How Disability Stereotypes Shape Memory for Personal Attributes
Abstract
Abstract. Two experiments examined effects of including an information about a disability in a person description on memory about that person’s traits. In Experiment 1, this information impaired correct recognition of traits of a person that had been described in correspondence to gender stereotypes. In Experiment 2, it induced false memories in accordance with stereotypes about people with disabilities. Participants’ false alarms for traits belonging to the dimension of warmth increased, whereas false alarms for traits belonging to the dimension of competence decreased. Thus, activating stereotypes through a disability prime influenced what could be recognized correctly or falsely was assumed to be recognized about a person.
Stereotypes comprise beliefs about members of social groups. Some stereotypes are so prevalent that almost everybody is aware of their content, for example, certain gender stereotypes. Other stereotypes may be more subtle but still impact on how members of a particular group are perceived and treated.
Research on stereotypes about people with disabilities suggests that the presence of disability by itself is a strong characteristic that serves social categorization. It is not so much the kind of disability that defines the respective social group but that there is some disability. When asked to imagine persons with physical or intellectual disabilities, these are described similarly in terms of personality traits (e.g. Fiske et al., 2002; Weinberg, 1976). These stereotypes are so strong that they can even eclipse gender stereotypes. In a study by Nario-Redmond (2010), participants were asked to picture a man and a woman with or without a disability. No more specific instructions were given, but participants were free to imagine whatever and whoever came to their mind spontaneously. Whereas descriptions of men and women without disability corresponded to gender stereotypes with frequent mentions of gender-typical traits, they described men and women with a disability very similarly, most frequently using traits expressing dependency, incompetence, and asexuality.
Person memory is influenced by stereotypes that also can be understood as schemata. Stereotype activation by naming a respective social group can enhance the accessibility of stereotype-consistent information learned about a person before because the memory schema provides additional retrieval routes (Snyder & Uranowitz, 1978). However, if stereotype-inconsistent information had been encoded, this may benefit as well due to its distinctiveness with regard to the schema (Rojahn & Pettigrew, 1992). In addition, a schema can strengthen a tendency for incorrectly guessing or falsely remembering consistent traits (Bellezza & Bower, 1981).
Here, we examined as to how mentioning a disability influences person memory. We assumed this mentioning to activate disability stereotypes that would influence memory for traits of a person described primarily in ways corresponding to gender roles (Experiment 1) as well as false memories with regard to stereotype-consistent or stereotype-inconsistent traits (Experiment 2).
Experiment 1
Participants studied two lists of traits of two fictitious persons (a man and a woman). These traits were either gender-typical or neutral. A recognition test then assessed memory for the studied items. We compared two groups. Only one group was informed that one of the two persons was paraplegic. We expected this disability prime to impair memory for that person.
Method
Participants
The sample comprised 90 participants who were recruited by students at the Ludwigsburg University of Education on campus as well as at the community as a part of a seminar on the psychology of physical disabilities (41 women, 49 men, Mage = 27.6). Participation was voluntary and not rewarded. 52% were university students, 6% were in nonuniversity education, and 41% were employed (one missing value). All participants provided informed consent. The sample size was determined assuming a medium effect size.
Design
Giving a disability prime was manipulated between participants. Only one group of participants was informed that one of the two persons they were learning about was paraplegic.
Material
Two item lists were constructed that comprised either five typically female traits (e.g., sensitive) or typically male traits (e.g., ambitious) as well as six neutral traits (e.g., polite) each (see Appendix A). The recognition tests comprised nine additional traits. Three of these were gender-typical, three were neutral, and three corresponded to disability stereotypes (e.g., brave).
Procedure
The experiment was conducted online. Participants received a hyperlink from the experimenter who had informed them that they were about to participate in a memory experiment and was available for questions. On arriving on the website, initial instructions explained that the purpose of the present study was the investigation of processes in the context of person memory. Participants were to study a series of words characterizing a person and to retrieve the words in a later memory test. By clicking on a button, participants then came to the next page that shortly stated demographic information about either a fictitious man or woman (name, age, and current profession). In addition, it was stated that words would be displayed that represented traits used by coworkers to describe him or her. The items were shown for three seconds each, separated by blank screens of half a second. After all 11 items had been shown, a 2-min distractor phase followed (solving displayed arithmetic problem on a sheet of paper). Subsequently, the memory test started. Participants had to decide whether a displayed word had been studied or not by clicking one of two buttons (old or new). Then, the procedure repeated with the second fictitious person. Again, short demographic information was stated. In addition to this, one group was informed that the person was paraplegic. Study, distractor, and memory test followed in the same manner as before. Two sequences starting with the man or the woman were counterbalanced between participants. At the end, participants indicated their age, gender, and study major or profession.
