist: Assessing children’s homework difficulties. Behav. Assess. 9 179–185. 10.1155/2020/1250801 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Austin C. A. (1988). Homework as a parental involvement strategy to improve the achievement of first grade children: Dissertation abstracts international, 50/03, 622. Doctoral dissertation. Memphis, TN: Memphis State University. [ Google Scholar ]
• Beghetto R. A., Kaufman J. C. (2007). Toward a broader conception of creativity: A case for mini-c creativity. Psycho. Aesthetics Creat. Arts 1 73–79. [ Google Scholar ]
• Beversdorf D. Q. (2018). “ Stress, pharmacology, and creativity ,” in The cambridge handbook of the neuroscience of creativity , eds Jung R. E., Vartanian O. V. (Cambridge: Cambridge University Press.), 73–91. 10.1017/9781316556238.006 [ CrossRef ] [ Google Scholar ]
• Chang Y. (2019). An investigation on relationship between homework and creativity of elementary and middle school students. Master thesis. Liaoning Jinzhou: Bohai University. [ Google Scholar ]
• Chen F. F. (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Struct. Equ. Modeling 14 464–504. 10.1080/10705510701301834 [ CrossRef ] [ Google Scholar ]
• Cheung G. W., Rensvold R. B. (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Struct. Equ. Modeling 9 233–255. [ Google Scholar ]
• Clark L. A., Watson D. (1995). Constructing validity: Basic issues in objective scale development. Psychological Assessment 7 309–319. 10.1037/1040-3590.7.3.309 [ CrossRef ] [ Google Scholar ]
• Cooper H., Lindsay J. J., Nye B., Greathouse S. (1998). Relationships among attitudes about homework, amount of homework assigned and completed, and student achievement. J. Educ. Psychol. 90 70–83. 10.1037//0022-0663.90.1.70 [ CrossRef ] [ Google Scholar ]
• Cooper H., Robinson J. C., Patall E. A. (2006). Does homework improve academic achievement? A synthesis of research, 1987–2003. Rev. Educ. Res. 76 1–62. [ Google Scholar ]
• Cooper H., Steenbergen-Hu S., Dent A. L. (2012). “ Homework ,” in APA educational psychology handbook, Vol.3. Application to learning and teaching , eds Harris K. R., Graham S., Urdan T. (Washington DC: American Psychological Association; ), 475–495. [ Google Scholar ]
• De Jong R., Westerhof K. J., Creemers B. P. M. (2000). Homework and student math achievement in junior high schools. Educ. Res. Eval. 6 130–157. [ Google Scholar ]
• Dettmers S., Trautwein U., Lüdtke O. (2009). The relationship between homework time and achievement is not universal: Evidence from multilevel analyses in 40 countries. Sch. Effect. Sch. Improv. 20 375–405. 10.1080/09243450902904601 [ CrossRef ] [ Google Scholar ]
• Dolean D. D., Lervag A. (2022). Variations of homework amount assigned in elementary school can impact academic achievement. J. Exp. Educ. 90 280–296. 10.1080/00220973.2020.1861422 [ CrossRef ] [ Google Scholar ]
• Epskamp S., Stuber S., Nak J., Veenman M., Jorgensen T. D. (2022). semPlot: Path diagrams and visual analysis of various sem packages’ output. R package Version 1.1.5. Availabl eonline at: https://cran.r-project.org/web/packages/semPlot/index.html (accessed July 18, 2022). [ Google Scholar ]
• Fan H., Xu J., Cai Z., He J., Fan X. (2017). Homework and students’ achievement in math and science: A 30-year meta-analysis, 1986–2015. Educ. Res. Rev. 20 35–54. 10.1016/j.edurev.2016.11.003 [ CrossRef ] [ Google Scholar ]
• Ferketich S. (1991). Focus on psychometrics. Aspects of item analysis. Res. Nurs. Health 14 165–168. 10.1002/nur.4770140211 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Fernández-Alonso R., Álvarez-Díaz M., Suárez-Álvarez J., Muñiz J. (2017). Students’ achievement and homework assignment strategies. Front. Psychol. 8 : 286 . 10.3389/fpsyg.2017.00286 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Fernández-Alonso R., Suárez-álvarez J., Javier M. (2015). Adolescents’ homework performance in mathematics and science: Personal factors and teaching practices. J. Educ. Psychol. 107 1075–1085. 10.1037/edu0000032 [ CrossRef ] [ Google Scholar ]
• Fernández-Alonso R., Woitschach P., Álvarez-Díaz M., González-López A. M., Cuesta M., Muñiz J. (2019). Homework and academic achievement in latin america: A multilevel approach. Front. Psychol. 10 : 95 . 10.3389/fpsyg.2019.00095 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Fornell C., Larcker D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. J. Mark. Res. 18 39–50. 10.1177/002224378101800104 [ CrossRef ] [ Google Scholar ]
• Gajda A., Karwowski M., Beghetto R. A. (2017). Creativity and academic achievement: A meta-analysis. J. Educ. Psychol. 109 269–299. 10.1037/edu0000133 [ CrossRef ] [ Google Scholar ]
• Green S. B., Yang Y. (2015). Evaluation of dimensionality in the assessment of internal consistency reliability: Coefficient alpha and omega coefficients. Educ. Meas. Issues Pract. 34 14–20. 10.1111/emip.12100 [ CrossRef ] [ Google Scholar ]
• Guilford J. P. (1950). Creativity. Am. Psychol. 5 444–454. 10.1037/h0063487 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Guo L. (2018). The compilation of homework behavior questionnaire for junior middle school students. Master thesis. Liaoning Jinzhou: Bohai University. [ Google Scholar ]
• Guo L., Fan H. (2018). Analysis and prospect of homework instruments in primary and middle schools. Educ. Sci. Res. 3 48–53. [ Google Scholar ]
• Hattie J. (2009). Visible learning: A synthesis of over 800 meta-analyses relating to achievement. London: Routledge. [ Google Scholar ]
• Hu L. T., Bentler P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct. Equ. Modeling 6 1–55. 10.1080/10705519909540118 [ CrossRef ] [ Google Scholar ]
• Jepma M., Verdonschot R. G., van Steenbergen H., Rombouts S. A. R. B., Nieuwenhuis S. (2012). Neural mechanisms underlying the induction and relief of perceptual curiosity. Front. Behav. Neurosci. 6 : 2012 . 10.3389/fnbeh.2012.00005 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Kaiipob I. A. (1951). Pedagogy (Shen yingnan, Nan zhishan et al, translated into chinese). Beijing: People’s Education Press, 150–155. [ Google Scholar ]
• Kalenkoski C. M., Pabilonia S. W. (2017). Does high school homework increase academic achievement? Educ. Econ. 25 45–59. 10.1080/09645292.2016.1178213 [ CrossRef ] [ Google Scholar ]
• Kaufman J. C., Beghetto R. A. (2009). Beyond big and little: The Four-C model of creativity. Rev. Gen. Psychol. 13 1–12. 10.1037/a0013688 [ CrossRef ] [ Google Scholar ]
• Kaufman J. C., Glăveanu V. P. (2019). “ A review of creativity theories: What questions are we trying to answer? ,” in Cambridge handbook of creativity , 2nd Edn, eds Kaufman J. C., Sternberg R. J. (New York, NY: Cambridge University Press; ), 27–43. [ Google Scholar ]
• Kaufman J. C., Plucker J. A., Baer J. (2008). Essentials of creativity assessment. Hoboken, NJ: John Wiley & Sons. [ Google Scholar ]
• Karwowski M., Jankowska D. M., Brzeski A., Czerwonka M., Gajda A., Lebuda I., et al. (2020). Delving into creativity and learning. Creat. Res. J. 32 4–16. 10.1080/10400419.2020.1712165 [ CrossRef ] [ Google Scholar ]
• Kim K. H. (2011). The creativity crisis: The decrease in creative thinking scores on the torrance tests of creative thinking. Creat. Res. J. 23 285–295. 10.1080/10400419.2011.627805 [ CrossRef ] [ Google Scholar ]
• Kozbelt A., Beghetto R. A., Runco M. A. (2011). “ Theories of creativity ,” in The cambridge handbook of creativity , eds Kaufman J. C., Sternberg R. J. (New York, NY: Cambridge University Press; ), 20–47. [ Google Scholar ]
• Kupers E., van Dijk M., Lehmann-Wermser A. (2018). Creativity in the here and now: A generic, micro-developmental measure of creativity. Front. Psychol. 9 : e2095 . 10.3389/fpsyg.2018.02095 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Liu X.-L., Liu L., Qiu Y.-X., Jin Y., Zhou J. (2016). Reliability and validity of williams creativity assessment packet. J. Sch. Stud. 13 51–58. 10.3969/j.issn.1005-2232.2016.03.007 [ CrossRef ] [ Google Scholar ]
• Liu Y., Gong S., Cai X. (2013). Junior-high school students’ homework effort and its influencing factors. Adv. Psychol. Sci. 21 1422–1429. 10.3724/SP.J.1042.2013.01422 [ CrossRef ] [ Google Scholar ]
• Long H., Kerr B. A., Emler T. E., Birdnow M. (2022). A critical review of assessments of creativity in education. Rev. Res. Educ. 46 288–323. 10.3102/0091732X221084326 [ CrossRef ] [ Google Scholar ]
• Luo L., Arizmendi C., Gates K. M. (2019). Exploratory factor analysis (EFA) programs in R. Struct. Equ. Modeling 26 819–826. 10.1080/10705511 [ CrossRef ] [ Google Scholar ]
• Main K. J., Aghakhani H., Labroo A. A., Greidanus N. S. (2020). Change it up: Inactivity and repetitive activity reduce creative thinking. J. Creat. Behav. 54 395–406. 10.1002/jocb.373 [ CrossRef ] [ Google Scholar ]
• Martindale C., Anderson K., Moor K., West A. (1996). Creativity, oversensitivity and rate of habituation. Pers. Individ. Diff. 20 423–427. 10.1016/0191-8869(95)00193-X [ CrossRef ] [ Google Scholar ]
• Nie Y., Zheng X. (2005). A study on the developmental characteristics of children’s and adolescent’s creative personality. Psychol. Sci. 28 356–361. 10.16719/j.cnki.1671-6981.2005.02.024 [ CrossRef ] [ Google Scholar ]
• Niu W., Sternberg R. J. (2003). Societal and school influences on student creativity: The case of China. Psychol. Sch. 40 103–114. 10.1002/pits.10072 [ CrossRef ] [ Google Scholar ]
• Núñez J. C., Suárez N., Cerezo R., González-Pienda J., Valle A. (2013). Homework and academic achievement across Spanish Compulsory Education. Educ. Psychol. 35 1–21. 10.1080/01443410 [ CrossRef ] [ Google Scholar ]
• Nunnally J. C., Bernstein I. H. (1994). Psychometric theory , 3rd Edn. New York, NY: McGraw-Hill. [ Google Scholar ]
• OECD (2014). Does homework perpetuate inequities in education? Pisa in Focus, No. 46. Paris: OECD Publishing, 10.1787/5jxrhqhtx2xt-en [ CrossRef ] [ Google Scholar ]
