Using linguistic performance to measure problem-solving

The Accounting Education Change Commission (AECC) and the large international accounting firms have all emphasized that accountants must be able to work with unstructured problems to be successful in today's business environment. Measures of this ability are essential if accounting educators are to assist students in improving their abilities to work with unstructured problems. However, there appears to be no measure that has been widely accepted as being the ‘best’ measure. This study considers whether two linguistic performance measures might be usable by accounting educators for this purpose. We use data obtained from a student assessment centre to consider two measures of linguistic performance, idea density and grammatical complexity. We incorporate five criteria in deciding whether these measures could be usable: (1) whether the measures are related to students' performance when solving unstructured problems; (2) whether the measures distinguish between the ability to work with unstructured compared to structured problems; (3) ease of obtaining necessary information from students; (4) ease of scoring, and (5) robustness of the findings after considering other variables that may have an effect. Our results indicate that subjects' linguistic performance as measured by idea density meets these five criteria. However, grammatical complexity is not related to performance for either type of problem. These results were found even after controlling for the effects of other variables such as grade point average (GPA), experience, and personality variables. Because the measures can be obtained from virtually any written work produced by students, the information to be scored is relatively easy to obtain. Scoring is also relatively straightforward.