research on age and stereotype suppression indicates that

ORIGINAL RESEARCH article

The impact of naturalistic age stereotype activation.

• 1 Department of Psychology, University of the Pacific, Stockton, CA, United States
• 2 Department of Psychology, University of Florida, Gainesville, FL, United States

Almost self-fulfilling, commonly held negative stereotypes about old age and memory can impair older adults’ episodic memory performance, due to age-based stereotype threat or self-stereotyping effects. Research studies demonstrating detrimental impacts of age stereotypes on memory performance are generally conducted in research laboratories or medical settings, which often underestimate memory abilities of older adults. To better understand the “real world” impact of negative age and memory stereotypes on episodic memory, the present research tested story recall performance of late middle-aged and older adults ( N = 51) following a naturalistic age stereotype manipulation, wherein every day, newspaper-style materials (comics and puzzles) were either embedded with negative age and memory stereotype stimuli (stereotype group) or neutral stimuli (control group). Furthermore, all participants were tested in favorable, familiar environments. Potential moderators of the stereotype effects, e.g., metamemory beliefs, were assessed at baseline. Current memory evaluation and subjective age, as well as perceived stereotype threat and task-related anxiety, were assessed following the stereotype manipulation as potential mechanisms of the expected stereotype effects. Results suggested a contrast effect, as the stereotype group demonstrated superior story recall performance compared to the control group. Marginally significant moderation effects by age and perceived stereotype threat indicated that stereotype rejection was present for late middle-aged adults but not older adults, indicative of stereotype lift, and for individuals who reported low and average, but not high, levels of perceived stereotype threat. Additionally, a trend suggested more positive memory evaluation for those in the stereotype group who reported awareness of the stereotype stimuli than those who did not notice the stimuli. These results are consistent with other research demonstrating benefits to memory performance in adulthood based on motivational and contextual factors, such as using relevant memory materials and testing in favorable conditions. Moreover, the results of this study contribute to our understanding of individuals’ responses to different types of stereotype stimuli, and the differential impact of stereotype manipulations that are subtle versus blatant. Individuals were motivated to counteract negative stereotype effects when conditions were supportive, stereotype presentations were naturalistic, and personal beliefs were positive.

Introduction

Pervasive negative stereotypes about memory in aging and commonly held expectations of universal, inevitable, and irreversible senility in late adulthood ( Hummert, 2011 ) are partly based on social truths. Age-related deficits in episodic memory are well-documented in both longitudinal and cross-sectional studies ( Park and Festini, 2016 ; Cabeza et al., 2018 ). However, in adulthood, not all aspects of memory are characterized by steep declines, individuals vary greatly in their memory performance and change, and episodic memory performance can be improved via intervention and engaged lifestyles ( Hertzog et al., 2008 ; Gross et al., 2012 ; Strickland-Hughes and West, 2016 ; Nyberg and Pudas, 2019 ). Furthermore, age-related deficits in episodic memory performance rarely generalize to impairments in everyday cognitive functioning ( Salthouse, 2012 ; Barber, 2020 ), and laboratory and clinical settings consistently underestimate the cognitive competence of older adults ( Barber and Lui, 2020 ). Thus, current thinking emphasizes the importance of social, motivational, and other contextual factors–not just ability–in explaining memory performance in aging ( Hess and Emery, 2012 ; Hess, 2014 ; Wu and Strickland-Hughes, 2019 ).

For example, older adults may not be motivated to selectively engage their cognitive resources on abstract, laboratory memory tasks ( Hess, 2014 ). Some supporting research shows older adults might better remember characteristics about people than lists of words ( Sindi et al., 2013 ), information of high, versus low, social importance ( Hargis and Castel, 2017 ) or positive, rather than negative, information ( Mather and Carstensen, 2005 ). Older adults may also underperform on memory tests compared to their true competence level when tested in unfamiliar research laboratory and medical clinic settings, compared to familiar settings, such as community centers ( Hehman and Bugental, 2013 ; Sindi et al., 2013 ; Eich et al., 2014 ; Schlemmer and Desrichard, 2018 ). Ironically, pervasive negative stereotypes about memory and aging themselves might be one factor that disrupts older adults’ memory performance, functioning like a science fiction “causal loop temporal paradox”: Negative age and memory stereotypes, self-held or assumed to be held by others, might worsen memory performance, reinforcing the validity of the stereotypes.

Several theories can explain the impact of age stereotypes on older adults’ memory performance. Stereotype threat theory ( Steele, 1997 ) proposes that concern about performance judgments based on membership in a social group, and related fears of confirming the stereotypes, disrupt performance. This occurs possibly through “hot” cognitive mechanisms, such as increased anxiety, distracting thoughts, or demands on working memory ( Wheeler and Petty, 2001 ; Pennington et al., 2016 ). Stereotype threat effects may be more pronounced when individuals’ highly value the stereotyped domain, perceive that performance assessment will be related to the stereotype, and strongly identify with the stereotyped group. Each of these effects has been demonstrated in memory research ( Chasteen et al., 2011 ). Alternatively, a stereotype threat situation might elicit stereotype reactance , where individuals could experience increased motivational arousal and better (stereotype inconsistent) performance in response to presumed limitations on their freewill associated with being categorized into the stereotyped group ( Miron and Brehm, 2006 ). For example, women may demonstrate “better” negotiation behaviors when threatened explicitly with gender stereotypes ( Kray et al., 2004 ).

Individuals may also experience improved performance following the stereotype threat manipulation if they do not identify as part of the stereotyped group, due to downward social comparison. Stereotype lift refers to performance improvements resulting from negative stereotype activations related to a denigrated outgroup ( Walton and Cohen, 2003 ). Meta-analytic work using the control groups from stereotype threat research indicates that non-stereotyped groups (e.g., Caucasian males) perform better (e.g., on intelligence tests) when the negative outgroup stereotype is emphasized rather than nullified ( d = 0.34; Walton and Cohen, 2003 ). Importantly, several proposed mechanisms for stereotype lift are common to those for stereotype threat, such as anxiety (lower for lift), self-efficacy (higher for lift), and concerns about perceived judgment (fewer for lift).

Although stereotype threat can impact performance for many different social groups, age-based stereotype threat effects are specific to disruptions of older adults’ performance due to “old age” stereotypes ( Barber, 2020 ). Meta-analyses confirm that age-based stereotype threat effects can impair older adults’ performance on cognitive tasks in general ( d = 0.36; Lamont et al., 2015 ) and episodic memory specifically ( d = 0.25; Armstrong et al., 2017 ). Notably, stereotype threat requires identification with the stereotyped group. Yet, age stereotypes are unique from stereotypes about other social groups (e.g., race, gender) given the malleability of their self-relevance and the dynamic, multidimensional nature of defining “old.” For example, middle-aged and older adults generally feel significantly younger than their chronological age ( Rubin and Berntsen, 2006 ) and may further psychologically distance themselves from “old age” when presented with negative aging stereotypes ( Weiss and Freund, 2012 ; Weiss and Lang, 2012 ). On the other hand, individuals report feeling older immediately after memory testing ( Hughes et al., 2013 ). Further, one’s age salience and age identification may change responsively to social and contextual factors, and the transition from “middle-aged” to “old age” is defined culturally and socially (e.g., retirement, grandparenthood), not just chronologically ( Montepare, 2009 ; Brothers et al., 2017 ). Thus, the delineation of “ingroup” versus “outgroup” status for “old age” stereotypes–and potential for stereotype threat versus stereotype lift effects–may be “murky,” especially for late middle-aged adults (e.g., in their 50’s or early 60’s) or young-old adults (e.g., those in their late 60’s or early 70’s). Indeed, stereotype threat effects are generally more pronounced for young-old than old-old adults ( Hess et al., 2004 ; Eich et al., 2014 ). Additionally, although middle-aged adults are not commonly included in age-based stereotype threat research, an exemplar study ( Hess and Hinson, 2006 ) compared the impact of positive and negative age stereotype activations on memory performance of adults aged 24–86 years old. They report a stereotype lift effect for middle-aged adults and stereotype threat effect for older adults, with better and worse memory performance, respectively, following the negative age stereotype presentation compared to a positive age stereotype presentation. The importance of this transitional period is highlighted in the present research with the inclusion of people over 50, that is, late middle-aged adults.

Whereas age-based stereotype threat effects traditionally focus on concerns about stereotyping by others, which assumes awareness of the stereotypes or “threat in the air” ( Steele, 1997 ), self-stereotyping may also occur when individuals apply negative stereotypes about old age to themselves, possibly subconsciously; this can result in stereotype-consistent performance ( O’Brien and Hummert, 2006 ). In fact, a cold cognitive account suggests stereotype priming may also result in stereotype-consistent performance for outgroup members, purely because the presentation of the stereotype makes its content more cognitively accessible and influences behavior automatically ( Wheeler and Petty, 2001 ). Stereotype embodiment theory ( Levy, 2009 ) offers one explanation of self-stereotyping effects in aging, proposing that negative attitudes about old age and aging–learned early in life–become self-relevant later in life, when one identifies as old, and increase in salience in response to social and environmental cues. However, stereotype embodiment is an evolving process as there is no liminal chronological age that defines when one considers themselves old. Even outside of conscious awareness, negative attitudes can be self-fulfilling in terms of hindered performance through reduced expectations and lower self-efficacy. Further, given that we are not born into “old age,” but instead transition into this group, older adults may be ill-equipped to cope with common and sometimes socially acceptable ageist cues–a problem that might be pronounced for young-old adults (who are new to being “old”) and late middle-aged adults (who are anticipating and perhaps fearing being “old”). While positive stereotype primes could boost performance and negative stereotype primes might hinder performance, one meta-analysis suggests that the impact of negative age primes is much more influential than positive age primes ( Meisner, 2012 ). Another meta-analysis confirmed that negative age primes can impair older adults’ memory performance specifically ( d = 0.38; Horton et al., 2008 ). However, instead of internalizing negative old age stereotypes, older adults sometimes respond with age-group dissociation, distancing themselves from the stereotypes (e.g., feeling younger than same-aged peers), perhaps to protect their self-concept ( Weiss and Kornadt, 2018 ).

Different types of stereotype activation effects on older adults’ episodic memory, including age-based stereotype threat and self-stereotyping or priming effects, have been extensively documented using a variety of creative paradigms (for reviews and meta-analyses, see Hess, 2006 ; Horton et al., 2008 ; Chasteen et al., 2011 ; Meisner, 2012 ; Lamont et al., 2015 ; Armstrong et al., 2017 ; Barber and Lui, 2020 ). Common “threat” paradigms include explaining that the purpose of the study was to compare the performance of older adults to that of younger adults and emphasizing the age-sensitive nature of the test. Researchers may even explicitly state that older adults are not expected to do as well because they are old, may include a younger adult confederate in the testing session ( Kang and Chasteen, 2009 ; Popham and Hess, 2013 ; Swift et al., 2013 ; Fernández-Ballesteros et al., 2015 ; Mazerolle et al., 2017 ), or may ask participants to read news articles or watch videos describing age-related deficits in memory ( Hess et al., 2003 ; Hess and Hinson, 2006 ; Thomas and Dubois, 2011 ; Wong and Gallo, 2018 ; Marquet et al., 2019 ). Researchers have also manipulated task instructions to emphasize or de-emphasize the memory component of the testing ( Rahhal et al., 2001 ; Hess et al., 2004 ; Chasteen et al., 2005 ; Desrichard and Köpetz, 2005 ; Sindi et al., 2013 ; Bouazzaoui et al., 2016 ). Common “priming” paradigms include embedding stereotype stimuli into other tasks, such as lexical decision tasks (e.g., identifying words versus pronounceable non-words), sentence scramble tasks, or rapid presentation of this stimuli on a computer screen, just above or below participants’ perceptual thresholds ( Levy, 1996 ; Hess et al., 2004 ; Chasteen et al., 2005 ; Levy and Leifheit-Limson, 2009 ; Eibach et al., 2010 ).

Overall, research using these different paradigms demonstrates poorer performance under negative stereotype conditions than neutral or positive stereotype conditions. Stereotype effects may be more pronounced when the domain of the stereotype matches the performance outcome, such as pairing senile and forgetful with a memory test ( Levy and Leifheit-Limson, 2009 ). Some meta-analytic work suggests that the characteristics of stereotype manipulations matter greatly. For example, Lamont et al., 2015 ) found that age-based stereotype threat manipulations that emphasized opinion-based statements, rather than factual ones (e.g., news articles), were more threatening. Armstrong et al., 2017 ) found that older adults’ episodic memory performance was more sensitive to blatant age-based stereotype manipulations than subtle ones. However, Weiss and Kornadt (2018) argue that blatant age-stereotype activations may be more likely to promote age-dissociation, while the subtle age-stereotype manipulations may promote internalization of the stereotypes. Potentially, stereotype awareness could prompt age group dissociation (although attempts to counteract the stereotype could be cognitively demanding and thus disrupt memory performance further; Pennington et al., 2016 ).

Stereotype effects may also be moderated by pre-existing individual beliefs about memory and age. Evidence suggests memory self-efficacy and one’s evaluation of their memory at baseline may moderate stereotype effects, and that individuals with low memory self-efficacy and higher dementia worry are the ones most susceptible to stereotype effects in clinical settings ( Desrichard and Köpetz, 2005 ; Fresson et al., 2017 ; Schlemmer and Desrichard, 2018 ; but see Chasteen et al., 2005 ). Stereotype effects may also be moderated by memory achievement, or the value and importance a person places on their memory. For example, Hess et al. (2003) found that the inferior memory performance for participants in a negative stereotype condition–compared to positive and neutral stereotype conditions–was exaggerated as level of memory achievement increased, although they did not replicate this effect in a follow-up study ( Hess et al., 2004 ). Weiss (2018) reported poorer memory following a stereotype manipulation, but only for older adults who initially believed that decline with memory was inevitable. Stereotype threat may also be moderated by individuals’ attitudes toward old age and aging satisfaction, wherein the impact of stereotype threat on memory performance is worse for people who hold more negative age attitudes ( Fernández-Ballesteros et al., 2015 ) and older subjective age may relate to greater sensitivity to stereotype threat effects on memory ( Eibach et al., 2010 ). Alternatively, strong group identification with “old” may buffer against age-based stereotype threat effects ( Kang and Chasteen, 2009 ).

Related research suggests that personal beliefs about memory and age are also sensitive to stereotype effect manipulations. For example, Hess and Hinson (2006) found that a stereotype manipulation impacted several metamemory beliefs, such as a sense of personal control over one’s memory performance. In that research, changes in memory beliefs were a more important predictor of memory performance than the stereotype manipulation itself ( Bouazzaoui et al., 2016 ) demonstrated that older adults assigned to a stereotype threat condition (exposed to negative age stereotypes in a questionnaire) reported more subjective memory complaints and lower memory self-efficacy than a no-threat comparison group. Further, they found that the post-manipulation memory complaints and memory self-efficacy fully mediated the effect of the stereotype manipulation on memory performance. Beliefs about age may also be affected by stereotype manipulations, although the mixed results sometimes evidence assimilation effects (self-beliefs aligned with the stereotypes) and other times they represent age-dissociation effects. For example, after reading negative age stereotypes presented in a fake news article (such as in a traditional age-based stereotype threat manipulation), healthy middle-aged and older adults reported feeling older ( Kotter-Grühn and Hess, 2012 ). In contrast, older adults primed with loss-oriented negative age stereotypes in an “Aging Quiz,” compared to those completing the quiz with growth-oriented positive age stereotypes or neutral age stereotypes, felt younger, reported weaker identification with the “old age” group ( Weiss and Lang, 2012 ), and rated themselves as being more dissimilar to pictures of older people ( Weiss and Freund, 2012 ). Notably, many studies have established links between personal memory and age beliefs and subjective and objective cognitive performance ( Beaudoin and Desrichard, 2011 ; Lachman and Agrigoroaci, 2012 ; Stephan et al., 2014 , 2017 ). Thus, these findings underscore the value of assessing personal beliefs following stereotype manipulations.

Some controversy surrounds whether stereotype priming manipulations create stereotype threat, or which specific paradigms represent “true” age-based stereotype threat ( Chasteen et al., 2005 ; Kang and Chasteen, 2009 ), and stereotype threat and stereotype priming effects may occur concurrently ( Wheeler and Petty, 2001 ). A few studies have combined explicit threat-type manipulations with implicit priming-type manipulations (e.g., Hess et al., 2004 ; Bouazzaoui et al., 2016 ). Threat appraisal or perceived stereotype threat is often not assessed in studies of stereotype effects on memory. However, when perceived threat is measured, it generally does not differ between the stereotype groups and comparison groups as expected (but see Swift et al., 2013 ), but it is age-sensitive, negatively related to memory performance, and positively related to anxiety ( Chasteen et al., 2005 ; Kang and Chasteen, 2009 ; Popham and Hess, 2013 ; Marquet et al., 2019 ). Chasteen et al., 2005 ) reported that perceived stereotype threat fully mediated the effect of age on memory performance, and later found that perceived stereotype threat moderated their stereotype manipulation, wherein memory performance was poorer for those in the stereotype group who also reported higher perceived stereotype threat ( Kang and Chasteen, 2009 ). The variance in reported levels of perceived stereotype threat unrelated to experimental manipulations suggests that some older adults may feel threatened from other characteristics of the testing situation, separate from the stereotype manipulations. Often these studies are conducted in the laboratory or clinic with abstract memory tasks (but see Sindi et al., 2013 ). As such, they might not generalize to the “real world,” especially given the social and motivational impacts on memory in aging as described above. Concern regarding the practical everyday impact of stereotype threat is not specific to age-based stereotype threat effects, as it has been voiced as well by scholars of gender-based stereotype threat ( Stoet and Geary, 2012 ; Flore and Wicherts, 2015 ).

The present study aimed to increase our understanding of the potential “real world” impact of negative age and memory stereotypes on middle-aged and older adult’s episodic memory performance. Using a between-groups design, participants completed a story recall memory test after completing tasks with or without negative age and memory stereotype stimuli. The research was designed to be naturalistic and ecologically valid in two primary ways. First, familiar newspaper-like materials were used: Age and memory stereotype stimuli were subtly presented in comics and puzzles, and the memory performance was assessed via story recall, a meaningful, everyday memory task like retelling a news article. Second, participants were tested in familiar settings at preferred times. Testing occurred in participants’ homes and community meeting rooms, rather than a university research laboratory or clinic. Participants were able to schedule testing sessions at the times and days (including weekends) of their choice. Additionally, some participants were recruited from the extended social networks of research assistants (e.g., family friends and neighbors, shared religious and other community groups), which potentially could have encouraged a sense of personal connection to the research. In addition to better representing “real world” experiences, these testing characteristics might also be less stressful for late middle-aged and older adults ( Sindi et al., 2013 ).

The primary purpose of the study was to examine whether this naturalistic design would replicate established stereotype effects ( Horton et al., 2008 ; Lamont et al., 2015 ; Armstrong et al., 2017 ; Barber, 2020 ). Specifically, we expected that episodic memory performance would be poorer for participants exposed to negative old age and memory stereotype stimuli (stereotype group), compared to those exposed to neutral stimuli (control group). We also aimed to explore potential mediators and moderators of the stereotype effects. Given meta-analyses suggesting that subtle stereotype primes may be more disruptive to episodic memory than blatant ones, along with the results of studies that manipulated awareness of age stereotype presentations ( Hess et al., 2004 ; Lamont et al., 2015 ; Armstrong et al., 2017 ), we also expected the stereotype effect to be moderated by awareness of the stereotype stimuli. Specifically, we expected better memory performance for participants reporting awareness of the stereotyped stimuli, compared to participants who were unaware, with aware participants possibly showing stereotype lift effects.

Given the relative lack of stereotype research including explicit comparisons of late middle-aged adults and older adults for memory, it is not clear whether these two groups would react in the same way. Some past research has shown that young-old adults (e.g., in the 60’s) may be most vulnerable to impairments related to age stereotype presentations, whereas early middle-aged adults and older-old adults (e.g., those over 70 years old) may be resilient ( Hess et al., 2004 ; Eich et al., 2014 ). In this study, both groups might self-stereotype and show reductions in performance after exposure to naturalistic negative age and memory stimuli Alternatively, stereotype lift might be evident for the middle-aged group and not the older group, or for the least threatened among both age groups.

To address important issues from past literature, we explored whether reactions to stereotyped stimuli were moderated by pre-existing beliefs about age and memory. We also aimed to determine whether the post-test effects of the “real world” stereotype manipulation would be consistent with age-based stereotype threat theory ( Steele, 1997 ; Shapiro and Neuberg, 2007 ; Lamont et al., 2015 ) and/or with stereotype priming effects as in stereotype embodiment theory ( Horton et al., 2008 ; Levy, 2009 ). If the naturalistic stereotype presentation activated an age-based stereotype threat effect, then we would expect evidence of hot cognitive changes ( Wheeler and Petty, 2001 ), specifically the stereotype group to report greater perceived stereotype threat ( Chasteen et al., 2005 ; Kang and Chasteen, 2009 ; Swift et al., 2013 ) and higher levels of state anxiety ( Osborne, 2001 ; Swift et al., 2013 ; but see Hess et al., 2003 , 2004 , 2009 ; Hess and Hinson, 2006 ) than the control group, and these effects could mediate the impact of the stereotype manipulation on memory performance. If the stereotype manipulation acted more like stereotype priming, then participants might respond with self-stereotyping as in stereotype embodiment, or they might demonstrate age-group dissociation. In the case of self-stereotyping, the stereotype-exposed group might report worse memory evaluations ( Bouazzaoui et al., 2016 ; Chasteen et al., 2005 ; Hess and Hinson, 2006 ; Wong and Gallo, 2018 ) and older subjective ages ( Hess et al., 2003 ; Kang and Chasteen, 2009 ; Eibach et al., 2010 ; Kotter-Grühn and Hess, 2012 ; Fernández-Ballesteros et al., 2015 ; Marquet et al., 2019 ) than the control group, and these effects could also operate as mediators. Alternatively, reports of better memory evaluation and younger subjective ages for the stereotype group than the control group might represent an age group-dissociation.

Materials and Methods

Participants.

A convenience sample of community-dwelling adults were recruited primarily from word-of-mouth (e.g., family friends, community groups), as well as from existing university participant pools and advertisements (e.g., posted to Nextdoor app). Participants were compensated with their choice of $10 or a paper packet of compiled memory exercises and information about memory. Eligible participants were at least 50 years old (range: 50–88 years old; M age = 63.35 years; SD age = 9.92 years), reported hearing adequate to complete a telephone interview, and rated their overall English ability greater than or equal to 6 on a scale from 1 = poor to 10 = excellent . Participants were classified as late middle-aged (50–64 years old) or older (65+ years old).