• Pang W., Plucker J. A. (2012). Recent transformations in China’s economic, social, and education policies for promoting innovation and creativity. J. Creat. Behav. 46 247–273. 10.1002/jocb.17 [ CrossRef ] [ Google Scholar ]
• Pendergast L. L., Watkins M. W., Canivez G. L. (2014). Structural and convergent validity of the homework performance questionnaire. Educ. Psychol. 34 291–304. 10.1080/01443410.2013.785058 [ CrossRef ] [ Google Scholar ]
• Pelletier R. (2005). The predictive power of homework assignments on student achievement in grade three (Order No. 3169466). Available from proquest dissertations & theses global. (305350863). Available online at: http://search.proquest.com/docview/305350863?accountid¼12206 [ Google Scholar ]
• Peterson R., Kim Y. (2013). On the relationship between coefficient alpha and composite reliability. J. Appl. Psychol. 98 194–198. 10.1037/a0030767 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Plucker J. A., Makel M. C., Qian M. (2019). “ Chapter3: assessment of creativity ,” in The cambridge handbook of creativity , 2nd Edn, eds Kaufman J. C., Sternberg R. J. (Cambridge University Press: New York, NY; ), 44–68. [ Google Scholar ]
• Podsakoff P. M., Mac Kenzie S. B., Podsakoff N. P. (2012). Sources of method bias in social science research and recommendations on how to control it. Annu. Rev. Psychol. 63 539–569. 10.1146/annurev-psych-120710-100452 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Power T. J., Dombrowski S. C., Watkins M. W., Mautone J. A., Eagle J. W. (2007). Assessing children’s homework performance: Development of multi-dimensional, multi-informant rating scales. J. Sch. Psychol. 45 333–348. 10.1016/j.jsp.2007.02.002 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Putnick D. L., Bornstein M. H. (2016). Measurement invariance conventions and reporting: The state of the art and future directions for psychological research. Dev. Rev. 41 71–90. 10.1016/j.dr.2016.06.004 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Qian A. (2006). Research on the creative thought ability training in the language teaching material work system. Ph.D. thesis. Jiangsu Nanjing: Nanjing Normal University. [ Google Scholar ]
• R Core Team (2019). R: a language and environment for statistical computing. Vienna: R Foundation for Statistical Computing. [ Google Scholar ]
• Revelle W. (2022). Psych: Procedures for psychological, psychometric, and personality research. Evanston, IL: Northwestern University. [ Google Scholar ]
• Rosário P., Núñez J., Vallejo G., Cunha J., Nunes T., Mourão R., et al. (2015). Does homework design matter? The role of homework’s purpose in student mathematics achievement. Contemp. Educ. Psychol. 43 10–24. 10.1016/j.cedpsych.2015.08.001 [ CrossRef ] [ Google Scholar ]
• Rosseel Y. (2012). Lavaan: An R package for structural equation modeling. J. Stat. Softw. 48 : 97589 . 10.3389/fpsyg.2014.01521 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Said-Metwaly S., Fernández-Castilla B., Kyndt E., Van den Noortgate W., Barbot B. (2021). Does the fourth-grade slump in creativity actually exist? A meta-analysis of the development of divergent thinking in school-age children and adolescents. Educ. Psychol. Rev. 33 275–298. 10.1007/s10648-020-09547-9 [ CrossRef ] [ Google Scholar ]
• Smith J. K., Smith L. F. (2010). “ Educational creativity ,” in The cambridge handbook of creativity , eds Kaufman J. C., Sternberg R. J. (New York, NY: Cambridge University Press; ), 250–264. [ Google Scholar ]
• Soh K.-C. (2000). Indexing creativity fostering teacher behavior: A preliminary validation study. J. Creat. Behav. 34 118–134. 10.1002/j.2162-6057.2000.tb01205.x [ CrossRef ] [ Google Scholar ]
• Sternberg R. J. (2019). Measuring creativity: A 40+ year retrospective. J. Creat. Behav. 53 600–604. 10.1002/jocb.218 [ CrossRef ] [ Google Scholar ]
• Sternberg R. J., Karami S. (2022). An 8P theoretical framework for understanding creativity and theories of creativity. J. Creat. Behav. 56 55–78. 10.1002/jocb.516 [ CrossRef ] [ Google Scholar ]
• Sun M., Du J., Xu J. (2021). Are homework purposes and student achievement reciprocally related? A longitudinal study. Curr. Psychol. 40 4945–4956. 10.1007/s12144-019-00447-y [ CrossRef ] [ Google Scholar ]
• Sun L., Shafiq M. N., McClure M., Guo S. (2020). Are there educational and psychological benefits from private supplementary tutoring in Mainland China? Evidence from the China Education Panel Survey, 2013–15. Int. J. Educ. Dev. 72 : 102144 . [ Google Scholar ]
• Swank A. L. G. (1999). The effect of weekly math homework on fourth grade student math performance. Master of arts action research project. Knoxville, TN: Johnson Bible College. [ Google Scholar ]
• Tas Y., Sungur S., Oztekin C. (2016). Development and validation of science homework scale for middle-school students. Int. J. Sci. Math. Educ. 14 417–444. 10.1007/s10763-014-9582-5 [ CrossRef ] [ Google Scholar ]
• Trautwein U., Lüdtke O. (2007). Students’ self-reported effort and time on homework in six school subjects: Between-student differences and within-student variation. J. Educ. Psychol. 99 432–444. 10.1037/0022-0663.99.2.432 [ CrossRef ] [ Google Scholar ]
• Trautwein U., Lüdtke O. (2009). Predicting homework motivation and homework effort in six school subjects: The role of person and family characteristics, classroom factors, and school track. Learn. Instr. 19 243–258. 10.1016/j.learninstruc.2008.05.001 [ CrossRef ] [ Google Scholar ]
• Valle A., Piñeiro I., Rodríguez S., Regueiro B., Freire C., Rosário P. (2019). Time spent and time management in homework in elementary school students: A person-centered approach. Psicothema 31 422–428. 10.7334/psicothema2019.191 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Wang M., Lin X. (1986). Research on the revised williams creative aptitude test. Bull. Spec. Educ. 2 231–250. [ Google Scholar ]
• Welch W. W., Walberg H. J., Fraser B. J. (1986). Predicting elementary science learning using national assessment data. J. Res. Sci. Teach. 23 699–706. 10.1002/tea.3660230805 [ CrossRef ] [ Google Scholar ]
• Williams F. E. (1979). Assessing creativity across Williams “CUBE” model. Gifted Child Q. 23 748–756. 10.1177/001698627902300406 [ CrossRef ] [ Google Scholar ]
• Wilson J. L. (2010). The impact of teacher assigned but not graded compared to teacher assigned and graded chemistry homework on the formative and summative chemistry assessment scores of 11th-grade students with varying chemistry potential (Order No. 3423989). Available from proquest dissertations & theses global. (759967221). Available online at: https://www.proquest.com/docview/759967221 (accessed July 18, 2022). [ Google Scholar ]
• Xu J. (2006). Gender and homework management reported by high school students. Educ. Psychol. 26 73–91. 10.1080/01443410500341023 [ CrossRef ] [ Google Scholar ]
• Xu J. (2008). Validation of scores on the homework management scale for high school students. Educ. psychol. Meas. 68 304–324. 10.1177/0013164407301531 [ CrossRef ] [ Google Scholar ]
• Xu J. (2010). Homework purpose scale for high school students: A validation study. Educ. Psychol. Meas. 70 459–476. 10.1177/0013164409344517 [ CrossRef ] [ Google Scholar ]
• Xu J. (2017). Homework expectancy value scale for high school students: Measurement invariance and latent mean differences across gender and grade level. Learn. Individ. Diff. 60 10–17. 10.1016/j.lindif.2017.10.003 [ CrossRef ] [ Google Scholar ]
• Xu J. (2018). Reciprocal effects of homework self-concept, interest, effort, and math achievement. Contemp. Educ. Psychol. 55 42–52. 10.1016/j.cedpsych.2018.09.002 [ CrossRef ] [ Google Scholar ]
• Xu J. (2020). Longitudinal effects of homework expectancy, value, effort, and achievement: An empirical investigation. Int. J. Educ. Res. 99 : 101507 . 10.1016/j.ijer.2019.101507 [ CrossRef ] [ Google Scholar ]
• Xu J. (2021). Math homework purpose scale: Confirming the factor structure with high school students. Psychology in the Schools 58 1518–1530. 10.1002/pits.22507 [ CrossRef ] [ Google Scholar ]
• Xu J., Corno L. (2003). Family help and homework management reported by middle school students. Elem. Sch. J. 103 503–518. 10.1086/499737 [ CrossRef ] [ Google Scholar ]
• Xu J., Du J., Cunha J., Rosrio P. (2021). Student perceptions of homework quality, autonomy support, effort, and math achievement: Testing models of reciprocal effects. Teach. Teach. Educ. 108 : 103508 . 10.1016/j.tate.2021.103508 [ CrossRef ] [ Google Scholar ]
• Xu J., Du J., Liu F., Huang B. (2019). Emotion regulation, homework completion, and math achievement: Testing models of reciprocal effects. Contemp. Educ. Psychol. 59 : 101810 . 10.1016/j.cedpsych.2019.101810 [ CrossRef ] [ Google Scholar ]
• Xu J., Núñez J., Cunha J., Rosário P. (2020). Validation of the online homework distraction scale. Psicothema 32 469–475. 10.7334/psicothema2020.60 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Xu J., Yuan R., Xu B., Xu M. (2014). Modeling students’ managing time in math homework. Learn. Individ. Differences 34 33–42. 10.1016/j.lindif.2014.05.011 [ CrossRef ] [ Google Scholar ]
• Yang F., Tu M. (2020). Self-regulation of homework behavior: Relating grade, gender, and achievement to homework management. Educ. Psychol. 40 392–408. 10.1080/01443410.2019.1674784 [ CrossRef ] [ Google Scholar ]
• Yang F., Xu J. (2017). Homework expectancy value scale: Measurement invariance and latent mean differences across gender. J. Psychoeduc. Assess. 36 863–868. 10.1177/0734282917714905 [ CrossRef ] [ Google Scholar ]
• Zhai J., Fan H. (2021). “ The changes in primary and middle school students’ homework time in china: A cross-temporal meta-analysis ,” in Paper presented at the meeting of the 23rd national academic conference of psychology , Huhhot. [ Google Scholar ]
• Zheng Y. (2013). Problems and causes of China’s education. Beijing: China CITIC Press, 125. [ Google Scholar ]

