Stigma Consciousness Among Adolescents With Learning Disabilities: Considering Individual Experiences of Being Stereotyped

Site and Participants

In Table 1, we summarize the characteristics of participants in all three phases of research: two initial phases of iterative development (Study 1) and the full piloting of the new measure (Study 2). All participants were from urban and suburban areas in the northeastern United States. The first phase of development in Study 1 included six students (two boys and four girls) identified by their schools as having a specific LD, per the categories in federal special education legislation. Participants included two students in each of ninth, 10th and 11th grades at a single public comprehensive high school. LD status was determined based on review of each student’s individualized education program (IEPs), with confirmation from parent report. The second phase of Study 1 focused on testing a revised version of the SCQ-LD and the other measures to be used in Study 2. Ten students (four boys and six girls) who were identified by school administrators as having a specific LD were recruited from a second large high school.

OPEN IN VIEWER

Table 1.

Demographic Characteristics of Participants in Each Study Phase.

	Study 1 (Phase 1)	Study 1 (Phase 2)	Study 2 (Pilot)
Characteristic	n (%)	n (%)	n (%)
Gender
Male	2 (33)	4 (40)	31 (74)
Female	4 (67)	6 (60)	11 (26)
Grade
Eighth	0 (0)	0 (0)	4 (10)
Ninth	2 (33)	7 (70)	22 (52)
10th	2 (33)	2 (2)	13 (31)
11th	2 (33)	0 (0)	3 (7)
Race
White/European American	5 (83)	—	38 (90)
Asian, Hispanic, African American	1 (17)	—	4 (10)
Disability status
Language-based LD			13
Math-based LD			2
Language-based LD, math LD			3
Language-based LD, ADHD			12
Language-based LD, math LD, ADHD			5
LD, unspecified	6 (100)	10 (100)	7

Note. Race data were not collected in Study 1, Phase 2. Detailed information about the nature of students’ LD was not collected in either phase of Study 1. LD = learning disabilities; ADHD = attention deficit hyperactivity disorder.

Study 2, the more formal pilot testing, included 42 eighth- through 11th-grade students (31 male, 11 female). As in Study 1, students were identified by their schools as having a specific LD, per the categories in federal special education legislation. Most of the students in the sample (35 of 42 participants) were enrolled at private schools that restrict admission to students with this diagnosis. Including such a large proportion of specialized school students poses limitations in terms of the generalizability of the findings regarding prevalence of stigma consciousness, as described more fully in the “Discussion” section, but the stringent requirements for enrollment in these schools that serve only students with language-based LD indicate that students reliably meet the criteria for the diagnosis of specific LD.

As described in Table 1, the sample was predominantly White (90%), and all students in the sample were native English speaking. Seventy-four percent of participating students (n = 31) provided access to psychoeducational testing histories. Among this portion of the sample, all had average or above-average full scale intelligence quotient (M = 100, SD = 11.5). The specific nature of the learning challenges varied, with language-based deficits being most prevalent, as is the case in the population. Twenty-seven students were specifically identified by their schools as having dyslexia. Seventeen students had diagnoses of both LD and attention deficit hyperactivity disorder (ADHD), but students with other disabilities were excluded from the study. No students in the sample had diagnoses related to social or emotional challenges. This narrowed the sample to those students for whom experiences of stigma related to a LD were not likely clouded by other stigmatizing disability-related conditions. The specific nature of the LD was expected to be less critical in this research than in other LD-focused work because it is the label that was expected to be salient in the experience of stigma rather than the specific academic challenges. This is similar to the Higgins et al. (2002) findings in which the social and emotional effects of the LD were unrelated to the nature or severity of the difficulty.