Research assistants confirmed participant eligibility in a preliminary phone interview. In addition to general demographics and questions about health, participants answered questions about beliefs, specifically personal control beliefs ( Lachman and Weaver, 1998 ) and aging satisfaction ( Lawton, 1975 ). Cognitive tests administered by phone included assessments of working memory (backward digit span; Wechsler, 1997 ) and immediate and delayed recall of a word list ( Lezak, 1995 ) from the Brief Test of Adult Cognition by Telephone (BTACT); these measures are reliable and valid when administered by phone to middle-aged and older adults ( Lachman et al., 2014 ). Participants were free from cognitive impairment. Research assistants administered the Telephone Interview for Cognitive Status (TICS) ( Brandt et al., 1988 ) to the participants who recalled fewer than five of fifteen words on the immediate recall task ( Lezak, 1995 ; Lachman et al., 2014 ) and had difficulty following instructions during the phone interview. Four participants completed the TICS, and all scored above 31, the recommended cutoff score.

This manuscript reports data from 51 participants who completed both the phone interview and the in-person assessment. Sensitivity analyses in G ∗ Power (version 3.1.9.4; Faul et al., 2007 ) indicated that a sample size of 51 participants would have 80% power (1–β error probability) to detect a large effect size of Cohen’s f = 0.40 with α = 0.05 in a two-groups analysis of covariance (ANCOVA) with one covariate (the analysis for our primary research aim). Meta-analyses of stereotype activation effects have reported effect sizes of d = 0.34–0.38 ( Walton and Cohen, 2003 ; Horton et al., 2008 ; Lamont et al., 2015 ). One additional participant was excluded for inability to follow directions during the in-person assessment. Overall, participants were healthy ( M = 8.00, SD = 1.61, rated on a scale from 1 = poor to 10 = excellent ) and well-educated ( M y ea rs = 15.10, SD y ea rs = 2.79). Table 1 reports descriptive data by age group and experimental condition for health, cognition, beliefs, and basic demographic information.

Table 1. Demographics and background information by experimental condition and age group.

To test for baseline differences between groups randomly assigned to the stereotype condition or the control condition, we conducted a series of independent samples t tests for each of the continuous variables reported in Table 1 (e.g., years of education, working memory scores). None of these differences were significant, nor trended toward significance, all p s > 0.100. We also found no significant difference in the proportions of participants in the stereotype condition and in the control condition who were female, χ 2 = 0.04, p = 0.843, or who were retired, χ 2 (1,51) = 1.37, p = 0.242. This pattern of results suggests that random assignment to condition was successful in creating comparable groups.

We also conducted independent samples t tests to compare participants assigned to the two age categories. The results suggested, on average, a greater sense of perceived mastery (global personal control beliefs) for late middle-aged participants than older participants. The mean difference of 0.41, 95% CI [0.07, 0.75], was significant, t (49) = 2.42, p = 0.019, and represented a large effect, d = 0.80, consistent with other research ( Robinson and Lachman, 2017 ). Results also suggested marginally significant effects for hearing and working memory, consistent with typical age-related changes ( Li-Korotky, 2012 ; McCabe and Loaiza, 2012 ) with late middle-aged participants reporting better hearing, M diff = 0.90, 95% CI [−0.09,1.89], t (49) = 1.82, p = 0.073, d = 0.56, and scoring higher on the backward digit span task, M diff = 0.65, 95% CI [−0.13, 1.43], t (49) = 1.69, p = 0.098, d = 0.55. No other baseline differences between late middle-aged and older participants were significant, p s > 0.100.

Participants completed a 15-min preliminary phone interview and a 45- to 60-min in-person assessment. Most participants completed their in-person assessment in the same week as their phone interview. Research assistants followed detailed protocols for administering both interviews, which were audio recorded for quality control. The order of presentation of tests and measures for the phone interview and in-person assessment are summarized in Table 2 along with relevant reliability statistics and citations. All study procedures and materials were approved by the University of Florida Institutional Review Board 02 (#2015-U-0680).

Table 2. Presentation order of telephone interview and in-person assessment tests and measures.

Participants completed the in-person assessments in small groups with no more than four participants. Because the experimental stimuli (described below) were presented in paper-and-pencil packets, small groups often included participants randomly assigned to both conditions (stereotype and control). Importantly, in-person interviews were not conducted in an on-campus psychology laboratory setting, which itself may induce stereotype threat ( Barber and Lui, 2020 ). Rather, the in-person interviews took place in participants’ homes and familiar community meeting areas in Georgia and Florida that were quiet and free from distractions.

For the in-person interview, instructions were presented orally and on paper packets on which participants wrote their responses. Informed consent was followed by general instructions and information about the story recall task. Baseline memory beliefs (as potential moderators of stereotype effects) were then assessed prior to completion of the three naturalistic activities that either presented stereotypic age and memory stimuli or control stimuli: rating comics, completing a word search, and solving a word jumble. Following the stereotype manipulation, participants completed the story recall task and surveys on their state anxiety and perceived stereotype threat (to identify if the manipulation functioned like an age-based stereotype threat manipulation) and on their subjective age and memory evaluation (to identify if the stereotype presentation resulted in assimilation effects consistent with stereotype embodiment theory or in age-group dissociation effects). The interview concluded with a manipulation check and debriefing. Each of these specific measures are described below (section “Measures”).

Naturalistic Stereotype Activation

The independent variable in this research was experimental condition: Participants were randomly assigned to complete naturalistic newspaper-style activities (word search, word jumble, and rating comics) with or without embedded words that primed negative age and memory stereotypes (e.g., forgetful, weak). The comics for the stereotype condition included punchlines related to old age and/or memory, whereas the control comics included punchlines related to other age groups and other cognitive processes. To select the word stimuli for the stereotype and control versions of the word search and word jumble activities, we aggregated words used in past age-based stereotype threat and stereotype priming research ( Bargh et al., 1996 ; Levy, 1996 ; Chasteen et al., 2002 ; Hess et al., 2004 ), as well as designated old-age traits from aging attitudes research ( Schmidt and Boland, 1986 ; Hummert et al., 1994 ; Schwartz and Simmons, 2001 ; Grühn et al., 2011 ). Stereotypical words were selected and categorized as relating primarily to “old age” (e.g., frail , helpless , lonely ) or “memory” (e.g., senile , misplace , impaired ). These words were then pseudo-randomly assigned to lists for the activities (word search, word jumble, and examples for each) without replacement, so that each word could only be included in one item. For all three activities, two alternative versions were used for both the stereotype condition and the control condition, with the versions assigned by counterbalancing, to minimize stimulus-specific effects. The three activities are described in detail below.

Comics rating

For the comics rating activity, participants used an 11-point scale to rate the punchline (0 = not at all funny to 10 = very funny ), the extent to which each comic “shows real life” (0 = very untrue to 10 = very true ), and the tone of each comic (0 = very mean to 10 = very nice ). Each participant had to rate six total comics, including three neutral ones. The neutral comics were negatively valanced but absent of any themes related to age or cognition. For example, one neutral comic was a Robots Read News cartoon by Scott Adams where a robot newscaster says humans are awful “when they are awake.” All participants in both experimental conditions rated the three neutral comics, which permitted examination of potential response biases between the two conditions. For the stereotype condition, participants rated a comic related to old age, a comic related to memory, and a comic related to both old age and memory. For example, one of the old age and memory comics was a Pickles cartoon by Brian Crane that showed an older man misplacing his glasses, which were on his head. For the control condition, participants rated a comic related to age (but not old age), cognition (but not memory), and age and cognition (again, neither related to old age or memory). For example, one of the age and cognition comics was a Zits cartoon by Jerry Scott showing a teenager struggling with attention (his history text “went in one eye and out the other”).

Comics were identified and assigned to the alternative versions of the stereotype condition and control condition tasks based on a multi-stage rating process. First, research assistants curated a selection of 85 comics with themes related to age or cognition from current newspapers and online archives. Research assistants and their friends and family members rated these comics on 11-point scales for negativity (e.g., pleasant-unpleasant), realism (e.g., accurate-inaccurate), relatability (e.g., just like my life-nothing like my life), humor (e.g., not at all funny-very funny), annoyance (e.g., made me feel not at all bothered-extremely bothered), and age- and memory-salience. Based on these ratings, 15 comics were selected for additional review by an independent sample of middle-aged and older adults ( N = 34) using an online survey. The final comics selected for the research were similar in terms of overall negativity, realism, relatability, humor, and annoyance, as rated by that independent sample. Further, comics were matched for size and format across conditions (e.g., number of panes; horizontal arrangement).

Word search

Participants received instructions on how to solve the word search and were given a laminated reference sheet with additional “tips and tricks” (reference sheet included in Supplementary Materials ). To enhance their engagement with the stimulus words, participants were instructed to cross off words, once located, to help track progress. To ensure that the instructions were clear, participants practiced by completing a 6-by-6 letter grid sample word search that included three words to find (printed in a box at the bottom of the page). The main word search was a 10-by-10 letter grid with nine words to find (also printed at the bottom of the page). Participants had 3 min to solve the word search. In the control condition, all words for the sample and main word search were neutral and unrelated to old age or memory. In the stereotype condition, the words in the sample word search ( aged , recall , and watch ) included category age and memory terms to cognitively prime these categories during the subsequent task (such as in Hess et al., 2004 ). The main word search in the stereotype condition included three words related to memory, three words related to old age, and three neutral words, unrelated to age or memory. For the word search puzzles, the final four lists (two stereotype and two control sets of stimuli) were comparable in terms of mean word length (in letters) and the proportions of words of each length (ranging from four to eight letters long). All word search puzzles are included in Supplementary Materials .

Performance on the word search was not of central interest in this study. Rather, we designed the activity to maximize exposure to the stimuli. However, we noted that all participants found all three of the words in the sample, and half of the participants (51%) found all nine of the words in the main word search.

Word jumbles

The final activity in the naturalistic stereotype activation was completion of a word jumble. A word jumble has multiple components, which must be used to create a final word or phrase (often a pun) that fits a cartoon and its accompanying descriptor. To solve any jumble, participants first unscramble the letters for common words. In the solution box for each of those common words, some letters are circled. The circled letters can then be re-arranged to write the required final word or phrase. A list of words including correct answers and distractors were printed at the bottom of the page. This procedure deviated from the standard newspaper format but ensured exposure to the stimuli, even if participants did not solve the jumbles, and made the task easier. As with the word search, participants received instructions and completed an example before completing the main jumble. Participants were again encouraged to use a laminated reference sheet with “tips and tricks” (see Supplementary Materials ).

We created the jumbles by replacing words from real newspaper jumbles with our stimuli. The cartoons and the final correct responses were unchanged. The example was the same for all participants and included two four-letter scrambled words with four possible solutions at the bottom. Participants were given 5 min to complete the main jumble. All four versions of the main jumble (two stereotype condition and two control condition) used five scrambled common words (one 5-letters long, three six-letters long, and one eight-letters long), and showed ten words printed at the bottom of the page. The final two-word phrase required for each cartoon used eight to ten circled letters. For the stereotype condition, three of the five scrambled words and three additional words included as distractors were taken from the old age and/or memory stimuli lists described above. All words in the control condition were neutral, unrelated to old age nor memory.

Again, performance on this task was not important, although we encouraged all participants to solve the jumble. Indeed, 86% of participants correctly unscrambled all five words in the jumble.

Story Recall Test

We assessed episodic memory, our primary outcome measure, with performance on a story recall task. This everyday memory task was selected for its fit with the naturalistic stereotype stimuli, given the task’s similarity to reading and recollecting a newspaper article. Participants were given 1 min to encode the story. Two matched eight-sentence stories were counterbalanced by experimental condition [from Dixon et al. (1989) complication of structurally equivalent texts]. One story was about an older couple camping during the summer, and the other story was about a couple expecting the birth of a grandchild. Four minutes were allotted for recall, when participants were instructed to: Please write down everything that you remember from the story. You need to recall the story as precisely as possible . Story recall was calculated as the percent of words recalled from the story text, following procedures commonly used in our laboratory ( Smith, 2014 ; West et al., 2018 ).

Stereotype Stimuli Awareness

A manipulation check was conducted at the end of the in-person assessment. All participants were asked whether they noticed anything “in particular” about the (a) comics, (b) word search, and (c) word jumble. If they indicated “yes,” they were asked to describe what they noticed. Two independent raters evaluated each response and coded whether participants in the stereotype condition indicated being aware of the old age and/or memory theme(s) in the activities. Disagreements in coding were rare and were resolved through discussion.

Indicators of Age-Based Stereotype Threat and Stereotype Priming Effects

We assessed four measures following the stereotype manipulation and memory testing to assess whether responses reflected age-based stereotype threat (higher perceived stereotype threat and higher state anxiety) and/or stereotype embodiment (lower general memory evaluation and older subjective age). Responses to these measures for participants in the stereotype group may reflect their reaction to the stereotype stimuli and memory testing, whereas responses from the control group reflect reactions to the testing itself.

Perceived stereotype threat

Participants responded to four items about judgments of their personal memory ability from the Perceived Stereotype Threat measure ( Chasteen et al., 2005 ) following the story recall task. An example item is I often feel I have to prove to others that their perceptions of my memory ability are wrong . Items were rated from 1 = strongly disagree to 5 = strongly agree . Responses were averaged to calculate a composite score (α = 0.70). Higher scores indicated a greater overall feeling of age-based stereotype threat for memory.

State anxiety

State anxiety was assessed with an eight-item self-report survey ( Osborne, 2001 ; Abrams et al., 2006 ). Participants used a scale ranging from 1 = not at all to 7 = very much to rate the extent to which they felt under pressure, tense, nervous/jittery, confident, uneasy, calm, afraid of not doing well, and uncomfortable during the story recall task. A state anxiety score was calculated as the mean response to the eight items (α = 0.83) that ranged from one to seven. Responses were reverse coded so that higher scores represented greater level of anxiety.

General memory evaluation

The General Memory Evaluation survey ( West et al., 2003 , 2008 ) was administered following the stereotype manipulation and story recall task. Responses to three items on a seven-point Likert-type scale (e.g., 1 = very unsatisfied to 7 = very satisfied ) were averaged to calculate an index score (α = 0.73; theoretical range: 1–7). The items concern evaluation of one’s recent memory performance, comparison of one’s memory to that of same-age peers, and overall satisfaction with recent memory performance. Higher values indicate higher perceived recent memory ability.

Subjective age identity

Subjective age identity reflects how old people feel ( Montepare, 2009 ). Participants responded to five subjective age identity questions assessing their felt age (e.g., At this moment, how old do you feel? ) by providing ages in years ( Strickland-Hughes et al., 2016 ). Following standard procedures ( Rubin and Berntsen, 2006 ), we averaged the responses to these five questions to compute a subjective age score (α = 0.83). Then, we calculated proportional subjective age identity scores by dividing the difference between participant’s subjective age and chronological age by their chronological age. Proportional subjective age scores are interpretable as the percentage older (positive scores) or younger (negative scores) one feels, relative to their “actual” age.

Baseline Beliefs About Memory and Age

We assessed some personal beliefs about memory aging prior to the stereotype manipulation and the memory testing, consistent with past research (e.g., Hess et al., 2003 , 2004 ). Each of these measures represent general beliefs based on long-term experiences, so we assessed them before the stereotype manipulation as potential moderators.

Metamemory in adulthood

We administered four subscales of the Metamemory in Adulthood questionnaire (MIA; Dixon et al., 1988 ), a reliable and valid survey assessing beliefs about memory. A five-point response scale from agree strongly to disagree strongly was used for all items. Many items were reverse coded. Subscale scores for MIA Achievement (motivation to perform well; α = 0.73), MIA Anxiety (impact of stress on one’s memory; α = 0.80), MIA Capacity (confidence in one’s ability; α = 0.75), and MIA Control (effect of one’s own effort on memory; α = 0.72) were calculated as means of responses to all items in the subscale, with theoretical ranges from 1 to 5. Higher scores suggest higher levels of those beliefs. For additional survey details, see Dixon et al. (1988) .

Aging satisfaction

Aging satisfaction was assessed with the Attitudes toward Aging subscale of the Philadelphia Geriatric Center Morale Scale ( Lawton, 1975 ). We averaged responses to eight statements (α = 0.84) about one’s aging experience (e.g., The older I get , the more useless I feel , and I have as much pep as I did last year ) that participants made on a six-point Likert-type scale from 1 = strongly disagree to 6 = strongly agree . Items were reverse coded so that higher scores represented greater satisfaction with one’s aging experience.

The primary goal of this research was to test whether naturalistic stereotype presentations impacted late middle-aged and older adult’s memory performance. We further aimed to evaluate whether the this expected relationship was moderated or mediated by personal factors (e.g., chronological age, awareness of age stereotypes, beliefs). Additionally, we explored whether the stereotype manipulation impacted perceived stereotype threat, anxiety, memory evaluation, and subjective age, as factors that could provide insight to participants’ responses to the manipulation. Our analytic approach included linear models, e.g., analyses of covariance and multiple linear regressions. For the reported results, all assumptions for the analyses (e.g., no outliers or influential cases, independent errors, homoscedasticity) were met or appropriate techniques or corrections were applied (e.g., weighted least squares regression, bootstrapped confidence intervals). Consistent with other studies ( Swift et al., 2013 ), we included age as a covariate in most analyses in order to test for effects while controlling for the relationship with age, as many of our variables are age-sensitive.

Impact of Naturalistic Stereotype Manipulation

Effect on story recall performance.

To address the primary issue about condition effects, we conducted an ANCOVA to compare mean story recall scores between the stereotype condition and the control condition while accounting for age. We included age as a covariate because chronological age is generally related to episodic memory performance ( Park and Festini, 2016 ; Cabeza et al., 2018 ). Age was a significant predictor of story recall, F (1,47) = 4.17, p = 0.047, r = 0.22. An increase in 1 year of age was associated with a 0.40% decrement in story recall performance. There was also a significant effect of experimental condition on story recall, F (1,47) = 4.43, p = 0.041, partial η 2 = 0.09. Results suggested superior memory performance for the stereotype condition ( EM = 59.45%, se = 2.74%, 95% CI [53.94%, 64.95%]) compared to the control condition ( EM = 51.25%, se = 2.74%, 95% CI [45.74%, 56.76%]). Story recall for participants in the stereotype condition was about 1.16 times that of those in the control condition. This effect is illustrated in Figure 1 .

Figure 1. Story recall performance by condition. Error bars represent 95% confidence intervals. Covariate age (in years) evaluated at 63.12. * Indicates significant mean difference, p < .05.

Moderation by Age Group

For our first exploratory question, we tested whether the effect of the stereotype manipulation on memory performance was moderated by age. First, we conducted a 2 condition (stereotype, control) × 2 age group (late middle-aged, older) independent groups analysis of variance (ANOVA) to test whether the impact of the stereotype manipulation was similarly evidenced in both age groups or moderated by age group. The main effect of condition was not significant, F (1,46) = 2.46, p = 0.125, partial η 2 = 0.05, although it was significant above, when controlling for an age covariate. The main effect of age group was marginally significant, F (1,46) = 3.15, p = 0.082, partial η 2 = 0.082, as late middle-aged participants ( M = 57.96%, SD = 14.94%) trended toward better performance on the story recall task than older participants ( M = 51.08%, SD = 14.33%). The condition × age group interaction effect was not significant, F (1,46) = 0.98, p = 0.329, partial η 2 = 0.02. However, as illustrated in Figure 2 , pairwise comparisons suggested that for late middle-aged participants, performance was significantly better for the stereotype condition ( M = 63.24%, SD = 12.13%) than the control condition ( M = 53.01%, SD = 15.98%), M diff = 10.23, p = 0.044, 95% CI [0.31, 20.16]. In contrast, no statistical difference between the stereotype condition ( M = 52.18%, SD = 11.67%) and the control condition ( M = 49.86%, SD = 11.67%) was observed for older participants ( M diff = 2.33, p = 0.714, 95% CI [−15.02, 10.37]) using this more sensitive statistical test.

Figure 2. Story recall performance by age group and condition. Error bars represent 95% confidence intervals. * Indicates significant mean difference, p < .05.

Awareness of Stereotype Stimuli

Of the 26 participants randomly assigned to the stereotype condition, 13 indicated awareness of the age and/or memory themes in the comics and puzzles (“aware” condition), and 13 did not indicate awareness of the themes (“unaware” condition). Even with this small sample size, the results suggested that age was related to awareness of the stereotype stimuli. A marginally significant independent samples t test, t (24) = 1.93, p = 0.065, suggested that those who were aware of the stereotype stimuli were younger ( M = 61.65, SD = 8.61) than those who were unaware of the stereotype stimuli ( M = 68.55, SD = 9.79). The mean difference of 6.98 years, 95% CI [−0.48, 14.44], represented a large effect, d = 0.72. It is also important to know whether awareness affected outcomes. Results of an ANCOVA suggested no significant difference in story recall between the “aware” group ( M = 61.25%, SD = 14.09%) and “unaware” groups ( M = 56.18%, SD = 11.59%), while accounting for age, F (1,23) = 0.96, p = 0.338, partial η 2 = 0.04.

Moderation by Baseline Memory and Age Beliefs

Similarly, we explored whether beliefs about memory (four subscales of the Metamemory in Adulthood questionnaire) and about age (aging satisfaction) moderated the impact of the naturalistic stereotype manipulation on story recall. We conducted a series of moderation analyses using PROCESS (version 3.5; Hayes, 2017 ). None of these beliefs were significant predictors of memory performance, nor were their interactions with condition, based on all p s > 0.10.

Evidence for Stereotype Threat or Stereotype Embodiment

A series of ANCOVAs compared outcomes administered following the stereotype manipulation and memory testing theoretically related to age-based stereotype threat effects (perceived stereotype threat and state anxiety) and stereotype priming effects (memory evaluation and subjective age) as reported by the stereotype and control groups, and accounting for age. For the two outcomes related to stereotype threat, the direction of the differences in the estimated means suggested lower perceived threat ( M diff = −0.05), lower anxiety ( M diff = −0.29) for the stereotype condition than the control condition. In terms of possible priming effects, the direction of the differences in the estimated means suggested more positive memory evaluation ( M diff = 0.35) and older proportional subjective ages ( M diff = 0.02) for the stereotype condition than the control condition. However, none of these differences were significant, p s > 0.100. We did not conduct further tests of mediation for these factors.

We followed these analyses with ANCOVAs comparing these same key outcomes between the “aware” and “unaware” stereotype groups, again accounting for age. Results suggested a trend toward a more positive memory evaluation in the “aware” condition ( EM = 5.25, se = 0.24, 95% CI [4.75, 5.76]) than the “unaware” condition ( EM = 4.62, se = 0.24, 95% CI [4.11, 5.12]), when controlling for age, F (1,23) = 3.16, p = 0.089, partial η 2 = 0.12. This effect is illustrated in Figure 3 . Age was not a significant predictor of memory evaluation in this model, F (1,23) = 0.12, p = 0.731. The estimated means for the “aware” condition suggested lower perceived stereotype threat ( M diff = −0.26) and higher anxiety ( M diff = 0.36) than the “unaware” condition, but these mean differences were not significant, p s > 0.10.

Figure 3. Evaluation of memory ability by stereotype awareness. Error bars represent 95% confidence intervals. Covariate age (in years) evaluated at 65.05. * Indicates significant mean difference, p < .05.

