reading intervention case study

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
• HHS Author Manuscripts

Effects of Reading Interventions Implemented for Upper Elementary Struggling Readers: A Look at Recent Research

Rachel e. donegan.

Northern Illinois University

Jeanne Wanzek

Vanderbilt University

In this study, we conducted a review of reading intervention research (1988–2019) for upper elementary struggling readers and examined intervention area (e.g., foundational, comprehension, or multicomponent) and intensity (e.g., hours of intervention, group size, and individualization) as possible moderators of effects. We located 33 studies containing 49 treatment-comparison contrasts, found small effects for foundational reading skills ( g = 0.22) and comprehension ( g = 0.21), and decreased effects when considering standardized measures only. For intervention area, only multicomponent interventions predicted significant effects for both comprehension and foundational outcomes. For intensity, we did not find systematic evidence that longer or individualized interventions were associated with larger effects. However, interventions implemented in very small groups predicted larger comprehension outcomes. Overall, more research examining the quality of school provided reading instruction and how the severity of reading difficulties may impact effects of more intensive interventions is needed.

The upper elementary grades represent a critical time in literacy development. There are several advancements expected of students as they transition from the early elementary to late elementary grades. Students are expected to read texts that increasingly include longer, more difficult words ( Hiebert, 2008 ). In addition, students are expected to read and comprehend expository text which they may have had little exposure to earlier in their schooling ( Duke, 2000 ) and which poses unique and increasing challenges to comprehension ( Duke & Roberts, 2010 ). These challenges that emerge as students transition from the primary to upper elementary grades can contribute to continued reading difficulties for students who struggled with reading in the primary grades and also may be associated with the appearance of new, late-emerging reading difficulties for other students leading to a substantial population of struggling readers as students progress to their last years in elementary school ( Compton, Fuchs, Fuchs, Elleman, & Gilbert, 2008 ; Lipka, Lesaux, Siegel, 2006 ).

Upper elementary struggling readers are unique from those found in the primary grades and consequently may require specialized and diverse intervention approaches. In contrast to struggling readers in the primary grades where word reading difficulties account for the majority of the reading problems ( Nation & Snowling, 1997 ; Shankweiler et al., 1999 ), older struggling readers may demonstrate difficulties in foundational skills such as word reading, or more advanced skills such as comprehension of text, or they may struggle in both domains ( Leach, Scarborough, Rescorla, 2003 ; Lipka et al., 2006 ). According to the Simple View of Reading, deficits in either domain can severely impact overall reading performance ( Gough & Tumner, 1986 ).

In order to address the diverse reading needs of this group of struggling readers, interventions that focus on foundational skills or comprehension as well as multicomponent interventions that include instruction in both domains may all be needed. Foundational skill interventions include instruction in phonemic awareness, phonics and word recognition, and/or fluency. Effective phonics instruction for older struggling readers should go beyond the decoding of single syllable words to include strategies for decoding multisyllabic such as using syllabication and word parts (e.g., roots and affixes) to break apart and decode longer words ( Boardman et al., 2008 ). Repeated reading, where students read the same passage multiple times while receiving feedback and teacher or peer support, and nonrepetitive reading, where students receive the same feedback and supports but read one or more texts without additional readings, are two types of fluency interventions that have both been shown effective for improving struggling readers’ foundational reading skills ( Lee & Yoon, 2017 ; Zimmerman, Reed, & Aloe, 2019 ).

Explicit vocabulary instruction and comprehension strategy instruction are two effective approaches for improving the reading comprehension of older struggling readers ( Elleman, Lindo, Morphy, & Compton, 2009 ; Kamil et al., 2008 ). Direct instruction in word meaning that goes beyond definition teaching to include deep processing activities including generating examples and nonexamples and creating semantic maps are effective for increasing the vocabulary knowledge of older struggling readers ( Bryant, Goodwin, Bryant, & Higgins, 2003 ). When it comes to comprehension strategy instruction, three particular areas that show consistent positive effects for older struggling readers are learning strategies for identifying the main idea, summarizing the text, and self-regulating or monitoring learning strategies while reading ( Wanzek, Wexler, Vaughn, & Cuillo, 2010 ; Solis et al., 2012 ).

Multicomponent interventions that include instruction in both foundational and comprehension skills have shown promise for students in the upper elementary grades according to a review by Wanzek et al. (2010) . These interventions combine effective instruction across components making them well-aligned to the needs of older struggling readers who tend to have difficulties across reading domains. However, there were few studies examining multicomponent interventions at the time of the review. In addition, since the Wanzek et al. review, more research examining multicomponent reading interventions for upper elementary students with reading difficulties has been conducted.

Intensive Interventions

It may also be important to consider features of intervention intensity such as implementation intensity (e.g., total duration and group size) and individualization when considering intervention approaches for older struggling readers since these readers may demonstrate persistent reading difficulties or present reading skills that are far below those of their peers.

There has been limited work examining the effects of reading interventions implemented for different durations and in different sized groups for older students and the results have been mixed. One recent study examining the effects of intervention duration for late elementary students, those randomly assigned to two years of intervention demonstrated stronger word reading skills than those randomly assigned to receive intervention for only one year; however, these effects did not extend to comprehension outcomes ( Miciak et al., 2018 ). Another study examining impact of group size on middle school struggling readers found a pattern of effects favoring students randomly assigned to smaller groups for decoding and spelling outcomes but these differences were not statistically significant ( Vaughn et al., 2010 ).

