other group members. The group task (to achieve a group culture of intimacy, acceptance, introspection, understanding, and interpersonal honesty) is, in essence, interpersonal. However, research with a wide variety of groups has demonstrated that satisfactory participation in the group task, regardless of its nature, is an important source of satisfaction for the group members. 76 Patients who cannot introspect, reveal themselves, care for others, or manifest their feelings will derive little gratification from participation in group activities. Such patients include many of the types discussed earlier: for example, the schizoid personality, patients with other types of overriding intimacy problems, the deniers, the somatizers, the organically impaired and mentally retarded. These patients are better treated in a homogeneous, problem-specific group that has a group task consonant with their abilities.
Members of many kinds of groups derive satisfaction from membership because the outside world regards their group as highly valued or prestigious. Therapy groups obviously are at a disadvantage in this regard, and generally this source of attraction to the group is not readily available. Therapy group members will, however, usually develop some pride in their group: for example, they will defend it if it is attacked by new members. They may feel superior to outsiders--to those "in denial," to individuals who are as troubled as they but lack the good sense to join a therapy group. If patients manifest extraordinary shame at membership and are reluctant to reveal their membership to intimate friends or even to spouses, then therapy group membership must appear to them dissonant with the values of other important anchor groups. It is not likely that such patients will become deeply attracted to the group. Occasionally, outside groups (family, military, or, more recently, industry) will exert pressure on the individual to join a therapy group. 77 Groups held together only by such coercion are tenuous at first, but the evolving group process may generate other sources of cohesiveness.
Selection for group therapy is, in practice, a process of deselection: group therapists exclude certain patients from consideration and accept all others. Though empirical outcome studies and clinical observation have generated few inclusion criteria, the study of failures in group therapy, especially those patients who drop out early in the course of the group, provides important exclusion criteria. Patients should not be placed in a group if they are likely to