Measuring problem solving in computer environments: current and future states☆

摘要

The first part of this article will review the status of technology and educational reform; the second will treat traditional approaches to problem-solving measurement common today in technological settings; and the third will propose how problem-solving assessment may change and, in particular, the role that authoring strategies may play in measuring problem-solving performance successfully in computer environments. We will suggest four areas that must be considered in which computers would increase the fidelity and validity of measures of complex problem solving: (1) the intentions and skills of assessment designers; (2) the range of performance that counts as problem solving; (3) the ways in which validity evidence can be sought; and (4) the degree to which the measurement produces results that generalize across tasks and contexts.