A Model of Statistics Performance Based on Achievement Goal Theory

Structural equation modeling techniques were used to test a model of statistics performance based on achievement goal theory. Data were collected after the midterm and final examinations in an introductory statistics course, and models were fit at each time point. Learning goals were positively related to the use of deep-processing strategies and to self-efficacy and were negatively related to test anxiety. Performance goals were positively related to disorganization in study strategies and to test anxiety. Both learning and performance goals affected achievement indirectly through study strategies, self-efficacy, and test anxiety. Use of deep-processing strategies was positively related to effort but displayed an unexpected negative relationship to achievement. Disorganization was a positive predictor of test anxiety. Implications of these findings for teaching and learning statistics are discussed.