Ideologies, idols (and graven images?): rejoinder to Gurman and Kniskern

Three major issues raised in Gurman and Kniskern's commentary are discussed. These are (a) the suitability of established research design criteria for studying the outcome of family therapy; (b) the impact of therapist relationship factors on therapy outcome; and (c) the place of concrete or objective change measures in psychotherapy outcome research. Areas of agreement and disagreement with Gurman and Kniskern's observations are identified.