Sarason has demonstrated that a model who discloses anxiety about performing on tests and at the same time outlines successful tactics for coping with this anxiety has a facilitative effect on the learning performance of high test anxious subjects. The effectiveness of the above model was determined with high test anxious female subjects who were psychologically stressed at the time that they observed the model. Using the number of correct responses per trial and the number of trials to criterion in a serial list learning task as a dependent measure. A 2 (stress - high and low) x 3 (model instruction order) x 16 (trials) analysis of variance failed to demonstrate the effectiveness of the videotape model on the learning performance regardless of whether the model was observed under conditions of stress, or prior to the stress manipulation. Assessment of the stress manipulation using groups who performed without the opportunity to observe a model indicated that the combination of ego-involving instructions plus later negative feedback to the subject about her performance midway through the task was effective, in that performance of high stress subjects later in the study was inferior to low stress subjects. Partial support was seen for earlier studies which indicated that evaluative instructions negatively affected high test anxious subjects. Some methodological differences between Sarason's 1975 study and the present study were discussed as possible sources of explanations of the differential results between the two studies. Future affects of research regarding the examination of the components of the stress manipulation, the use of live versus videotaped models, etc. were outlined. / Ph. D.