Clinical evaluation of student performance: the effects of criteria on accuracy and reliability.
Faculty from four baccalaureate nursing programs participated in testing the accuracy and reliability of faculty evaluation of student performance as seen on a videotape. Three videotapes each portrayed a nursing student performing an ordinary nursing activity at five different levels of competency. The study design included three fixed factors (use of criteria, faculty experience, and type of student activity) and a trial factor with repeated measures (the five levels of performance). The experimental group (n = 27) used a 5-point scale with criteria for rating the level of performance while the control group (n = 33) used the same scale without criteria. A computed score and an estimated score were collected from each subject for each level of performance. The experimental group that used the scale with criteria gave more accurate computed and estimated evaluation scores than the control group. Faculty experience was not a significant factor in accuracy, but type of student activity was significant with the highest scores given to affective behaviors and the lowest scores given to psychomotor behaviors. There were no significant interactions between the three fixed factors.