Moderation by Perceived Stereotype Threat

Because of its theoretical importance, we conducted a moderation analysis using PROCESS (version 3.5; Hayes, 2017 ) to test whether perceived stereotype threat (mean centered) moderated the impact of condition on story recall, with age included as a covariate. The overall model explained about 20% of the variance in story recall performance, F (4,45) = 4.83, p = 0.003, R 2 = 0.20. The main effect of condition was significant ( p = 0.050), although the main effect of perceived threat ( p = 0.934) and the condition × perceived threat interaction effect ( p = 0.130) were not. Given the near marginal significance of the interaction effect, we decomposed the effect to explore the differences in memory performance between the two conditions for varying levels of perceived threat. Story recall performance was higher for the stereotype condition than the control condition when perceived stereotype threat was low, b = −13.12, se = 4.30, 95% CI [−21.78, −4.45], t = −3.05, p = 0.004, or average, b = −7.52, se = 3.74, 95% CI [−15.05, 0.01], t = −2.01, p = 0.050. However, the difference in story recall performance between the two conditions was not significant when perceived stereotype threat was high, b = −1.92, se = 5.99, 95% CI [−13.98, 10.13], t = −0.32, p = 0.749. The Johnson-Neyman method indicated that the upper bound of the zone of significance was −0.001 (56% of the mean-centered perceived threat values were below this boundary). Results of the moderation analysis are summarized in Table 3 and illustrated in Figure 4 .

Table 3. Linear model of condition and perceived threat predicting memory performance.

Figure 4. Story recall performance by condition and perceived stereotype threat. Perceived threat is mean centered. Low is defined as –1 SD (0.89), and High is defined as +1 SD (0.89).

Impaired episodic memory performance related to negative age stereotype priming and age-based stereotype threat are generally, although not exclusively, evidenced in unfamiliar testing environments (e.g., Memory , Aging , and Dementia Lab at Research Intensive Medical University) and with laboratory-type abstract memory tests that might not be important to older adults ( Hess, 2014 ). This research aimed to test whether a naturalistic stereotype manipulation in a supportive testing environment disrupted the memory performance of late middle-aged and older adults, as past research on stereotype threat and stereotype priming has robustly evidenced for older adults ( Horton et al., 2008 ; Lamont et al., 2015 ; Armstrong et al., 2017 ; Barber and Lui, 2020 ). Crucially, late middle-aged is a time of transition, when individuals begin to anticipate moving into the “old” category, with age 65 as a typical cut-off point for recategorization.

Our results provided evidence for rejection of the stereotypes embedded in naturalistic materials. In general, late middle-aged and older adults exposed to the negative age and memory stereotype stimuli performed better on the story recall test than a comparison group exposed to neutral stimuli. Perhaps, when old age stereotypes are presented in favorable conditions (e.g., testing at home at preferred times of day) with familiar, naturalistic tasks, they are “unthreatening,” or the supportive environment and naturalistic materials may promote the requisite motivational arousal and persistent effort to reject behavioral assimilation to the old age categorization ( Miron and Brehm, 2006 ). These testing characteristics may be more supportive for late middle-aged and older participants compared to standard procedures in memory and aging studies ( Barber and Lui, 2020 ). Certainly, the present research did not provide evidence for the stereotype manipulation being “threatening” in ways consistent with an age-based stereotype threat effect. The stereotype group and control group reported comparable levels of perceived stereotype threat and task-related anxiety, suggesting the combination of presenting negative stereotype stimuli with memory testing was not more threatening than memory testing alone. Even though the power analysis suggested that this sample size was appropriate, a larger sample size might have detected such effects. It is important to note that other authors have similar findings. Although some age-based stereotype threat paradigms (e.g., emphasizing that older people are expected to perform worse than younger people) report greater perceived threat and higher anxiety following threat inductions, as compared to neutral or positive conditions ( Swift et al., 2013 ), our finding is consistent with research demonstrating that presentation of age stereotype stimuli in a lexical decision task did not relate to greater perception of age-based stereotype threat ( Chasteen et al., 2005 ) and studies failing to demonstrate a relationship between stereotype manipulations and anxiety ( Hess et al., 2003 , 2004 , 2009 ; Hess and Hinson, 2006 ).

Favorable testing conditions for memory performance may be critical here ( Hehman and Bugental, 2013 ). While they did not present stereotype stimuli, Sindi et al., 2012 ) demonstrated that older adults had better episodic memory performance and less stress (assessed with cortisol levels) when tested in “old favoring” conditions similar to those in our study (e.g., testing during the morning in a familiar environment with socially relevant stimuli and de-emphasizing memory), as opposed to “young favoring” conditions (e.g., testing during the afternoon in an unfamiliar environment with a word list test emphasizing memory). Furthermore, our testing situation may have been “favorable” because most participants experienced “success” completing a cognitive task prior to the memory test. That is, because we wanted to maximize participants’ exposure to the stereotype stimuli in the word search and jumble, we arranged those tasks so that most people would be able to complete both puzzles (e.g., we provided “tips and tricks” sheets and word banks). Inadvertently, the resultant “success” with the puzzles may have promoted memory performance, consistent with findings that older adults show better memory when tests are immediately preceded by successful completion of another cognitive task ( Geraci and Miller, 2013 ; Geraci et al., 2016 ).

Relatedly, our data suggested that stereotype lift was evidenced for participants reporting average or below average levels of perceived stereotype threat, but not for those with above average levels of perceived stereotype threat. This finding suggests that lower perception of perceived stereotype threat (as might be promoted in favorable testing conditions) could “set the stage” for rejection of negative age and memory stereotypes. We recommend that future research directly compare the naturalistic negative stereotype presentation to presentation of control stimuli in favorable and unfavorable testing conditions, such as by modifying the type of memory test and/or prior success in cognitive tasks. That work might also compare the naturalistic stereotype presentation to traditional age-based stereotype threat inductions, separately and in combination. Such detailed follow-up examinations would help “unpack” findings from this novel approach employing more naturalistic exposure to stereotypes.

However, the absence of a threat reaction to the naturalistic stereotype manipulation or potentially minimized “threat in the air” from the favorable testing conditions in this study cannot explain the superior memory performance of those in the stereotype group compared to the control group. A few trends in our data hint at factors that might partially explain or moderate the observed benefits related to the stereotype activation. First, the effect might represent stereotype lift if the participants do not identify “old” as self-relevant ( Walton and Cohen, 2003 ). This effect may be particularly likely for participants whom we classified as “late middle-aged” but may apply to “older adults” who do not identify as old. Although the age group × condition interaction was not significant, follow-up analysis suggested superior memory in the stereotype group compared to the control group for late middle-aged participants, but no difference in performance between the two conditions for older participants. In some research, the young-old have been most vulnerable to age stereotypes with middle-aged and older-old adults showing resilience ( Hess et al., 2004 ; Eich et al., 2014 ). In this case, we speculate that late middle-aged participants may have been motivated to perform better following presentation of negative age and memory stimuli. That is, middle-aged participants’ performance may have been boosted from the positive downward social comparison of themselves to those who are old. In support of this stereotype lift interpretation, and consistent with other research ( Rubin and Berntsen, 2006 ), participants in this study reported subjective ages that were, on average, 21% younger than their actual ages (equivalent to a 65-year-old feeling 51 years old). The rejection of personal relevance of the stereotype stimuli may reflect age-group dissociation reported by others in terms of feeling younger and having lower identification with “old” following stereotype manipulation ( Weiss and Freund, 2012 ; Weiss and Lang, 2012 ). One interesting avenue for future research would be systematic comparison of negative versus neutral age stereotype activation on memory performance for adults of varied chronological ages who self-identify as “old” or “middle-aged” prior to the experiment. Unlike many other personal features that are stereotyped, this would be particularly valuable for understanding age stereotype activation given that the age identification process evolves over time.

The observed age-group dissociation effects might suggest the naturalistic stereotype manipulation functioned as a stereotype priming effect, promoting stereotype assimilation or embodiment. However, we did not observe differences between the stereotype group and control group in reported subjective age and memory evaluation following the stereotype manipulation and memory testing, which we proposed would suggest stereotype embodiment. Instead, we noted that participants who reported awareness of the stereotype stimuli trended toward a more positive evaluation of their memory than those who were unaware of the stimuli, a possible sign of stereotype lift or age-group dissociation. That is, when aware of the stereotype presentations, participants may have rejected their self-relevance, reporting more positive memory evaluations (inconsistent with old age stereotypes) rather than reporting poorer memory evaluations, which would suggest stereotype assimilation. This evidence complements findings of other research emphasizing the difference in impact of subtle or implicit versus blatant or explicit stereotype manipulations ( Horton et al., 2008 ; Lamont et al., 2015 ). For example, the impact of stereotype stimuli presented blatantly (e.g., highlighted in yellow) in a lexical decision task on memory performance was lesser than the impact of the same stimuli presented less obviously (e.g., not highlighted), and age differences in memory performance (between younger and older adults) were exaggerated following subtle, rather than blatant, presentation of negative age stereotypes ( Hess et al., 2004 ). Furthermore, our finding aligns with Weiss and Kornadt (2018) observations that blatant stereotype manipulations may promote age-group dissociation whereas subtle stereotype manipulations promote stereotype internalization.

In sum, our findings suggest that the presentation of negative age and memory stereotypes could bolster the memory performance of late middle-aged adults, and perhaps older adults, specifically when (a) memory testing follows a “success” experience for a different cognitive task, (b) conditions of the testing situation are generally favorable, and (c) participants’ overall perception of stereotype threat is not high. The findings of this research are notable because they evidence better behavioral performance following a negative stereotype condition, rather than a positive stereotype condition, as in Swift et al. (2013) or Meisner (2012) . Collectively, our results did not support the idea that the naturalistic stereotype presentation could induce age-based stereotype threat. However, it is possible for older adults–who had similar performance in the stereotype condition and control condition–that the supportive conditions of our testing situation nullified potential detrimental impact of the stereotype presentation. This notion is partially supported by the marginally significant evidence for moderation of the stereotype effect by perceived stereotype threat. Given our relatively selective sample (e.g., healthy and well-educated) a direct replication of our findings by others is warranted. We also recommend that future research test dispositional and state perceptions of stereotype threat, as in Kang and Chasteen (2009) , to separately evaluate threat felt in response to manipulations and individual differences in overall sensitivity to stereotype threat.

Furthermore, we suggest additional research examine the role of memory beliefs as potential moderators of reactions to stereotype presentations. In this study, participants who reported awareness of the stereotype stimuli might have been motivated to disprove the stereotypes (as suggested by hints of more positive memory evaluation compared to the unaware group), or exert their “free will” as in stereotype reactance theory ( Miron and Brehm, 2006 ). Perhaps memory beliefs could help explain responses consistent with stereotype assimilation versus those aligned with group dissociation. For example, memory beliefs may moderate responses to blatant age stereotypes: If one has high confidence in their memory or believes their memory performance is within their control, they may be more emboldened to “prove wrong” (or prove personally irrelevant) stereotypes they notice about senility in aging. Yet, these beliefs might not be able to protect individuals from subtly presented stereotypes that they do not consciously notice. Certainly, the results here do not suggest that negative age and memory stereotypes should be promoted to benefit late middle-aged people. Such benefits might be short-lived or cognitively taxing, and adoption of negative attitudes about aging earlier in life may be related to a host of negative aging outcomes later in life, when the stereotypes eventually become self-relevant, and result in more problems later ( Levy, 2009 ; Weiss and Kornadt, 2018 ).

Instead, an interesting series of follow-up studies might emphasize trying to train awareness and responses to the stereotype primes, while also promoting more positive attitudes toward aging and better memory beliefs. The latter component may be especially important given that stereotype priming can promote stereotype-consistent behavior even in outgroups, especially if they personally associate the stereotyped group and the characteristics ( Wheeler and Petty, 2001 ). To better tease out how age is related to type of stereotyped stimuli and testing environments, more stereotype activation work should aim to test for the real world, practical impact of age and memory stereotypes on performance “in the wild.” Such examinations might explicitly compare such results to those observed in environments with varying degrees of possible threat (i.e., number of opportunities for success before memory testing, emphasis on knowledge or expertise, degree of familiarity of stimuli in which stereotypes are embedded, ecological validity of specific memory tests). Systematic follow-up studies, using a wide range of age groups, would further help researchers to identify the specific mechanisms controlling stereotype lift effects.

Data Availability Statement

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

Ethics Statement

The studies involving human participants were reviewed and approved by University of Florida Institutional Review Board 02 (approval code #2015-U-0680). The patients/participants provided their written informed consent to participate in this study.

Author Contributions

CS-H took the lead in analysis and interpretation of data and drafting of the manuscript in consultation with RW. Both authors conceived and planned the study, supervised the data collection, provided critical feedback, and helped to shape the research, analysis and manuscript.

This work was supported by the Department of Psychology at University of Florida, including a Jacquelin Goldman Foundation research grant awarded to CS-H.

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.

Acknowledgments

We thank all our research assistants in the Aging and Cognitive Training Laboratory at the University of the Pacific and the Memory and Aging Laboratory at the University of Florida for their assistance with participant scheduling, data collection, and data entry, including Christopher Andrews, Mercedes Ball, Charlie Bisbee, Karina Chang, Sarah Charles, Robyn Cotney, Albert Dizon, Priscilla Hu, Carmen Huang, Mary Johnson, Sonia Koul, Andrew Leyva, Samantha Loscalzo, Mohima Meera, Brianna Parlette, Rezwana Parveen, Kimberly Santo, Matthew Tineo, Nguyen (Nancy) Vo, Michael Yeber, and Guiying (Angel) Zhong. Special thanks are owed to Vineet Polineni for his lab supervision of research assistants and the recruitment and phone interview procedures.

Supplementary Material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg.2021.685448/full#supplementary-material

Supplementary Data Sheet 1 | Word search and word jumble puzzles used for the stereotype activation.

Supplementary Data Sheet 2 | Reference sheet provided to aid completion of the word search and word jumble puzzles.

Abrams, D., Eller, A., and Bryant, J. (2006). An age apart: The effects of intergenerational contact and stereotype threat on performance and intergroup bias. Psychol. Aging 21, 691–702. doi: 10.1037/0882-7974.21.4.691

PubMed Abstract | CrossRef Full Text | Google Scholar

Armstrong, B., Gallant, S. N., Li, L., Patel, K., and Wong, B. I. (2017). Stereotype threat effects on older adults’ episodic and working memory: A meta-analysis. Gerontologist 57, S193–S205. doi: 10.1093/geront/gnx056

Barber, S. J. (2020). The applied implications of age-based stereotype threat for older adults. J. Appl. Res. Mem. Cogn. 9, 274–285. doi: 10.1016/j.jarmac.2020.05.002

CrossRef Full Text | Google Scholar

Barber, S. J., and Lui, K. (2020). “Stereotype threat and the cognitive performance of older adults,” in Handbook of Cognitive Aging: A Life Course Perspective , eds A. Thomas and A. Gutchess (Cambridge: Cambridge University Press).

Google Scholar

Bargh, J. A., Chen, M., and Burrows, L. (1996). Automaticity of social behavior: direct effects of trait construct and stereotype activation on action. J. Personal. Soc. Psychol. 71, 230–244. doi: 10.1037/0022-3514.71.2.230

Beaudoin, M., and Desrichard, O. (2011). Are memory self-efficacy and memory performance related? A meta-analysis. Psychol. Bull. 137, 211–241. doi: 10.1037/a0022106

Bouazzaoui, B., Follenfant, A., Ric, F., Fay, S., Croizet, J. C., Atzeni, T., et al. (2016). Ageing-related stereotypes in memory: When the beliefs come true. Memory 24, 659–668. doi: 10.1080/09658211.2015.1040802

Brandt, J., Spencer, M., and Folstein, M. (1988). The telephone interview for cognitive status. Neuropsychiatry Neuropsychol. Behav. Neurol. 1, 111–117.

Brothers, A., Miche, M., Wahl, H.-W., and Diehl, M. (2017). Examination of associations among three distinct subjective aging constructs and their relevance for predicting developmental correlates. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 72, 547–560. doi: 10.1093/geronb/gbv085

Cabeza, R., Albr, A., Belleville, S., Craik, F. I. M., Duarte, A., Grady, C., et al. (2018). Maintenance, reserve, and compensation: the cognitive neuroscience of healthy ageing. Nat. Neurosci. Rev. 19, 701–710. doi: 10.1038/s41583-018-0068-2

Chasteen, A. L., Bhattacharyya, S., Horhota, M., Tam, R., and Hasher, L. (2005). How feelings of stereotype threat influence older adults’ memory performance. Exper. Aging Res. 31, 235–260. doi: 10.1080/03610730590948177

Chasteen, A. L., Kang, S. K., and Remedios, J. D. (2011). “Aging and stereotype threat,” in Stereotype threat: Theory, process, and application , 1st Edn, eds M. Inzlicht and T. Schmader (New York: Oxford University Press), 202–216.

Chasteen, A. L., Schwarz, N., and Park, D. C. (2002). The activation of aging stereotypes in younger and older adults. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 57, 540–547. doi: 10.1093/geronb/57.6.p540

Desrichard, O., and Köpetz, C. (2005). A threat in the elder: the impact of task-instructions, self-efficacy and performance expectations on memory performance in the elderly. Eur. J. Soc. Psychol. 35, 537–552. doi: 10.1002/ejsp.249

Dixon, R. A., Hultsch, D. F., and Hertzog, C. (1988). The Metamemory in Adulthood (MIA) questionnaire. Psychopharmacol. Bull. 24, 671–688. doi: 10.1037/t02356-000

Dixon, R. A., Hultsch, D. F., and Hertzog, C. (1989). A manual of twenty-five three-tiered structurally equivalent texts for use in aging research. Victoria, B.C: Collaborative Research Group on Cognitive Aging.

Eibach, R. P., Mock, S. E., and Courtney, E. A. (2010). Having a “senior moment”: induced aging phenomenology, subjective age, and susceptibility to ageist stereotypes. J. Exper. Soc. Psychol. 46, 643–649. doi: 10.1016/j.jesp.2010.03.002

Eich, T. S., Murayama, K., Castel, A. D., and Knowlton, B. J. (2014). The dynamic effects of age-related stereotype threat on explicit and implicit memory performance in older adults. Soc. Cogn. 32, 559–570. doi: 10.1521/soco.2014.32.6.559

Faul, F., Erdfelder, E., Lang, A.-G., and Buchner, A. (2007). G*Power 3: a flexible statistical power analysis program for the soical, behavioral, and biomedical sciences. Behav. Res. Methods 39, 175–191. doi: 10.3758/bf03193146

Fernández-Ballesteros, R., Bustillos, A., and Huici, C. (2015). Positive perception of aging and performance in a memory task: compensating for stereotype threat? Exper. Aging Res. 41, 410–425. doi: 10.1080/0361073x.2015.1053757

Flore, P. C., and Wicherts, J. M. (2015). Does stereotype threat influence performance of girls in stereotyped domains? A meta-analysis. J. School Psychol. 53, 25–44. doi: 10.3389/fpsyg.2020.604978

Fresson, M., Dardenne, B., Geurten, M., and Meulemans, T. (2017). The effect of stereotype threat on older people’s clinical cognitive outcome: Investigating the moderating role of dementia worry. Clin. Neuropsychol. 31, 1306–1328. doi: 10.1080/13854046.2017.1307456

Geraci, L., Hughes, M. L., Miller, T. M., and De Forrest, R. L. (2016). The effect of prior task success on older adults’ memory performance: examining the influence of different types of task success. Exper. Aging Res. 42, 365–381. doi: 10.1080/0361073X.2016.1191860

Geraci, L., and Miller, T. M. (2013). Improving older adults’ memory performance using prior task success. Psychol. Aging 28, 340–345. doi: 10.1037/a0030332

Gross, A. L., Parisi, J. M., Spira, A. P., Kueider, A. M., Ko, J. Y., Saczynski, J. S., et al. (2012). Memory training interventions for older adults: A meta-analysis. Aging Mental Health 16, 722–734. doi: 10.1080/13607863.2012.667783

Grühn, D., Gilet, A.-L., Studer, J., and Labouvie-Vief, G. (2011). Age-relevance of person characteristics: Persons’ beliefs about developmental change across the lifespan. Dev. Psychol. 47, 376–387. doi: 10.1037/a0021315

Hargis, M. B., and Castel, A. D. (2017). Younger and older adults’ associative memory for social information: The role of information importance. Psychol. Aging 32, 325–330. doi: 10.1037/pag0000171

Hayes, A. F. (2017). Introduction to mediation, moderation, and conditional process analysis , 2nd Edn. New York City: Guilford Publications.

Hehman, J. A., and Bugental, D. B. (2013). “Life stage-specific” variations in performance in response to age stereotypes. Dev. Psychol. 49, 1396–1406. doi: 10.1037/a0029559

Hertzog, C., Kramer, A. F., Wilson, R. S., and Lindenberger, U. (2008). Enrichment effects on adult cognitive development: can the functional capacity of older adults be preserved and enhanced? Psychol. Sci. Public Interest 9, 1–65. doi: 10.1111/j.1539-6053.2009.01034.x

Hess, T. M. (2006). “Attitudes toward aging and their effects on behavior,” in Handbook of the psychology of aging , 6th Edn, eds J. E. Birren and K. W. Schaie (San Diego: Academic Press), 379–406.

Hess, T. M. (2014). Selective engagement of cognitive resources: Motivational influences on older adults’ cognitive functioning. Perspect. Psychol. Sci. 9, 388–407. doi: 10.1177/1745691614527465

Hess, T. M., Auman, C., Colcombe, S. J., and Rahhal, T. A. (2003). The impact of stereotype threat on age differences in memory performance. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 58B, 3–11. doi: 10.1093/geronb/58.1.p3

Hess, T. M., and Emery, L. J. (2012). “Memory in context: The impact of age-related goals on performance,” in Memory and Aging: Current Issues and Future Directions , eds M. Naveh-Benjamin and N. Ohta (New York: Psychology Press, Taylor & Francis), 183–214.

Hess, T. M., and Hinson, J. (2006). Age-related variation in the influences of aging stereotypes on memory in adulthood. Psychol. Aging 21, 621–625. doi: 10.1037/0882-7974.21.3.621

Hess, T. M., Hinson, J. T., and Hodges, E. A. (2009). Moderators of and mechanisms underlying stereotype threat effects on older adults’ memory performance. Exper. Aging Res. 35, 153–177. doi: 10.1080/03610730802716413

Hess, T. M., Hinson, J. T., and Statham, J. A. (2004). Explicit and implicit stereotypes activation effects on memory: Do age and awareness moderate the impact of priming? Psychol. Aging 19, 495–505. doi: 10.1037/0882-7974.19.3.495

Horton, S., Baker, J., Pearce, G. W., and Deakin, J. M. (2008). On the malleability of performance: Implications for seniors. J. Appl. Gerontol. 27, 446–465. doi: 10.1177/0733464808315291

Hughes, M. L., Geraci, L., and De Forrest, R. L. (2013). Aging 5 years in 5 minutes: The effect of taking a memory test on older adults’ subjective age. Psychol. Sci. 24, 2481–2488. doi: 10.1177/0956797613494853

Hummert, M. L. (2011). Age stereotypes and aging. In K. W. Schaie & S. L. Willis, Handbook of the psychology of aging , 7th Edn. London: Elsevier, 249–262.

