In order to test the effect of implicit theories on short-term stress following social adversity, assessed via the daily diary survey, we will conduct two different classes of models.
First, we will average the overall secondary appraisals for stress for the week (daily judgments of handling stress) and the average valence of the negative social events listed on the daily diary survey. After standard data cleaning (e.g., transforming variables if necessary), then we will examine whether the incremental theory moderates the relationship between daily adversity and secondary appraisals--specifically, whether the correlation between daily adversity and secondary appraisals is larger in the control than in the treatment. We will consider the following possible covariates measured at baseline: test scores, optimism, stress (the PSS scale by Cohen et al.), and baseline implicit theories, in addition to demographics (sex, race, special ed status, poverty). Covariates that do not improve model fit will be discarded unless, due to chance, there are baseline differences between experimental conditions in terms of a given variable. Altogether, these will be standard OLS or Anova models.
Next, we will explore the possibility of nested models (e.g., mixed-effects regressions in LMER or HLMs) assessing the within-person correlation between daily adversity and secondary appraisals. The same covariates will be tested.
Potential baseline moderators include:
- Sex, race, or special ed status
- Baseline implicit theories
- Baseline stress measures (assessed via the PSS or the baseline daily diary)
- Perceived popularity with peers (self-reported)
- Bullying (self-reported)
- Scores on the "cognitive sensitivity to hierarchy" battery--this is under construction.