Chapter 8. The Less Than Awesome Power of Expectations to Distort Information-Seeking

Abstract

This chapter presents a critical analysis of claims that perceivers seek social information in ways that constrain targets’ reactions in such a manner as to almost insure expectancy confirmation.  First, demonstrates that the early research never actually demonstrated anything remotely supporting such a strong conclusion.  Second, it reviews the subsequent research, which generally shows that: 1) People do not seek social information in ways that constrain targets’ behavior or responses; 2) People generally seek social information in ways maximally conducive to fairly diagnosing or testing their expectations, rather than ways highly biased towards confirming those expectations; 3) People do have a very slight tendency to ask questions to which a “yes” answer is more likely to confirm than disconfirm their expectations; and 4) This tendency combines with acquiescence (a response bias whereby people are more likely to respond “yes” to any question), to create a small bias in social interaction towards targets giving responses that confirm perceivers’ expectations.  However, the extent to which even such responses lead to behavioral confirmation of self-fulfilling prophecies is likely to be modest.
EXCERPT:
    Do people’s expectations bias the manner in which they seek social information?  Do people rig the interaction, intentionally or not, to get what they expect?  The answer is a resoundingly clear “Ahh, well, uh, kinda sorta maybe a little.”

The following conclusions are justified by the research on the role of expectations in social information-seeking: 1) People almost never spontaneously ask the type of biased, constraining questions that Snyder & Swann (1978) required perceivers to use in the first study of lay social hypothesis testing; 2) In general, people greatly prefer diagnostic questions and information over confirmatory information and questions; 3) There is a slight tendency for people to prefer questions to which a “yes” answer confirms the hypothesis over questions to which a “no” answer confirms the hypothesis; 4) From a scientific or logical standpoint, such questions are often, though not always, highly diagnostic and appropriate; 5) The combination of perceivers’ use of a positive test strategy with targets’ tendency to give more yes than no answers (acquiescence) may ultimately lead perceivers to obtain social information in a manner somewhat more likely to confirm than disconfirm their hypotheses.
       Bottom line: expectancy effects, both self-fulfilling prophecy and bias, are real, but people are not completely out to lunch.  In fact, they are hardly out to lunch at all.  Mostly, they are minding the store quite effectively.  They do not seek to blindly confirm their prior beliefs.  Although biases do creep in, people prefer accurate information and do not rigidly resist disconfirming information.