THE DISTINCTIVENESS OF AFFECTS IN SPECIFIC SCHOOL SUBJECTS - AN APPLICATION OF CONFIRMATORY FACTOR-ANALYSIS WITH THE NATIONAL EDUCATIONAL LONGITUDINAL-STUDY OF 1988

Does academic affect generalize across different school subjects, or i s it specific to particular subjects? Substantively, this study consid ers the distinctiveness of affects associated with different school su bjects and critically evaluates this distinctiveness in relation to sc hool grades and standardized test scores. Methodologically, the study describes problems related to combining responses to single-item self- rating scales, adapts confirmatory factor analysis (CFA) models of mul titrait-multimethod (MTMM) data to address this problem, and provides guidelines for more effective use of the National Educational Longitud inal Survey of 1988 (NELS88) data. A large, nationally representative sample of eighth-grade students (N = 24,599) rated three affects (look ing forward to, perceived usefulness, anxiety) in each of four school subjects (mathematics, science, social studies, English). The CFA mode ls showed that simple scale scores were inappropriate. MTMM models ind icated that ratings in different school subjects were very distinct, a nd extended models incorporating school grades and test scores support ed this subject specificity of academic affect.