Using confirmatory factor analysis to validate the Chamberlin affective instrument for mathematical problem solving with academically advanced students.

Background: Student affect plays a considerable role in mathematical problem solving performance, yet is rarely formally assessed. In this manuscript, an instrument and its properties are discussed to enable educational psychologists the opportunity to assess student affect.
Aims: The study was conducted to norm the CAIMPS (instrument) with gifted students. In so doing, educational psychologists are informed of the process and the instrument's properties.
Sample: The sample was comprised of 160 middle-grade (7 and 8) students, identified as gifted, in the United States.
Methods: After completing one of four model-eliciting activities (MEAs), all participants completed the CAIMPS (Chamberlin Affective Instrument for Mathematical Problem Solving). Data were analysed using confirmatory factor analysis to ascertain the number of factors in the instrument. The normed fit index (0.6939), non-normed fit index (0.8072), and root mean square error approximation (.076) were at or near the acceptable levels. Alpha levels for factors were also robust (.637-.923).
Results and Conclusion: Data suggest that the instrument was a good fit for use with mathematics students in middle grades when solving problems. Perhaps the most impressive characteristic of the instrument was that the four factors (AVI: anxiety, value, and interest), SS (self-efficacy and self-esteem), ASP (aspiration), and ANX (anxiety) did not correlate highly with one another, which defies previous hypotheses in educational psychology.
(© 2017 The British Psychological Society.)