Previous work has demonstrated individualized interventions that include instruction designed to address individual student need may be effective for elementary and secondary students who have failed to respond adequately to previous interventions ( Denton et al., 2013 ; O’Connor, Fulmer, Harty, & Bell, 2005 ; Pyle & Vaughn, 2012 ). However, there hasn’t been conclusive evidence that individualized interventions have added value over standardized interventions where instruction is consistent across students. For example, a study by Vaughn et al. (2011) compared reading outcomes of middle school struggling readers randomly assigned to a standardized intervention, individualized intervention, or business as usual comparison group and found both intervention groups similarly outperformed the comparison group on decoding, fluency and comprehension.

In summary, fourth and fifth grade struggling readers present diverse reading needs. Due to these diverse needs, a broad range of interventions have been developed. When examining the effects of interventions, in addition to considering instructional content, a closer look at intensity when considering intervention effects may be needed, especially for students with intensive reading needs.

Several researchers have examined reading intervention research for older students, including students in the upper elementary grades. Flynn, Zheng, & Swanson (2012) synthesized the effects of reading interventions for students in Grades 5 to 9 with reading disabilities and reported small to moderate effects across outcomes but noted few studies examined comprehension or multicomponent interventions. Outcomes did not differ significantly by type of intervention or duration of intervention. Scammacca, Roberts, Vaughn, & Stuebing (2015) used meta-analytic techniques to analyze the effects of reading interventions for older students (grades 4–12) and found a mean effect size of 0.49 for all measures as well as an effect of 0.21 for standardized reading outcomes. The authors noted that the effects of interventions decreased across year of publication with more recent studies demonstrating smaller effects, suggesting a trend of diminishing effects. Scammacca and colleagues also reported some moderation of outcomes by intervention type, with vocabulary interventions resulting in higher effects than other types of interventions. Comprehension interventions were also found to have a significantly higher mean effect size than fluency or multicomponent interventions. The number of hours of intervention was not a significant moderator of outcomes for more recent studies.

Wanzek et al. (2013) examined the research on interventions with 75 or more sessions for students with reading difficulties in Grades 4 to12. Overall effects were reported as 0.10 for comprehension outcomes, 0.16 for reading fluency outcomes, 0.16 for word fluency outcomes, and 0.15 for spelling outcomes with only comprehension outcomes having significant variation across studies. The authors reported no significant moderation of the comprehension outcomes by group size or number of hours of intervention.

These research syntheses combined research on interventions for students in the upper elementary grades with the middle and secondary grades. Although no moderation by grade level was reported, outcomes and moderators of outcomes can vary dramatically across this wide range of grade levels ( Pearson, Palinscar, Biancarosa, & Berman, 2020 ), making it difficult to ascertain the implications specifically for upper elementary students. Wanzek et al. (2010) conducted a comprehensive review of school-based reading intervention research for students in Grades 4 and 5. They located 13 studies that used between group treatment-comparison designs. Out of these 13 studies, five focused on comprehension interventions with medium to very large mean effects reported, six studies focused on foundational skills with small to moderate effects reported for word study interventions specifically, and two studies focused on multicomponent interventions with medium to large effects reported. Overall, there were positive effects for reading interventions provided to upper elementary struggling readers. However, there were a small number of studies available overall, and the authors noted limitations of some of the studies including the use of unstandardized, researcher-developed measures to evaluate the effects of comprehension and vocabulary interventions. These types of measures are often associated with larger effects when compared to standardized, norm-referenced outcomes ( Edmonds et al., 2009 ; Scammacca et al., 2015 ) and only provide information on outcomes closely aligned with the intervention rather than more generalizable outcomes. In addition, Wanzek et al. reported there were few studies on multicomponent interventions.

Since Wanzek’s et al.’s (2010) earlier review, several large-scaled randomized control trials focusing on late elementary struggling readers have been completed. This more robust research base along with Scammacca et al.’s (2015) finding of a trend of diminishing effects for more recent studies, and the unique and changing needs of struggling readers as they transition to the late elementary grades warrant an updated review of the research on reading interventions for late elementary students. The purpose of this review is to update and extend the earlier synthesis by Wanzek and colleagues (2010) for upper elementary struggling readers. To do this, we conducted updated searches to identify reading intervention research with late elementary students published since 2007. To extend the work, we carried out a statistical analysis of effects of intervention on reading outcomes, examining how intervention areas and intensity may have moderated effects. We sought to answer the following research questions:

• What are the effects of reading interventions for fourth and fifth grade struggling readers?
• How are effects moderated by intervention areas and intensity?

