| Abstract/Results: | ABSTRACT:
In a psychotherapy session, it is very difficult to decompose the flux of information in its parts. Such non-decomposable task, favors holistic and intuitive judgment processes (Hammond et al, 1987), which constrains the interpretation of subsequent information along the lines of the firstly activated schemes (Eyal et al., 2011). Thus, we hypothesize that nondecomposable
tasks favor primacy effects and often lead to confirmatory biases. In contrast, a task decomposed and analysed in its parts is expected to elicit a more deliberative reasoning, counteracting primacy effects and leading more often to disconfirmatory strategies of hypothesis testing.
To test these hypotheses, in two studies, we presented participants with an audio excerpt of a fictional client describing her depression symptoms, manipulated the decomposability of the excerpt. In the non-decomposable condition, participants were presented a case (without interruptions) and asked a final global clinical judgment (risk to develop psychopathology based in all behaviors). In the decomposable condition, participants heard the case in 6 smaller parts, completing, after each excerpt, an interim judgment about the likelihood of developing psychopathology based in that specific behavior. After the 6 excerpts,
participants made the same final global clinical judgment. Following this, participants rated the likelihood of three possible diagnoses. We manipulated within-participants the order of the presentation of depression symptoms in the beginning (depression scheme activated) vs. end (no activation of depression scheme) of the case.
Results show that when a scheme is activated (depression symptoms presented in the beginning), understanding the case in a non-decomposable way leads to higher ratings only to depression, than participants in the decomposable condition, which gave higher ratings to two diagnosis. This suggests that making a non-decomposable task lead to a more confirmatory bias. Implications to therapy session will be discussed.
|