Validating a measure of STEM Students' SE for a Mixed Methods Research Design*

Self-efficacy, crucial for academic achievement in STEM, is typically assessed using pre-post measurements from surveys. However, developing a way to assess in-the-moment impacts to students' self-efficacy would create many opportunities for further study and intervention. With this goal, we devised a mixed-methods approach combining the Experience Sampling Method (ESM) with individualized daily journal prompts. While rich, this design poses validation challenges. To address these challenges, we analyzed three ESM survey questions, indicative of task-level self-efficacy, for skewness. Deviations from normal distributions suggest that traditional validation techniques may not apply. We further examined the Pearson's coefficient of skewness for each item and participant that does not assume the normal distribution of data. We would expect that if all ESM survey items are aligned with self-efficacy then the skewness for each individual should be consistent across those items. In this [talk/poster], we will present the shifts in skewness between items and discuss the data from validation interviews that expand on our quantitative findings.