Inclusion Criteria

We adapted Wanzek et al.’s (2010) inclusion criteria for this review. In order to be included in this review, studies had to meet the following criteria:

• A reading intervention targeting literacy in English for students struggling with reading was provided. Reading interventions provided to all students as a part of the general education curriculum were excluded.
• The reading intervention included instruction in word study, fluency, vocabulary, comprehension, or a combination of these components and was provided in a school setting. Programs conducted in home, clinic, or camp settings were excluded.
• Participants in the reading intervention were described by study authors as below grade level in reading, at risk for reading disabilities or difficulties, or identified with reading disabilities or disaggregated data for these readers was provided. Studies that included participants with sensory or intellectual disabilities were excluded.
• More than 50% of participants were in fourth or fifth grade (9 to 11 years old) or disaggregated data for these participants was provided.
• Interventions were provided for a minimum of 15 sessions.
• Study outcomes included at least one measure of reading.
• The study used an experimental or quasi-experimental design where effects of intervention were compared across groups, and data for calculating effect sizes were provided. Quasi-experimental studies had to either include evidence of baseline equivalence or use a group assignment procedure to equate group (e.g., matched-pairs).
• The study was published in English in a peer-reviewed journal.

Search and Screening Procedures

In order to identify studies for this review, we both conducted updated searches and identified studies included Wanzek’s et al., (2010) earlier review that met our inclusion criteria. In order to identify articles published since Wanzek and colleagues (2010) review, we followed a three-step search process. First, we conducted a search of three electronic databases, ERIC, Education Full Text, PsycINFO. We used search terms combining key words for reading ( read* ), struggling readers (reading difficult*, learning disabilit*, dyslexi*, learning problem*, reading failure, language impair*, reading disability*, low-achieving ) and reading intervention (remedial reading, teaching methods, reading NEAR/4 intervention*, reading education, special education ). Since the purpose of this search was to identify studies published since the original review, we limited results to documents from 2007 to 2019. Second, to identify recently published articles, we conducted a hand search of major journals that frequently publish reading intervention research ( Exceptional Children, Journal of Educational Psychology, Journal of Learning Disabilities, Journal of Special Education, Learning Disabilities Quarterly, Learning Disabilities Research and Practice, Reading and Writing, Remedial and Special Education, Scientific Studies of Reading ) for articles published online or in print in 2018 and 2019. Finally, ancestral searches were conducted of identified articles that met all inclusion criteria.

Using these search procedures, 5,100 articles were identified. Articles identified from the electronic and hand search were screened using a two-step process (see Figure 1 ). First, abstracts of identified articles were screened. After conducting the abstract screening, the full text of the remaining articles was screened. Common reasons articles were eliminated included during the abstract screen and full text screen were the majority of participants were outside fourth or fifth grades or a reading intervention was not provided. In total, 25 articles published since the original review were identified for inclusion.

Prisma diagram.

Note . *One article described two studies

Next, we screened articles included in the original review for inclusion. Out of the 12 articles detailing treatment/comparison design studies included in the original review, seven articles describing eight studies met all inclusion criteria and were included. Three articles were excluded because they did not include information to calculate effect sizes and two articles were excluded because the intervention provided targeted literacy in a language other than English. In total, this left 33 studies identified for inclusion in this review.

In order to systematically capture study information, we developed a code sheet that included information on participants; study design and quality; measures including the component skill measured (foundational or comprehension) and if the measure was a norm-referenced, standardized reading outcome or an unstandardized measure such as an experimenter created measure, a skill inventory, or informal diagnostic measure closely aligned to the intervention; intervention areas including if the intervention included instruction in foundational reading skills only (phonemic awareness, phonics, or fluency), comprehension only (vocabulary and comprehension) or was multicomponent including instruction across both foundational skills and comprehension; intervention intensity including total hours of intervention, group size, and individualization; and reading outcomes. Studies were identified as including individualized instruction if any of the instructional targets or materials used in the intervention were initially planned and adjusted during the course of the intervention based on individual student progress. Due to reporting by some study authors of group size and hours of intervention as a range, these study characteristics were coded categorically. For group size, studies were coded according to the maximum group size reported (or average group size if maximum size was not reported) by the authors and were coded as implemented in group sizes of one or two, group sizes of four to seven, or groups of nine or more. For hours of intervention, studies were coded as having an intervention of 15 or fewer hours, 16 to 30 hours, or more than 30 hours. If the information described above was not included in the article, the first author of the article was contacted to request the information.

All studies were coded by the first author. To assess interrater reliability, two advanced doctoral students were trained in screening and coding procedures. Training included reviewing inclusion criteria and identifying articles that met did not meet inclusion criteria, reviewing the codebook and coding sample articles to reach a minimum of 90% reliability. Then, we randomly selected 20% of articles from both the abstract screen and full text screen to be double screened. Interrater reliability was 93% for the abstract screening and 92% for the full text screening. We randomly selected 13% of included articles to be double coded. Interrater reliability was 100% across all sections of the code sheet.

Meta-analytic Methods

Several of the identified articles included multiple treatment groups, resulting in a total of 49 treatment/comparison contrasts. These were analyzed separately for foundational and comprehension outcomes. Effect sizes were calculated with post-test mean and standard deviations using the Campbell Collaborative Calculator ( Wilson, n.d. ). The calculator provided Cohen’s d effect sizes. For one study ( Reed, Aloe, Reeger, & Folsom, 2019 ), we used author reported effect sizes and variances in the absence of post-test means and standard deviations. All effect sizes were converted to Hedges g using the following formula

where n 1 and n 2 correspond to the number of participants in each group and df equals the total number of participants minus the number of groups. For studies with multiple effect sizes, we computed the average effect size for foundational and comprehension outcomes, weighting the averages by the sample size.

