Examining the interrelationships among students' personological characteristics, attitudes toward the Unified Modeling Language, self-efficacy, and multiple intelligences with respect to student achievement in a software design methods course.

摘要

The purpose of this study was to investigate the interrelationships among student's demographics, attitudes toward the Unified Modeling Language (UML), general self-efficacy, and multiple intelligence (MI) profiles, and the use of UML to develop software. The dependent measures were course grades and course project scores. The study was grounded in problem solving theory, self-efficacy theory, and multiple intelligence theory. The sample was an intact class of 18 students who took the junior-level Software Design Methods course, CSE 3421, at Florida Institute of Technology in the Spring 2008 semester. The course incorporated instruction in UML with Java. Attitudes were measured by a researcher-modified instrument derived from the Computer Laboratory Survey by Newby and Fisher, and self-efficacy was measured by the Generalized Self-Efficacy Scale developed by Schwarzer and Jerusalem. MI profiles, which were the proportion of Gardner's eight intelligences, were determined from Shearer's Multiple Intelligence Developmental Assessment Scales.;Results from a hierarchical multiple regression analysis showed that only the collective set of MI profiles was significant, but none of the individual intelligences were significant. The study's findings supported what one would expect to find relative to problem solving theory, but were contradictory to self-efficacy theory. The findings also supported Gardner's concept that multiple intelligences must be considered as an integral unit and the importance of not focusing on an individual intelligence. The findings imply that self-efficacy is not a major consideration for a software design methods class that requires a transition to problem solving strategy and suggest that the instructor was instrumental in fostering positive attitudes toward UML. Recommendations for practice include (1) teachers should not be concerned with focusing on a single intelligence simply because they believe one intelligence might be more aligned to a specific topic or discipline (e.g., logical-mathematical for a mathematics course), (2) teachers should examine how MI can be addressed in a computer science course (e.g., a three-dimensional, animated UML tool that added sound when UML tools were selected or moved from one location from another), and (3) the instructor should model an appropriate attitude toward the problem solving concept the course presents.