Youth Reactions to Self-Report Measures and Their Relation to Student and Teacher Ratings of Aggression

Introduction. Although researchers frequently depend upon youth self-report measures to assess a variety of constructs, little is known about youth’s reactions to completing these measures. Previous studies of risk behavior indicate that up to 12% of adolescents may respond in inconsistent or invalid ways (Cornell et al., 2012). Moreover, those with inconsistent or invalid responses tend to over-report risk behavior and substance use (Cornell et al., 2012; Farrell et al. 1991). Further research is needed to identify attitudes toward surveys that influence inconsistent responding and over-reporting of risk behavior. Identifying these factors could inform efforts to improve the accuracy of self-report data.

Method. Middle school students (N = 852) completed a computer-assisted battery of measures as part of a study of a community-based youth violence prevention strategy. The sample was 46% male with a mean age of 12.6 years. The majority was African-American (71%) or Hispanic/Latino (15%). At the end of each survey, participants completed a ten-item Reactions to Survey Scale. Students’ frequency of aggression was measured by both self-report and by teacher ratings on the Problem Behavior Frequency Scale (Farrell et al., 2000).

Results. Exploratory factor analysis of the Reactions to Survey Scale suggested three factors representing Positive Reactions to the Survey, Negative Reactions to the Survey, and Validity of Responses on the Survey. Students reporting positive reactions were more likely to also indicate that they gave valid responses (r = .59, p < .01). Negative reactions were not related to youth’s report of the validity of their responses (r = -.04, p = .33). Students reporting higher levels of aggression were more likely to have lower ratings on both the positive reactions and valid response factors (rs = -.11 and -.12, respectively). This same pattern was found in correlations with teacher ratings of aggression, indicating that this finding was not an artifact of response accuracy (both rs = -.09). In contrast, negative reactions to the survey were not related to aggression.

Discussion. Although school administrators and IRBs often express concerns about how students will respond to surveys containing sensitive items, the majority of students reported favorable reactions to completing the surveys and indicated that their responses accurately reflected their experiences. Although students who reported less positive reactions to the survey tended to also report higher levels of aggression, collateral data from teacher ratings suggested that this was not simply an artifact of students’ over-reporting. A surprising finding was that negative reactions to the survey did not appear to influence the validity of responses or related to aggression.