We examined the effect sizes for the presence of outliers by constructing stem and leaf plots. In order to assess the presence and magnitude of heterogeneity in effect sizes, we calculated Q , τ 2, , I 2 statistics. In order to analyze effects, we used a weighted regression using a random effects model. We calculated weights for each effect size using the following formula

where σ 2 refers to the subject sampling variance estimate and τ 2 refers to the between study variance. We applied these weights when conducting the regression. We assessed the impact of the following moderators: type of measure (standardized or unstandardized), intervention components (foundational, comprehension, or multicomponent), and intensity (group size, hours of intervention, individualization).

Participants

The 33 studies included 4565 total participants. Participants included readers with and at risk for disabilities and students generally classified as struggling readers and therefore included a broad range of reading abilities (see Table 1 ). Across studies, on average 55% of participants were male and 45% of participants were female.

Intervention Descriptions and Outcomes

Note . T = Treatment; C = Comparison; Avg. = Average; Found. = Foundational; PA = Phonological awareness; WR & S = Word reading and spelling; Comp. = Comprehension; Vocab. = Vocabulary; GMRT = Gates-MacGinitie Reading Test; WRMT = Woodcock Reading Mastery Test; TOWRE = Test of Word Reading Efficiency; TOSREC = Test of Silent Reading Efficiency and Comprehension; PPVT = Peabody Picture Vocabulary Test;

Forty-seven interventions were implemented across 49 treatment/comparison contrasts in the included studies (see Table 1 ). Out of the 49 contrasts, 38 contrasted a reading intervention with typical school reading instruction only. Typical school reading instruction consisted of core reading instruction for all students and also may have consisted of school provided intervention to some students in treatment and/or comparison groups as schools deemed necessary. In three contrasts, a researcher-provided intervention was compared with an active control condition ( Keller, Ruthruff, & Keller, 2019 ) or an alternative researcher-provided intervention ( Ring, Barefoot, Avrit, Brown, & Black, 2013 ; Torgesen et al., 2001 ). In three contrasts, a researcher-provided intervention was compared to a school-provided intervention ( Kim, Capotosto, Hartry, & Fitzgerald, 2011 ; Kim, Samson, Fitzgerald, & Hartry, 2010 ). In two contrasts, a teacher provided intervention was compared to a different teacher provided intervention ( Das, Hayward, Georgiou, Janzen, & Boora, 2008 ; Xin & Rieth, 2001 ). In two contrasts, a researcher-provided intervention was compared to a previously received intervention ( Das, Mishra, & Pool, 1995 ).

Intervention Implementation

Most interventions were implemented by researchers ( n = 33). The remaining were implemented by in-service or pre-service teachers ( n = 10), paraeducators ( n = 2), or volunteers ( n = 1). The majority of the interventions ( n = 20) were multicomponent, 16 focused on foundational skills only, and 11 included instruction in comprehension only. In order to assess the intensity of the implemented intervention, we analyzed the total hours of intervention, group size, and if any aspect of the intervention was individualized (see Table 2 ). Most interventions ( n = 17) included more than 30 hours of instruction. Out of the five interventions implemented for the most hours, three were after-school or summer programs ( Kim et al., 2010 ; Kim et al., 2011 ; Reed et al., 2019 ), one included two daily intervention sessions ( Torgesen et al., 2001 ), and one implemented a reading intervention across two school years ( Miciak et al., 2018 ).

Study Characteristics

Note . Q = Quasi-experimental; E = Experimental; SWD = Students with disabilities; Indiv. = Individualized

Group size varied widely across the interventions. Most interventions ( n = 22) were implemented in small groups with a maximum group size between four to seven students. Two studies that implemented interventions with the largest groups also included limited time for practice and instruction in smaller groups ( Kim et al., 2011 ; Reed et al., 2019 ).

Eleven interventions included at least some individualization of instruction during the intervention. Out of these, only one intervention included individualization throughout instruction ( Thames et al., 2008 ). The remaining interventions included individualization in specific components of the intervention but not for the full intervention. These included computerized multicomponent practice ( Kim et al., 2010 ; Kim et al., 2011 ; Roberts et al., 2018 ), word study instruction ( Miciak et al., 2018 [two interventions]; Vaughn, Solís, Miciak, Taylor, & Fletcher, 2016 ), fluency instruction (Young, Mohr, Rasinski, 2015), brief multicomponent small group instruction (provided in addition to larger group instruction that was not individualized; Reed et al., 2019 ), or individualized texts based on student performance ( O’Connor et al., 2002 ; Therrien, Wickstrom, & Jones, 2006 ).

Meta-Analytic Results

In this section, we report the meta-analytic results, including mean effects and variance explained, both overall for foundational and comprehension outcomes and for each moderator. We initially interpret effect sizes as small, medium, or large using benchmarks suggested by Cohen (1988) though more nuanced interpretation of these effects in comparison to effects from similar studies are found in the discussion.

Foundational outcomes.

There were 46 effects for foundational outcomes. We constructed a stem and leaf plot to detect outliers (see Figure 2 ) and found two effect sizes at the top end of the distribution from unstandardized reading outcomes. We retained these effect sizes and used a moderator analysis to further investigate differences in effects between standardized and unstandardized reading outcomes.