• Study Guides
• Homework Questions

Negative Effects of Social Media

• Communications

This paper is in the following e-collection/theme issue:

Published on 9.4.2024 in Vol 26 (2024)

Moderating Effect of Coping Strategies on the Association Between the Infodemic-Driven Overuse of Health Care Services and Cyberchondria and Anxiety: Partial Least Squares Structural Equation Modeling Study

Authors of this article:

Original Paper

• Richard Huan Xu 1 , PhD   ; 
• Caiyun Chen 2 , PhD  

1 Department of Rehabilitation Sciences, Faculty of Health and Social Sciences, Hong Kong Polytechnic University, Hung Hom, China (Hong Kong)

2 Nanjing Academy of Administration, Nanjing, China

Corresponding Author:

Richard Huan Xu, PhD

Department of Rehabilitation Sciences

Faculty of Health and Social Sciences

Hong Kong Polytechnic University

11 Yuk Choi Rd

China (Hong Kong)

Phone: 852 27664199

Email: [email protected]

Background: The COVID-19 pandemic has led to a substantial increase in health information, which has, in turn, caused a significant rise in cyberchondria and anxiety among individuals who search for web-based medical information. To cope with this information overload and safeguard their mental well-being, individuals may adopt various strategies. However, the effectiveness of these strategies in mitigating the negative effects of information overload and promoting overall well-being remains uncertain.

Objective: This study aimed to investigate the moderating effect of coping strategies on the relationship between the infodemic-driven misuse of health care and depression and cyberchondria. The findings could add a new dimension to our understanding of the psychological impacts of the infodemic, especially in the context of a global health crisis, and the moderating effect of different coping strategies on the relationship between the overuse of health care and cyberchondria and anxiety.

Methods: The data used in this study were obtained from a cross-sectional web-based survey. A professional survey company was contracted to collect the data using its web-based panel. The survey was completed by Chinese individuals aged 18 years or older without cognitive problems. Model parameters of the relationships between infodemic-driven overuse of health care, cyberchondria, and anxiety were analyzed using bootstrapped partial least squares structural equation modeling. Additionally, the moderating effects of coping strategies on the aforementioned relationships were also examined.