Measures

The primary measure was the SCQ-LD, the development and final version of which is described in the “Results” section below. To address Research Question 3 about the relationship between this new measure and existing validated measures, we also administered the Self-Consciousness Scale for Children (SCS-C; Higa, Phillips, Chorpita, & Daleiden, 2008) and the widely used Self-Perception Profile for Learning Disabled Students (Harter, 2012). The SCS-C was used to test the hypothesized positive relationship between stigma consciousness and concern about self, relative to the thoughts and feelings of other people. The SCS-C is a 19-item self-report measure requiring respondents to rate on a 5-point scale (1 = strongly disagree; 5 = strongly agree) aspects of their personality (e.g., “I care about the way others think I look”). It has strong internal consistency (Cronbach’s α = .86) and was purposefully developed for use with this age group. The Self-Perception Profile for Learning Disabled Students (Harter, 2012) was used to measure self-perception, which was expected to be negatively associated with stigma consciousness on the SCQ-LD. This 15-item self-report measure asks students to indicate perceptions of themselves along several dimensions: global self-worth, intellectual ability, and social competence. Responses indicate how “true for me” descriptions of teens are (e.g., “Some teens are sure they are pretty smart in school. But other teens are not so sure they are all that smart in school.”). Reliability, reported via Cronbach’s alpha, is strong across subscales (global self-worth = .83; intellectual ability = .81; and social acceptance = .81).

Procedure

Research Question 1: Is the Method for Revising the SCQ That Has Been Used in Previous Adaptations of This Measure Appropriate for a Focus on Adolescents With LD (Readability, Understanding, Accessibility)?

The first round of adaptation of the SCQ for students with LD revealed several critical challenges with the typical approach to adapting the SCQ. As described above, we adopted the simple replacement approach and sought to maintain person-first language. This approach was problematic for two reasons. First, the language of the survey became syntactically cumbersome. For example, the reverse-scored item “I almost never think about the fact that I am female when I interact with men” would become “I almost never think about the fact that I am a person with a learning disability when I interact with people without learning disabilities.” Similarly, a simple term to describe people who are prejudiced against individuals with LD does not exist. The item “I often think that men are unfairly accused of being sexist” had no straightforward translation for our target population. To respond to these challenges, each item was adjusted to maintain the intent of the original but to use more appropriate language and syntax. This resulted, for example, in changing “Most men have a lot more sexist thoughts than they actually express” to “Most people have negative views about kids with learning disabilities even though they don’t say them out loud.” The original item “I never worry that my behaviors will be viewed as stereotypically female” became “I never worry that people think I act like someone with a learning disability.”

After pilot testing, we identified two other challenges that required revision to the survey. First, we observed that all of the participants struggled with the usability of the survey. In particular, the number of response choices was unnecessarily complex, significantly slowing participants as they recorded their responses. To address this challenge, we revised the SCQ-LD to provide only four response options (strongly disagree, disagree, agree, strongly agree), removing the midpoint option and simplifying the scale from the original SCQ.

Second, we found that many students were unfamiliar with key terms on the SCQ-LD. Most notably, they were not consistently able to explain the word “stereotypes,” which was essential to understanding two of the items. We adjusted the language of these items to convey the intent of the original without relying on this vocabulary. In addition, we found substantial variation in students’ definitions of the term “learning disability.” We ultimately decided to keep this as the focal term (instead of “learning difference,” or “learning difficulty”) because of a desire to focus on the potential stigma associated with the label itself, but we sought a means to capture differences in interpretation. This led to the addition of two open-ended questions at the beginning of the measure: “What does having a ‘learning disability’ mean to you?” and “In your experience, what do you think most people think about kids with learning disabilities?” This gave respondents an opportunity to voice their understandings and experiences and added contextual depth to future findings.

The final SCQ-LD used in Study 2 (see Table 2 for SCQ-LD items) consisted of 17 self-report items, administered via computer with consistent read-alouds. This version included two open-response items and 15 statements describing concerns about judgments other people may make about individuals with LD; respondents indicated their level of agreement with each statement, from 1 (strongly disagree) to 4 (strongly agree).

OPEN IN VIEWER

Table 2.

The SCQ-LD Items, Their Means and Standard Deviations, and Factor Loadings for Exploratory Factor Analysis With Varimax Rotation.