Stem and leaf plot of foundational outcome effect sizes

The estimated mean effect across all studies was g = 0.22 ( p < .001, 95% CI [0.10, 0.33]) indicating a small positive and significant effect of reading interventions on foundational reading outcomes. The Q statistic indicated heterogeneity was present ( Q total = 159.44, df = 45, p < .001). Further analysis indicated large heterogeneity ( I 2 = 71.78, τ 2 = 0.10). In order to determine if there was a difference in effects for standardized versus unstandardized reading outcomes, we entered outcome type into the regression equation as a predictor. Outcome type explained a significant amount of heterogeneity ( Q between = 21.02, df = 1, p < .001) and thus was a predictor of effects. Unstandardized outcomes were associated with large, positive effects ( g = 0.83, SE = .014, p < .001; β = 0.73, p < .001). For standardized outcomes, the mean effect was smaller and nonsignificant (see Table 3 ).

Foundational Outcomes

Next, we determined the mean effect size and significance for each moderator when entered into the regression separately (see Table 3 ). Then we entered all moderators into the equation together to determine which moderators predicted unique variance in effect sizes. For intervention components, small, positive effects were found for multicomponent interventions ( g = 0.16, SE = 0.08, p = .04) and were significantly different from outcomes for foundational and comprehension interventions ( β = 0.16, SE = 0.08, p = .04). The effect sizes for foundational interventions, though also small and positive ( g = 0.36, SE = .13, p < .01), did not predict unique effects when compared to interventions with other components. For group size, interventions implemented in instructional groups of four to seven predicted small effects ( g = 0.25, SE = 0.08, p = .002; β = 0.25, p = .002). Interventions implemented for 16 to 30 hours were associated with small, positive effects ( g = 0.35, SE = 0.14, p = .01), although these effects were not unique from interventions of different durations. Finally, interventions that were not individualized predicted small, positive effects ( g = 0.25, SE = 0.07, p < .001; β = 0.25, p < .001) that were more consistent than effects from individualized interventions.

When all predictors were entered together into the regression equation, a significant portion of heterogeneity was explained ( Q between = 24.00, df = 12, p = .02). Outcome type was the only predictor of unique effects on foundational outcomes with larger effects for unstandardized outcomes ( β = 0.68, SE = 0.17, p < .001).

Comprehension outcomes.

There were 53 effect sizes for comprehension outcomes. We used the same procedure to detect outliers (see Figure 3 ). We did not detect any outliers but followed the same procedure and used a moderator analysis to further investigate differences in effects.

Stem and leaf plot of comprehension outcome effect sizes

The estimated mean effect was g = 0.21 ( p < .001, 95% CI [0.12, 0.30]) indicating a small positive, significant effect of reading interventions on comprehension reading outcomes. The Q statistic indicated significant heterogeneity ( Q total = 113.41, df = 52, p <.001). Further analysis indicated medium heterogeneity ( I 2 = 54.15, τ 2 = 0.05). We used the same procedure to determine if there was a difference in effects by outcome type and found a significant amount of heterogeneity explained ( Q between = 9.05, df = 1, p = .003). Outcome type was a significant predictor of effects. Both outcome types predicted significant effects with unstandardized outcomes predicting larger effects ( g = 0.44, SE = 0.10, p < .001; β = 0.30, p = .003) than standardized outcomes ( g = 0.13, SE = 0.05, p = .009; β = 0.13, p = .009).

We used the same procedure to calculate mean effect sizes and determine significance for each moderator (see Table 4 ). For intervention components, multicomponent interventions predicted small, positive effects ( g = 0.17, SE = 0.06, p = .004; β = 0.17, p = .004). Though small, positive effects were also noted for comprehension interventions ( g = 0.29, SE = 0.10, p = .003), these interventions did not predict unique effects when compared to interventions with different components. Group size predicted effects. Interventions implemented in groups of one or two students predicted the largest effects with medium, positive effects noted ( g = 0.50, SE = 0.10, p < .001; β = 0.41, p < .001) while mean effect sizes for larger groups were nonsignificant. Interventions that included 15 or fewer hours of instruction predicted unique effects on comprehension outcomes with medium, positive effects noted ( g = 0.44, SE = 0.12, p < .001; β = 0.28, p = .02); however, this finding is limited due to five of the 14 effect sizes for this moderator related to unstandardized outcomes. Interventions that more than 30 hours predicted unique effects with small, positive significant effects noted ( g = 0.16, SE = 0.07, p = .01; β = 0.16, p = .01). Interventions that were not individualized predicted small, positive effects on comprehension outcomes that were more consistent than effects from individualized interventions ( g = 0.25, SE = 0.05, p < .001; β = 0.25, p < .001).

Comprehension Outcomes

When all predictors were entered together into the regression equation, a significant portion of heterogeneity was explained ( Q between = 28.86, df = 12, p = .004). Group size was the only unique predictor of comprehension outcomes with very small groups (one or two students) predicting positive, larger effects ( β = 0.44, SE = 0.12, p < .001).

Year of publication.