Results: A total of 986 respondents completed the web-based survey. The mean scores of the Generalized Anxiety Disorder-7 and Cyberchondria Severity Scale-12 were 8.4 (SD 3.8) and 39.7 (SD 7.5), respectively. The mean score of problem-focused coping was higher than those of emotion- and avoidant-focused coping. There was a significantly positive relationship between a high level of infodemic and increased overuse of health care (bootstrapped mean 0.21, SD 0.03; 95% CI 0.1581-0.271). The overuse of health care resulted in more severe cyberchondria (bootstrapped mean 0.107, SD 0.032) and higher anxiety levels (bootstrapped mean 0.282, SD 0.032) in all the models. Emotion (bootstrapped mean 0.02, SD 0.008 and 0.037, SD 0.015)- and avoidant (bootstrapped mean 0.026, SD 0.009 and 0.049, SD 0.016)-focused coping strategies significantly moderated the relationship between the overuse of health care and cyberchondria and that between the overuse of health care and anxiety, respectively. Regarding the problem-based model, the moderating effect was significant for the relationship between the overuse of health care and anxiety (bootstrapped mean 0.007, SD 0.011; 95% CI 0.005-0.027).

Conclusions: This study provides empirical evidence about the impact of coping strategies on the relationship between infodemic-related overuse of health care services and cyberchondria and anxiety. Future research can build on the findings of this study to further explore these relationships and develop and test interventions aimed at mitigating the negative impact of the infodemic on mental health.

Introduction

Covid-19–related mental health problems.

In today’s technologically advancing society, widespread and rapid digitization has led to a substantial increase in the use of social media and the internet. This, in turn, has facilitated the rapid dissemination of all types of information. Although this can be beneficial in filling information gaps quickly, it has its drawbacks. A prominent drawback is the amplification of harmful messages, which can have negative effects on individuals [ 1 , 2 ]. The World Health Organization (WHO) acknowledged the presence of an infodemic during the COVID-19 pandemic and subsequent responses. WHO defines an infodemic as an excessive amount of information, including both accurate and inaccurate content [ 3 ]. This abundance of information makes it difficult for individuals to distinguish reliable sources from unreliable sources and to find trustworthy guidance when they need it.

Excessive use of health care services can have adverse effects on individuals and the overall sustainability of health care systems. Although challenges associated with the overuse of health care services were evident before the COVID-19 pandemic [ 4 , 5 ], the urgent need for sustainable health care systems was exacerbated by the pandemic. Because large portions of the population were instructed to self-isolate at home and had limited access to health care professionals during the pandemic, the internet became the primary source of information for numerous individuals seeking answers to health-related questions. However, the abundance of web-based information, including both true and false content, can leave individuals feeling overwhelmed and struggling to make informed choices. This information overload can lead to depression because individuals bombarded with conflicting messages may feel unsure of what to believe [ 6 - 10 ].

Besides depression, cyberchondria has also emerged as a significant public health challenge since the onset of the COVID-19 pandemic. This refers to the repeated and excessive search for health-related information on the internet, leading to a significant increase in distress or anxiety [ 11 ]. Although the global emergency caused by the COVID-19 pandemic is over, telehealth remains a growing trend. An increasing number of studies have indicated that telehealth can improve health care access, outcomes, and affordability by offering a bridge to care and an opportunity to reinvent web-based care models [ 12 ]. However, increasing internet exposure increases the risk of cyberchondria, especially under conditions of uncertainty and increased risk, due to the large volume of information it contains. Thus, it is crucial to understand how to provide support and guidance to help people adopt appropriate strategies for using web-based resources safely in the context of an infodemic.

Current Research on the COVID-19–Related Infodemic

The harms of infodemic are well documented. An Italian study suggested developing early warning signals for an infodemic, which can provide important cues for implementing effective communication strategies to mitigate misinformation [ 13 ]. Other studies have shown that successful use of coping strategies can help individuals manage stressful events and reduce negative emotions during a pandemic. For example, Yang [ 14 ] found a positive correlation between emotion-focused coping and cyberbullying and depression during the COVID-19 pandemic. A large-scale UK study indicated that supportive coping was associated with a faster decrease in depression and anxiety symptoms [ 15 ]. Shigeto et al [ 16 ] emphasized the importance of training young adults to develop resilience, flexibility, and specific coping skills to offset the psychological effects of significant lifestyle changes resulting from pandemics or other health crises in the future. A recent study used machine learning technology to enhance the accuracy and efficiency of automated fact-checking and infodemic risk management at a strategic level [ 17 ]. However, the impact of coping strategies on the relationship among the infodemic, cyberchondria, and anxiety at an individual level during the COVID-19 pandemic is still unknown.

Importance of Coping Strategies

The ability of individuals to discern and adopt appropriate coping strategies can have a profound impact on their mental health, particularly in relation to conditions such as depression and anxiety. The ability to select and implement coping strategies is not uniform across all individuals, and these differences can significantly influence the trajectory of their mental health outcomes. For some, the ability to effectively choose and implement coping strategies can serve as a protective factor, mitigating the severity of the symptoms of depression or anxiety and promoting overall health and well-being. Conversely, for others, inability or difficulty in selecting and implementing effective coping strategies can exacerbate mental health conditions, leading to increased severity of depression and anxiety. This, in turn, can have detrimental effects on individuals’ overall health and well-being. Therefore, understanding the factors that influence individuals’ ability to select and implement effective coping strategies is of paramount importance in the field of mental health research and intervention [ 18 ].

Research has demonstrated the importance of appropriate coping mechanisms in managing mental health problems. Coping strategies, which are essential for dealing with stress or challenging situations, can be categorized into 3 primary types: emotion focused, problem focused, and avoidant focused [ 19 ]. Emotion-focused strategies are centered around managing and regulating emotions. They serve as a means to cope with stress or difficult situations. These strategies might involve seeking emotional support from others, using relaxation techniques, or practicing mindfulness. In contrast, problem-focused strategies actively address the problem or stressor. These strategies might encompass problem-solving, devising a plan of action, or seeking information and resources to effectively tackle the situation. Avoidant-focused strategies involve evading or distancing oneself from the stressor or problem. These strategies might include denial, distraction, or engaging in activities to escape or avoid contemplating the issue [ 18 ]. The effectiveness of different coping strategies can vary depending on the situation. Individuals often use different or a combination of strategies, tailoring their approach to their circumstances.

Coping Strategies in the COVID-19–Related Infodemic

From a social perspective, this study underscores the importance of mental health in the context of public health emergencies such as the COVID-19 pandemic. It highlights the need for society to recognize and address the mental health burden that such emergencies can place on individuals, particularly in relation to the phenomenon of cyberchondria, which is the unfounded escalation of concerns about common symptoms based on reviews of web-based literature and resources.

Practically, this study provides valuable insights for policy makers and practitioners. It emphasizes the need for the development of effective coping strategies and programs to manage the negative impact of an overload of misinformation and disinformation on mental health. This is particularly relevant in the digital age, where individuals have access to a plethora of information, not all of which is accurate or reliable. Policy makers and practitioners can use the findings of this study to design interventions that not only provide accurate information but also equip individuals with the skills to distinguish reliable sources from unreliable sources and to cope with the anxiety that misinformation can cause. From a research standpoint, this study fills a gap in the literature by assessing the impact of the infodemic on cyberchondria and the moderating effect of coping strategies in this relationship. It opens up new avenues of research into the complex interplay among public health emergencies, infodemic, cyberchondria, and coping strategies. Future research could build on the findings of this study to further explore these relationships and develop and test interventions aimed at mitigating the negative impact of infodemic on mental health.

Objective of the Study

Currently, the association between the overuse of health care services and mental health problems in the context of an infodemic remains unclear, as is the moderating effect of different coping strategies on this association. Thus, this study investigated the moderating effect of coping strategies on the relationship between the infodemic-driven misuse of health care and depression and cyberchondria.