Results
A 2 (first person, second person) × 2 (gender-typical traits, neutral traits) × 2 (group: disability prime, no disability prime) ANOVA with repeated measures on the first and second factor examined hits for studied items. A significant main effect reflected more hits for traits of the first person, F(1, 88) = 24.47, p < .001, ηp2 = .22. However, a significant interaction, F(1, 88) = 7.08, p = .009, ηp2 = .08, showed that this difference was only reliable in the group with a disability prime for the second person, F(1, 88) = 28.94, p < .001, ηp2 = .25, whereas the group without disability prime did not have significantly fewer hits for the second person, F(1, 88) = 2.61, p = .110 (see Figure 1). A further significant interaction, F(1, 88) = 4.53, p = .036, ηp2 = .05, reflected that slightly more neutral traits than gender-typical traits of the first person were correctly recognized, F(1, 88) = 1.63, p = .205, but slightly more gender-typical traits of the second person, F(1, 88) = 2.10, p = .151. No other interaction or main effect was significant, F < 1.
A 2 (first person, second person) × 3 (gender-typical traits, neutral traits, disability-related traits) × 2 (group: disability prime, no disability prime) ANOVA with repeated measures on the first and second factor examined false alarms. A significant main effect, F(2, 176) = 20.11, p < .001, ηp2 = .19, reflected that there were more false alarms for gender-typical traits than for neutral traits (p < .001) or disability-related traits (p < .001) that did not differ significantly from each other (p = .085). There were not significantly more false alarms for the second person than for the first person, F(1, 88) = 3.29, p = .073, or other significant main effects or interactions, F < 2.57, p > .112.
Discussion
The mentioning of paraplegia impaired memory for traits. This finding suggests that the disability prime conflicted with memory for a person explicitly described as possessing several gender-typical attributes. Research on disability stereotypes has shown that mentioning disability eclipses gender stereotypes (Nario-Redmond, 2010). With regard to stereotypes acting as memory schemata, this implies that gender-consistent schemata could not be used as efficiently anymore after a disability prime. Person memory suffered when disability was mentioned before studying traits of that person.
This impairment was not moderated by trait type, that is, it did not only affect gender-typical traits. However, we hesitate to interpret these results as indicating that memory suffered irrespective of the content. Although this is possible, it could also be that the correspondence with gender stereotypes of the gender-typical items did not preclude participants from associating neutral traits with them as well. Importantly, there were no inconsistent traits nor traits associated to a different stereotype. Moreover, the lack of an effect of stereotype activation on false alarms suggests that it did not involve a general, unspecific memory impairment. The only effect on false alarms was item content. Probably, there were more false alarms for gender-typical traits because of their associations to studied gender-typical traits. This can be considered a kind of false memory. It has been shown before that the activation of gender stereotypes can induce stereotype-consistent false memories (Lenton et al., 2001). Perhaps, a disability prime did not affect false memory, because associations between new and studied words were too strong, whereas disability-related words were not associated with any studied items or simply were not specific enough regarding disability stereotypes. We examined whether an adapted approach might allow to observe an effect of a disability prime on false memories in a further experiment.
Experiment 2
According to the two-dimensional stereotype-content model by Fiske et al. (2002), stereotypes about many social groups can be described according to how high group members are rated on traits belonging to the dimensions of competence and warmth. People with disabilities are characterized as low in competence but high in warmth. Using implicit measures and comparing different contexts, a study by Rohmer and Louvet (2018) demonstrated that persons with disability were associated with less competence when priming a work context. This finding is in line with discriminatory behavior against people with disability in the workplace.
We followed the distinction of warmth and competence and chose traits representing the two dimensions as false memory lures. Participants read two biographical sketches, one including a disability prime, and then received a recognition test. We assumed the disability prime to entail more false memories in correspondence with stereotypically lower competence and higher warmth.
