Solving Dilemmas by Computer or Counselor

Abstract

Results of studies in which either a human counselor or a computer was used to solve personal dilemmas are compared. Subjects were 415 undergraduate students. To increase stability of comparisons the counselor groups are aggregated as are the computer groups and the control groups. Comparison tests are then made among the aggregated groups. Two types of comparisons are made, significance tests and comparisons for size of effects. Computer counseling was as effective as standard counseling for dilemma-type problems. Computer counseling was significantly more effective than the control condition as was standard counseling. Comparisons of size of effect indicate that the average person in the computer groups was moved to the 51st percentile of the counselor groups and to the 78th percentile of the control groups. Size of effect for general psychotherapy and counseling indicate movement to the 75th percentile of the control groups. Thus, the effect of computer counseling resembles the over-all effect for general therapy. Future comparison of computer and counselor methods, effectiveness, and reactions is discussed.