Hypotheses of the Study

The study used a hypothesis-driven format. Specifically, there are five hypotheses: (1) a positive relationship exists between infodemic and the misuse of health care, (2) a positive relationship exists between the misuse of health care and depressive disorders, (3) a positive relationship exists between the misuse of health care and cyberchondria, (4) coping strategies mitigate the negative effect of the misuse of health care on depression, and (5) coping strategies mitigate the negative effect of the misuse of health care on cyberchondria. Hypotheses 2-5 are separately evaluated for the three types of coping strategies: problem focused (H2.1), emotion focused (H2.2), and avoidant focused (H2.3).

Study Design and Sample Size

The data used in this study were obtained from a cross-sectional and web-based survey conducted between April and May 2023 in China.

There is no gold standard for sample estimation in partial least squares structural equation modeling (PLS-SEM). Following Hair et al [ 20 ], we set the significance level at 5% and the minimum path coefficients to between 0.05 and 0.1. Based on these criteria, a minimum sample size of 619 was determined.

Data Source and Collection

A professional surveying company, WenJuanXing, was invited to collect the data through its web-based panel. The panel of WenJuanXing consists of 2.6 million members, with an average of over 1 million questionnaire respondents daily. At the beginning of the project, a survey manager collaborated with the research team to screen and recruit participants using the company’s internal social network platform. All of the eligible panel members received a survey invitation, and a voluntary response sampling method was used. The survey manager checked the data quality using WenJuanXing’s artificial intelligence data quality control system to ensure that respondents met our inclusion criteria and provided valid responses, thus ensuring a high level of data accuracy and integrity. The inclusion criteria were (1) aged older than 18 years, (2) able to understand and read Chinese, and (3) agreed to provide informed consent. All eligible respondents were invited to participate in a web-based survey. The first section of the survey was the informed consent, which the participants were required to read and agree to before proceeding. All the participants who agreed to participate in the survey were asked to complete six questionnaires covering (1) demographics and socioeconomic status, (2) COVID-19 information–related questions, (3) a cyberchondria questionnaire, (4) an eHealth literacy questionnaire, (5) an anxiety questionnaire, and (6) a coping strategy questionnaire. The English translations of the questionnaires are presented in Multimedia Appendix 1 . To ensure data quality, we collaborated with the survey company and implemented various indicators. We monitored completion time, excluding responses that took less than 6 minutes. We also tracked ID addresses, ensuring that each ID address could only complete the questionnaire once. To minimize random errors, we used an artificial intelligence formula developed by the survey company to identify and filter any response patterns that appeared to be generated in parallel.

Ethical Considerations

The study protocol and informed consent process were approved by the institutional review board of the Hong Kong Polytechnic University (HSEARS20230502006). Informed consent was collected from all participants. The survey was conducted anonymously, and no personally identifiable information was collected. No compensation was provided by the research team.

Instruments

Cyberchondria severity scale-12.

The Cyberchondria Severity Scale-12 (CSS-12), derived from the 33-item CSS, was used to measure the severity of cyberchondria. The CSS-12 exhibited equally good psychometric properties as the original version and has been validated in Chinese populations [ 21 ]. The CSS-12 items are scored on a Likert-type scale ranging from 1=“never” to 5=“always,” giving total scores ranging from 12 to 60. A higher score indicates a higher severity of suspected cyberchondria. The psychometric properties of the Chinese version of the CSS-12 were reported by Peng et al [ 22 ].

Generalized Anxiety Disorder Assessment

The Generalized Anxiety Disorder Assessment-7 (GAD-7) was used to screen for generalized anxiety disorder and related anxiety disorders [ 23 ]. This scale consists of 7 items designed to assess the frequency of anxiety symptoms during the 2 weeks preceding the survey. The GAD-7 score is calculated by assigning scores of 0, 1, 2, and 3 to the response categories of “not at all,” “several days,” “more than half the days,” and “nearly every day,” respectively. The scores of the 7 questions are then summed, giving a total ranging from 0 to 21, with higher scores indicating a higher severity of anxiety disorders. Many studies have reported the psychometric properties of the GAD-7 in Chinese populations, such as that conducted by Sun et al [ 24 ].

Coping Orientation to Problems Experienced Inventory

The Coping Orientation to Problems Experienced Inventory (Brief-COPE) is a 28-item self-report questionnaire used to measure effective and ineffective strategies for coping with a stressful life event [ 25 ]. The Brief-COPE assesses how a person deals with stressors in their daily life. The questionnaire measures 3 coping strategy dimensions: problem focused, emotion focused, and avoidant focused [ 26 ]. Each item is rated on a 4-point scale. The scores for the 3 overarching coping styles are calculated as average scores. This is done by dividing the sum of the item scores by the number of items. These average scores indicate the extent to which the respondent engages in each coping style. A higher score indicates that the respondent does not have many coping skills. The Chinese version of the Brief-COPE and its psychometric properties in Chinese populations were reported by Wang et al [ 27 ].

Infodemic- and Misinformation-Driven Overuse of Health Care Services

The COVID-19–related infodemic and misinformation-driven medical misbehavior were assessed using 2 self-developed items. The first item was “Do you believe there is an excessive amount of information regarding the COVID virus and vaccine on a daily basis?” The second item was “Has misinformation or disinformation about COVID-19 led you to engage in the overuse of health care services (eg, frequently visiting the doctor/psychiatrist or buying unnecessary medicine)?” The respondents were required to indicate their response to these 2 questions by selecting 1 of 2 options presented dichotomously: yes or no.

Statistical Analysis

Descriptive statistics were used to describe the participants’ background characteristics. Continuous variables (eg, age) were calculated as means and SDs. Categorical variables (eg, sex) were calculated as frequencies and proportions. The Pearson correlation coefficient ( r ) was used to examine the association between measures, where  r ≥0.3 and  r ≥0.5 indicated moderate and large effects, respectively [ 28 , 29 ].

In this study, we used PLS-SEM to estimate the research model parameters, as it works efficiently with small samples and complex models. Compared with covariance-based structural equation modeling, PLS-SEM has several advantages, such as the ability to handle non-normal data and small samples [ 30 ]. Unlike covariance-based structural equation modeling, which focuses on confirming theories, PLS-SEM is a causal-predictive approach that explains variance in the model’s dependent variables [ 31 ]. To improve the model fit, we used the bootstrapping method with 10,000 replications to obtain the estimates of the mean coefficients and 95% CIs [ 32 ]. Composite reliability rho_a (>0.7), composite reliability rho_c (>0.7), and average variance extracted (>0.5) were used to examine the model performance.

PLS-SEM encompasses measurement models that define the relationship between constructs (instruments) and indicator variables and a structural model. The structural model used in this study is presented in Figure 1 . We hypothesized that the infodemic significantly affects misinformation-driven medical misbehavior, resulting in cyberchondria and high anxiety levels. Furthermore, we speculated that coping strategies significantly modify this relationship. To test these hypotheses, we used 3 models that used the full sample to separately investigate the moderating effect of the 3 types of coping strategies (problem focused, emotion focused, and avoidant focused). We analyzed the data and estimated the PLS-SEM parameters using the “SEMinR” package in R (R Foundation for Statistical Computing). A P value of ≤.05 was considered statistically significant.

Background Characteristics of Participants

A total of 986 respondents completed the web-based survey and provided valid responses, resulting in a response rate of 84%. Among the participants, 51.7% (n=510) were female, approximately 95% (n=933) had completed tertiary education or above, and 71.2% (n=702) resided in urban areas. The participants’ background characteristics are listed in Table 1 .

a A currency exchange rate of 7.23 CNY=US $1 applies.

Mean Scores and Frequency of Responses

The mean score of the GAD-7 was 8.4 (SD 3.8), while the mean score of the CSS-12 was 39.7 (SD 7.5). Problem-focused coping had a higher mean score than emotion- and avoidant-focused coping. Respondents with active employment reported statistically significantly higher mean scores on the GAD and avoidant-focused coping subscale compared to those with nonactive employment. A higher proportion of respondents with chronic diseases experienced an infodemic and exhibited the overuse of health care services relative to those without chronic diseases ( Table 2 ). The correlations between all of the measures are presented in Multimedia Appendix 2 .

a GAD-7: Generalized Anxiety Disorder Assessment-7.

b CSS-12: Cyberchondria Severity Scale-12.

c COPE: Coping Orientation to Problems Experienced Inventory.

g P <.001.