Method
Participants
The sample comprised 38 participants who again were recruited by students at the Ludwigsburg University of Education on campus as well as at the community as a part of a seminar on the psychology of physical disabilities (25 women, 13 men, Mage = 27.5). Participation was voluntary and not rewarded. 71% were university students, 26% were employed, and one person was unemployed. All participants provided informed consent. The sample size was determined assuming a medium effect size.
Design
Stereotype activation was manipulated within participants. One of two biographical sketches read by participants included a disability prime.
Material
Two biographical sketches described two men in about 250 words. Both stated where the person was born and grew up. Their education and current job were described, as well as their family situation and favorite leisure activities. Whereas the structure was the same, the content differed in every respect, that is, the two biographical sketches did not share any details. Moreover, the first person was described as professionally more successful. One biographical sketch included a disability prime: The person had to use a wheelchair because of a deformity of the spine. The assignment of the disability prime to the two biographical sketches was counterbalanced between participants.
Procedure
The experiment was conducted online. The experimenter was connected in a video call through Cisco Webex Meetings. First, participants received a text file with the biographical sketches via e-mail. There was only one sequence of the two persons described, but the assignment of the disability prime to either the first or the second person was counterbalanced between participants. Participants were instructed to carefully read the text and were informed about the upcoming memory test. When they had read the complete text, they closed the file and the experimenter sent another text file comprising the memory test. Questions about the first person came first. There were six questions asking about a previously studied fact (e.g. “How old is he?”) together with four options, one being correct. The proportion of correct responses on these six questions was analyzed in addition to false alarms on another test item. This item asked to identify which traits described his personality according to other people. Six trait words were listed that had not been included in the biographical sketch and belonged either to the dimension of warmth (empathetic, sensitive, understanding) or competence (confident, determined, persistent). At the end, participants indicated their age, gender, and study major or profession.
Results
A 2 (disability prime: person with disability, person without disability) × 2 (trait dimension: warmth, competence) × 2 (sequence: first person with disability, second person with disability) ANOVA with repeated measures on the first and second factor and sequence as a control factor examined false alarms for trait words. A significant main effect of trait dimension reflected overall more false alarms for competence traits, F(1, 36) = 6.04, p = .019, ηp2 = .14. A significant interaction, F(1, 36) = 5.77, p = .022, ηp2 = .14, showed that there were significantly more false alarms for warmth traits of the person with disability, F(1, 36) = 9.21, p = .004, ηp2 = .20, but marginally fewer false alarms regarding competence traits, F(1, 36) = 3.25, p = .080, ηp2 = .08 (see Figure 2). In addition, a significant interaction of sequence and disability prime, F(1, 36) = 11.96, p = .001, ηp2 = .25, and a significant three-way interaction, F(1, 36) = 61.03, p < .001, ηp2 = .63, reflected that the content of the two different biographical sketches influenced false alarms in addition to the disability prime, proving the necessity of counterbalancing.
When the professionally more successful person was described with a disability, the mean false alarm rate for warmth traits was .25 (SD = 0.29) for that person, whereas it was .44 (SD = 0.14) for the other person. The mean false alarm rate for competence traits was .63 (SD = 0.27), whereas it was .26 (SD = 0.45) for the other person. When the professionally less successful person was described with a disability, the mean false alarm rate for warmth traits was .51 (SD = 0.12) for that person, whereas it was .07 (SD = 0.14) for the other person. The mean false alarm rate for competence traits was .05 (SD = 0.23), whereas it was .68 (SD = 0.24) for the other person. All pairwise comparisons were significant (p ≤ .001).
In contrast to false alarms, correct responses did not differ reliably between the person with or without disability, │t│ < 1. The mean proportion of correct responses for the person with disability was .85 (SD = 0.14), and the mean proportion of correct responses for the person without disability was .83 (SD = 0.19).
Discussion
The overall high false alarm rates show that participants obviously exhibited false memories. False alarms did not reflect random guessing but matched the content of the biographical sketches. A disability prime caused them to go up for warmth traits and to go down for competence traits, in correspondence with disability stereotypes (e.g. Fiske et al., 2002; Kittson et al., 2013; Rohmer & Louvet, 2012). There was also an association with details about the two persons. More false memories for competence and fewer for warmth traits occurred for the person described as professionally more successful. However, the possibility of item-specific effects due to item selection must be taken into account when considering interpretations of differences between the two biographical sketches.