Hummert, M. L., Garstka, T. A., Shaner, J. L., and Strahm, S. (1994). Stereotypes of the elderly held by young, middle-aged, and elderly adults. J. Gerontol. 49, 240–249. doi: 10.1093/geronj/49.5.P240

Kang, S. K., and Chasteen, A. L. (2009). The moderating role of age-group identification and perceived threat on stereotype threat among older adults. Int. J. Aging Hum. Dev. 69, 201–220. doi: 10.2190/ag.69.3.c

Kotter-Grühn, D., and Hess, T. M. (2012). The impact of age stereotypes on self-perceptions of aging across the adult lifespan. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 67, 563–571. doi: 10.1093/geronb/gbr153

Kray, L. J., Reb, J., Galinsky, A. D., and Thompson, L. (2004). Stereotype reactance at the bargaining table: The effect of stereotype activation and power on claiming and creating value. Personal. Soc. Psychol. Bull. 30, 399–411. doi: 10.1177/0146167203261884

Lachman, M. E., and Agrigoroaci, S. (2012). Low perceived control as a risk factor for episodic memory: The mediational role of anxiety and task interference. Mem. Cogn. 40, 287–296. doi: 10.3758/s13421-011-0140-x doi:

Lachman, M. E., Agrigoroaei, S., Tun, P. A., and Weaver, S. L. (2014). Monitoring cognitive functioning: Psychometric properties of the brief test of adult cognition by telephone. Assessment 21, 404–417. doi: 10.1177/1073191113508807

Lachman, M. E., and Weaver, S. L. (1998). Sociodemographic variations in the sense of control by domain: Findings from the MacArthur studies of midlife. Psychol. Aging 13, 553–563. doi: 10.1037/0882-7974.13.4.553

Lamont, R. A., Swift, H. J., and Abrams, D. (2015). A review and meta-analysis of age-based stereotype threat: Negative stereotypes, not facts, do the damage. Psychol. Aging 30, 180–193. doi: 10.1037/a0038586

Lawton, M. P. (1975). The philadelphia geriatric center morale scale: A revision. Gerontology 30, 85–89. doi: 10.1093/geronj/30.1.85

Levy, B. (2009). Stereotype embodiment. Curr. Direc. Psychol. Sci. 18, 332–336. doi: 10.1111/j.1467-8721.2009.01662.x

Levy, B. R. (1996). Improving memory in old age through implicit self-stereotyping. J. Personal. Soc. Psychol. 71, 1092–1107. doi: 10.1037/0022-3514.71.6.1092

Levy, B. R., and Leifheit-Limson, E. (2009). The stereotype-matching effect: Greater influence on functioning when age stereotypes correspond to outcomes. Psychol. Aging 24, 230–233. doi: 10.1037/a0014563

Lezak, M. (1995). Neuropsychological assessment , 3rd Edn. New York: Oxford University Press.

Li-Korotky, H.-S. (2012). Age-related hearing loss: quality of care for quality of life. Gerontologist 52, 265–271. doi: 10.1093/geron6t/gnr159

Marquet, M., Missotten, P., Dardenne, B., and Adam, S. (2019). Interactions between stereotype threat, subjective aging, and memory in older adults. Aging Neuropsychol. Cogn. 26, 121–143. doi: 10.1080/13825585.2017.1413166

Mather, M., and Carstensen, L. L. (2005). Aging and motivated cognition: the positivity effect in attention and memory. Trends Cogn. Sci. 9, 496–502. doi: 10.1016/j.tics.2005.08.005

Mazerolle, M., Régner, I., Barber, S. J., Paccalin, M., Miazola, A.-C., Huguet, P., et al. (2017). Negative aging stereotypes impair performance on brief cognitive tests used to screen for predementia. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 72, 932–936. doi: 10.1093/geronb/gbw083

McCabe, D. P., and Loaiza, V. M. (2012). “Working memory,” in The Wiley-Blackwell handbook of adulthood and aging , eds S. K. Whitebourne and M. J. Sliwinski (Oxford, UK: Wiley-Blackwell), 154–173.

Meisner, B. A. (2012). A meta-analysis of positive and negative age stereotype priming effects on behavior among older adults. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 67B, 13–17. doi: 10.1093/geronb/gbr062

Miron, A. M., and Brehm, J. W. (2006). Reactance theory – 40 years later. Zeitschrift Für Sozialpsychol. 37, 9–18. doi: 10.1024/0044-3514.37.1.3

Montepare, J. M. (2009). Subjective age: Toward a guiding lifespan framework. Int. J. Behav. Dev. 33, 42–46. doi: 10.1177/0165025408095551

Nyberg, L., and Pudas, S. (2019). Successful memory aging. Annu. Rev. Psychol. 70, 219–243. doi: 10.1146/annurev-psych-010418-103052

O’Brien, L. T., and Hummert, M. L. (2006). Memory performance of late middle-aged adults: Contrasting self-stereotyping and sterotype threat accounts of assimilation to age stereotypes. Soc. Cogn. 24, 338–358. doi: 10.1521/soco.2006.24.3.338

Osborne, J. W. (2001). Testing stereotype threat: does anxiety explain race and sex differences in achievement? Contempor. Educ. Psychol. 26, 291–310. doi: 10.1006/ceps.200.1052

Park, D. C., and Festini, S. B. (2016). Theories of memory and aging: A look at the past and a glimpse of the future. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 72, 82–90. doi: 10.1093/geronb/gbw066

Pennington, C. R., Heim, D., Levy, A. R., and Larkin, D. T. (2016). Twenty years of stereotype threat research: A review of psychological mediators. PloS one 11:e0146487. doi: 10.1371/journal.pone.0146487

Popham, L. E., and Hess, T. M. (2013). Age differences in the underlying mechanisms of stereotype threat effects. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 70, 223–232. doi: 10.1093/geronb/gbt093

Rahhal, T. A., Hasher, L., and Colcombe, S. J. (2001). Instructional manipulations and age differences in memory: now you see them, now you don’t. Psychol. Aging 16, 697–706. doi: 10.1037/0882-7974.16.4.697

Robinson, S. A., and Lachman, M. E. (2017). Perceived control and aging: A mini-review and directions for future research. Gerontology 63, 435–442. doi: 10.1159/000468540

Rubin, D. C., and Berntsen, D. (2006). People over forty feel 20% younger than their age: subjective age across the lifespan. Psychon. Bull. Rev. 13, 776–780. doi: 10.3758/BF03193996

Salthouse, T. (2012). Consequences of age-related cognitive declines. Annu. Rev. Psychol. 63, 201–226. doi: 10.1146/annurev-psych-120710-100328

Schlemmer, M., and Desrichard, O. (2018). Is medical environment detrimental to memory? A test of a white coat effect on older people’s memory performance. Clin. Gerontol. 41, 77–81. doi: 10.1080/07317115.2017.1307891

Schmidt, D. F., and Boland, S. M. (1986). Structure of perceptions of older adults: evidence for multiple stereotypes. Psychol. Aging 1, 255–260. doi: 10.1037/0882-7974.1.3.255

Schwartz, K., and Simmons, J. P. (2001). Contact quality and attitudes toward the elderly. Educ. Gerontol. 27, 127–137. doi: 10.1080/03601270151075525

Shapiro, J. R., and Neuberg, S. L. (2007). From stereotype threat to stereotype threats: Implications of a multi-threat framework for causes, moderators, mediators, consequences, and interventions. Personal. Soc. Psychol. Rev. 11, 107–130. doi: 10.1177/1088868306294790

Sindi, S., Fiocco, A. J., Juster, R. P., Ruessner, J., and Lupien, S. J. (2013). When we test, do we stress? Impact of the testing environment on cortisol secretion and memory performance in older adults. Psychoneuroendocrinology 38, 1388–1396. doi: 10.1016/j.psyneuen.2012.12.004

Sindi, S., Fiocco, A. J., Juster, R.-P., Pruessner, J., and Lupien, S. J. (2012). When we test, do we stress? Impact of the testing environment on cortisol secretion and memory performance in older adults. Psychoneuroendocrinology 38, 1388–1396. doi: 10.106/j.psyneuen.2012.12.004

Smith, K. A. (2014). Changing Expectancies: Self-Efficacy, Outcome Expectancy, and Memory Performance in Older Adults [Unpublished master’s thesis]. Gainesville, FL: University of Florida.

Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectual identity and performance. Am. Psychol. 52, 613–629. doi: 10.1037/0003-066x.52.6.613

Stephan, Y., Caudroit, J., Jaconelli, A., and Terracciano, A. (2014). Subjective age and cognitive functioning: A 10-year prospective study. Am. J. Geriatr. Psychiatry 22, 1180–1187. doi: 10.1016/j.jagp.2013.03.007

Stephan, Y., Sutin, A. R., Luchetti, M., and Terracciano, A. (2017). Subjective age and changes in memory in older adults. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 71, 675–683. doi: 10.1093/geronb/gbv010

Stoet, G., and Geary, D. C. (2012). Can stereotype threat explain the gender gap in mathematics performance and achievement? Rev. General Psychol. 16, 93–102. doi: 10.1037/a0026617

Strickland-Hughes, C. M., and West, R. L. (2016). “Memory: Training methods and benefits,” in Encyclopedia of Geropsychology , ed. N. Pachana (Singapore: Springer).

Strickland-Hughes, C. M., West, R. L., Smith, K. A., and Ebner, N. C. (2016). False feedback and beliefs influence name recall in younger and older adults. Memory 25, 1072–1088. doi: 10.1080/09658211.2016.1260746

Swift, H. J., Abrams, D., and Marques, S. (2013). Threat or boost? Social comparison affects older people’s performance differently depending on task domain. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 68, 23–30. doi: 10.1093/geronb/gbs044

Thomas, A. K., and Dubois, S. J. (2011). Reducing the burden of stereotype threat eliminates age differences in memory distortion. Psychol. Sci. 22, 1515–1517. doi: 10.1177/0956797611425932

Walton, G. M., and Cohen, G. L. (2003). Stereotype lift. J. Exper. Soc. Psychol. 39, 456–467. doi: 10.1016/S0022-1031(03)00019-2

Wechsler, D. (1997). Wechsler memory scale-third edition: Administration and scoring manual. San Antonio, TX: The Psychological Corporation.

Weiss, D. (2018). On the inevitability of aging: essentialist beliefs moderate the impact of negative age stereotypes on older adults’ memory performance and physiological reactivity. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 73, 925–933. doi: 10.1093/geronb/gbw087

Weiss, D., and Freund, A. M. (2012). Still young at heart: Negative age-related information motivates distancing from same-aged people. Psychol. Aging 27, 173–180. doi: 10.1037/a0024819

Weiss, D., and Kornadt, A. E. (2018). Age-stereotype internalization and dissociation: Contradictory processes or two sides of the same coin? Curr. Direc. Psychol. Sci. 27, 477–483. doi: 10.1177/0963721418777743

Weiss, D., and Lang, F. (2012). “They” are old but “I” feel younger: Age-group dissociation as a self-protective strategy in old age. Psychol. Aging 27, 153–163. doi: 10.1037/a0024887

West, R. L., Bagwell, D., and Dark-Freudeman, A. (2008). Self-efficacy and memory aging: The impact of a memory intervention based on self-efficacy. Aging Neuropsychol. Cogn. 15, 302–329. doi: 10.1080/13825580701440510

West, R. L., Strickland-Hughes, C. M., and Smith, K. A. (2018). Age differences in self-set goal effects for memory. Aging Neuropsychol. Cogn. 25, 484–499. doi: 10.1080/13825585.2017.1327013

West, R. L., Thorn, R. M., and Bagwell, D. (2003). Memory performance and beliefs as a function of goal setting and aging. Psychol. Aging 18, 111–125. doi: 10.1037/0882-7974.18.1.111

Wheeler, S. C., and Petty, R. E. (2001). The effects of stereotype activation on behavior: A review of possible mechanisms. Psychol. Bull. 127, 797–826. doi: 10.1037//0033-2909.127.6.797

Wong, J. T., and Gallo, D. A. (2018). Activating aging stereotypes increases source recollection confusions in older adults: Effect at encoding but not retrieval. J. Gerontol. B Psychol. Sci. Soc. Sci. 74, 633–641. doi: 10.1093/geronb/gbx103

Wu, R., and Strickland-Hughes, C. M. (2019). Adaptation for growth as a common goal throughout the lifespan: Why and how. Psychol. Learn. Motiv. 71, 387–414. doi: 10.1016/bs.plm.2019.07.005

Keywords : age-based stereotype threat, story recall, memory, perceived threat, subjective age, memory beliefs, metamemory, stereotype priming

Citation: Strickland-Hughes CM and West RL (2021) The Impact of Naturalistic Age Stereotype Activation. Front. Psychol. 12:685448. doi: 10.3389/fpsyg.2021.685448

Received: 25 March 2021; Accepted: 14 June 2021; Published: 09 July 2021.

Reviewed by:

Copyright © 2021 Strickland-Hughes and West. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Carla M. Strickland-Hughes, [email protected]

Disclaimer: 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.

• Search Menu
• Advance articles
• Author Guidelines
• Submission Site
• Open Access
• Call for Papers
• About Work, Aging and Retirement
• Editorial Board
• Advertising and Corporate Services
• Journals Career Network
• Self-Archiving Policy
• Dispatch Dates
• Journals on Oxford Academic
• Books on Oxford Academic

Article Contents

• < Previous

Age Stereotypes Do Matter: Looking Through the Lens of the Attraction–Selection–Attrition Model

• Article contents
• Figures & tables
• Supplementary Data

David M Cadiz, Grant M Brady, Lale M Yaldiz, Sara Zaniboni, Donald M Truxillo, Age Stereotypes Do Matter: Looking Through the Lens of the Attraction–Selection–Attrition Model, Work, Aging and Retirement , Volume 8, Issue 4, October 2022, Pages 339–342, https://doi.org/10.1093/workar/waac009

• Permissions Icon Permissions

Murphy and DeNisi (2021) suggest that the real-world effects of age stereotypes on personnel decisions are weak, null, or inconsistent. However, we know that both conscious and unconscious age stereotypes exist ( Fiske, 2017 ; Posthuma & Campion, 2009 ), and both seem to affect people’s hiring decisions ( Zaniboni et al., 2019 ). For instance, a field experiment in Sweden in which 6,000 fictitious resumes were sent to open positions found that applicants over 40 received fewer callbacks and that callbacks decreased with applicant age ( Carlsson & Eriksson, 2019 ). Moreover, the implementation of age discrimination laws worldwide suggests that age discrimination is generally recognized as an issue for workers and job applicants.

In this commentary, we draw on the attraction–selection–attrition (ASA) model to illustrate when age stereotypes do affect personnel decisions throughout the employee lifecycle, acting “under the radar” of current organizational research. Through this lens, we suggest new avenues and approaches to understand both the mechanisms by which stereotypes work and their true effects. In doing so, we provide theory-based recommendations for researchers to better uncover when and how age stereotypes do influence personnel decisions (see Table 1 ).

Examples of research recommendations for age stereotyping research based on the ASA model.

The ASA model incorporates three phases that explain how organizational cultures are formed and maintained over time ( Schneider et al., 1995 ) and how these cultures influence employee and prospective employee behavior. Attraction includes the attraction of people to an organization whose members and values seem similar to their own personal characteristics and values. The selection phase indicates that cultures are sustained and reinforced by who is (and is not) selected into the organization. Finally, attrition refers to how employees that do not “fit” are more likely to leave an organization, either voluntarily or involuntarily. Ultimately, the ASA cycle is argued to lead to a strong organizational culture built on, and reinforcing, a relatively homogeneous workforce. The resulting organizational culture, reflected in observable artifacts, values, and assumptions ( Schein, 1990 ), can influence the stereotypes of which characteristics (“who”) best fit the organization: This affects who is attracted during recruitment, who is selected, and who stays or leaves because of fit. We illustrate how age stereotypes can influence personnel decisions at each ASA stage and where more research is needed.

Attraction incorporates the recruitment process and is most relevant to variables such as which employee characteristics the organization signals as desirable; the applicant’s perceived fit with the job, organization, or industry; and applicant decisions to apply or accept a job offer. Age stereotypes may be at work during attraction, but the current research may not capture these effects. Specifically, extant research does not include people who choose not to apply for a job because of signaling from the organization (e.g., Lievens & Highhouse, 2003 ; Lievens & Slaughter, 2016 ). Therefore, how the job is presented by the organization can—both intentionally and unintentionally, consciously and unconsciously—be affected by age stereotypes. For instance, recruitment materials and job ads may include terms such as “digital native” or exclude images of older workers. Such materials (i.e., observable artifacts) reflect explicit or implicit stereotypes of recruiters embedded within the organizational culture, and they are likely to deter older applicants. Moreover, organizations often engage in conscious attempts to exclude workers based on age. For instance, until 2019, Facebook’s (now Meta’s) job ad platform allowed targeted job ads by applicant age group, and many large employers (e.g., Amazon, Goldman Sachs, Target, and Verizon to name a few) took advantage of this function. For instance, Verizon limited some job ads only to those between ages 25 and 36, and Amazon limited their warehouse ads to 18- to 50-year-old users ( Angwin et al., 2017 ). In such cases, it is difficult to see another plausible explanation for restricting the potential applicant age aside from conscious or unconscious age stereotypes. Notably, this process can also work against younger workers. For positions in which “stability” and industry “wisdom” are desired, organizations may seek older applicants (and deter younger ones) because of the positive stereotypes of older workers on these factors. Indeed, this was observed in the targeted recruiting efforts by Verizon, which excluded workers under age 25.

Another aspect of attraction is an organization’s navigation of employment laws regarding age in different countries. For instance, although posting age range preferences in job ads is illegal in the United States, it is relatively common practice in some countries (e.g., Turkey, Italy). Similarly, in many countries (e.g., Turkey) requesting applicants to upload their picture with their resume is relatively common, but this practice is illegal in the United States as photographs could serve as a mechanism for purposely (or incidentally) introducing bias into the screening process. Again, the desire to restrict a job ad to those within a given age range is seemingly tied to age-based stereotypes, and the extent to which these practices are legal may influence age-based stereotypes and expectations. Taken together, attraction efforts by organizations may already be biased, such that workers are excluded from the applicant pool based on age, attenuating our observed effects of age stereotypes on personnel decisions.

Within the attraction phase, research should examine recruiters’ and hiring managers’ conscious and unconscious age stereotypes (cf., Zaniboni et al., 2019 ), how these stereotypes play out in both formal (e.g., job ads) and informal (e.g., interview behavior) communications, how applicants react to this communication to self-select, and how this may subsequently limit the age variability of organizational membership downstream. We thus suggest the integration of age stereotyping concepts (e.g., the stereotype content model; Cuddy & Fiske, 2002 ) into the recruitment, signaling, and employer image literatures (e.g., Lievens & Slaughter, 2016 ). For example, research could use signaling theory to examine how recruiters develop recruitment materials and how these decisions affect applicants of different ages. This includes recruiters’ propensity to incorporate biased language in job ads or include images of older workers in recruitment materials, and the extent to which the presence of such language and images are driven by conscious or unconscious age stereotypes. Research could also incorporate qualitative research designs to see whether recruiters intentionally use signals to target older or younger applicants (e.g., Wilhelmy et al., 2016 ). Finally, drawing on the ASA framework, researchers should investigate the broader, organization-level culture and the extent to which age stereotypes held across the organization influence who is (and is not) attracted to specific organizations and roles.

During the selection phase, hiring decisions are the most relevant personnel decision. In fact, decisions made in this phase may manifest in the difficulty older workers face in becoming reemployed ( Wanberg et al., 2016 ). The relevance of age stereotypes in selection likely depends on the level of information available to decision-makers. In the context of selecting external candidates, personnel decision-makers have minimal information, and as such, are more likely to rely on surface traits such as age to make decisions (e.g., Finkelstein et al., 1995 ). In fact, recruiters, who are often the first filter during the hiring process, are thought to spend as little as 7 s reviewing an individual resume ( Ladders, 2018 ). Moreover, even for candidates receiving an interview, the available information is greatly limited, particularly when compared with an internal applicant for whom performance data may be available. Thus, in terms of uncovering when age stereotypes do affect personnel decisions, research on the relationship between age stereotypes and age discrimination should focus on personnel decisions where limited information is available (i.e., external selection).

Furthermore, research should examine the separate, differential influences of explicit (conscious) and implicit (unconscious) age stereotypes on selection decisions. For example, a study by Zaniboni et al. (2019) found that explicit stereotypes positively affected the evaluations of younger applicants’ resumes, with no effect on those of older applicants. In contrast, implicit stereotypes negatively affected the evaluation of older workers’ resumes. This initial evidence indicates that implicit and explicit age stereotypes differentially affect the evaluations of older and younger job applicants and operate through different pathways. We argue that researchers should examine these dual, conscious and unconscious pathways, well accepted within social psychology (e.g., Strack & Deutsch, 2015 ), which may lead to a better understanding of the causes of workplace age discrimination and how to reduce it. For instance, the research by Zaniboni et al. (2019) could be extended to the interview context and whether implicit and explicit stereotypes differentially affect evaluations of job applicants.

Finally, cultures associated with different jobs and industries may affect the degree to which age stereotypes affect selection decisions. Consider older employees in the high-tech industry: Stereotypes of older workers as unmotivated and unable to learn ( Posthuma & Campion, 2009 ) run counter to those of a high-tech worker, who must constantly learn and adapt to new technologies. Thus, in the tech industry, we would expect age stereotypes to favor younger workers. However, extant age stereotype research lacks direct comparisons of jobs and industries in terms of how they activate age stereotypes to affect hiring decisions (e.g., Reeves et al., 2021 ). Future research should actively compare job and industry cultures for effects on the activation of age stereotype to influence personnel decisions.

In the attrition phase, the most relevant age-related personnel decisions include turnover (voluntary and involuntary), early retirement, and layoffs. But little research examines the effect of age stereotypes on more proximal factors that likely lead to attrition. For instance, how low-intensity age-related microaggressions (e.g., subtle purposeful or unintentional behaviors and messages that may devalue someone based on age and result in social exclusion) may mediate the relationship between age stereotypes and attrition, including an employer’s decision to retain an employee or an employee’s decision to stay or leave the organization. In other words, a gap in our current knowledge is how age stereotypes may indirectly influence attrition-related decisions through age-related microaggressions. For example, if a supervisor stereotypes older workers’ as being less willing to learn, they may exclude older workers from training opportunities (a microaggression), which may affect their career progress, promotion chances, or job performance. As a result, these employees may choose to leave the organization, or they may be fired or forced to retire. A recent meta-analysis of the incivility (a type of microaggression) literature found that incivility had a negative relationship with job performance, positive relationship with withdrawal behaviors, and that age was related to experienced incivility (younger workers reported more incivility; Han et al., 2021 ), which provides some initial evidence for the integration of these literatures.

