There are few studies referencing student opinions concerning testing formats. Therefore, the purpose of the current study was to develop a parsimonious logistical regression model for predicting student test preference for either multiple-choice tests or other tests. Participants were 150 mid-western university undergraduates (64 males, 86 females), between 18 and 55 years (M = 21.49, SD = 5.49). The initial set of predictors included four dimensions of personality type, and two total score measurements of test anxiety and need for cognition. Forward stepwise logistic regression was conducted at entry level alpha .10, with the initial model including all six predictors. The final model for predicting test preference indicated (a) need for cognition and personality type are significant predictors, (b) personality type variables were dropped, (c) odds of preferring multiple-choice tests for given levels of the predictor variables, and (d) test anxiety and need for cognition share an inverse relationship.According to Ornstein (1992), answering essay questions involves the following: 1. ... two types of essay testing formats that require students to compose written answers: restricted response and extended response (Criswell aamp; Criswell, 2004).
|Title||:||Can Student Test Preference be Predicted by Need for Cognition, Test Anxiety Or Personality Type?|
|Author||:||Mary Ellen Farmer|
|Publisher||:||ProQuest - 2007|