Predicting performance of information specialists

Limited research findings are available on the information‐processing behavior of users of scientific and technical information. Moreover, the information specialist who functions as intermediary between information sources and users has been the subject of even less empirical inquiry. Recognizing the potential usefulness of research findings in this area, empirical data obtained from an exploratory study of the task‐oriented information‐processing behavior of information specialists were analyzed in seeking to identify useful predictors of job performance. Forty tasks assigned to 18 subjects were analyzed, and a multiple discriminate function was developed using four categories of predictor variables: individual, task, individual‐task interaction, and information processing. The multiple correlation coefficient for the discriminate relationship consisting of 16 predictor variables and one criterion variable (job performance) was 0.869. The function correctly classified all 40 tasks into low and high performance categories. The five most important variables in the relationship were: (1) evaluation time for rejected information elements, (2) image state for task, (3) task result rating, (4) number of elements processed, and (5) intelligence. Implications of the study relative to selection, training, and supervision are discussed, and suggestions for further research are indicated.