Future research integrating incivility and microaggression frameworks with age stereotypes may provide more insight into how these phenomena ultimately result in age discrimination-related attrition, including organizational and individual decisions like turnover, layoffs, and retirement. To accomplish this, methodologies such as daily-diary studies focused on age-based microaggressions, which have not commonly been used in the age-stereotype literature, are likely to be needed. In addition, the current literature lacks a comprehensive list of age-specific microaggressions (e.g., exclusion of older workers from training opportunities, use of ageist language like “old timers,” “millennials,” or “okay boomer”), which would be a necessary first step to launch this line of inquiry. Finally, research should also examine whether age-related microaggressions are indicators of or related to a negative age culture or climate (e.g., Boehm et al., 2014 ). In other words, do these microaggressions become so pervasive among organizational members that they influence organizational climate and culture? This may provide some insights into how organizations may need to change to reduce age-related discriminatory attrition decisions.

Murphy and DeNisi (2021) provide a wakeup call to the field, pointing out the lack of consistent evidence that age stereotypes influence personnel decisions. We argue that age stereotypes clearly exist and that they do seem to affect personnel decisions—if we dig a bit more deeply into the phenomenon. Drawing on the ASA model, we identify new pathways for examining when age stereotypes matter for personnel decisions throughout the employment lifecycle and summarize future research questions ( Table 1 ). Addressing these research questions at each phase of ASA would help us understand when and why age stereotypes matter and how to overcome their effects.

Angwin , J. , Scheiber , N. , & Tobin , A. ( 2017 , December 20). Dozens of companies are using Facebook to exclude older workers from job ads . Pro Publica . https://www.propublica.org/article/facebook-ads-age-discrimination-targeting

Google Scholar

Google Preview

Boehm , S. A. , Kunze , F. , & Bruch , H. ( 2014 ). Spotlight on age-diversity climate: The impact of age-inclusive HR practices on firm-level outcomes . Personnel Psychology , 67 , 667 – 704 . doi: 10.1111/peps.12047

Carlsson , M. , & Eriksson , S. ( 2019 ). Age discrimination in hiring decisions: Evidence from a field experiment in the labor market . Labour Economics , 59 , 173 – 183 . doi: 10.1016/j.labeco.2019.03.002

Cuddy , A. J. , & Fiske , S. T. ( 2002 ). Doddering but dear: Process, content, and function in stereotyping of older persons. In T. Nelson (Ed.), Ageism: Stereotyping and prejudice against older persons (pp. 3 – 26 ). The MIT Press .

Finkelstein , L. M. , Burke , M. J. , & Raju , N. S. ( 1995 ). Age discrimination in simulated employment contexts: An integrative analysis . Journal of Applied Psychology , 80 , 652 – 663 . doi: 10.1037/0021-9010.80.6.652

Fiske , S. T. ( 2017 ). Prejudices in cultural contexts: Shared stereotypes (gender, age) versus variable stereotypes (race, ethnicity, religion) . Perspectives on Psychological Science , 12 , 791 – 799 . doi: 10.1177/1745691617708204

Han , S. , Harold , C. M. , Oh , I. -S. , Kim , J. K. , & Agolli , A. ( 2021 ). A meta-analysis integrating 20 years of workplace incivility research: Antecedents, consequences, and boundary conditions . Journal of Organizational Behavior , 43, 497–5 23 . doi: 10.1002/job.2568

Ladders . ( 2018 ). Eye-tracking study [White Paper] . https://www.theladders.com/static/images/basicSite/pdfs/TheLadders-EyeTracking-StudyC2.pdf

Lievens , F. , & Highhouse , S. ( 2003 ). The relation of instrumental and symbolic attributes to a company’s attractiveness as an employer . Personnel Psychology , 56 , 75 – 102 . doi: 10.1111/j.1744-6570.2003.tb00144.x

Lievens , F. , & Slaughter , J. E. ( 2016 ). Employer image and employer branding: What we know and what we need to know . Annual Review of Organizational Psychology and Organizational Behavior , 3 , 407 – 440 . doi: 10.1146/annurev-orgpsych-041015-062501

Murphy , K. R. , & DeNisi , A. S. ( 2021 ). Do age stereotypes predict personnel decisions? The state of the evidence . Work, Aging and Retirement , 8(4), 323 – 330 . doi: 10.1093/workar/waab019

Posthuma , R. A. , & Campion , M. A. ( 2009 ). Age stereotypes in the workplace: Common stereotypes, moderators, and future research directions . Journal of Management , 35 , 158 – 188 . doi: 10.1177/0149206308318617

Reeves , M. D. , Fritzsche , B. A. , Marcus , J. , Smith , N. A. , & Ng , Y. L. ( 2021 ). “Beware the young doctor and the old barber”: Development and validation of a job age-type spectrum . Journal of Vocational Behavior , 129 , 103616 . doi: 10.1016/j.jvb.2021.103616

Schein , E. H. ( 1990 ). Organizational culture . American Psychologist , 45 , 109 – 119 . doi: 10.1037/0003-066X.45.2.109

Schneider , B. , Goldstein , H. , & Smith , D. B. ( 1995 ). The ASA framework: An update . Personnel Psychology , 48 , 747 – 773 . doi: 10.1111/j.1744-6570.1995.tb01780.x

Strack , F. , & Deutsch , R. ( 2015 ). The duality of everyday life: Dual-process and dual system models in social psychology. In M. Mikulincer and P. R. Shaver (Eds.), APA handbook of personality and social psychology, volume 1: Attitudes and social cognition . (pp. 891 – 927 ). American Psychological Association .

Wanberg , C. R. , Kanfer , R. , Hamann , D. J. , & Zhang , Z. ( 2016 ). Age and reemployment success after job loss: An integrative model and meta-analysis . Psychological Bulletin , 142 , 400 – 426 . doi: 10.1037/bul0000019

Wilhelmy , A. , Kleinmann , M. , König , C. J. , Melchers , K. G. , & Truxillo , D. M. ( 2016 ). How and why do interviewers try to make impressions on applicants? A qualitative study. Journal of Applied Psychology , 101 , 313 – 332 . doi: 10.1037/apl0000046

Zaniboni , S. , Kmicinska , M. , Truxillo , D. M. , Kahn , K. , Paladino , M. P. , & Fraccaroli , F. ( 2019 ). Will you still hire me when I am over 50? The effects of implicit and explicit age stereotyping on resume evaluations . European Journal of Work and Organizational Psychology , 28 , 453 – 467 . doi: 10.1080/1359432X.2019.1600506

Email alerts

Citing articles via.

• Recommend to your Library

Affiliations

• Online ISSN 2054-4650
• Copyright © 2024 Oxford University Press
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

To help or not: negative aging stereotypes held by younger adults could promote helping behaviors toward older adults

• Published: 18 February 2023
• Volume 43 , pages 1041–1051, ( 2024 )

Cite this article

• Gu Ma 1 , 2 ,
• Zizhuo Chen 1 , 2 ,
• Wanhua Zou 1 , 2 &
• Xin Zhang   ORCID: orcid.org/0000-0001-9061-6930 1 , 2  

1621 Accesses

Explore all metrics

Aging stereotypes affect older adults’ behaviors, however, it is unclear whether and how (negative) aging stereotypes influence younger adults’ behaviors toward older adults. Two possibilities arose, such that aging stereotypes would reduce helping behaviors according to TMT and SIT; while based on the BIAS map, we would expect the opposite. The present study aimed to further compare the two possibilities by examining the effect of negative aging stereotypes on younger adults’ helping behaviors, and testing which theory would fit the data better. In a cross-sectional study (Study 1), 112 Chinese younger adults ( M  = 22.67, SD  = 2.56) were recruited. Aging stereotypes were measured by the Ambivalent Ageism Scale and the abbreviated ageism questionnaire. And their prosocial behaviors were measured by the modified third-party punishment task. The results revealed that high benevolent ageism would increase helping behaviors toward older adults. In the following experiment with aging stereotype priming (positive, neutral vs. negative) among 130 Chinese younger adults ( M  = 26.82, SD  = 3.70), we confirmed the influence of negative aging stereotypes on prosocial behaviors measured by both third-party punishment and Social Value Orientation tasks. Study 2 further demonstrated that pity might mediate the association between negative aging stereotypes and behaviors. Our results indicated that younger adults’ negative aging stereotypes could increase their prosociality toward older adults through pity in line with BIAS maps. It also had significant theoretical and practical implications for future research. For example, with more education and intergenerational contact in younger generation which could evoke pity feelings for older adults, could help to build harmonious intergenerational relations.

Similar content being viewed by others

Social Identity Theory

The impact of social media on social comparison and envy in teenagers: the moderating role of the parent comparing children and in-group competition among friends.

Peerayuth Charoensukmongkol

Positive Psychology: An Introduction

Avoid common mistakes on your manuscript.

The entire world is aging, and aging stereotypes (especially the negative ones) have been extensively studied by gerontologists. Numerous studies have found that negative aging stereotypes could influence older adults heavily on their physical, cognitive, and behavioral outcomes (see for review, Meisner, 2012 ). However, the influences of aging stereotypes on younger adults, for example, whether negative aging stereotypes could promote more helping behavior or more hostility toward older adults are largely unknown. Such issues also yielded practical importance nowadays, given the current circumstances that younger and older adults have more chances to interact with each other. Hence, the present study would investigate how negative aging stereotypes could influence younger adults’ helping behaviors toward older adults in the context of intergenerational interaction, from two distinct perspectives, i.e., Terror Management Theory (Greenberg et al., 1997 ) vs. BIAS map (Cuddy et al., 2007 ). Two competing hypotheses were proposed and tested on the basis of these theories in the present study to consolidate the association between negative attitudes and helping behaviors toward older adults.

Negative aging stereotypes and their consequences

Stereotypes are usually defined as a stable psychological tendency held by individuals towards a certain group. More specifically, aging stereotypes are specific beliefs held by individuals towards older adults (Eagly & Chaiken, 1993 ; Levy et al., 2000 ), and it is also suggested that there are more negative aging stereotypes than positive ones (e.g., Löckenhoff et al., 2009 ). For example, old people are usually considered weak and useless. Content analyses based on tweets implied that the life of older adults is undervalued, and younger adults made jokes about older adults during Covid-19 (Xiang et al., 2021 ). According to the Stereotype Content Model (Fiske et al., 2002 ), older adults fell into the High Warmth but Low Competence stereotype category. The consequences of negative aging stereotypes (as well as ageism) on older adults have been extensively studied by gerontologists, and it is found that negative aging stereotypes held by older adults can have a detrimental effect on the physical and psychological well-being of the individual, and even mortality (e.g., Levy, 2003 ; Levy et al., 2022 ; Moser et al., 2011 ; Wurm et al., 2007 ; Zhang et al., 2020 ).

However, there is still one question that remained unsolved, i.e., how negative aging stereotypes could influence younger adults, especially in the context of intergenerational interaction. Although, there have been studies reporting that negative attitudes toward aging and anxiety about death were positively correlated in younger adults, and the ageist attitudes held by younger adults are positively correlated with their risk-taking behaviors (DePaola et al., 2003 ; Popham et al., 2011 ). Fewer studies have directly investigated how younger adults’ aging stereotype influences their behaviors toward older adults.

To help or not

To better understand the role of negative aging stereotypes on younger adults’ behavior toward older adults, two perspectives could be utilized. On the one hand, Terror Management Theory (TMT; Greenberg et al., 1997 ) postulates that an individual’s awareness of mortality could lead to a higher level of anxiety, which in turn can lead to an enhanced motive in self-esteem protection and worldview defense. While Social Identity Theory (SIT; Tajfel, 1981 ) also suggests that self-esteem protection could be achieved by ingroup favoritism and outgroup discrimination, for instance, ageism could be regarded as a form of protection of younger adults’ self-esteem (Bodner, 2009 ). In summary, Terror Management Theory and Social Identity Theory conjointly predicted that negative aging stereotypes could lead to a lower level of prosocial behaviors in younger adults. Indeed, Bergman and Bodner ( 2015 ) revealed that younger adults’ ageism was associated with reduced compassion toward incapable older adults. People with high negative aging stereotypes were more likely to keep distance from and less likely to help older adults. Other evidence also suggested that if older adults were not included in younger adults’ self-group (i.e., be considered as out-group members), they would receive more hostile ageism and less helping behaviors (Chen & Zhang, 2022 ; Tasdemir, 2020 ). Spaccatini and colleagues (Spaccatini et al., 2022 ) reported that during the pandemic, the younger adults’ endorsement of ageist attitudes positively affected the support for selective lockdown on the older population only. Younger adults with higher level of ageist attitudes believed that it was wrong asking young people to sacrifice their social life staying at home meanwhile it would be enough to isolate the older people.

On the other hand, however, according to the Stereotype Content Model (Fiske et al., 2002 ) as well as its extension – the BIAS map (Behaviors from Intergroup Affect and Stereotypes map, Cuddy et al., 2007 ), we might predict the opposite. Older adults are perceived as warm and incompetent in terms of commonly held stereotypes. Cuddy et al. ( 2007 ) further suggested that such a stereotype could evoke the emotion of pity, which could eventually lead to increased helping behaviors towards members in such stereotyped group (i.e., older adults in our case). Supportive evidence showed that benevolent ageism was positively associated with containment behaviors, including protection for vulnerable older people, during the COVID-19 pandemic (Visintin, 2021 ); and feelings of pity and compassion could also facilitate prosocial behaviors toward others (Chen et al., 2022 ). A piece of indirect evidence, also supporting our argument, showed that males with high benevolent sexism were more likely to protect their female friends from sexual and relationship violence at a party because they believed that those women were deserving of protection and they should be the “White Knight” (Leone et al., 2020 ).

Measuring prosocial behaviors: the modified third-party punishment game

To better capture prosocial behaviors in the present study, a modified third-party punishment game (Fehr & Fischbacher, 2004 ) was used. Traditionally, prosocial behaviors could be measured by the dictator game or the second-party punishment game, both of which could be subject to social desirability, and it is suggested that compared with second-party punishment, sanctions involving the third party are more stable and effective in measuring prosociality (Bendor & Swistak, 2001 ; Zhou et al., 2017 ). Research has shown that about 60% of the third-party participants will punish the violations of social norms and follow the egalitarian distribution norm (Fehr & Fischbacher, 2004 ).

Besides, in the original third-party punishment game, the participant could punish the unfair proposer at the cost of his own benefit. In the modified version, the compensation component was also introduced for a more comprehensive understanding of both punishment and compensation (Hu et al., 2015 ; Leliveld et al., 2012 ) in the circumstances of social interactions.

Present study

In the present study, the core research question we would like to investigate is whether and how negative aging stereotypes could influence younger adults’ (prosocial) behaviors. Based on two different theoretical frameworks, two competing hypotheses could be proposed.

Competing Hypothesis 1: negative aging stereotypes could lead to a lower level of prosociality toward older adults, from the perspective of TMT and SIT;

Competing Hypothesis 2: negative aging stereotypes could lead to increased prosocial behaviors toward older adults, according to BIAS maps.

Two experiments were designed to examine these competing hypotheses. Study 1 was a correlation study to investigate the association between self-reported (negative) aging stereotypes and their prosociality toward older adults. While for the second study, aging stereotypes were experimentally manipulated, and we are interested in testing how different manipulations could lead to differences in their prosocial behaviors toward older adults as well as the underlying mechanisms. The results of experiments and the underlying mechanisms revealed could help us consolidate the effect of negative aging stereotypes on youngers adults’ prosocial behaviors toward older adults.

Study 1: associations between aging stereotypes and prosociality

In the first study, we used a modified third-party intervention task to measure participants’ prosocial behaviors, while we also measured their dispositional aging stereotypes, in order to test the two competing hypotheses. It is expected that if competing hypothesis 1 stands, a negative correlation between negative aging stereotype and prosocial social behaviors toward older adults would be observed; however, if competing hypothesis 2 is correct, the opposite association between negative aging stereotype and prosociality would be expected.

Participant

One hundred and twenty-nine younger adults (69.6% male, M age = 22.67, SD  = 2.56) were recruited via Wenjuanxing ( www.wjx.cn ), a Chinese professional online questionnaire survey and evaluation platform. All the participants received a random reward of 4 to 6 yuan for participation. Demographic information including age, sex, education, individual income, and health level was collected.

Materials and measurements

Third-party intervention task.

The task was modified from the third-party punishment paradigm (Fehr & Fischbacher, 2003 ). A dictator game (DG) situation was first presented to the participant who was told to be a third-party jury. In the dictator scenario, proposer A should decide how to allocate 100 tokens he shared with receiver B, and B can only choose to accept his proposal. Meanwhile, the third-party participant had 50 tokens, which could be allowed to punish A or compensate A, or the participant could choose just to do nothing and keep all 50 tokens in each trial.

Negative aging stereotypes

Negative Aging Stereotypes were measured by two self-report questionnaires. Namely, the Ambivalent Ageism Scale (AAS) developed by Cary et al. ( 2017 ) and the Abbreviated Stereotype Questionnaire, adapted from Fiske and colleagues ( 2002 , Study 3).

AAS is a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree), including 9 items measuring benevolent ageism (e.g., “Older people need to be protected from the harsh realities of society”) and 4 items measuring hostile subscale (e.g., “Old people are too easily offended”). A higher score represents a higher level of negative aging stereotypes. The scale yielded good internal consistencies as indicated by Cronbach’s α = 0.89 for benevolent ageism and 0.84 for hostile ageism, respectively.

The Abbreviated Stereotype Questionnaire is a 5-point Likert Scale (1 = not at all, 5 = extremely), asking participants to rate their attitudes toward a certain group in four dimensions (Competence, Warmth, Status, and Competition) with two questions for each dimension. Here we aimed to test the attitudes toward the older adult, so we chose two trait dimensions (Competence and Warmth), and the adapted question was like “How confident are the elderly?“. A higher score in warmth represents more positive attitudes toward older adults. The scale also yielded acceptable internal consistencies as indicated by significant correlations in the warmth dimension ( r  = .45, p  < .001) and competence dimension ( r  = .29, p  = .002).

After getting participants’ formal consent, they were first introduced to finish the modified third-party intervention task. In the present study, to better capture prosocial behaviors toward older adults, three different age combinations of proposer A and receiver B were developed, such as young proposer A and young receiver B (baseline condition), young proposer A and old receiver B (young proposer condition), and old proposer A and young receiver B (old proposer condition). In each condition, there are 11 proposer-receiver proposal levels (i.e., 100-0, 90 − 10, 80 − 20, 70 − 30, 60 − 40, 50–50, 40–60, 30–70, 20–80, 90 − 10, 0-100), representing extremely prosocial (offering all money to receiver) to extremely selfish (keeping all money to self). Participants would play the modified third-party intervention task with different age combinations and proposer-receiver proposals (which would yield a total of 3 × 11 trials), and each age combination condition and proposer-receiver levels were presented in random order. In order to encourage the participants to consider both their own interests and social norms, they were informed that one turn would be selected randomly at the end of the experiment and the remaining tokens in this turn could be their extra bonus of the experiment. For better understanding of the task, two practice trials were presented and participants were asked to act following the instruction. Later, the subjective indifference point (SIP) was calculated as an indicator of participants’ prosocial behaviors.

After the task, several self-report scales were measured, including the Ambivalent Ageism Scale (Cary et al., 2017 ) and the Abbreviated Stereotype Questionnaire (Fiske et al., 2002 ). Finally, demographics including sex, age, education level (1- primary school and below; 2- junior high; 3- senior high; 4- bachelor and above), self-reported health (from 1 = very poor to 5 = excellent) and subjective SES (from 1 = lower than 10th percentile, to 10 = above the 90th percentile) recorded. After participants accomplished all the tasks and measurements, they were thanked and compensated for their participation.

Results and discussion

Data preparation and descriptive statistics.

In the present experiment, the least square approach was used to model the linear association between the distribution amount from the proposer and punishment (and compensation) made by the participant with MATLAB R2017b (Fehr & Fischbacher, 2004 ) in each age combination condition. To be more specific, the independent variable is the number of tokens receiver B got (which is also the number of tokens distributed by proposer A), and the dependent variable is tokens spent by the participant to compensate (and punish) the proposer A (a negative value represents punishment while a positive value represents compensation). In theory, tokens spent by the third party to punish the proposer are supposed to reach the highest level when the proposer distributes 0 tokens to the receiver, and monotonically decrease close to 0 as the proposal seems fairer and fairer (Fehr & Fischbacher, 2004 ). Similarly, tokens spent to compensate the proposer would be the highest if the proposer distributed 100 tokens to the receiver. In other words, the number of tokens distributed by the proposer and the tokens spent to compensate (or punish) the proposer should be a linear (and positive) association. Three linear regression models (for each age combination condition) were conducted for each participant. And based on these regression models, two indices were also extracted, namely, the Subjective Indifference Point, which is the intersection point (of each regression line) with the horizontal axis, representing the (in)tolerance to unfair proposals; and the Strength of intervention, which is the average amount of tokens that the third-party participants are willing to use to punish (or reward) the proposer based on his proposals. See Fig. 1  for details.

Linear regressions of three age groups

Meanwhile, 3 (age conditions) x 2 (types of intervention) linear regressions were conducted. Participants were then screened on the following criteria: (1) the participants who could not fit the linear regression more than twice among those six regressions, and (2) having negative regression coefficients for slope, indicating that the participant was rewarding the receiver when punishment should be made and vice versa. Participants who met any of the criteria were removed, and a total of 17 participants were excluded with 112 participants remaining Footnote 1 . Table  1 depicted the demographic information of 112 participants.

Hypothesis testing

Subjective indifference point.

Repeated-measure ANOVA with age combination condition (Younger proposer – Younger receiver, Younger proposer – Older receiver, vs. Older proposer – Younger receiver) as the within-subject factor was conducted. A significant age combination main effect was found, F (2,216) = 25.30, p  < .001, \({\eta }_{p}^{2}\) = 0.19. Post hoc tests further revealed that the SIP of the Older proposer - Younger receiver group ( M  = 46.56, SD  = 7.96) were significantly smaller than those of the Younger proposer – Younger receiver group ( M  = 50.47, SD  = 9.76), which is also significantly smaller than the Younger proposer – Older receiver group ( M  = 52.99, SD  = 7.47), indicating that participants are more tolerant to unfair proposals by older adults, while harsher to unfair proposals made by younger adults, especially when the receiver is an older adult.

Then, linear regression for SIP and ageism-related variables was conducted. The dependent variables were SIPs of the younger-receiver group and older-receiver group. Age, gender, education level, health level, family income level, and baseline SIP were controlled as covariates. Benevolent ageism and hostile ageism, as well as competence and warmth, were entered as independent variables. It was found that the SIP of the younger-receiver group can be predicted by benevolent ageism, β = -2.11, SE  = 0.89, p  < .05 and competence, β = 2.69, SE  = 1.01, p  < .01, respectively (Table  2 ), indicating that people would like to give a few privileges to weak older adults. However, privileges would no longer in existence when older adults were perceived as competent.