Item	M (SD)	Factor
		Stigma consciousness	Societal perceptions
Items included in final version
16. I never worry about feeling judged because of my learning disability. (R)	2.29 (0.94)	.82	.10
6. I worry about people judging me because I have a learning disability.	2.05 (0.99)	.69	.02
10. In my experience, many people think that there is something wrong with kids with learning disabilities.	2.64 (0.91)	.52	.44
11. I never worry that people think I act like someone with a learning disability. (R)	2.21 (0.95)	.45	.09
3. Some people treat me differently because of my learning disability.	2.52 (0.83)	.38	.26
9. People assume I’m just like all the other kids with learning disabilities.	2.57 (0.77)	.32	–.01
4. Most people don’t judge someone else just because that person has a learning disability. (R)	2.36 (0.85)	.03	.72
7. Most people have negative views about kids with learning disabilities even though they don’t say them out loud.	2.86 (0.78)	.12	.75
12. Most people think they’re better than me just because of my learning disability.	2.17 (0.91)	.20	.14
13. People make assumptions about kids with learning disabilities, but that doesn’t affect me. (R)	2.24 (0.96)	.27	.01
15. I don’t usually think about my learning disability when I’m with kids who don’t have one. (R)	1.93 (0.92)	.09	–.08
17. Most people know that people with learning disabilities can be smarter or even more creative than other people. (R)	2.10 (1.01)	.21	.15
Items not included in final version
5. People with learning disabilities may learn differently than people without learning disabilities, but really they are just as smart.	1.19 (0.40)
8. Other kids with learning disabilities don’t treat me differently because I have a learning disability too.	1.62 (0.70)
14. People with learning disabilities are generally less intelligent than people without learning disabilities.	1.33 (0.69)
Percentage of variance		.28	.24
Eigenvalue		4.50	1.50

Note. Factor loadings >.30 are in boldface. Range is 1 (strongly disagree) to 4 (strongly agree). SCQ-LD = Stigma Consciousness Questionnaire–Learning Disabilities. (R) indicates items that were reverse coded.

Research Question 2: Do Tests of Reliability of the SCQ-LD (Internal Consistency, Factor Analysis, and Test–Retest) Indicate a Single, Stable Construct of Stigma Consciousness?

Initial analysis of the psychometric properties of the SCQ-LD focused on calculating Cronbach’s alpha with the 15 quantitative items. Three items (five, eight, and 14; see Table 2) were notably discrepant from the others. Correlations between each of these items and the rest were negative or zero. After removing these items, Cronbach’s alpha was excellent at both Time 1 and Time 2 (.82 and .84, respectively), indicating high internal consistency.

We next conducted an exploratory factor analysis of the remaining 12 items, using the first administration for all participants. An exploratory factor analysis is appropriate to verify unidimensionality beyond Cronbach’s alpha because reliability can be high even when several dimensions are assessed by a given measure (Fabrigar & Wegener, 2012). We note that the sample size (N = 42) could be considered small for factor analysis. Guidelines around sample size in factor analysis typically do not consider the specific conditions, including the number of factors, how correlated they are with each other, and the method of rotation (Fabrigar & Wegener, 2012; MacCallum, Widaman, Zhang, & Hong, 1999). When these characteristics are considered, sample sizes for effective factor analysis can vary dramatically; with 12 items, our sample-to-item ratio is 3.5, within the range of 3 to 6 as recommended values (MacCallum et al., 1999). A relatively low ratio was considered appropriate here because we expected a single factor that would be overdetermined, and we expected high communality between items. High communality was indeed found in the analysis we describe next, with the final main factor demonstrating an average communality of .59, approximating the recommended guideline of .60 for small samples (Worthington & Whittaker, 2006).

Using varimax rotation and an orthogonal solution, as recommended to facilitate future replication of this initial validation study (Henson & Roberts, 2006), we hypothesized a single factor reflecting students’ overall level of stigma consciousness related to have a LD. Kaiser’s criterion instead suggested the presence of two factors (eigenvalues = 3.69 and 1.30). A scree test similarly suggested two factors. Both factors were conceptually interpretable, leading us to retain two factors that combined to explain 55% of the variance.

