The purpose of this study was to investigate different methods of evaluating clinical decision making used by undergraduate nursing students using a computer simulation program. A nonprobability sample of 65 senior level nursing students completed a clinical-based computer simulation on the care of a client with acute complications of Chronic Obstructive Pulmonary Disease.

Nursing students generated three different types of data with the use of the computer simulation. The first type was the end scores, a summation of the points obtained from all the multiple choice questions the student responded to on the simulation. The second type was the individual decision choices, the options selected by the student in response to multiple choice questions. Process oriented data was the third type. The student reported his/her own thoughts while making a clinical decision. End scores and decision choices are types of data routinely collected from a computerized simulation. Process oriented data is a qualitative component, in the student's own words that is typically not obtained during a simulated experience.

Faculty raters evaluated the students' decision making ability after directly working with them in a hospital setting and after computer simulation data were generated by the students. Other measurement tools included the Watson Glaser Critical Thinking Appraisal, students' self evaluations, and faculty confidence ratings. Conflicting results were found regarding the computer simulation program as an alternate method of evaluating clinical decision making. The simulations' actual end scores were moderately correlated with the clinical evaluations (r = 0.38, p =.003, and r = 0.49, p $<$.001). However, the raters' evaluations from the simulation data were not significantly correlated with the clinical evaluation. They were significantly correlated with the Watson Glaser Appraisal (r =.27, p =.04 and r = 0.37, p =.005). Faculty raters' high degree of confidence in their ability to evaluate students' decision making via the computer simulation was reported, but only if the data included the process oriented data as well as end scores and decision choices.