Measurement Models

Tables 3 - 5 present the performance of the measurement models for the 3 coping strategies. The values of rho_C and rho_A were above 0.7, indicating acceptable construct reliability. All 3 constructs had Cronbach α values exceeding the cutoff of 0.7, indicating adequate reliability. Table 2 presents the models’ convergent validity. All the bootstrapped item loadings exceeded 0.3 and were significant at <.05 for the problem- and avoidant-focused models. However, for cyberchondria and the Brief-COPE, none of the average variance extracted values were above 0.5, indicating unsatisfactory model convergent validity.

a AVE: average variance extracted.

b GAD-7: Generalized Anxiety Disorder-7.

d HC: health care.

b GAD-7: Generalized Anxiety Disorder.

Structural Models

The structural model analysis involved estimating path coefficients for the conceptual model. We performed PLS-SEM on the research model 3 times to estimate path coefficients for the models with different coping strategies. We found that H1 was supported. A significant and positive relationship was observed between a high level of infodemic exposure and increased overuse of health care services (coefficient=0.212, 95% CI 0.151-0.271). In addition, the overuse of health care services was correlated with more severe cyberchondria and higher anxiety levels in all the 3 models, supporting H2 and H3. The effect of the overuse of health care services on cyberchondria was larger than its effect on anxiety. All these relationships were statistically significant ( Tables 3 - 5 ).

Moderating Effects

In our moderation analyses ( Figure 2 and Tables 6 and 7 ), we found that emotion- and avoidant-focused coping strategies significantly moderated the relationship between the overuse of health care services and cyberchondria and that between the overuse of health care services and anxiety, respectively, supporting H5 and H6. For the problem-based model (H4), the moderating effect was not significant for the relationship between the overuse of health care services and cyberchondria (coefficient=0.002, 95% CI −0.011 to 0.006), indicating that H4.1 was not supported. Compared with the direct effects on the relationship between the overuse of health care services and cyberchondria or anxiety, a strong ability to cope with difficulties can effectively mitigate the negative effects of the infodemic-driven overuse of health care services on cyberchondria and anxiety.

a HC: health care.

b GAD: Generalized Anxiety Disorder Assessment.

b CS: coping strategy.

c GAD: Generalized Anxiety Disorder Assessment.

Principal Findings

We performed a series of PLS-SEM analyses to examine the relationships between the infodemic-driven overuse of health care services and cyberchondria and anxiety and determine the moderating effects of 3 types of coping strategies on these relationships. We observed that the individuals who were exposed to an overload of COVID-19–related information were more likely to seek and use extra and unnecessary health care services during the pandemic. Such behavior may lead to a considerable wastage of health resources that are particularly limited during a public health crisis. Although some studies have indicated that during the COVID-19 pandemic individuals with increasing mental health symptoms rarely used mental health services [ 33 - 35 ], we found that the overuse of health care services may contribute to higher levels of depression and cyberchondria during a pandemic. This finding has never been reported before. However, we did not differentiate between the types of health care services, either physical or mental, that the individuals overused during the pandemic. This limitation may affect the implications of our findings for policy making purposes.

Comparisons With Previous Studies

We observed that enhanced coping strategies can mitigate the adverse effects of overusing health care on depression and cyberchondria. Studies have confirmed the association between pandemics and depression, have identified several sources of depression [ 6 , 7 , 10 , 36 , 37 ], and have determined the relationship between depression and cyberchondria [ 38 ]. However, few studies have investigated the relationship between depression or cyberchondria and the infodemic-driven overuse of health care services. Our findings demonstrate that the adverse effects of the pandemic are diverse and require the investigation of individuals’ health from multiple perspectives (ie, infodemic in health communication, the use of health care in health service research, and depression in psychiatry). These effects might not be immediately apparent, but they are all linked to each other and collectively cause harm. Thus, policy makers should develop a comprehensive and cost-effective strategy to address the potential adverse effects of pandemics on people’s health and well-being and better prepare for the next public health crisis.

This study offers new insights into the role of coping strategies in mediating the relationship between health care overuse and depression or cyberchondria during the COVID-19 pandemic. Overall, individuals with strong coping abilities were more likely to report lower levels of depression or cyberchondria than those with weak coping abilities. However, the moderating effects of different coping strategies varied slightly. We discovered that problem-focused coping strategies resulted in lower levels of depression and cyberchondria than avoidant-focused coping strategies. Additionally, emotion-focused coping strategies led to lower levels of depression than the other 2 types of coping strategies. These findings partially align with previous studies. For instance, Li [ 39 ] demonstrated that using both problem-focused and emotion-focused coping strategies was beneficial for psychological well-being. However, previous studies have reported mixed findings. For example, AlHadi et al [ 40 ] indicated that emotion-focused coping strategies were associated with increased depression, anxiety, and sleep disorders in the Saudi Arabian population. Few studies have examined the effect of avoidant-focused coping strategies. In this study, we found that respondents who reported living with chronic diseases exhibited a higher ability to use avoidant-focused coping. This finding is partially consistent with a previous study that found a positive relationship between avoidance-focused coping strategies and mental health in women with heart disease [ 41 ]. Individuals with medical conditions are more likely to adopt avoidant coping strategies. Firouzbakht et al [ 42 ] explained that avoidance is an effective strategy for handling short-term stress and is more likely to be adopted by certain patient groups.

We found that individuals who favor emotion-focused coping strategies to overcome difficulties are able to effectively mitigate the adverse effects of excessive health care use on depression and cyberchondria relative to those who opt for the other 2 coping strategies. This finding is not entirely surprising or unexpected. It is, in fact, quite reasonable when one considers that scholars and researchers in the field have previously indicated that people have a tendency to adopt emotion-focused strategies, especially when they find themselves in situations that are uncontrollable or unpredictable, such as the ongoing global pandemic [ 43 ]. Some studies have found that age can have a significant impact on an individual’s coping strategy preferences. For instance, younger adults were more likely to use emotion-focused coping strategies during the acute phase of the SARS outbreak, whereas older adults used this particular strategy several months after the outbreak had initially occurred [ 44 ]. This suggests that the coping strategies adopted by individuals can vary greatly depending on their age and the stage of the crisis they are experiencing. However, in the context of this study, we did not observe any significant differences in the coping strategy preferences of the different age groups. This could be due to a variety of factors, but a possible explanation is that our model incorporated the COVID-19 infodemic. In this context, it is understandable that providing emotional support might be more important than providing real solutions. This is particularly true in the current digital age, where the internet offers unlimited information sources for people to explore, which can often lead to information overload and increased anxiety. Therefore, emotion-focused coping strategies could be more beneficial in helping individuals navigate the sea of information and manage their emotional responses effectively.

In this study, we used self-developed items to measure the infodemic and overuse of health care services. While this approach allowed us to collect data that were directly related to the research questions, it may have introduced some potential issues. First, self-developed items may have less validity and reliability than standardized questionnaires. This could affect the accuracy of measurements and the validity of findings. Second, using self-developed items may limit comparability with other studies that use standardized questionnaires. Standardized questionnaires allow for easy comparison across studies and populations. The lack of a common metric may make it challenging to compare the findings of this study to other studies or to aggregate them in future meta-analyses. Finally, self-developed items may be more susceptible to response bias. They may not have considered factors like social desirability bias or acquiescence bias as standardized questionnaires do. This could have skewed the responses and affected the accuracy of the findings. Despite these limitations, the study’s findings provide valuable insights and pave the way for future research in this area.

