The researcher's own therapy allegiances: A "wild card" in comparisons of treatment efficacy

This report examines a possible distortion in the results of comparative tr eatment studies due to the association of the researcher's treatment allegi ances with outcomes of those treatments. In eight past reviews a trend appe ared for significant associations between the researcher's allegiance and o utcomes of treatments compared. In a new review of 29 studies of treatment comparisons, a similar trend appeared. Allegiance ratings were based not on ly on the usual reprint method, but also on two new methods: ratings by col leagues who knew the researcher well, and self-ratings by the researchers t hemselves. The two new allegiance methods intercorrelated only moderately, but each allegiance measure correlated significantly with outcomes of the t reatments compared, and when combined, the three measures explained 69% of the variance in outcomes! Such an association can distort comparative treat ment results. Our report concludes with how the researcher's allegiance may become associated with treatment outcomes and how studies should deal with these associations.