The studies included in the review were published over more than 3 decades (1988–2019), and other investigators have found year of publication moderated effects ( Austin et al., 2019 ; Scammacca et al., 2015 ). Thus, we completed an exploratory analysis on the effect of year of publication on both foundational and comprehension outcomes by entering it into the regression equation as a predictor. We found year of publication explained a significant portion of heterogeneity and was a significant predictor of effect size for comprehension outcomes only with more recent studies associated with smaller effects ( Q between = 5.20, df = 1, p = .02; β = −0.02, SE = 0.01, p = .02).

Publication bias.

In order to assess publication bias, we used analysis of visual displays of effect sizes and a statistical analysis. First, we constructed funnel plots by plotting each study’s effect size against an estimate of the study size and visually analyzed the plots for symmetry. Then we used Egger’s linear regression method to statistically evaluate the symmetry of the plots by regressing the standardized effect sizes on their precisions ( Egger, Smith, Schneider, & Minder, 1997 ). An intercept including zero and symmetrical plots with an equal dispersal of studies on both sides of the overall mean effect would indicate no publication bias. An intercept not including zero and asymmetry in the plots would indicate the possible presence of publication bias. For foundational outcomes, the funnel plot did not show significant asymmetry ( p = .27) indicating a lack of evidence of publication bias for these outcomes. For comprehension outcomes, the funnel plot did show significant asymmetry ( p < .001). Visual inspection showed smaller studies smaller studies with low or negative effects missing indicating some evidence of publication bias. Thus, the reported effect size for comprehension outcomes may be inflated.

Wanzek et al. (2010) synthesized the results of reading intervention research for upper elementary students with reading difficulties or disabilities conducted from 1988 to 2007 and found small to moderate effects for foundational interventions, medium to large effects for comprehension interventions, and medium to large effects for multicomponent interventions. The goal of this study was to update and extend Wanzek’s et al.’s earlier review by conducting a meta-analytic review, including studies published since 2007, and examining intervention components and intensity as potential moderators of effects.

Overall, we found positive effects of upper elementary reading interventions for foundational and comprehension reading outcomes, but effects that were smaller than those described in Wanzek et al.’s (2010) review. When it comes to interpreting the magnitude of effects, Lipsey et al. (2012) recommend comparing effects to those obtained in similar interventions to interpret magnitude. The mean effects of our study may best be compared to the results of Scammacca et al.’s (2015) findings for reading interventions studies with fourth to 12 th grade students conducted from 2005–2011 since the majority of our effects come from studies published since 2005. Scammacca et al. report mean effects of g = 0.23 across all reading outcomes, g = 0.21 across all reading comprehension outcomes, and g = 0.13 across all standardized reading outcomes. We found effects of similar magnitude with overall mean effects from g = 0.21 to 0.22 for all reading outcomes and mean effects from standardized outcomes ranging from g = 0.09 to 0.13.

Also similar to Scammacca et al. (2015) , we noted a trend of diminishing effects where the effects in this examination were smaller than those in past reviews focused on similar age students ( Wanzek et al., 2010 ; Flynn et al., 2012 ). In fact, several recent, large scale, high quality studies included in this review reported null effects for standardized reading outcomes ( Connor et al., 2018 ; Roberts et al., 2018 ; Vaughn et al., 2016 ; Vaughn, Roberts, Miciak, Taylor, & Fletcher, 2019 ; Wanzek & Roberts 2012 ). To further investigate this trend of diminishing effects, we examined if more recent publications were associated with smaller effects by analyzing year of publication as a potential moderator. We found more recent publications associated with smaller effects for comprehension outcomes only. These diminishing effects for comprehension may be due to increased study quality in more recent investigations ( Austin et al., 2019 ).

However, we did not find a similar relation between year of publication and effect size for foundational outcomes. Descriptively, we did note several differences in participants from the that may be important to note alongside our findings as possible explanations. Across the studies from Wanzek’s earlier review, on average participants earned a standard score of 81 in foundational reading skills at pretest. In the more recent studies published since Wanzek’s earlier review, participants on average demonstrated stronger foundational reading skills with an average standard score of 90 at pretest. It’s possible that this improved performance was due to students receiving better reading instruction or more students receiving reading intervention from their schools. Others have noted improved classroom reading instruction as a potential explanation of diminishing effects ( Lemons, Fuchs, Gilbert, & Fuchs, 2014 ). However, information included in studies on classroom reading instruction and school-based reading interventions was often limited.

Another possible explanation considers the population targeted in the more recent studies. Not only did the students in the more recent studies enter with better foundational reading skill, less of them were also identified with disabilities. Of the eight studies included in Wanzek’s earlier review, seven included a majority of participants with disabilities (see Table 2 ). Out of the 17 recent studies that gave information regarding participant disability status, only 3 studies included a majority of participants with identified disabilities.

It is critical to note, these explanations are speculative. More detailed examinations of school-provided interventions, core reading instruction, and more complete reporting on student disability or at-risk status are needed and may shed light on the reasons behind the diminishing effects for standardized reading outcomes in comparison to effects reported in previous meta-analytic reviews.

