ANALYTICAL REASONING SKILLS INVOLVED IN GRADUATE STUDY: PERCEPTIONS OF FACULTY IN SIX FIELDS

This study was intended to provide information on the role of analytical abilities in graduate study. Graduate faculty in six fields of study (chemistry, computer science, education, English, engineering, and psychology) were asked to judge:(a) the importance for academic success of a wide variety of analytical skills (b) the seriousness of various reasoning errors (c) the degree to which a variety of “critical incidents” had affected their estimations of students' analytical abilities. Faculty members were generally able to discriminate among the various skills, errors, and incidents they were asked to consider, although the vast majority of skills, for example, were rated to be at least moderately important on average. Some skills were viewed as extremely important in all disciplines. More typically, however, disciplines varied quite markedly with respect to faculty perceptions of the importance, seriousness, and impact of these skills, errors, and incidents. Several relatively interpretable dimensions were found to underlie particular clusters of reasoning skills. These dimensions or factors involved (a) the analysis and evaluation of arguments, (b) the drawing of inferences and the development of conclusions, (c) the definition and analysis of problems, (d) the ability to reason inductively, and (e) the generating of alternative explanations or hypotheses. These dimensions were judged to be differentially important for success in the six disciplines included in the study. The study results would appear to have definite implications for developing future editions of the GRE analytical ability measure. For instance, some reasoning abilities that were perceived as very important are well represented in the current version of the analytical measure but others are not. That some abilities appear to be much more important in some fields than in others would imply the need to carefully balance the measurement of various reasoning skills to ensure fairness to test takers from different fields of study. Fairness should be aided by including for measurement those skills rated as consistently important across all disciplines. Finally, the several dimensions underlying various clusters of reasoning skills should be applicable to extending the test specifications for the current version of the analytical measure.