This study aimed to identify elements of personal knowledge that were hypothesized to underlie within-person, across-context variations in students' appraisals of self-efficacy for coping with challenges encountered in Science, Technology, Engineering, and Mathematics (STEM) education. Freshman in a college of engineering completed assessments of (a) 4 elements of personal knowledge regarding themselves and their educational resources; (b) their subjective beliefs about links between these 4 personal and social qualities and each of 32 specific educational challenges; and (c) perceived self-efficacy for successfully handling each of the 32 challenges. Individual students' self-efficacy appraisals commonly varied substantially within-person, across contexts. This variability was predictable. Students displayed relatively high (low) self-efficacy within subsets of situations they subjectively linked to positively (negatively) valenced knowledge that they possessed, a finding consistent with the knowledge-and-appraisal model of personality architecture that guided the research. Additional analyses demonstrated that students with similar overall academic self-efficacy scores may display markedly different profiles of self-efficacy appraisal across context. Students' narrative accounts enriched understanding of these profile patterns. Educational Impact and Implications StatementThis study introduces novel methods for assessing the beliefs of individual STEM students and the specific social contexts in which these beliefs come into play. Findings show that our assessments predict both consistency and variability in student's confidence in their ability to handle challenges encountered in STEM education. Because our assessment methods provide richer, more nuanced information about individual students than the assessments commonly found in the scientific literature, they may be of particular value to school counselors and advisors who need to understand idiosyncrasies in the beliefs and social experiences of the individual student.

Language and the Data of Personality………………………………………..…. 1 C. The Big Five Model……………………………………………………………... 3 D. Interpreting Factor Analytic Results…………………………………………….. 4 E. Assumptions of Factor Analysis……………………………………………….... 8 F. Using Computer Simulations to Test the Influence of Inter-Item Connections upon Factor Analytic Results……………………………………... II. STUDY 1: SUFFICIENCY IN THE ABSENCE OF STRUCTURE ……………………… A. Overview………………………………………………………………………... B. Method………………………………………………………………………….. 1. Construction of the matrix containing non-ignorable relationships between items within the NEO-FFI……………………………………… a. Identifying item connections within the NEO-FFI …………….. b. Evaluation of the Type, Strength, and Direction of each Relationship…………………………………………………….. 2. Applying semantic and logical connections between items to otherwise random data……………………………………………………19 B. Results…………………………………………………………………………… C. Discussion………………………………………………………………………...26 III. STUDY 2: OBSCURING AN EXISTING STRUCTURE …………...………………….... A. Overview……………………………………………………………………..…. B. Method………………………………………………………………………….. 1. Identifying James' constructs within the NEO-FFI items………………... 2. Creating a tough-and tender-minded population………………………… C. Results…………………………………………………………………………... D. Discussion………………………………………………………………………. IV. GENERAL DISCUSSION………………………………………………………………….

No abstract