ABSTRACT
In 1966, Rosenthal’s book, Experimenter Effects in Behavioral Research, appeared and shortly thereafter a frequently embarrassing question about almost any data presentation reared its head: “How did you control for observer bias?” In 1967 Scott, Burton, and Yarrow reported that recording differences were noted between the senior author and observers who were not informed of the experimental hypotheses. In fact, data recorded by the senior author were more confirming of the experimental hypothesis than were data recorded by naive observers, and these differences were attributed to an expectation effort or observer bias. The study, however, was plagued by insufficient observer training and very poor reliabilities (range, – .09-.91). Nonetheless, the study was important in that it increased awareness of the possibility of observer bias – particularly in naturalistic settings.