Foundational reading

Foundational reading interventions predicted small, though not unique, effects ( g = 0.36) on all foundational reading outcomes. Across the studies, there were mixed effects for interventions focused solely on fluency with three interventions predicting small to medium, positive effects and 3 predicting negative effects. There were more consistent, positive effects for foundational interventions that included a combination of foundational reading components. Keller et al., (2019) reported large, positive effects for unstandardized reading outcomes for an intervention package that included computerized and researcher-provided instruction in phonemic awareness, phonics, spelling, fluency, text reading strategies along with mindfulness training. Torgesen et al., (2001) reported small, positive effects for standardized reading outcomes for an intervention that included researcher provided instruction focused on phonemic awareness, encoding, and decoding at the word level.

Interventions including decoding focused on multisyllabic words such as structural analysis (breaking a word down into decodable parts by using affixes or syllables) have been found by others to be effective for improving the reading outcomes of older struggling readers ( Edmonds et al., 2009 ) and also showed promise in our review. Toste, Capin, Williams, Cho, & Vaughn (2019) implemented a rigorous high-quality study examining the effects of a foundational reading intervention focused on multi-syllabic word reading and showed the largest effects among the foundational reading interventions (ES = 0.51 to 0.75) and word reading outcomes (ES = 0.45 to 0.80). In addition, several multicomponent interventions that showed the strongest, positive effects for foundational reading outcomes included decoding instruction focused on the word-level and expanded to the text level ( Miciak et al., 2018 ; O’Connor et al., 2002 ; Ring et al., 2012; Vaughn et al., 2019 ). This systematic integration of decoding skills is also recommended as an effective approach for younger struggling readers as it helps build generalization of skills to other contexts (Gersten et al., 2009).

Comprehension

Comprehension interventions also predicted small, but not unique, effects on comprehension outcomes ( g = 0.29). Six interventions ( Cirino et al., 2017 ; Fuchs et al., 2018 ; Mason, Davison, Hammer, Miller, & Glutting, 2013 ) where small to medium positive effects were noted for comprehension outcomes all focused on strategy instruction including main idea and summarization. Strategy instruction, where comprehension strategies are taught using direct and explicit instruction has shown strong evidence of effectiveness for improving students’ reading comprehension ( Kamil et al., 2008 ; Shanahan et al., 2010 ). Main idea and summarization, two specific strategies, have also been widely and effectively used to improve the comprehension of older struggling readers specifically ( Solis et al., 2012 ).

Similar to Wanzek et al.’s (2010) earlier findings, multicomponent interventions predicted significant, positive and unique effects for both foundational and comprehension outcomes. Unlike Wanzek et al.’s earlier finding, the mean effect was small across both outcomes ( g = 0.16 for foundational; g = 0.17 for comprehension). This finding differs from Scammacca et al.’s (2015) finding where comprehension interventions predicted the strongest effects.

The density of comprehension and vocabulary instruction in several intervention studies included in this review provides a possible explanation ( Kim et al., 2011 ; Reed et al., 2019 ; Vadasy & Sanders, 2008 ; Wanzek et al., 2017 ). These multicomponent interventions may be better equipped to address the needs of the current population of struggling readers. In contrast to the studies from Wanzek’s earlier review where students tended to show more substantial deficits in foundational reading skills than comprehension, students in the more recent studies tended to show more substantial difficulties in comprehension than in foundational skills. In regards to the instruction included in the multicomponent interventions, although these interventions included instruction in foundational reading skills, there was also substantial, explicit and systematic comprehension and vocabulary instruction. For example, when instruction focused on text reading fluency, it frequently occurred alongside comprehension instruction. Therefore, the balance of components of instruction in these interventions may be better suited to the older struggling reader

In order to examine the effects of intervention intensity on reading outcomes, we examined hours of intervention, group size, and individualization as moderators. When examining effect of intervention duration and group size, we had mixed findings.

Group size was a significant moderator of comprehension outcomes. The smallest groups had the largest mean effects. This may be especially important as other investigators have noted comprehension as a difficult outcome to impact for older students (e.g., Vaughn et al., 2016 , Vaughn et al., 2019 , Miciak et al., 2018 ). Smaller groups may have afforded more opportunities for teacher and student conversations, more student specific feedback, and differentiation of instruction to meet individual student needs. Previous findings have also been mixed in the area of instructional group size. Wanzek et al. (2013) reported no moderation of student reading outcomes based on group size; however, group size could only be examined categorically with studies implementing instructional groups of 5 or less compared to studies implementing instructional groups of 6 or more. Flynn et al. (2012) noted small group instruction had a positive correlation with student outcomes, but the authors also noted interpretation is difficult because all of the studies with small groups also had other instructional components that may be related to student outcomes.

Longer duration interventions predicted significant effects for comprehension outcomes only, and these effects were small. In addition, this finding may be explained by intervention components as 15 out of the 19 interventions included were multicomponent. Recent syntheses of intervention research with students in older grade levels have also found the duration of intervention did not moderate student outcomes in the interventions (e.g., Flynn et al., 2012 ; Wanzek et al., 2013 ; Scammacca et al., 2015 ; though Scammacca found that older studies of shorter duration demonstrated higher effect sizes).

