Empirical Studies of Psychoanalytical Theories, Volume I: Edited by Joseph Masling. Hillsdale, N.J.: The Analytic Press, 1983. 298 pp.

The purpose of this new series, as set forth in the preface to Volume I, is to present efforts at experimental testing of clinical hypotheses. While the editor notes that contributors all accept the richness and insights of psychoanalytic hypotheses emerging from a clinical setting, the clinicians who propose them are criticized as writing "ex cathedra," with only clinical evidence to support their views. In this perspective, since clinical data are observed, interpreted, and transformed into sweeping generalizations by one and the same person, there are inevitable problems. Proponents of each school of psychoanalytic thought generate evidence for their theories. The basic premise of this series is that experimental testing is the only way to resolve disputes between competing theories and to "separate that which is useful in psychoanalytic thought from that which is unproductive and mistaken" (p. x). It is a premise worthy of examination in relation to the contents of the volume.

Dianna E. Hartley and Hans H. Strupp present a study of therapeutic alliance in psychotherapy which attempts to test the influence of such a factor in outcome. The behavior of therapist and patient and their interaction are rated with regard to lists of items that are significant in identifying the quality of the alliance in a therapeutic session as high or low optimal activity. Using such a descriptive measure, the find much overlap in distribution of therapeutic alliance among good and poor outcome groups. In patients who quit treatment there is a high measure of activity in the therapeutic alliance scale based on the therapist's efforts to influence the situation.

The authors note the difficulties inherent in engaging negativistic, defensive, withdrawn patients and conclude that the therapeutic process is itself significant in facilitating a therapeutic