Strength of intervention

A 2 (intervention: punishment and reward) × 3 (age conditions) repeated-measure ANOVA was conducted. A significant main effect of age conditions was found, F (2, 222) = 21.81, p  < .001, \({\eta }_{p}^{2}\) = 0.16, qualified by a significant intervention x age condition interaction, F (2, 222) = 34.11, p < .001, \({\eta }_{p}^{2}\) = 0.24. Simple main effect analysis revealed that the punishment to the younger proposer ( M  = 24.13, SE  = 0.92) was significantly higher than baseline ( M  = 19.40, SE  = 0.85) and the older proposer conditions ( M  = 18.42, SE  = 0.82), but the latter two did not differ significantly. And the reward to the older proposer ( M  = 22.72, SE  = 1.02) was higher than the baseline ( M  = 18.17, SE  = 0.98), and the younger proposer conditions ( M  = 19.66, SE  = 0.96), but no significant difference was found between the latter two.

Similar linear regressions were conducted to further test how attitudes toward older adults could affect younger participants’ prosocial behaviors. The dependent variables were punishment (or reward) to the younger proposer and older proposer. Age, gender, education level, health level, family income level, and baseline punishment (or reward) were controlled as covariates. Benevolent ageism and hostile ageism as well as competence and warmth were entered as independent variables. It is found that the reward to older altruistic proposers can be positively predicted by benevolent ageism, β = 2.73, SE  = 0.82 (please also see Table  2 ), suggesting that participants would like to reward the older adults especially when the old were perceived weak.

In summary, Study 1 revealed that people were lenient to older adults, even under the circumstances of them offering unfair proposals. Further, the results from multiple regression also provided support to the argument made by BIAS maps, that the tolerance (of unfair proposals from older adults) was driven by the stereotypes of low competence and help-needing. In other words, negative aging stereotypes would indeed promote more helping behaviors toward older adults.

Study 2: manipulated incompetence leads to increased prosociality

Preliminary results from our first study revealed that indeed negative attitudes toward older adults could promote younger adults’ prosocial behaviors toward older adults. However, as the first study is correlational in nature, no causal relations could be inferred. In our second study, we sought to establish the causality between attitudes toward older adults and prosocial behaviors, by manipulating the different types of aging stereotypes. Besides, we also would like to explore the underlying mechanism involved. According to the BIAS map, a stereotype of high warmth but low competence is associated with the emotion of pity, which would eventually lead to more helping behaviors toward such stereotyped groups. Hence, in the present study, we also measured perceived pity toward older adults, to see whether pity could mediate the effect of negative attitudes toward older adults, and directly associate with increased prosocial behaviors toward older adults as predicted by the BIAS map.

In the present study, in addition to the third-party intervention task used in Study 1, we also utilized another widely used task to capture prosocial behaviors, i.e., the Social Value Orientation (SVO) task (Murphy et al., 2011 ), to further consolidate our results.

One hundred and sixty-four adults (53.8% male, M age = 26.82, SD  = 3.70) were recruited via Credamo ( www.credamo.com ), a Chinese one-stop intelligent research platform. All participants received a monetary reward of 15 to 17 Chinese Yuan for participation. Demographic information including age, sex, education, individual income, and health level was collected. Thirty-four participants were excluded because they could not pass the attention check or manipulation check set in the experiment. Demographic information of the rest 130 participants was shown in Table  1 .

• Prosocial behaviors

Besides the third-party intervention task used in Study 1, Social Value Orientation task developed by Murphy et al. ( 2011 ) was also adopted. The original task consists of 6 trials with 9 different monetary distribution proposals in each trial, and participants are required to choose one proposal to distribute money between themselves and a partner for each trial (similar to a dictator game). Then an SVO score could be calculated and used to represent the participant’s prosociality (Murphy et al., 2011 ). We again modified the task to fulfill the purpose of the present study by manipulating the age of the partner, such that participants are introduced that they are about to distribute some money with a younger adult (ranging from 20 to 30) or they are about to distribute the money with an older adult (ranging from 65 to 75).

Attitudes and feelings toward older adults

An 8-item self-report questionnaire adapted from (Zhang et al., 2016 ) was used to measure participants’ attitudes toward older adults (i.e., Warmth and Competence) as well as their emotions toward older adults (Pity and Envy), with a 7-point-Likert scaling (from 1 = strongly disagree, to 7 = strongly agree). The scale also yielded acceptable internal consistencies as indicated by significant correlations in Competence ( r  = .69, p  < .001), Warmth ( r  = .84, p  < .001), Pity ( r  = .84, p  < .001), and Envy ( r  = .62, p  < .001).

Experimental design and procedures

After getting participants’ informed consent, they were randomly assigned to three different aging stereotype priming conditions (i.e., positive aging stereotypes, negative aging stereotypes vs. control condition). Participants in each condition were asked to read a particular material accordingly. After reading the materials, two manipulation check questions were asked, to make sure they can understand the material correctly. For details, refer to the supplementary material .

Participants who failed at least one manipulation check question were dropped from the analysis. Then, participants were asked to provide their attitudes and feelings toward older adults. Next, participants finished the SVO task and third-party intervention task in a fixed order (the order of trials in each task was random). Finally, demographic information including age, sex, education level, self-reported health, and subjective income was also collected.

Similar to Study 1, we processed participants’ responses in the third-party intervention task using MATLAB and got their subjective indifference points and strength of intervention for each age combination condition. Following the advice from Murphy et al. ( 2011 ), participants’ average SVO scores for different-aged partners were also calculated.

A larger SVO score represents a higher level of prosocial behavior toward that partner.

Effectiveness of manipulation

One-way ANOVAs were conducted with priming conditions (positive, negative vs. control condition) as the between-subject factor on warmth, competence, pity, and envy, respectively. Significant main effects of priming were found for warmth, F (2, 122) = 16.62, p  < .001, \({\eta }_{p}^{2}\) = 0.21; competence, F (2, 122) = 22.04, p < .001, \({\eta }_{p}^{2}\) = 0.27; and pity, F (2, 122) = 5.06, p < .01, \({\eta }_{p}^{2}\) = 0.21; but not envy, F (2, 122) = 1.94, p = .15 \({\eta }_{p}^{2}\) = 0.03. Further post-hoc analyses revealed that participants in the positive priming condition indeed expressed the most positive attitudes toward older adults, while participants in the negative priming condition expressed the most negative attitudes. Moreover, participants in the negative priming condition exhibited a higher level of pity feelings toward older adults than did participants in the positive priming condition (Table  S1 in supplementary material).

Hypothesis testing – attitudes toward older adults predict prosocial behaviors

According to SCM (Fiske et al., 2002 ), older adults are seen as warm but incompetent, hence a stereotype score was calculated by subtracting competence from warmth to capture the stereotypes of high warmth but low competence. Besides, we also calculated SVO (towards the old) minus SVO (towards the young) as SVO difference representing to what extent participants were more friendly to older adults than to the young.

Linear regression with SVO difference as the dependent variable was conducted. Age, gender, education level, health level, family income level and were controlled as covariates. Stereotype score was the independent variable. It was found that SVO difference can be predicted by stereotype score, β = 3.02, SE  = 1.33 p  = .03 (please refer to Table  3 ). Similar linear regressions with SIPs (both younger-receiver group and older-receiver group) as the dependent variables were also conducted. It was found that only the SIP of the older-receiver group can be positively predicted by stereotype score, β = 1.17, SE  = 0.51, p  = .02 (see Table  4 ). These results suggested that participants would be more altruistic and lenient if they perceived older adults as more warm-but-incompetent.

Hypothesis testing – the mediation role of pity

Using PROCESS (v 3.4 by Andrew F. Hayes) in SPSS, the mediation effects of pity on SVO difference and SIPs were analyzed (controlling covariates, bootstrap = 5000). Starting with the main model, it was found that SVO difference increased with higher Stereotype scores (B = 3.02, 95% CI = [0.38, 5.65]), suggesting that participants holding a more negative aging stereotype (high warmth but low competence) were more friendly to older than to younger adults. In the mediation model, hypotheses regarding the mediation effect of perceived pity on the association between attitudes and prosocial behaviors toward the old were tested. Adding pity into the model attenuated the effect of stereotype on SVO, such that the stereotype score became insignificant (B = 1.95, 95% CI = [-0.73, 4.63]), but perceived pity could significantly predict both higher SVO differences (B = 2.92, 95% CI = [0.83, 5.00]), see Table  3 . The results of the mediation model indicated that pity fully mediated the effect of aging stereotypes on prosocial behaviors toward older adults, making younger adults allocate more resources to older partners (vs. younger partners).

Similar mediation analyses with SIP of younger-receiver and older-receiver groups were conducted. It was found that there was no association between stereotype score and SIP of the younger-receiver group in neither the main model (B = -0.52, 95% CI = [-1.75, 0.71]), nor the mediation model (B = -0.10, 95% CI = [-1.35, 1.15]). However, in the mediation model, it was found that a higher stereotype score positively predicted pity (B = 0.33, 95% CI = [0.11, 0.55]), and higher perceived pity predicted lower SIP (B = -1.26, 95% CI = [-2.27, -0.25]), suggesting that higher level of pity toward older adults, would make participants more tolerance to unfair proposals made by older adults. For the SIP of the older-receiver group, it was found that SIP increased with a higher stereotype score (B = 1.17, 95% CI = [0.16, 2.19]), suggesting that people who held a higher warm-but-incompetent stereotype of older adults would be less tolerant to unfair proposals made by younger adults. Furthermore, the attenuated and insignificant effect of stereotype on SIP was found (B = 0.81, 95% CI = [-0.22, 1.84]), after pity was entered as a mediator. Significant effects of pity on SIP (B = 1.08, 95% CI = [0.25, 1.92]) indicated that pity fully mediated the association between stereotype and SIP, suggesting that, people holding a more warm-but-incompetent attitude had a higher level of pity toward older adults, which would result in less tolerant to unfair offers made by younger adults.

In summary, the results revealed the mediation effect of pity on the association between aging stereotypes and helping behaviors to older adults, such that the warm-but-incompetent stereotype toward older adults would evoke feelings of pity in younger adults, which could, in turn, increase younger adults’ prosociality toward older adults.

General discussion

The present study aimed to compare two competing hypotheses regarding the relationship between (negative) aging stereotypes and prosocial behaviors based on two different perspectives (Cuddy et al., 2007 ; Greenberg et al., 1997 ; Tajfel, 1981 ). Our results indicate that the prosocial behaviors toward older adults would increase as the negative aging stereotypes increase, which provides support to the Competing Hypothesis 2 based on the BIAS map.

In Study 1, it was found that high benevolent ageism could predict more tolerance to selfish older adults and more rewards to altruistic older adults, which is in line with the competing hypothesis 2, suggesting that people with high benevolent ageism would be more likely to help or protect the weak (i.e., older adults in the present study; see Leone et al., 2020 ; Visintin, 2021 ). In Study 2, we manipulated the participants’ aging stereotypes and introduced another paradigm to mutually validate our findings. The experiments showed that as the young participants perceived the older adults as warm-but-incompetent, they would have more prosocial behaviors toward the older than toward the younger adults as indicated by SVO, besides, their standard of fairness would be stricter to younger adults and looser to old adults as indicated by SIP. These results again supported the competing hypothesis 2 that negative aging stereotypes lead to increased prosocial behaviors toward older adults.

Moreover, the findings from study 2 further confirmed that pity could mediate the association between negative aging stereotypes and prosociality as suggested by the BIAS map (Cuddy et al., 2007 ). Such results are also consistent with empirical evidence that sympathy and compassion, which come with feelings of warmth, concern, and positive affect, could increase prosocial behaviors (Chen et al., 2022 ; Chierchia & Singer, 2017 ; Leiberg et al., 2011 ). However, they are partly contrary to existing research suggesting that ageism is associated with reduced compassion and efficacy to help incapable older adults (Bergman & Bodner, 2015 ). One possible explanation is that a high level of hostile ageism toward older adults indeed triggers younger adults’ negative feelings and trends to keep their distance. However, the ambivalent ageism (e.g., benevolent but ageist attitudes toward older adults) caused by perceived warm-but-incompetent might lead to more helping behaviors though unwanted, according to Cary and colleagues ( 2017 ). In general, with two studies, we provided consistent support for the BIAS map, revealing general negative ageism could indeed promote prosocial behaviors in younger adults in the economic domain via the mediation of pity feelings toward older adults. Such findings also yielded important practical implications. For example, we could educate people about aging in a nonpatronizing way to evoke their pity and make them realize that the old need help. It could be helpful for building harmonious intergenerational relations and social environments. We could also encourage the younger generation to have more intergenerational interaction (Verhage et al., 2021 ), to build up a proper image that older adults need and deserve help.

Limitations and future directions

Several limitations should also be acknowledged before we make any conclusion. First of all, In the present study, Chinese participants were recruited, making the generalizability of the present study less clear, as there has been evidences suggesting that Eastern Asians hold more positive attitudes toward older adults than Westerners because the cultural values such as filial piety and collectivism make Eastern Asians value and respect the old more (Ackerman & Chopik, 2021 ; Boduroglu et al., 2006 ; Sung, 2001 ). However, on the other hand, meta-analysis and empirical studies reveal that Eastern Asians indeed exhibited more negative attitudes (Huang, 2013 ; North & Fiske, 2015 ), or similar attitudes (Zhang et al., 2016 ) toward older adults compared with Westerners. The evidence suggested that ageism might be domain-specific (Vauclair et al., 2017 ; Voss et al., 2018 ), and no clear pattern of cultural differences would emerge. According to Zhang et al. ( 2016 ), personal values rather than cultural values had a significant influence on ageism attitudes. Moreover, considering the increasing aging population, changes in the social economic environment, and cultural values (e.g., the younger generation becoming more individualistic, Tan et al., 2021 ) in China, we believe that our findings could be generalized to other societies. Nevertheless, replication studies are indeed necessary in the future. Second, we manipulated participants’ aging stereotypes in Study 2, but we did not find behavioral differences between priming conditions. It might be because attitudes are implicit and have individual differences within groups. Third, although competing hypothesis 1 was not supported in the present study, we still cannot conclude that negative aging stereotypes could solely promote prosocial behaviors rather than hostility. As there is still evidence showing that negative stereotypes reduce prosocial behaviors through outgroup discrimination (Spaccatini et al., 2022 ; Stepanikova et al., 2011 ). Potential moderators might affect the association. For example, Wlodarczyk et al. ( 2014 ) found that realistic threats would increase ingroup favoritism and decrease prosocial behaviors. However, in the present study, older adults are not perceived as threatening, so there would be no increased ingroup favoritism and decreased prosocial behaviors. Nevertheless, future studies might be needed to test the boundary conditions for such association.

In conclusion, our results indicate that younger adults’ negative aging stereotypes could lead to more helping behaviors toward older adults, and pity plays an important mediation role. Such finding provides support to the BIAS map and also has practical implications for building harmonious intergenerational relations.

Data Availability

The datasets analysed during the current study are available from the corresponding author on reasonable request.

t tests showed that there was no significant difference between remained and excluded participants on demographic variables.

Ackerman, L. S., & Chopik, W. J. (2021). Cross-cultural comparisons in implicit and explicit age bias. Personality & Social Psychology Bulletin, 47 (6), 953–968. https://doi.org/10.1177/0146167220950070

Article   Google Scholar  

Bendor, J., & Swistak, P. (2001). The evolution of norms. The American Journal of Sociology, 10 (6), 1493–1545. https://doi.org/10.1086/321298

Bergman, Y. S., & Bodner, E. (2015). Ageist attitudes block young adults’ ability for compassion toward incapacitated older adults. International Psychogeriatrics, 27 (9), 1541–1550. https://doi.org/10.1017/S1041610215000198

Article   PubMed   Google Scholar  

Bodner, E. (2009). On the origins of ageism among older and younger adults. International Psychogeriatrics, 21 (6), 1003–1014. https://doi.org/10.1017/S104161020999055X

Boduroglu, A., Yoon, C., Luo, T., & Park, D. C. (2006). Age-related stereotypes: a comparison of american and chinese cultures. Gerontology (Basel), 52 (5), 324–333. https://doi.org/10.1159/000094614

Cary, L. A., Chasteen, A. L., Remedios, J., & Pruchno, R. (2017). The ambivalent ageism scale: developing and validating a scale to measure benevolent and hostile ageism. The Gerontologist, 57 (2), e27–e36. https://doi.org/10.1093/geront/gnw118

Chen, Z., & Zhang, X. (2022). We were all once young: reducing hostile ageism from younger adults’ perspective. Frontiers in Psychology, 13 , 793373–793373. https://doi.org/10.3389/fpsyg.2022.793373

Article   PubMed   PubMed Central   Google Scholar  

Chen, Y., Wang, Z., Zhang, Q., Dong, W., Xu, J. H. C., Wu, S. J., Zhang, X., & Chen, C. (2022). Compassion, discrimination, and prosocial behaviors: young diasporic chinese during the COVID-19 pandemic. Frontiers in Psychology, 13 , 814869–814869. https://doi.org/10.3389/fpsyg.2022.814869

Chierchia, G., & Singer, T. (2017). The neuroscience of compassion and empathy and their link to prosocial motivation and behavior. Decision neuroscience: An integrative perspective (pp. 247–257). https://doi.org/10.1016/B978-0-12-805308-9.00020-8

Cuddy, A. J. C., Fiske, S. T., & Glick, P. (2007). The BIAS map: behaviors from intergroup affect and stereotypes. Journal of Personality and Social Psychology, 92 (4), 631–648. https://doi.org/10.1037/0022-3514.92.4.631

DePaola, S. J., Griffin, M., Young, J. R., & Neimeyer, R. A. (2003). Death anxiety and attitudes toward the elderly among older adults: the role of gender and ethnicity. Death Studies, 27 (4), 335–354. https://doi.org/10.1080/07481180302904

Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes . Harcourt Brace Jovanovich College Publishers.

Fehr, E., & Fischbacher, U. (2003). The nature of human altruism. Nature, 425 (6960), 785–791. https://doi.org/10.1038/nature02043

Article   CAS   PubMed   ADS   Google Scholar  

Fehr, E., & Fischbacher, U. (2004). Third-party punishment and social norms. Evolution and human behavior , 25(2), 63–87 https://doi.org/10.1016/S1090-5138(04)00005-4

Fiske, S. T., Cuddy, A. J. C., Glick, P., & Xu, J. (2002). A model of (often mixed) stereotype content: competence and warmth respectively follow from perceived status and competition. Journal of Personality and Social Psychology, 82 (6), 878–902. https://doi.org/10.1037/0022-3514.82.6.878

Greenberg, J., Solomon, S., & Pyszczynski, T. (1997). Terror management theory of self-esteem and cultural worldviews: empirical assessments and conceptual refinements. Advances in Experimental Social Psychology, 29 (C), 61–139. https://doi.org/10.1016/S0065-2601(08)60016-7

Hu, Y., Strang, S., & Weber, B. (2015). Helping or punishing strangers: neural correlates of altruistic decisions as third-party and of its relation to empathic concern. Frontiers in Behavioral Neuroscience, 9 , 24–24. https://doi.org/10.3389/fnbeh.2015.00024

Huang, C. (2013). Undergraduate students’ knowledge about aging and attitudes toward older adults in east and west: a socio-economic and cultural exploration. International Journal of Aging & Human Development, 77 (1), 59–76. https://doi.org/10.2190/AG.77.1.d

Article   ADS   Google Scholar  

Leiberg, S., Klimecki, O., & Singer, T. (2011). Short-term compassion training increases prosocial behavior in a newly developed prosocial game. PLoS One1, 6 (3), e17798–e17798. https://doi.org/10.1371/journal.pone.0017798

Article   CAS   Google Scholar  

Leliveld, M. C., Dijk, E., & Beest, I. (2012). Punishing and compensating others at your own expense: the role of empathic concern on reactions to distributive injustice: punishing and compensating others at your own expense. European Journal of Social Psychology, 42 (2), 135–140. https://doi.org/10.1002/ejsp.872

Leone, R. M., Schipani-McLaughlin, A. M., Haikalis, M., & Parrott, D. J. (2020). The “white knight” effect: benevolent sexism accounts for bystander intervention in party situations among high status men. Psychology of Men & Masculinity, 21 (4), 704–709. https://doi.org/10.1037/men0000314

Levy, B., Ashman, O., & Dror, I. (2000). To be or not to be: the effects of aging stereotypes on the will to live. Omega: Journal of Death and Dying, 40 (3), 409–420. https://doi.org/10.2190/Y2GE-BVYQ-NF0E-83VR

Levy, B. R. (2003). Mind matters: cognitive and physical effects of aging self-stereotypes. The Journals of Gerontology Series B Psychological Sciences and Social Sciences, 58 (4), P203–P211. https://doi.org/10.1093/geronb/58.4.P203

Levy, B. R., Chang, E., Lowe, S. R., Provolo, N., & Slade, M. D. (2022). Impact of media-based negative and positive age stereotypes on older individuals’ mental health. The Journals of Gerontology Series B Psychological Sciences and Social Sciences, 77 (4), e70–e75. https://doi.org/10.1093/geronb/gbab085

Löckenhoff, C. E., De Fruyt, F., Terracciano, A., McCrae, R. R., De Bolle, M., Costa, P. T., Aguilar-Vafaie, M. E., Ahn, C., Ahn, H., Alcalay, L., Allik, J., Avdeyeva, T. V., Barbaranelli, C., Benet-Martinez, V., Blatný, M., Bratko, D., Cain, T. R., Crawford, J. T., Lima, M. P., & Yik, M. (2009). Perceptions of aging across 26 cultures and their culture-level associates. Psychology and Aging, 24 (4), 941–954. https://doi.org/10.1037/a0016901

Meisner, B. A. (2012). A meta-analysis of positive and negative age stereotype priming effects on behavior among older adults. The Journals of Gerontology Series B Psychological Sciences and Social Sciences, 67B (1), 13–17. https://doi.org/10.1093/geronb/gbr062

Moser, C., Spagnoli, J., & Santos-Eggimann, B. (2011). Self-perception of aging and vulnerability to adverse outcomes at the age of 65–70 years. The Journals of Gerontology Series B Psychological Sciences and Social Sciences, 66B (6), 675–680. https://doi.org/10.1093/geronb/gbr052

Murphy, R. O., Ackermann, K. A., & Handgraaf, M. J. J. (2011). Measuring social value orientation. Judgment and Decision Making, 6 (8), 771–781.

North, M. S., & Fiske, S. T. (2015). Modern attitudes toward older adults in the aging world: a cross-cultural meta-analysis. Psychological Bulletin, 141 (5), 993–1021. https://doi.org/10.1037/a0039469

Popham, L. E., Kennison, S. M., & Bradley, K. I. (2011). Ageism and risk-taking in young adults: evidence for a link between death anxiety and ageism. Death Studies, 35 (8), 751–763. https://doi.org/10.1080/07481187.2011.573176

Spaccatini, F., Giovannelli, I., & Pacilli, M. G. (2022). “You are stealing our present”: younger people’s ageism towards older people predicts attitude towards age-based COVID‐19 restriction measures. Journal of Social Issues, 78 (4), 769–789. https://doi.org/10.1111/josi.12537

Stepanikova, I., Triplett, J., & Simpson, B. (2011). Implicit racial bias and prosocial behavior. Social Science Research, 40 (4), 1186–1195. https://doi.org/10.1016/j.ssresearch.2011.02.004

Sung, K. (2001). Elder respect: exploration of ideals and forms in east asia. Journal of Aging Studies, 15 (1), 13–26. https://doi.org/10.1016/S0890-4065(00)00014-1

Tajfel, H. (1981). Human groups and social categories: studies in social psychology . Cambridge University Press.

