ClmicalP~_~rholo~~ R&u,, \'ol. 7, pp. 525-538. 1987 0272-7358/87 $3.00 + .OO Printed in the CSX. All rights rrservrd. Copyright @ 1987 Pergamon Journals Ltd. zyxwvutsr A SSESSING THE EFFEC TS O F BEHA VIO RA L MA RITA L THERA PY: A SSUMPTIO NS A ND MEA SUREMENT STRA TEG IES Thomas N. Bradbury Frank D. Fincham University of Illinois at Urbana-Champaign zyxwvutsrqponmlkjihgfedcbaZY A BSTRA CT. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA An important consideration in examining the effects of behavioral marital therapy (BM T) is the measurement strategies that are used to assess treatment outcome. The present paper identtfies and evaluates several important assumptions underlying the two primary means of assessing BMT outcome, viz., the use Of self-report measures alone and the use of self-report measures in combination with behavioral observation. Based upon this analysis, it is concluded that (a) contrary to common ,@ actice, the use of both self-report and observational measures is the more appropriate of the two alternativesfor assessing EMT outcome; (6) currently used self-report measures are not optimalfor assessing marital satisfaction; and(c) greater attention should begiven in future outcome research to assumptions and models underlying assessment strategies. A central concern for marital researchers and practitioners is the need to establish a scientifically defensible form of marital therapy. This goal involves determining which measurement strategy is most appropriate for assessing thera- py outcome (see Baucom & Hoffman, 1986). Two approaches for assessing change in behavioral marital therapy (BMT) have been the subject of recent debate (see Gottman, 1985; Jacobson, 1985a, 1985b, 1985c, 1985d; Jacobson, Follette, 8r Elwood, 1984; Kniskern, 1985; Weiss & Frohman, 1985). The first favors the use of self-reported marital satisfaction as the primary measure of BMT outcome, with the rationale being that marital satisfaction is “the final common pathway This article was written while the second author was supported as a Faculty Scholar by the zyxwvuts W. T. Grant Foundation and by grant #l-5-32169 from the National Institute of Child Health and Human Development. The authors thank Gregory Miller for his detailed review of an earlier draft of the paper. The helpful comments of Karen Glickman, Fred Kanfer, Dan O’Leary, and Robert Weiss are also gratefully acknowledged. Requests for reprints should be addressed to Thomas N. Bradbury or Frank D. Fin- cham, Department of Psychology, University of Illinois at Urbana-Champaign, 603 East Daniel Street, Champaign, IL 61820.