As described in Table 2, Factor 1, which was dominant, loaded to .30 or higher on six of the 12 items, with Item 16 (“I never worry about feeling judged because of my learning disability”) and Item 6 (“I worry about people judging me because I have a learning disability”) contributing the highest loadings. We consider this factor the expected measure of stigma consciousness. Factor 2 loaded to .30 or higher on only three items (Item 7: “Most people have negative views about kids with learning disabilities even though they don’t say them out loud”; Item 4: “Most people don’t judge someone else just because that person has a learning disability” [reverse coded]; Item 10: “In my experience, many people think that there is something wrong with kids with learning disabilities.”). All relate not to one’s personal experiences or concerns about stigmatization, but to society’s views of learning disability, leading us to name this factor “societal perceptions.” As confirmation, a parallel factor analysis was conducted with the same 12 items from the second administration. These again loaded on two factors. ²

Subsequent analysis focused on the first factor, which is most central to the focus of this validation work. Rather than limiting the stigma consciousness score to the items with high factor loadings, we calculated stigma consciousness scores for each student by averaging across the 12-item scale. Higher scores indicate greater stigma consciousness. Scores ranged from 1.33 to 3.33 (M = 2.32, SD = 0.51) on a 4-point scale. We chose to maintain all items because we recognize the relatively small sample size and exploratory nature of this work; we expect future studies to confirm whether deletion of items is warranted. Sensitivity analyses with only the high-loading items suggest results below were unaffected by retaining all items.

To assess the test–retest reliability of the SCQ-LD, we calculated a correlation between stigma consciousness on our first and second assessments, 4 to 6 weeks apart. The coefficient was r(30) = .80, p < .01, indicating good stability over time.

Research Question 3: Does the SCQ-LD Correlate in Expected Ways to Previously Validated Measures of Self-Perception and Self-Consciousness?

We next examined the relationships between stigma consciousness and other measures to determine construct validity. We found significant correlations consistent with hypothesized directions, indicating validity of the stigma consciousness construct. Specifically, stigma consciousness was positively correlated with students’ overall self-consciousness, r(42) = .33, p < .05, and negatively correlated with each of the three subscales of the Self-Perception Profile for Learning Disabled Students: intellectual ability, r(42) = –.37, p < .05; social competence, r(42) = –.33, p < .05; global self-worth, r(42) = –.35, p < .05. All are moderately correlated, as would be expected for related but not identical constructs. Descriptive statistics for these measures are provided in Table 3.

OPEN IN VIEWER

Table 3.

Summary of Measures.

Measure	M	SD	Minimum	Maximum
Stigma consciousness	2.33	0.52	1.33	3.33
Self-Perception Profile for Learning Disabled Students
General intellectual ability	2.91	0.76	1.40	4.00
Social acceptance	3.08	0.67	1.20	4.00
Global self-worth	3.25	0.63	2.00	4.00
Self-Consciousness Scale for Children	1.43	0.51	0.37	2.37

Note. LD = learning disabilities.

Research Question 4: Do Students With LD Report Stigma Consciousness Related to the LD Label, and to What Degree?

Thirty-eight percent of participants’ average stigma consciousness scores fell in the “agree” to “strongly agree,” range indicating they were conscious of stigma related to their LD. In addition, 67% of respondents either agreed or strongly agreed with at least four items on the scale, indicating that a significant majority had some consciousness of stigma related to their LD. Stigma consciousness did not differ by gender, grade, or school.

In addition to the quantitative findings, we were able to qualitatively explore student responses to the two open-response questions about definition of “learning disability” and perceptions of others’ views of LD. As with the earlier qualitative analysis during measure development, we systematically reviewed student statements using constant comparison and typological analysis (LeCompte, Preissle, & Tesch, 1993). Our analysis revealed great variability in both how students view and understand their LD, and how they perceive others’ views and understanding of LD. When asked what having a LD means to them personally, many students responded that it means learning differently than other people and that certain tasks may be difficult or require extra time or help: “It means I need to learn differently than my peers and struggle a bit more to get the same amount of progress.” Many students expressed that although they learn differently, they are just as capable of achieving what others can: “It means that it will be harder to learn but not impossible and if you persevere you can do anything.” However, one student responded that having a LD means “that you are going to have a hard time in school and in real life.” Several students described a LD as a gift or advantage: “I see a learning disability more as an advantage, I think we think differently than everyone else, but it’s a gift. I think it can be very difficult to handle it sometimes, because it makes certain things more difficult.” When asked what other people think about kids with LD, most students responded that other people think kids with LD are less smart, stupid, or dumb: “Most people think that they are just plain out dumb and will never be as smart as people without them.” In contrast, some students said that other people do not see or treat them any differently: “I don’t think that they think anything different of them.”

