Thesis (Ph. D.)--University of Washington, 1994. / Vita. Includes bibliographical references (leaves [166]-178).

Research has consistently shown group psychotherapy to be an effective form of treatment. Group cohesiveness, the quantity and magnitude of the strength of the bonds within the group, has received considerable attention. This study describes the integration of the cohesion literature using a random effects meta-analysis to ascertain the magnitude of the relationship between cohesion and outcome as well as explore the effect specific group variables have on this association. Studies were included which measured the relationship between groups comprised of at least three members, included both an outcome and cohesion measure convertible to an effect size, and were published in English. 40 published studies, comprised of 3,323 participants, were analyzed and yielded an omnibus effect size of r = .25, suggesting a positive, moderate relationship exists between cohesion and outcome. Four group variables (group size, treatment duration, treatment setting, and group focus) and a client variable (age) demonstrated markedly stronger relationships with cohesion. The study also examined the most frequently used cohesion measures and definitions as they relate to their relationship to outcome and frequency of use. Clinical implications for group therapists are identified and future research suggested.