Main Contributions of This Study

The importance of preparedness, prevention, and emergency response to infodemiology is highly encouraged by the WHO [ 45 ]. This study makes a significant contribution by exploring and empirically evaluating the relationship between the infodemic, the overuse of health care services, cyberchondria, and anxiety in the context of the COVID-19 pandemic. It provides empirical evidence supporting the assertion that a high level of infodemic can lead to the increased overuse of health care services, resulting in more severe cyberchondria and heightened anxiety levels. This finding adds a new dimension to our understanding of the psychological impacts of the infodemic, especially in the context of a global public health crisis. Additionally, this study highlights that adopting appropriate coping strategies can potentially reduce the severity of cyberchondria and anxiety, even among people facing high levels of the infodemic and the overuse of health care services.

Future Research

The study’s findings emphasize the importance of coping strategies in reducing the negative effects of the infodemic and the excessive use of health care. Future research could focus on developing and testing interventions to improve coping skills, such as cognitive-behavioral, mindfulness-based, or psychoeducational approaches. Additionally, other factors like social support, personality traits, or health literacy may moderate the relationship between infodemic, health care overuse, cyberchondria, and anxiety. Future research could further explore these variables. This study’s findings may not apply to all populations, so future research could investigate these relationships in different groups, including those with pre-existing mental health conditions, health care professionals, or diverse cultural contexts. By pursuing these future directions, researchers could build on this study’s findings, thereby enhancing our understanding of the psychological impact of infodemic and developing effective interventions.

Limitations

This study has several limitations that need to be addressed. A primary limitation is that the data were cross-sectional and self-reporting, which can introduce several biases. Social desirability bias may occur when respondents provide answers they believe are socially acceptable rather than truthful. Recall bias may also be present, as the respondents were asked to recall experiences from months or even a year ago. The data are also prone to response bias, as respondents may agree or disagree with statements regardless of their content. These biases may have affected the accuracy of the findings. In the future, we will try to collect data at multiple time points to reduce the biases and identify changes over time. Second, the data used in this analysis were obtained from a web-based survey, which excluded individuals who are not familiar with web-based surveys or do not have access to the internet. This could have resulted in selection bias. Additionally, due to the nature of the web-based survey, the demographic information of our sample was highly skewed. The majority of the respondents were young and highly educated and were frequent internet users who may have experienced more infodemic effects than older and less educated individuals. This may have affected the reliability of our findings. A quota sampling method could be used in future studies to improve the representativeness of the sample. Third, the study was conducted in China; thus, it is important to consider the unique context of China when interpreting the results. It is necessary to conduct further research in different cultural and regional contexts to determine the generalizability of the results. Finally, the evaluation of health care service overuse and strength of the infodemic relied on 2 self-developed items, which may have affected the measurement properties and limited the reliability of our findings. The development of standardized questionnaires to measure the infodemic and the overuse of health care services during a pandemic would be a valuable contribution to future research in this field.

Conclusions

This study is the first to demonstrate a significant correlation between the infodemic-driven overuse of health care services and high levels of depression and cyberchondria in the Chinese population during the COVID-19 pandemic. We find that 3 types of coping strategies can effectively mitigate the adverse effects of infodemic-driven health care overuse on depression and cyberchondria. Among them, emotion-focused coping strategies have stronger moderating effects than the other 2 types of coping strategies. These findings provide empirical evidence that can guide policy makers in developing strategies to reduce cyberchondria, provide accurate information about public health crises, and promote adaptive coping strategies to effectively manage future public health crises.

Data Availability

The data sets generated and analyzed during this study are available from the corresponding author on reasonable request.

Authors' Contributions

RHX contributed to developing the study concept and design, data analysis and interpretation, software, writing the original draft, and review and editing. CC contributed to data collection, software, and review and editing. Both authors approved the submitted version.

Conflicts of Interest

None declared.

English-translated questionnaire.

Correlations between measures.