General Discussion
Two experiments showed different ways in which disability stereotypes can shape memory. They impaired memory about a person that had been described in terms consistent with gender stereotypes. They influenced false memories for traits in a manner consistent with the content of disability stereotypes. The mentioning of disability changed what was remembered or was believed to be remembered about the respective person.
The influence of disability stereotypes was schema-consistent in both experiments. The results suggest that schema activation guided retrieval and perhaps storage of the read information about the person described with mentioning a disability. False memories indicate a schema effect on retrieval processes because they consider items that had not been encoded, of course. The correspondence of trait words with the schema made them seem as presented or probably presented before. We did not assess introspective reports on how strongly participants felt about actually having seen items. However, research on false memories shows that an experience of recollection is usually more typical than just a vague sense of familiarity (Brainerd et al., 2001). Schema activation might also have blocked access to items when participants attempted to retrieve them in the test phase of Experiment 1 because the items did not match the schema. However, it is also possible that the schema already hampered encoding processes, and items were not stored in memory as well as without schema activation (Sklenar et al., 2021; van Kesteren et al., 2012).
False memories are ubiquitous (Chang & Brainerd, 2021; Loftus & Pickrell, 1995). Our mind completes what we remember in accordance with our knowledge. Stereotypes are one kind of schematic knowledge and comprise beliefs about how people are that belong to a certain group. The present findings show that disability stereotypes can create memory illusions about a person with disability. In addition, memory for what was actually said about a person with disability suffered. Disability is a highly salient person feature. Its mere mentioning influences how persons are perceived, what is presumed of them, and what is remembered.
References
(1981). Person stereotypes and memory for people. Journal of Personality and Social Psychology, 41(5), 856–865. 10.1037/0022-3514.41.5.856
(2001). Conjoint recognition and phantom recollection. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27(2), 307–327. 10.1037/0278-7393.27.2.307
(2021). Semantic and phonological false memory: A review of theory and data. Journal of Memory and Language, 119,
Article 104210 . 10.1016/j.jml.2020.104210(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. 10.1037/0022-3514.82.6.878
(2013). The effect of video observation on warmth and competence ratings of individuals with a disability. Psychology of Sport and Exercise, 14(6), 847–851. 10.1016/j.psychsport.2013.07.003
(2001). Illusions of gender: Stereotypes evoke false memories. Journal of Experimental Social Psychology, 37(1), 3–14. 10.1006/jesp.2000.1426
(1995). The formation of false memories. Psychiatric Annals, 25(12), 720–725. 10.3928/0048-5713-19951201-07
(2010). Cultural stereotypes of disabled and non-disabled men and women: Consensus for global category representations and diagnostic domains. British Journal of Social Psychology, 49(3), 471–488. 10.1348/014466609X468411
(2012). Implicit measures of the stereotype content associated with disability. British Journal of Social Psychology, 51(4), 732–740. 10.1111/j.2044-8309.2011.02087.x
(2018). Implicit stereotyping against people with disability. Group Processes & Intergroup Relations, 21(1), 127–140. 10.1177/1368430216638536
(1992). Memory for schema-relevant information: A meta-analytic resolution. British Journal of Social Psychology, 31(2), 81–109. 10.1111/j.2044-8309.1992.tb00958.x
(2021). Person memory mechanism underlying approach and avoidance judgments of social targets. Social Cognition, 39(6), 747–772. 10.1521/soco.2021.39.6.747
(1978). Reconstructing the past: Some cognitive consequences of person perception. Journal of Personality and Social Psychology, 36(9), 941–950. 10.1037/0022-3514.36.9.941
(2023). Data for “How disability stereotypes shape memory for personal attributes”. https://osf.io/4qzvk/
(2012). How schema and novelty augment memory formation. Trends in Neurosciences, 35(4), 211–219. 10.1016/j.tins.2012.02.001
(1976). Social stereotyping of the physically handicapped. Rehabilitation Psychology, 23(4), 115–124. 10.1037/h0090911
Appendix A
Item Type
Appendix B
Response Type