Tan, T. X., Yi, Z., Camras, L. A., Cheng, K., Li, Z., Sun, Y., & Chen, N. (2021). The effect of academic performance, individualistic and collectivistic orientation on chinese youth’s adjustment. Social Psychology of Education, 24 (5), 1209–1229. https://doi.org/10.1007/s11218-021-09650-x

Tasdemir, N. (2020). Young group identification and motives as predictors of ageism, aging anxiety, and life satisfaction. The Journal of Genetic Psychology, 181 (5), 375–390. https://doi.org/10.1080/00221325.2020.1783195

Vauclair, C., Hanke, K., Huang, L., & Abrams, D. (2017). Are asian cultures really less ageist than western ones? It depends on the questions asked. International Journal of Psychology, 52 (2), 136–144. https://doi.org/10.1002/ijop.12292

Verhage, M., Schuurman, B., & Lindenberg, J. (2021). How young adults view older people: exploring the pathways of constructing a group image after participation in an intergenerational programme. Journal of Aging Studies, 56 , 100912–100912. https://doi.org/10.1016/j.jaging.2021.100912

Visintin, E. P. (2021). Contact with older people, ageism, and containment behaviours during the COVID-19 pandemic. Journal of Community & Applied Social Psychology, 31 (3), 314–325. https://doi.org/10.1002/casp.2504

Voss, P., Kornadt, A. E., Hess, T. M., Fung, H. H., & Rothermund, K. (2018). A world of difference? Domain-specific views on aging in China, the US, and Germany. Psychology and Aging, 33 (4), 595–606. https://doi.org/10.1037/pag0000237

Wlodarczyk, A., Basabe, N., & Bobowik, M. (2014). The perception of realistic and symbolic threat and its influence on prejudice, ingroup favouritism and prosocial response: the native population in the face of immigration. Revista De Psicologia Social, 29 (1), 60–89. https://doi.org/10.1080/02134748.2013.878574

Wurm, S., Tesch-Römer, C., & Tomasik, M. J. (2007). Longitudinal findings on aging-related cognitions, control beliefs, and health in later life. The Journals of Gerontology Series B Psychological Sciences and Social Sciences, 62 (3), P156–P164. https://doi.org/10.1093/geronb/62.3.P156

Xiang, X., Lu, X., Halavanau, A., Xue, J., Sun, Y., Lai, P. H. L., Wu, Z., SLAC National Accelerator Lab., Menlo Park, CA (United States). (2021). Modern senicide in the face of a pandemic: an examination of public discourse and sentiment about older adults and COVID-19 using machine learning. The Journals of Gerontology Series B Psychological Sciences and Social Sciences, 76 (4), e190–e200. https://doi.org/10.1093/geronb/gbaa128

Zhang, X., Xing, C., Guan, Y., Song, X., Melloy, R., Wang, F., & Jin, X. (2016). Attitudes toward older adults: A matter of cultural values or personal values? Psychology and Aging, 31 (1), 89–100. https://doi.org/10.1037/pag0000068

Zhang, X., Kamin, S. T., Liu, S., Fung, H. H., & Lang, F. R. (2020). Negative self-perception of aging and mortality in very old chinese adults: the mediation role of healthy lifestyle. The Journals of Gerontology Series B Psychological Sciences and Social Sciences, 75 (5), 1001–1009. https://doi.org/10.1093/geronb/gby136

Zhou, Y., Jiao, P., & Zhang, Q. (2017). Second-party and third-party punishment in a public goods experiment. Applied Economics Letters, 24 (1), 54–57. https://doi.org/10.1080/13504851.2016.1161709

Download references

Author information

Authors and affiliations.

School of Psychological and Cognitive Sciences, Peking University, Beijing, People’s Republic of China

Gu Ma, Zizhuo Chen, Wanhua Zou & Xin Zhang

Beijing Key Laboratory of Behavior and Mental Health, Peking University, Beijing, People’s Republic of China

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Xin Zhang .

Ethics declarations

Conflict of interest.

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Additional information

Publisher’s note.

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

Supplementary information

Below is the link to the electronic supplementary material.

(DOCX 27.9 KB)

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Ma, G., Chen, Z., Zou, W. et al. To help or not: negative aging stereotypes held by younger adults could promote helping behaviors toward older adults. Curr Psychol 43 , 1041–1051 (2024). https://doi.org/10.1007/s12144-023-04371-0

Download citation

Accepted : 03 February 2023

Published : 18 February 2023

Issue Date : January 2024

DOI : https://doi.org/10.1007/s12144-023-04371-0

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

• Aging stereotypes
• Third-party punishment
• Find a journal
• Publish with us
• Track your research

• Research Note
• Open access
• Published: 29 August 2023

The role of perspective-taking in suppressing stereotypes about mathematics

• Mana Yamamoto 1 &
• Takashi Oka 2  

BMC Research Notes volume  16 , Article number:  189 ( 2023 ) Cite this article

667 Accesses

Metrics details

When people attempt to suppress stereotypes, they often end up making stereotypical judgments. The adverse effects of this form of suppression are called “paradoxical effects.” This study examined the effect of perspective-taking as a strategy to reduce the paradoxical effects related to stereotype suppression. Specifically, this study addressed stereotypes within the context of women’s mathematical abilities, with Japanese university students as participants. It was predicted that when participants suppressed the stereotype of a woman, those who engaged in perspective-taking toward that woman would make less stereotypical judgments of other women, compared with those who did not. Moreover, as this study focuses on gender stereotypes, an exploratory analysis was conducted to investigate whether the effects of engaging in perspective-taking about women vary depending on the participants’ gender.

Although no significant effect was observed and the hypothesis was not supported, and while the results of this study were statistically inadequate, they suggest that among the female participants, those who did not engage in perspective-taking showed the paradoxical effects of stereotype suppression. However, those paradoxical effects were not observed among those who performed perspective-taking.

Peer Review reports

Introduction

When people suppress stereotypes, their stereotypical thinking is often activated, making them more likely to make stereotypical judgments. The adverse effects of this suppression are called paradoxical effects (e.g., [ 1 ]; as a review, [ 2 ]). For instance, participants who were instructed to suppress stereotypic thoughts produced more stereotypic descriptions and altered their behavior toward members of the target social category [ 1 ].

Previous studies have shown that handy replacement thoughts (e.g., if a white bear comes to mind, think of a red Volkswagen) may decrease paradoxical effects (e.g., [ 3 ]). It has been shown that replacement thoughts that are person-related in content and have high accessibility, and do not require cognitive resources, may be effective in stereotype suppression (e.g., [ 4 ]).

We propose that engaging in perspective-taking when people suppress stereotypes may facilitate the use of effective replacement thoughts. Perspective-taking is the process of imagining oneself in another person’s shoes and envisioning the world from their perspective [ 5 ]. Previous studies have shown that during perspective-taking, an individual’s self-concept gets activated (e.g., [ 5 ]); this activation does not require cognitive resources, as it is an unconscious process [ 6 ]. The present study predicted that when people engage in perspective-taking while suppressing stereotypes, their self-concept becomes a highly accessible replacement thought that does not require cognitive resources; this causes paradoxical effects of a lower degree than when they do not engage in perspective-taking.

Aims and hypotheses

The objective of this study was to examine the effectiveness of perspective-taking as a strategy to decrease the paradoxical effects of stereotype suppression. Galinsky and Moskowitz [ 7 ] conducted a comparison of stereotype suppression and perspective-taking conditions; however, they did not examine the effects of perspective-taking during suppression directly. Therefore, the present study examined the role of perspective-taking during stereotype suppression.

This study focused on gender stereotypes, specifically assessing the stereotype that “women can’t do math” (e.g., [ 8 , 9 ]). Previous research has demonstrated that gender stereotypes can vary across cultures (e.g., [ 10 ]). The current study targeted Japanese university students because Japan, as a cultural context, has demonstrated the existence of stereotypes related to female mathematical competence (e.g., [ 9 ]). Furthermore, it is noteworthy that Japan ranks low on the Global Gender Gap Report [ 11 ].

Therefore, this study examined whether engaging in perspective-taking about a target woman when suppressing stereotypes about the woman’s mathematical ability would decrease the extent of subsequent stereotypical judgments. We hypothesized that the paradoxical effects would be reduced when perspective-taking is performed during stereotype suppression, compared to when it is not performed. Footnote 1 Furthermore, as this study addresses stereotypes about women, we also conducted an exploratory examination of whether the effects of engaging in perspective-taking about women differ by gender.

Design. This study had a 2 (stereotype suppression: suppression vs. non-suppression) × 2 (perspective-taking: taking vs. non-taking) factor between-participant design. The dependent variables were ratings of a woman’s mathematic ability and estimate of her mathematics test score.

Participants. The participants included 329 Japanese university students. However, one participant who did not follow the task instructions for the manipulation of the independent variables was excluded from the analysis. Therefore, 328 participants were finally included in the analysis (196 males, 126 females, and 6 did not disclose their gender; M age = 19.20, SD  = 1.30).

This study was approved by the Research Ethics Committee of the College of Humanities and Sciences, Nihon University (Approval Number: 02–53); it was conducted in 2021 and 2022. Informed consent was obtained from all participants. All methods were performed in accordance with the relevant guidelines and regulations.

Procedure. Participants completed an online questionnaire and were randomly assigned to one of four conditions: (2 [stereotype suppression: suppression vs. non-suppression] × 2 [perspective-taking: taking vs. non-taking]). Following Macrae et al. [ 1 ], the task of this experiment included presenting a photograph of the target woman to the participants and asking the participant to respond to items regarding the woman’s mathematical abilities. First, the participants completed a task for manipulating the independent variables. In this task, the participants were presented with a photograph of a woman, asked to imagine this person taking a math class, and then asked to freely describe what they imagined. For the stereotype suppression condition, participants were asked to suppress any stereotypes about mathematical ability, whereas no such instruction was given for the non-suppression condition. For the perspective-taking condition, participants were asked to engage in perspective-taking about the person in the photo—to put themselves in their shoes—whereas no such instruction was given for the non-perspective-taking condition. Participants then answered questions to a manipulation check. They responded to one item for the manipulation check regarding stereotype suppression (“When responding about the person in the photo, did you make an effort to avoid thinking in terms of prejudice?”), and to two items for a manipulation check of perspective-taking (“Did you imagine how you would feel if you were in that position?” and “Did you try to put yourself in that person’s shoes?”). Participants responded to the questions using a 7-point scale.

Next, the participants completed a task used to measure the dependent variables. They were presented with a picture of a different woman from the previous task and asked to rate the degree to which she “appears to be unable to do mathematical calculations,” “appears to be bad at mental arithmetic,” and “appears to be weak with numbers,” using a 7-point scale. Participants were also asked to estimate the woman’s math test score on a scale from 0 to 100. Finally, they were asked to indicate their age and gender.

Table  1 shows the means and standard deviations of the variables and correlations among the variables. The rating score of the woman’s mathematical ability included the mean of the three items: “appear to be unable to do mathematical calculations,” “appears to be bad at mental arithmetic,” and “appears to be weak with numbers” (Cronbach’s  α  = .833).

First, to confirm that the stereotype suppression and perspective-taking instructions were effective, we conducted Welch’s t -test on the manipulation check score for stereotype suppression. The score was significantly higher in the stereotype suppression condition ( M  = 5.14, SD  = 1.56) than in the non-suppression condition ( M  = 3.84, SD  = 1.94, t (311.17) = 6.68, p  < .001, d  = 0.74). The manipulation check score for perspective-taking was the mean for the two items ( r  = .536, p  < .001). The score for the perspective-taking condition ( M  = 5.15, SD  = 1.37) was significantly higher than that for the non-perspective-taking condition ( M  = 4.19, SD  = 1.59, t (323.62) = 5.89, p  < .001, d  = 0.65). These results indicate that both the stereotype suppression and perspective-taking manipulations were successful.

Next, we examined the effect of perspective-taking on stereotype suppression by conducting a 2 (stereotype suppression: suppression vs. non-suppression) × 2 (perspective-taking: taking vs. non-taking) factor between-participant analysis of variance (two-way ANOVA) with the rating score of the woman’s mathematical ability as the dependent variable. The results showed no significant effects. We also conducted a similar ANOVA with estimated score of the woman’s mathematics test score as the dependent variable, but found no significant effect.

To determine whether the effect of perspective-taking differed by gender, we added gender as a factor and conducted a 2 (stereotype suppression: suppression vs. non-suppression) × 2 (perspective-taking: taking vs. non-taking) × 2 (participant gender: male vs. female) factor between-participant analysis of variance (three-way ANOVA). Analysis of the rating score of the woman’s mathematical ability as the dependent variable showed no significant effects. As the results of the analysis with estimated score of the woman’s mathematics test score as the dependent variable showed a significant interaction effect among the three factors (Table  2 , F (1, 314) = 4.15, p = .043, η 2 p = .013), a simple interaction test of stereotype suppression × perspective-taking was conducted on each female and male participant. As the results showed a significant trend in the interaction effect for stereotype suppression × perspective-taking among females ( F (1, 314) = 2.91, p = .089, η 2 p = .023), a simple-simple main effect test revealed that in the non-taking condition for females, the score was lower in the stereotype suppression condition than in the non-suppression condition ( F (1, 314) = 4.09, p = .044, η 2 p = .058). Among males, the simple interaction effect of stereotype suppression × perspective-taking was not significant.

This study examined whether perspective-taking reduces paradoxical effects when suppressing stereotypes. The results of our analysis showed no significant effect, and the hypothesis was not supported. There are two possible reasons why the hypothesis was not supported in this study. First, explicit instructions were provided for stereotype suppression and measurement. Specifically, it is conceivable that the effects of stereotype suppression manipulation could have persisted in the subsequent stereotype measurement task. In the future, it will be necessary to enhance the segregation between the suppression task and subsequent stereotype measurement task through the incorporation of filler tasks and refined instructions. Moreover, it is plausible that the measurement of stereotype-based judgments could have been subject to an avoidance of negative responses due to social desirability concerns. Moving forward, it is imperative to incorporate measurement tasks for stereotype-based judgments that minimize the impact of social desirability. Indeed, previous research has shown that explicit and implicit gender stereotypes have been shown to make different predictions regarding performance related to mathematics (e.g., [ 14 , 15 ]). Therefore, in the future, it becomes essential to utilize assessment tasks such as lexical decision tasks or the Implicit Association Test (IAT) that are less susceptible to the influence of participants’ intentions and consciousness. Furthermore, a comparative analysis between explicit and implicit responses would be necessary, facilitating a more comprehensive understanding by delving into latent indicators less influenced by participants’ awareness and intentions. Second, this study presented facial photographs of the target individuals using online forms. Compared to face-to-face contact, it is possible that the activation of stereotypes and engagement in perspective-taking were less likely to occur.

As the present study addressed stereotypes about women, we also conducted an exploratory analysis of whether the effects of perspective-taking about women differed by gender. The results did show a significant trend; the estimated math test scores of the female participants who did not perform perspective-taking were lower when they engaged in stereotype suppression than when they did not. This result is consistent with the previous finding that stereotype suppression leads to paradoxical effects (e.g., [ 1 ]). However, among the female participants who engaged in perspective-taking, there was no significant difference between the stereotype suppression and non-suppression conditions. These results suggest that, for women, perspective-taking may reduce the paradoxical effects of suppressing stereotypes about women’s mathematical abilities. Although the results of this study showed a significant trend, distinct findings emerged for male and female participants. Future studies should examine how the role of perspective-taking differs depending on the relationship between the target group and the perspective taker.

Limitations

This study has two notable limitations. First, it proposed that replacement thoughts would arise from the self-concept activated by perspective-taking; however, our results do not specify the content of those replacement thoughts. In future studies, the content of the replacement thoughts actually used should be examined. Second, this study presented outcomes exclusively derived from participants in Japan. Given the potential for stereotypes to exhibit cultural variations (e.g., [ 10 ]), it is crucial to account for cultural factors when interpreting and extending the implications of the results. In the future, it is necessary to conduct investigations that also consider cultural factors.

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

In this study, self-esteem was measured as an individual difference in self-concept, using the 10-item Japanese version of Rosenberg’s [ 12 ] Self-Esteem Scale [ 13 ]. Analysis of covariance was performed with the self-esteem scores set as a covariate, but the results were similar to those of this study, which did not set a covariate.

Abbreviations

Analysis of variance

Implicit Association Test

Macrae CN, Bodenhausen GV, Milne AB, Jetten J. Out of mind but back in sight: Stereotypes on the rebound. J Pers Soc Psychol. 1994;67(5):808–17. https://doi.org/10.1037/0022-3514.67.5.808 .

Article   Google Scholar  

Monteith MJ, Sherman JW, Devine PG. Suppression as a stereotype control strategy. Pers Soc Psychol Rev. 1998;2(1):63–82. https://doi.org/10.1207/s15327957pspr0201_4 .

Article   CAS   PubMed   Google Scholar  

Wegner DM, Schneider DJ, Carter SR, White TL. Paradoxical effects of thought suppression. J Pers Soc Psychol. 1987;53(1):5–13. https://doi.org/10.1037/0022-3514.53.1.5 .

Yamamoto M, Oka T. The role of counter-stereotypes and non-dominant stereotypes in suppressing stereotypes. The Japanese J Psychol. 2018;89(1):12–21. https://doi.org/10.4992/jjpsy.89.16020 .

Galinsky AD, Ku G, Wang CS. Perspective-taking and self-other overlap: Fostering social bonds and facilitating social coordination. Group Process Intergroup Relat. 2005;8(2):109–24. https://doi.org/10.1177/1368430205051060 .

Davis MH, Conklin L, Smith A, Luce C. Effect of perspective taking on the cognitive representation of persons: A merging of self and other. J Pers Soc Psychol. 1996;70(4):713–26. https://doi.org/10.1037/0022-3514.70.4.713 .

Galinsky AD, Moskowitz GB. Perspective-taking: Decreasing stereotype expression, stereotype accessibility, and in-group favoritism. J Pers Soc Psychol. 2000;78(4):708–24. https://doi.org/10.1037/0022-3514.78.4.708 .

Else-Quest NM, Hyde JS, Linn MC. Cross-national patterns of gender differences in mathematics: A meta-analysis. Psychol Bull. 2010;136(1):103–27. https://doi.org/10.1037/a0018053 .

Article   PubMed   Google Scholar  

Morinaga Y. Women can’t do math: A review of the social psychological literature on the effects of negative stereotypes on both women and girls. Japanese Psychol Rev. 2017;60(1):49–61. https://doi.org/10.24602/sjpr.60.1_49 .

Moè A, Hausmann M, Hirnstein M. Gender stereotypes and incremental beliefs in STEM and non-STEM students in three countries: Relationships with performance in cognitive tasks. Psychol Res. 2021;85:554–67. https://doi.org/10.1007/s00426-019-01285-0 .

World Economic Forum. The Global Gender Gap Report. 2023. https://www3.weforum.org/docs/WEF_GGGR_2023.pdf . Accessed 7 Aug 2023.

Rosenberg M. Society and the adolescent self-image. Princeton, NJ: Princeton University Press; 1965.

Book   Google Scholar  

Yamamoto M, Matsui Y, Yamanari Y. The structure of perceived aspects of self. The Japanese J Educ Psychol. 1982;30(1):64–8. https://doi.org/10.5926/jjep1953.30.1_64 .

Nosek BA, Banaji MR, Greenwald AG. Math = male, me = female, therefore math ≠ me. J Pers Soc Psychol. 2002;83(1):44–59. https://doi.org/10.1037/0022-3514.83.1.44 .

Guizzo F, Moè A, Cadinu M, Bertolli C. The role of implicit gender spatial stereotyping in mental rotation performance. Acta Psychol. 2019;194:63–8. https://doi.org/10.1016/j.actpsy.2019.01.013 .

Download references

Acknowledgements

Not applicable.

This work was supported by the Japan Society for the Promotion of Science KAKENHI (20K14141).

Author information

Authors and affiliations.

College of Commerce, Nihon University, Tokyo, Japan

Mana Yamamoto

College of Humanities and Sciences, Nihon University, Tokyo, Japan

Takashi Oka

You can also search for this author in PubMed   Google Scholar

Contributions

MY and TO conceived and designed the study. MY conducted data gathering, performed statistical analyses, and wrote the first draft of the manuscript. MY and TO contributed to manuscript revision and approved the final draft.

Corresponding author

Correspondence to Mana Yamamoto .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Ethics approval and consent to participate

This study was approved by the Research Ethics Committee of the College of Humanities and Sciences, Nihon University (Approval Number: 02–53). Informed consent was obtained from all participants.

Consent for publication

Additional information, publisher’s note.

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

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/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Yamamoto, M., Oka, T. The role of perspective-taking in suppressing stereotypes about mathematics. BMC Res Notes 16 , 189 (2023). https://doi.org/10.1186/s13104-023-06460-6

Download citation

Received : 03 May 2023

Accepted : 14 August 2023

Published : 29 August 2023

DOI : https://doi.org/10.1186/s13104-023-06460-6

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

• Stereotype suppression
• Paradoxical effects
• Replacement thoughts
• Perspective-taking

BMC Research Notes

ISSN: 1756-0500

• Submission enquiries: [email protected]
• General enquiries: [email protected]

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
• Int J Environ Res Public Health

Societal Age Stereotypes in the U.S. and U.K. from a Media Database of 1.1 Billion Words

1 Lee Kuan Yew School of Public Policy, National University of Singapore, Singapore 119260, Singapore; gs.ude.sun@gnrpps

2 Lloyd’s Register Foundation Institute for the Public Understanding of Risk, National University of Singapore, Singapore 119260, Singapore

Associated Data

Data are publicly available online at https://www.english-corpora.org (accessed on 22 July 2021).

Recently, 194 World Health Organization member states called on the international organization to develop a global campaign to combat ageism, citing its alarming ubiquity, insidious threat to health, and prevalence in the media. Existing media studies of age stereotypes have mostly been single-sourced. This study harnesses a 1.1-billion-word media database comprising the British National Corpus and Corpus of Contemporary American English—with genres including spoken/television, fiction, magazines, newspapers—to provide a comprehensive view of ageism in the United Kingdom and United States. The US and UK were chosen as they are home to the largest media conglomerates with tremendous power to shape public opinion. The most commonly used synonym of older adults was identified, and its most frequently used descriptors were analyzed for valence. Such computational linguistics techniques represent a new advance in studying aging narratives. The key finding is consistent, though no less alarming: Negative descriptions of older adults outnumber positive ones by six times. Negative descriptions tend to be physical, while positive ones tend to be behavioral. Magazines contain the highest levels of ageism, followed by the spoken genre, newspapers, and fiction. Findings underscore the need to increase public awareness of ageism and lay the groundwork to design targeted societal campaigns to tackle ageism—one of our generation’s most pernicious threats.