Limitations

This study provides an initial step and useful tool for the field to pursue stigma consciousness as a new direction for understanding students’ experiences, but we recognize the limitations of this work and the need for continued research. A primary limitation is clearly the nature of the sample, which was small, demographically homogeneous, and primarily from private school settings. This presents a serious challenge, particularly in determining the prevalence of stigma consciousness, but we suggest that it does not invalidate the development and validation work; the SCQ-LD is reliable, valid, and usable. We hope that it will be used for research in other contexts with many more students so that the field can continue to understand both the most effective measure design and the nature of stigma consciousness in the population of adolescents with LD as a whole. The factor analysis was exploratory and on a small sample; as such, the solution is our current understanding about how the scale organizes, but it could be that with a larger, more diverse sample, a more complex factor structure will emerge. This initial work follows in the same vein as the Frostig Center study (Goldberg, Higgins, Raskind, & Herman, 2003; Higgins et al., 2002) in which all participants had attended the private school for students with LD; the field benefited from understanding their experiences and outcomes to then investigate similar trends with a broader population.

In addition, we did not investigate what contributed to students’ reported experiences of stigmatization. Future work needs to consider the role of, for example, societal-level influences like those presented in the media, relationships with key individuals such as parents, teachers, and peers, and instructional or school-related contextual factors. A developmental perspective would additionally contribute to understanding the dynamics of stigma experiences, perhaps particularly in relation to time of diagnosis and types of academic experiences at different grade levels. When focusing on students in high school, investigation of the relative contribution to perceptions of stigma from childhood experiences compared with those during adolescence would contribute to the body of life-course perspectives (McNulty, 2003). This would more effectively inform the identification of high-leverage opportunities for intervention.

Implications for Practice

This work has potential implications for practice, as future work could investigate the potential to use the SCQ-LD as a component of assessing school climate. Although the measure has not yet been used in partnership with educators, administrators with whom we partnered expressed concern about the stigma experiences of their students with LD and a desire for support in this regard. Overall, the prevalence of high levels of stigma consciousness supports previous calls for a focus on supporting the social and emotional health of students with LD, including for adolescents (Al-Yagon, 2016).

Addressing stigma consciousness is complex. Interventions with other populations have taken different perspectives. One approach would be to encourage a group identity. Stronger identification with a group may be associated with high levels of stigma consciousness, and also provide benefits through increased likelihood of fighting against stereotyping and discrimination (Pinel, 1999). Recent campaigns like “say dyslexia” (http://www.saydyslexia.org/) and Project Eye-to-Eye’s work on “owning the label” (http://eyetoeyenational.org) suggest that some in the LD field have embraced this identity salience strategy. Students with LD are encouraged to recognize, embrace, and appreciate their LD as a core part of themselves and a reason for pride rather than concern. Recent intervention work to address the negative effects of stigma consciousness and stereotype threat with other vulnerable populations take a different perspective, suggesting that harm from awareness and experience of stigma manifest primarily when stigmatization becomes a critical aspect of one’s identity (Cohen, Garcia, Purdie-Vaughns, Apfel, & Brzustoski, 2009; Pinel & Bosson, 2013). These instead focus on strengthening self-perception regarding other aspects of identity. For example, interventions emphasize strengths and values without attention to group membership. Understanding more richly the experiences of individual students may be able to inform the design of interventions that effectively leverage a group identity and avoid some of the psychological risks associated with it.