Self-Identified Race, Test Anxiety, Reading Motivation, and Reading Comprehension: A Case for Serial Mediation

• In Event: Scenario-Based Assessment (SBAs) in the Age of Gen-AI: Developing Tools for Postsecondary Instructors

Abstract

Students who self-identified as Black/African-American have been shown to underachieve on tests of academic skills compared to White and Asian groups (e.g., NAEP, 2019). This trend persists among postsecondary students Black/AA students (e.g., Espinosa et al., 2019). We explored the roles of test anxiety and reading motivation on reading test performance through the lens of stereotype threat in a sample of post-secondary. The sample had sufficient diversity to explore differences across participants who self-identified as Black/AA and White. Given prior research, we hypothesized explanatory (i.e., mediational) relationships, wherein 1) students who self-identify as Black face higher levels of stereotype threat, 2) these students have increased test anxiety as a result, 3) test anxiety and stereotype threat negatively affect student’s motivation and engagement, and 4) their RC performance is negatively impacted.

Methods
Participants: 90 students from 4-year institution; ~39% self-identified as White; ~61% as Black/AA.

Measures
Reading Comprehension (RC) (Sabatini et al., 2019) consists of short passages and multiple-choice questions (~α = .70).

Multidimensional Test Anxiety Scale (MTAS)
An adapted version of the MTAS measures two cognitive (Worry and Cognitive Interference; ~α=0.86) and two affective-physiological dimensions (Tension and Physiological Indicators; ~α=0.96) of test anxiety. We used average of subset “in-the-moment” test anxiety items to ground them in the task just completed (Tonks et al., 2021).

Situated Reading Motivation Measure (SRMM)
We used the Avoidance Reading Motivation subscale (sample α of 0.85) of the Situated Reading Motivation Measure (SRMM; Tonks et al., 2021) to assess a reader’s “in-the-moment” reading motivation.

Procedure
Participants completed RC assessment, MTAS, SRMM and a demographic questionnaire (in that order), online out-of-class as assignment or extra credit.

Results
To examine the research questions, a serial mediation analysis (SMA) model was tested (Hayes, 2017). In this analysis, the relations between each variable are tested along with the indirect serial pathway (i.e., self-identified race → test anxiety → reading motivation → reading comprehension performance).
Results were complex. To summarize:
1. Self-identified race was a significant predictor of test anxiety (95%CI[.11, .97], p < .05), with students identifying as Black having higher anxiety scores.
2. Test anxiety was a significant, positive predictor of situated reading avoidance motivation (95%CI[-6.72, .-1.09], p < .05).
3. The indirect, serial effect (i.e., self-identified race → test anxiety → reading motivation → reading comprehension performance) indicated that the indirect coefficient was significant (95%CI[-2.04, -0.04]);
4. All predictors accounted for ~ 25% of the variance in RC scores.

Discussion/Conclusion

Consistent with the Serial Mediation Hypothesis, results suggested that both test anxiety and reading avoidance motivation partially mediated the relationship between self-identified race and reading comprehension performance. Like others, we argue that identifying causal and correlational factors associated with low or underachievement among minority groups is an essential first step in rectifying such issues. In the session, we relate how we are applying these results into the design and research of the SBAs and classroom practices.

Full references available upon request.

Authors

• Daniel Feller, University of Memphis

• John P. Sabatini, University of Memphis