While there hasn’t been evidence that interventions with smaller groups or longer durations produce larger effects for older struggling readers broadly, there is evidence that these factors may matter for students with the most severe reading difficulties and disabilities, especially in accelerating foundational reading outcomes (e.g., Donegan, Wanzek, & Al Otaiba, 2020 ; Miciak et al., 2018 ; Sanchez & O’Connor, 2015 ). A few studies included in this review that implemented interventions for the longest durations (e.g., 60 hours or more) reported positive effects for students with severe reading difficulties and disabilities ( Miciak et al., 2018 ; Reed et al., 2019 ; Torgesen et al., 2001 ). Our findings related to the intensity moderators may have been different had there been enough studies to examine this population only.

Interventions that were not individualized showed more consistent effects than those that were individualized for both foundational and comprehension outcomes. Interventions described as individualized vary across the literature, both in how the individualization occurs and how much of the instruction is individualized. In order to capture the effects of different types and levels of individualization, we chose to define it broadly. Reading interventions were identified as individualized if any aspect of instruction was initially planned and adjusted throughout the intervention based on student learning. Using this definition, we identified 11 interventions as individualized, the vast majority of which described only specific components of instruction (e.g., word reading instruction, fluency instruction) as individualized and none of which used data-based individualization, one specific method of systematic and iterative intervention adjustment based on student data shown to be effective for improving student reading outcomes ( Jung, Mcmaster, Kunkel, Shin, & Stecker, 2018 ). It should be noted that student participants in four out of the 11 studies with individualized instruction demonstrated foundational reading skills at or below the 16 th percentile ( Miciak et al., 2018 ; O’Connor et al., 2002 ; Therrien et al., 2006 ; Vaughn et al., 2016 ), and two studies included more than 20% of participants with disabilities ( Kim et al., 2010 ; Reed et al., 2019 ). Therefore, it’s possible that these interventions were targeted to students with more significant needs which may account for the smaller effects.

Limitations, Implications and Future Directions for Research

As often is the case with research, many questions remain. First, it appears a trend of diminishing effects is continuing with smaller effects for standardized outcomes found in this review of recent reading intervention studies than in past reports. Improvements in school-provided core reading instruction and intervention and changing reading profiles of participating students offer possible explanations. However, more detailed examinations are needed to invest in either of these hypotheses. Second, there did not seem to be a systematic relationship between intervention duration and effects. Longer interventions may have been targeted to students with more intensive needs which may have impacted effects. One study ( Miciak et al., 2018 ) did offer preliminary evidence that longer interventions may be more effective than shorter ones for students with severe reading difficulties and disabilities. More experimental investigations of the impact of intervention duration for students in the grade range are needed in order for conclusions to be drawn. In addition, we found little evidence that individualized components or interventions offer added value. However, our investigation did not consider if these effects may have differed for students with the most severe reading difficulties and none of our studies used data-based individualization, one particular method of individualization shown to be effective. Finally, we did find some evidence of publication bias for comprehension outcomes. Although it appeared minimal, this may have resulted some inflation of mean effects for comprehension outcomes.

In summary, the results of this review suggest overall positive effects of reading intervention for upper elementary students; smaller, nonsignificant effects for standardized foundational reading outcomes; and small significant effects for standardized comprehension outcomes. According to the results of this review, multicomponent interventions show promise for improving foundational and comprehension reading outcomes. In addition, interventions implemented in very small groups may be an effective way to increase intensity and improve comprehension outcomes.

This research was supported in part by Grant H325H140001 from the Office of Special Education Programs, U.S. Department of Education and by Award Number R01HD091232 from the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the U.S. Department of Education.

Contributor Information

Rachel E. Donegan, Northern Illinois University.

Jeanne Wanzek, Vanderbilt University.

* Denotes studies included in this meta-analytic review

• < Previous

Home > ACADEMIC-UNITS > College of Education > EDUCATION_ETD > 144

Education Dissertations and Projects

A case study of the impact of reading intervention in early elementary school grade levels.

Bonnie S. Smith , Gardner-Webb University

Date of Award

Document type.

Dissertation

Degree Name

Doctor of Education (EdD)

Committee Chair

This case study examined the impact of small class size using small group instruction with ability grouping for early literacy intervention. In this study, data were analyzed to determine the impact small class size, paired with small group instruction for intervention, had on students in terms of attaining grade-level proficiency standards. DIBELS data and facilitator interviews were analyzed from first and second grade intervention and traditional classes to determine the effectiveness of small class size and ability grouping for students to achieve grade-level proficiency in literacy. Results show that using direct instruction in small groups with Title I inclusion helped students grow in their reading abilities by increasing their self-efficacy. The percentage of first-grade students reaching grade-level proficiency was much higher than that of second-grade students. Interventions were successful; however, it appeared that second-grade students were much further behind in attaining grade-level proficiency.

Creative Commons License

Citation Information

Smith, Bonnie S., "A Case Study of the Impact of Reading Intervention in Early Elementary School Grade Levels" (2015). Education Dissertations and Projects . 144. https://digitalcommons.gardner-webb.edu/education_etd/144

Since October 26, 2016

Included in

Curriculum and Instruction Commons , Educational Methods Commons , Elementary Education and Teaching Commons

• Collections
• Disciplines

Advanced Search

• Notify me via email or RSS

Author Corner

• Submit Research
• Copyright and Publishing Info
• College of Education homepage
• Dover Library

Home | About | FAQ | My Account | Accessibility Statement

Privacy Copyright