• Joseph AM, Fernandez V, Kritzman S, Eaddy I, Cook OM, Lambros S, et al. COVID-19 misinformation on social media: a scoping review. Cureus. 2022;14(4):e24601. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Farsi D, Martinez-Menchaca HR, Ahmed M, Farsi N. Social media and health care (Part II): narrative review of social media use by patients. J Med Internet Res. 2022;24(1):e30379. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Health topic - Infodemic. World Health Organization. 2023. URL: https://www.who.int/health-topics/infodemic#tab=tab_1 [accessed 2023-05-14]
• Gogol M, Siebenhofer A. [Choosing wisely—against overuse in healthcare systems—activities in Germany and Austria in geriatric medicine]. Wien Med Wochenschr. 2016;166(5-6):155-160. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Morgan DJ, Leppin AL, Smith CD, Korenstein D. A practical framework for understanding and reducing medical overuse: conceptualizing overuse through the patient-clinician interaction. J Hosp Med. 2017;12(5):346-351. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Alnazly E, Khraisat OM, Al-Bashaireh AM, Bryant CL. Anxiety, depression, stress, fear and social support during COVID-19 pandemic among Jordanian healthcare workers. PLoS One. 2021;16(3):e0247679. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Choi EPH, Hui BPH, Wan EYF. Depression and anxiety in Hong Kong during COVID-19. Int J Environ Res Public Health. 2020;17(10):3740. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Tomasoni D, Bai F, Castoldi R, Barbanotti D, Falcinella C, Mulè G, et al. Anxiety and depression symptoms after virological clearance of COVID-19: a cross-sectional study in Milan, Italy. J Med Virol. 2021;93(2):1175-1179. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Tran HTT, Nguyen MH, Pham TTM, Kim GB, Nguyen HT, Nguyen NM, et al. Predictors of eHealth literacy and its associations with preventive behaviors, fear of COVID-19, anxiety, and depression among undergraduate nursing students: a cross-sectional survey. Int J Environ Res Public Health. 2022;19(7):3766. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Bareeqa SB, Ahmed SI, Samar SS, Yasin W, Zehra S, Monese GM, et al. Prevalence of depression, anxiety and stress in China during COVID-19 pandemic: a systematic review with meta-analysis. Int J Psychiatry Med. 2021;56(4):210-227. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Starcevic V, Berle D. Cyberchondria: towards a better understanding of excessive health-related internet use. Expert Rev Neurother. 2013;13(2):205-213. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Smith AC, Thomas E, Snoswell CL, Haydon H, Mehrotra A, Clemensen J, et al. Telehealth for global emergencies: implications for coronavirus disease 2019 (COVID-19). J Telemed Telecare. 2020;26(5):309-313. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Gallotti R, Valle F, Castaldo N, Sacco P, De Domenico M. Assessing the risks of 'infodemics' in response to COVID-19 epidemics. Nat Hum Behav. 2020;4(12):1285-1293. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Yang F. Coping strategies, cyberbullying behaviors, and depression among Chinese netizens during the COVID-19 pandemic: a web-based nationwide survey. J Affect Disord. 2021;281:138-144. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Fluharty M, Bu F, Steptoe A, Fancourt D. Coping strategies and mental health trajectories during the first 21 weeks of COVID-19 lockdown in the United Kingdom. Soc Sci Med. 2021;279:113958. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Shigeto A, Laxman DJ, Landy JF, Scheier LM. Typologies of coping in young adults in the context of the COVID-19 pandemic. J Gen Psychol. 2021;148(3):272-304. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Hang CN, Yu PD, Chen S, Tan CW, Chen G. MEGA: Machine Learning-Enhanced Graph Analytics for infodemic risk management. IEEE J Biomed Health Inform. 2023;27(12):6100-6111. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Lazarus RS. Coping theory and research: past, present, and future. Psychosom Med. 1993;55(3):234-247. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Wechsler B. Coping and coping strategies: a behavioural view. Appl Anim Behav Sci. 1995;43(2):123-134. [ FREE Full text ] [ CrossRef ]
• Hair JF, Hair J, Hult GTM, Ringle CM, Sarstedt M. A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM). Thousand Oaks, CA. SAGE; 2022.
• McElroy E, Kearney M, Touhey J, Evans J, Cooke Y, Shevlin M. The CSS-12: development and validation of a short-form version of the cyberchondria severity scale. Cyberpsychol Behav Soc Netw. 2019;22(5):330-335. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Peng XQ, Chen Y, Zhang YC, Liu F, He HY, Luo T, et al. The status and influencing factors of cyberchondria during the COVID-19 epidemic. a cross-sectional study in Nanyang City of China. Front Psychol. 2021;12:712703. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Spitzer RL, Kroenke K, Williams JBW, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006;166(10):1092-1097. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Sun J, Liang K, Chi X, Chen S. Psychometric properties of the Generalized Anxiety Disorder Scale-7 Item (GAD-7) in a large sample of Chinese Adolescents. Healthcare (Basel). 2021;9(12):1709. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Carver CS. You want to measure coping but your protocol's too long: consider the brief COPE. Int J Behav Med. 1997;4(1):92-100. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Poulus D, Coulter TJ, Trotter MG, Polman R. Stress and coping in Esports and the influence of mental toughness. Front Psychol. 2020;11:628. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Wang AWT, Cheng CP, Chang CS, Chen DR, Chen ST, Shieh V, et al. Does the factor structure of the brief COPE fit different types of traumatic events? Eur J Psychol Assess. 2018;34(3):162-173. [ FREE Full text ] [ CrossRef ]
• Brown TA. Confirmatory Factor Analysis for Applied Research, Second Edition. New York, NY. Guilford Publications; 2015.
• Nunnally JC. Psychometric Theory, 3rd Edition. New York, NY. Tata McGraw-Hill Education; 1994.
• Hair JF, Sarstedt M, Pieper TM, Ringle CM. The use of partial least squares structural equation modeling in strategic management research: a review of past practices and recommendations for future applications. Long Range Plann. 2012;45(5-6):320-340. [ FREE Full text ] [ CrossRef ]
• Streukens S, Leroi-Werelds S. Bootstrapping and PLS-SEM: a step-by-step guide to get more out of your bootstrap results. Eur Manag J. 2016;34(6):618-632. [ FREE Full text ] [ CrossRef ]
• Hair JF, Hult GTM, Ringle CM, Sarstedt M, Danks NP, Ray S. Partial Least Squares Structural Equation Modeling (PLS-SEM) Using R: A Workbook. Cham, Switzerland. Springer International Publishing; 2021.
• Liberman JN, Bhattacharjee S, Rui P, Ruetsch C, Rothman B, Kulkarni A, et al. Impact of the COVID-19 pandemic on healthcare resource utilization in individuals with major depressive disorder. Health Serv Res Manag Epidemiol. 2022;9:23333928221111864. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Saunders R, Buckman JEJ, Leibowitz J, Cape J, Pilling S. Trends in depression and anxiety symptom severity among mental health service attendees during the COVID-19 pandemic. J Affect Disord. 2021;289:105-109. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Lee J, Jeong HJ, Kim S. Stress, anxiety, and depression among undergraduate students during the COVID-19 pandemic and their use of mental health services. Innov High Educ. 2021;46(5):519-538. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Luceño-Moreno L, Talavera-Velasco B, García-Albuerne Y, Martín-García J. Symptoms of posttraumatic stress, anxiety, depression, levels of resilience and burnout in Spanish health personnel during the COVID-19 pandemic. Int J Environ Res Public Health. 2020;17(15):5514. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Chen J, Liu X, Wang D, Jin Y, He M, Ma Y, et al. Risk factors for depression and anxiety in healthcare workers deployed during the COVID-19 outbreak in China. Soc Psychiatry Psychiatr Epidemiol. 2021;56(1):47-55. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Jungmann SM, Witthöft M. Health anxiety, cyberchondria, and coping in the current COVID-19 pandemic: which factors are related to coronavirus anxiety? J Anxiety Disord. 2020;73:102239. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Li Q. Psychosocial and coping responses toward 2019 coronavirus diseases (COVID-19): a cross-sectional study within the Chinese general population. QJM. 2020;113(10):731-738. [ FREE Full text ] [ CrossRef ] [ Medline ]
• AlHadi AN, Alarabi MA, AlMansoor KM. Mental health and its association with coping strategies and intolerance of uncertainty during the COVID-19 pandemic among the general population in Saudi Arabia: cross-sectional study. BMC Psychiatry. 2021;21(1):382. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Sidhu A, Arora AK. A study of mental health and coping among women with coronary heart disease and women without coronary heart disease. Int J Adv Res Dev. 2017;2(6):1-3. [ FREE Full text ]
• Firouzbakht M, Rahmani N, Nia HS, Omidvar S. Coping strategies and depression during the COVID-19 pandemic in pregnant women: a cross sectional study. BMC Psychiatry. 2022;22(1):153. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Chew QH, Wei KC, Vasoo S, Chua HC, Sim K. Narrative synthesis of psychological and coping responses towards emerging infectious disease outbreaks in the general population: practical considerations for the COVID-19 pandemic. Singapore Med J. 2020;61(7):350-356. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Yeung DYL, Fung HH. Age differences in coping and emotional responses toward SARS: a longitudinal study of Hong Kong Chinese. Aging Ment Health. 2007;11(5):579-587. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Wilhelm E, Ballalai I, Belanger ME, Benjamin P, Bertrand-Ferrandis C, Bezbaruah S, et al. Measuring the burden of infodemics: summary of the methods and results of the Fifth WHO Infodemic Management Conference. JMIR Infodemiology. 2023;3:e44207. [ FREE Full text ] [ CrossRef ] [ Medline ]

Abbreviations

Edited by G Eysenbach, T de Azevedo Cardoso; submitted 05.10.23; peer-reviewed by K Wang, J Chen, CN Hang, E Vashishtha, D Liu; comments to author 06.11.23; revised version received 14.11.23; accepted 22.03.24; published 09.04.24.

©Richard Huan Xu, Caiyun Chen. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 09.04.2024.

This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on https://www.jmir.org/, as well as this copyright and license information must be included.

Watch CBS News

Can the eclipse impact your astrological sign? An astrologer weighs in

By Kristine Johnson

Updated on: April 5, 2024 / 6:14 PM EDT / CBS New York

NEW YORK —  We're just a few days away from the April 8 solar eclipse , and everyone will be looking skyward to watch this amazing cosmic phenomenon,  with the proper glasses  of course. 

Astrologers who look to the stars say that if you feel a little different during an eclipse, there's a reason. People are impacted by this alignment of the sun, the moon and the earth, according to astrologer Rebecca Gordon.

Gordon acknowledged to WCBS that, for some, astrology is all smoke and mirrors, but she encourages keeping an open mind. She says not all signs will be affected equally during the eclipse.

"Aries will definitely be impacted. Also, their opposite sign of Libra will be. And the signs of Cancer and Capricorn, as well, will certainly be impacted," she said. "Let's just say that every member of the zodiac will be impacted. Nobody gets out of an eclipse without impact." 

Gordon believes the planets always assert an influence over events and our behavior, and eclipses only magnify what is happening. 

"You look at patterns and cycles," Gordon said. "Big things happening on eclipses. That you simply cannot argue with." 

Gordon points to what she calls patterns of significant historical events that have occurred at key astrological times. And she says the impacts of the moon's path are significant in astrology.

"When the moon is full, the tides are high. All crustaceans are born on full moons. Did you know jellyfish are often born on full moons? So essentially, the waters of the ocean swell. What is your body made up of?" Gordon asked. 

"You are sort of in the eye of the needle of eclipse season. That is when there might be a bit of chaos and confusion. So in your life, there might be - why did that job let me go, that relationship just began, that relationship ended? There can be a whole lot of change... You want to not plan too much, leave space open, because you will need that space to react thoughtfully." 

CBS News New York will have complete coverage of the eclipse on April 8, from 2 to 4 p.m. 

Kristine Johnson currently co-anchors the 5 p.m. & 11 p.m. news at the CBS owned and operated station in New York City. She joined CBS 2 in the fall of 2006. Since then, she has been the recipient of several Emmy awards.

More from CBS News

Sacramento-area students, families safely stare at sky to see rare solar eclipse

Why do you need special glasses to watch a solar eclipse? Doctors explain.

Is it safe to take pictures of the solar eclipse with your phone?

Thousands gather in Sacramento to celebrate end of Ramadan