1. Introduction

Proposed by Butler [ 1 ], the concept of ageism was originally conceived as comprising three distinct but interrelated components: prejudicial attitudes, discriminatory behavior, and institutional policies. Today, the term ‘ageism’ is more commonly understood as the stereotyping, prejudice and discrimination of people on the grounds of age [ 2 ]. In 2016, 194 World Health Organization (WHO) member states called on the international organization to develop a global campaign to combat ageism, citing its alarming ubiquity and insidious threat to health [ 3 ]. However, the ongoing COVID-19 crisis has brutally exposed the reality of ageism in society. Both older and younger people have been subject to various stereotypes in mainstream discourse; the constant homogenization of older adults as frail and vulnerable [ 4 ] has spawned distasteful references to the virus as a ‘Boomer Remover’ [ 5 ], while young people have been depicted as selfish and even blamed for spikes in COVID-19 cases [ 6 ].

Although ageism can be directed at people of all ages, the present study focuses on ageism towards older adults. Inaccurate and negative stereotypes of older adults have unfortunately resurfaced and intensified during the outbreak. Seeking to quantify the economic cost of ageism, Levy and colleagues [ 7 ] combined the impact of predictors of ageism—negative age stereotypes and self-perceptions of aging—with healthcare spending data for the eight most expensive health conditions ailing Americans aged 60 and above. They found that the one-year cost associated with ageism in the United States was $63 billion. It is therefore important to understand the types of age stereotypes that exist in society. Against this backdrop, the present study leverages a corpus of 1.1 billion words from multiple sources—newspapers, magazines, books, and television/radio transcripts—and applies computational linguistics to analyze the kinds of age stereotypes that prevail in the media.

Stereotypes refer to widely held beliefs about members belonging to a particular social group. Age stereotypes often include oversimplified generalizations about how people at, below or over a certain age ought to behave without any consideration of individual differences [ 8 ]. Stereotypes of old age can be positive, neutral, or negative. Positive age stereotypes refer to favorable beliefs about older adults (e.g., wise, kind, etc.), while negative age stereotypes refer to unfavorable beliefs about older people (e.g., slow, cranky, etc.). Past research has established that older adults who embrace positive stereotypes about aging are more likely to have better health outcomes than those who hold negative age stereotypes [ 9 ].

The role of the media in perpetuating age stereotypes has been widely acknowledged [ 10 ]. As a major agent of socialization, the media serves as a powerful avenue to understand how old age is socially constructed. Cultivation theory indicates that repeated exposure to the media can shape individuals’ perceptions of reality [ 11 ]. Portrayals of older adults in the media have the potential to reinforce stereotypes and shape individuals’ attitudes toward the aging process, ultimately predicting health outcomes in later life [ 12 ]. Media representations can be either visual or discursive in nature. The concept of ‘visual ageism’ refers to the practice of visually underrepresenting older adults or presenting them in a prejudiced manner [ 13 ]. Examples of visual ageism include the portrayal of older adults as lacking positive traits, as well as the non-realistic or exaggerated portrayal of this cohort [ 13 ]. Our study focuses on the discursive representations of older adults. It is well established in the social sciences that prejudice contains a strong discursive dimension [ 14 ]. A study on the discursive representations of older people will provide insight into the kinds of socially shared patterns of thought and behavior [ 14 ] that govern attitudes toward members of this group.

Our study is significant in several ways. Conceptually, this is one of the first known studies to use large-scale and multi-sourced databases to provide a comprehensive view of ageism in the US and UK. Previous content analyses of societal ageism are mostly singled-sourced (e.g., poetry or fiction)—more details elaborated on below. We chose the US and UK as they are home to the world’s largest media conglomerates [ 15 ] and therefore have tremendous power to shape public opinion [ 11 ]. Practically, the detailed and timely content analysis of ageism enables targeted messaging to counteract it. Of broader significance, this study creates an unprecedented platform for social scientists and policy makers to study societal perceptions that complements traditional methods of focus groups and surveys. We synthesize the literature on the impact of age stereotypes on well-being and the conceptual framework of age stereotypes.

1.1. Impact of Aging Stereotypes in Physical and Mental Well-Being

Levy and Langer [ 16 ] investigated attitudes towards aging and the effects of negative aging stereotypes on memory loss in American deaf and older adults in China who were previously unexposed or minimally exposed to negative age stereotypes from the media, and hearing Americans who are constantly exposed to negative aging stereotypes through various media channels [ 17 ]. As predicted, the Chinese participants reported the most positive attitudes toward aging, following by the American deaf and the American hearing sample. The same rank order was also observed in the memory scores of these three groups. These results show that negative aging stereotypes have a detrimental effect on memory.

A survey of older adults from 65–95 years found that negative age stereotypes were associated with decreasing self-reported sense of responsibility toward others, subjective health, adherence to medical appointments, participation in community social activities, and regular physical activity [ 18 ]. Overall, negative aging stereotypes were associated with lower overall quality of life for older adults.

On the other hand, positive aging stereotypes resulted in a variety of positive outcomes. Researchers primed either negative or positive stereotypes in a group of older adults above 62 years and put them through mathematical and verbal tasks [ 19 ]. The positive stereotype group recorded significantly lower cardiovascular stress as evidenced by lower systolic and diastolic blood pressure, heart rate, and skin conductance. Another study found that older adults with more positive age stereotypes had better screened hearing after 36 months [ 20 ]. Positive age stereotypes were also linked to 7.5 years of life among participants of the Ohio Longitudinal Study of Aging and Retirement compared to those who espoused negative views [ 21 ]. In essence, these landmark studies found that negative age stereotypes are linked to poorer physical and psychological health, highlighting the importance of studying the media origins of these stereotypes to counteract them.

1.2. Conceptual Framework and Measurement of Age Stereotypes

We synthesize the extant literature on the conceptual framework and measurement of age stereotypes to highlight the need for more rigorous datasets and methods that our study provides. We build on the conceptual framework that the multiple dimensions of age stereotypes can be grouped by valence and theme [ 22 , 23 ]. With regard to valence, age stereotypes can either be positive or negative. For example, travel advertisements typically portray negative images of grumpy and listless older adults who could be transformed into beaming “golden agers” (positive image) basking in the Florida sun by buying their packages [ 24 ]. With respect to theme, age stereotypes often involve physical and/or behavioral attributes [ 25 ]. For example, companies often portray older individuals as suffering from physical problems, such as the inability to perform activities of daily living, to market their supplements and solutions [ 26 ].

On the measurement front, direct and indirect strategies are used. Direct methods include interviews and surveys methods where participants are asked directly about their perceptions through telephone interviews [ 27 ], established scales like the Palmore Fact of Aging Quiz [ 28 ], and Implicit Association Tests (IAT) to understand stereotypical associations at the level of automatic cognition [ 29 ]. A key study across 26 countries/territories surveyed 3435 college students and found that negative perceptions of older adults are similar across different countries in the domains of physical attractiveness, ability to perform everyday tasks, and learning new things [ 30 ].

While Direct Method studies have contributed significantly to understanding age stereotypes, there are limitations. Firstly, survey studies sometimes use convenience sampling of college students to draw country-level conclusions—the ecological fallacy [ 31 , 32 ]. Secondly, survey studies typically use validated scales that were designed years ago, capturing the content of ageism at the time they were designed, and may not reflect the current reality. The broader point is the need for indirect methods that take a bottom-up approach of capturing the content of age stereotypes, to complement the top-down approach of validated scales.

Indirect methods to investigate ageism include ethnographic approaches [ 33 ], analysis of fiction [ 34 ], images from literature [ 35 ], poetry [ 36 ], and art [ 37 ]. For example, a study examined 150 age-specific birthday cards and found that textual messages tend to be more ageist than pictures [ 38 ]. Of the cards with textual messages, 66.7 percent represented ageing negatively. There are allusions to cognitive decline like “Oh you’re 50? You don’t even remember what I’m talking about do you? Oh well, happy 50th birthday anyway”. Other cards, though whimsical, sarcastically scorned physical decline “As you’re entering your 50s new doors will open for you. Geriatric crisis center, cosmetic surgery clinic, office of ageing.” In-depth interviews also found that ageism existed in the advertising industry, with older workers perceived by young workers as “dead wood” [ 39 ]. Overall, the results are remarkably similar: Negative stereotypes are highly prevalent, compared to positive stereotypes. Against this backdrop, we hypothesize that negative age stereotypes will outnumber positive/neutral ones (Hypothesis 1), as shown in previous studies, albeit single sourced.

There are nuances across sources. Older men are hardly present in top men’s magazines in the US; women’s magazines feature mainly younger women, although older women form the majority of their readers [ 40 , 41 ]. In the UK, advertisements in magazines that target an older audience typically associate older adults with food items and hearing aids. However, positive stereotypes surface occasionally in magazines; for instance, older adults were portrayed as sexually attractive with ‘optimal sexual engagement’ in the dating sections of Canadian magazines [ 42 ], though magazines typically portrayed older adults negatively in both content [ 43 ] and advertisements [ 44 ]. We hypothesize that among popular genres (spoken/TV, fiction, popular magazines, newspapers), magazines contain the highest ageism scores (Hypothesis 2).

These indirect and bottom-up methods are helpful in understanding societal ageism as cultivation theory states that different forms of media reflect societal perceptions they seek to portray [ 45 ]. However, most studies are single-sourced, anecdotal, and run the risk of cherry-picking sources to support a scientific agenda [ 46 ]. Importantly, single-sourced studies do not provide strong evidence for a targeted media campaign at the societal level. The focus on single sources is likely due to the lack of suitable large-scale datasets and methods to interrogate them. Recent grants by the National Endowment for the Humanities (NEH) and the National Science Foundation (NSF) provided unprecedented corpora, and scholars have adapted computational linguistics methods to analyze these databases for societal stereotypes [ 47 , 48 , 49 , 50 , 51 , 52 ], which we will leverage for the current study.

2. Materials and Methods

2.1. datasets.

The British National Corpus (BNC) is compiled by Oxford University’s Computing Services. It is a 100-million-word corpus that spans the 1980s to 1993 and represents a wide cross-section of British English, both spoken and written. The written part of the BNC (90%) includes extracts from regional and national newspapers, specialist periodicals and journals for all ages and interests, academic books, popular fiction, published and unpublished letters and memoranda, school and university essays, among many other kinds of text. The spoken part (10%) consists of orthographic transcriptions of unscripted informal conversations (recorded by volunteers selected from different age, region, and social classes in a demographically balanced way) and spoken language collected in different contexts, ranging from formal business or government meetings to radio shows and phone-ins. Work on building the corpus began in 1991, culminating in the latest third edition BNC XML in 2007.

The Corpus of Contemporary American English (COCA) is the largest, balanced corpus of contemporary American English. It contains more than one billion words of text, including 20 million words each year from 1990, and it is equally divided among spoken, fiction, popular magazines, newspapers, and academic texts. The even distribution of various sources gives the corpus its balance. The corpus is also updated every six to nine months, therefore serving as a unique record of linguistic changes in American English [ 53 ].

2.2. Measurement of Age Stereotypes

We analyzed the prevalence of various synonyms in the datasets. ‘Elderly’ is the most commonly used synonym with the highest prevalence of 22.2 per million. Other synonyms evidenced a markedly lower prevalence, for example, ‘old people’ had a prevalence of 0.72 per million. This imbalance corroborated previous findings that older adults are labelled negatively in the media [ 51 ]. While we acknowledge the negative connotation of ‘elderly’, and the movement against it [ 54 ], it was used as a target synonym as it provided sufficient data for meaningful analysis. This however presents limitations that we address in the paper.

We compiled the top 20 words that co-occurred most frequently with the target word, known as collocates, with the following inclusion criteria: (a) Lexical Proximity: collocate present within four words prior or after the target word. Articles such as ‘the’, ‘a’ were not included in the six-word lexical span. If the target noun was the first word of a sentence, the collocates from the prior sentence were excluded.; (b) Relevant context: collocate referred to specifically to an old person (checked by two raters); (c) Mutual Information (MI) Score of three and above: collocate had a stronger association with the respective synonym than other words in the corpus indicating semantic bonding [ 50 , 55 ]. MI score estimates word association norms directly from the corpus. It is calculated via sentiment analysis, which shows the mutual information between collocates and target words. The higher the MI value, the closer the relationship between the collocate and target word. The MI value is calculated using the formula:

‘ A ’ indicates the possibility of the target word A appearing, which is calculated by the frequency of the target word. ‘ B ’ indicates the possibility of the collocate B appearing, which is calculated by the frequency of word B. ‘ C ’ indicates the possibility of ‘ A ’ and ‘ B ’ appearing together, which is calculated by the frequency of collocate B appearing near the target word A. ‘ SizeCorpus ’ refers to the size of corpus or the number of words. Span is the span of words (e.g., if there are 6 words to the left and 6 words to the right of the target word, span = 12). log (2) = 0.30103. This is a well-established application of computational linguistics to study stereotypes in other studies [ 47 , 48 , 49 , 50 , 51 , 52 ].

Thereafter, each collocate that met the study criteria was rated in binary fashion: 1 (negative) and 0 (positive/neutral) to test Hypothesis 1. Each collocates was also rated on a scale from 1 (very positive) to 5 (very negative) for Hypothesis 2—a method found to be valid and reliable to measure age-stereotype associated words [ 16 ]. For example, ‘frail’ and ‘dementing’ are rated as very negative while ‘caring’ and ‘happy’ are rated as very positive. Ratings were done by two independent gerontology researchers—consistent with similar corpus studies [ 47 , 48 , 49 , 50 , 51 , 52 ]—and the inter-rater reliability using Cronbach’s alpha was 0.972 (95% CI: 0.946, 0.986). Age stereotype scores were created by calculating the mean of all scores for the respective corpora using this mixed-method approach that integrated the rigor of text analytics and qualitative depth as exemplified in the coding process. Table 1 presents sample descriptors of older adults that are negative, neutral, or positive.

Examples of Negative, Neutral and Positive Descriptors of Older Adults in the Media.

2.3. Analytic Strategy

To test Hypothesis 1—negative age stereotypes will outnumber positive/neutral ones—we ran a chi-square goodness-of-fit to show that the actual distribution of negative age stereotypes was significantly different from the expected distribution. Hypothesis 2 stated that among popular genres (spoken/TV, fiction, magazines, newspapers), magazines contain the most negative age stereotype scores. We ran a two-way ANOVA to test the interaction between corpus and genre; specifically, the main effect of genre will be used to test Hypothesis 2. All data pre-processing, text analytics and statistical analyses were conducted in Python 3.7. and OriginPro 2019b.

The size of conversations in the UK about older adults (48.96 words per million) is significantly higher compared to the US (27.16 words per million), T 7 = −3.77, p = 0.007. In the UK, the highest prevalence of conversations related to older adults appeared in newspapers at 60.96 words per million while in the US, it was magazines at 30.99 words per million. As hypothesized, a chi-square goodness-of-fit achieved significance, X 2 (1) = 5.44, p = 0.02, showing that the actual distribution of negative age stereotypes was significantly different from the expected distribution. Across the genres, there are six times more negative descriptors of older adults compared to positive/neutral ones, supporting Hypothesis 1. Table 2 shows the top 10 descriptors/collocates of older adults.

Top 10 Descriptors of Elders in the UK and US in a 1.1-billion-word corpus of popular media.

Notes : 1 Top collocates of “elderly” in the British National Corpus ranked by mutual information score. 2 Top collocates of “elderly” in the Corpus of Contemporary American Corpus ranked by mutual information score.

The country and genre interaction effect did not reach significance. US evidenced significantly higher negative stereotypes than UK in the corpus, F (1, 28) = 4.75, p = 0.038. As hypothesized, ageism across genre also reached significance, F (3, 28) = 4.69, p = 0.009, with the highest ageism in magazines (M = 4.80, SD = 0.45), followed by spoken (M = 4.60, SD = 0.74) and newspapers (M = 4.35, SD = 1.33), with the least in fiction (M = 3.00, SD = 1.27), providing support for Hypothesis 2.

4. Discussion

This study used large-scale (1.1 billion words) and multi-sourced genres of spoken/TV, fiction, magazines, newspapers to provide a comprehensive view of ageism in the US and UK and employed a technique that represents new advances in the study of aging narratives and stereotypes. Collectively, both countries are home to the world’s largest media conglomerates [ 15 ] and have outsized power to shape public opinion [ 11 ]. Specifically, we contributed to the indirect method of analyzing age stereotypes that have been primarily single-sourced. The key finding is consistent, though no less alarming: Negative age stereotypes outnumber positive ones by six times. Negative stereotypes tend to be physical (e.g., frail), while positive ones tend to be behavioral (e.g., caring). Magazines contained the highest levels of ageism, followed by the spoken genre (e.g., TV and radio talk shows).

The content of negative age stereotypes in the present study, replicated findings from many single-sourced studies such as fiction [ 34 ], art [ 37 ], and poetry [ 36 ]. The prevalence of negative age stereotypes could be a result of coverage bias where negative events and narratives tend to gain more attention than positive ones [ 56 ]. In any case, the prevalence of negative age stereotypes across all media genres of our dataset is instructive of society’s perceptions of ageing as the media reflect societal perceptions they seek to portray [ 40 , 45 ].

Our study made the following contributions. First, negative social realities of ageing are widespread—across multiple media genres—and not only confined within advertisements or fiction. Ageism in the traditional media like newspapers could have an amplification effect as the negative stereotypes are funneled through the echo chambers of social media platforms that draw widely from these news sources. Drilling further, the prevalence of ageism across magazines and newspapers signals the possibility of an implicit age bias in journalists. While scholars have looked at how implicit racial [ 57 ] and gender bias [ 58 ] can emerge in the reporting and writing process, little is known about whether journalists carry implicit age biases—an area that warrants further scrutiny in research. Nonetheless, evidence suggests that journalists who are aware of their cognitive biases tend to produce less biased work than those who are not [ 58 ]. Self-awareness exercises could be held to ensure that journalists make the conscious effort to curb their age biases to prevent inaccurate stereotypes from being perpetuated. This is particularly crucial given that newspapers and magazines hold a discursive power to propagate ageism.

The fact that magazines contain the highest levels of ageism has serious consequences for the older demographic. A recent study revealed a strong discrepancy in print magazine readership by age, with individuals aged 65 and above twice as likely to read a print magazine as those aged from 18 to 24 in the United States [ 59 ]. The constant exposure to damaging views of aging is likely to result in self-limiting views of the aging process, which will affect the health of older adults [ 12 ].

Thus far, interventions to reduce ageism have been designed at both the interpersonal and societal level. Significant progress has been made at the interpersonal level [ 59 ]. However, at the societal level, campaigns to combat the overwhelmingly negative portrayals of older adults in the media are few, with the most recent being ‘Disrupt Aging Collection’. This campaign presented realistic and positive images of aging and was launched by the American Association of Retired Persons (AARP) in collaboration with Getty Images [ 60 ].

The effectiveness of such laudable media campaigns ultimately depends on the ability to target specific erroneous and negative stereotypes. However, the current state of research regarding media portrayals of older adults is typically single-sourced and does not provide strong evidence for targeted campaigns. Furthermore, one of the most common and insidious ways that ageism is expressed is through language [ 61 ]. As exemplified by the viral phrase ‘Ok, Boomer’, age stereotypes constantly take on new forms—be it through new slurs or lingo. Prejudice cannot be alleviated or understood fully without paying attention to how stereotypes are embedded in language [ 62 ]. As such, it is imperative that policy makers are cognizant of how age stereotypes evolve and manifest in new ways.

Practically, interventions and awareness campaigns could focus on educating journalists as their prose wields significant power to influence societal attitudes. Of broader significance, the study provides a roadmap on where and how to focus anti-ageism campaigns. Prior research has established that educational programs have the potential to reduce ageism [ 63 ]. Both mass media and societal campaigns could therefore include an educational component to equip members of the public with a more accurate understanding of older adults and aging in general. Societal campaigns could target magazines and TV where ageism levels are the highest by framing aging in a way that takes into account both the gains and losses inherent in the aging process [ 52 ]. Although a certain number of older adults are at risk of becoming frail during their later years, there is also evidence indicating that overall physical functioning of older adults has improved over time [ 64 ]. Likewise, even as old age poses a severe risk for COVID-19 mortality, there have been occasions where older adults have recovered from the virus and younger people have not [ 4 ]. Thus, both positive and negative narratives of older people should be featured in mainstream discourse to ensure a more balanced portrayal of the aging process.

From a methodological standpoint, our study provided a proof-of-concept for scholars and policy makers to study societal narratives using corpus data, which complements traditional methods of focus groups and surveys [ 65 , 66 , 67 , 68 , 69 , 70 , 71 , 72 , 73 ]. There is potential for future studies using our datasets. For example, in this COVID-19 pandemic, techniques described in this study could be used to track the societal narratives [ 74 ] around vulnerable groups for ageism and racism. Understanding the content of ageism and racism during a pandemic provides important evidence for designing a targeted campaign to counteract it.

While we sought to circumvent the weaknesses of other studies by providing a multi-sourced platform of 1.1 billion words, this study is not without weaknesses. We did not include social media in our corpus as most major platforms (e.g., Facebook) no longer have open and publicly available data, leaving limited and outdated datasets that may not be comprehensive. Nevertheless, this is a significant weakness as ageism cannot be compared across media genres. A future iteration of the study could include social media for comparative analysis to facilitate advocacy campaigns.

Another issue is the use of ‘elderly’ as a target search term. The term has negative connotations and would therefore attract negative collocates. While we are cognizant of the negative connotations and wholeheartedly support the movement against it [ 54 ], our decision was based on data considerations. ‘Elderly’ was the most common synonym with the highest prevalence of 22.2 per million, over 30 times more than ‘old people’ with a low prevalence of 0.72 per million. Choosing the latter would not generate a sufficient dataset for analysis. Nevertheless, it is unfortunate that a derogatory synonym is the most commonly used term for an older adult and future studies should focus on strategies to reframe ageing positively in the media.

5. Conclusions

In conclusion, this study contributed by detailing how ageism manifests across different genres of media. We found support for our hypotheses that negative age stereotypes in the media outnumber those that are positive or neutral, and that magazines contain the most ageist content. There is an urgent need to study the content of ageism such that societal campaigns can be designed to combat it. We hope to have laid the groundwork to design interventions to tackle societal ageism—one of our generation’s most insidious threats.

Acknowledgments

N. Indran for important research assistance.

We gratefully acknowledge the support by the Social Science Research Council SSHR Fellowship (MOE2018-SSHR-004), and the Lloyd’s Register Foundation IPUR Grant (IPUR-FY2019-RES-03-NG). The funders had no role in study design, data collection, analysis, or writing.

Data Availability Statement

Conflicts of interest.

No conflict of interest to declare.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.