Glossary
Expectation Bias
What is Expectation Bias?
Expectation bias, also called expectancy bias, describes ways prior beliefs can shape attention, interpretation or measurement. You expect a result, and ambiguous evidence starts to look like confirmation.
What is expectancy bias in psychology?
The term expectancy bias is also used when expectations influence judgments. In research, observer expectations can affect how an outcome is assessed. This needs to be distinguished from a self-fulfilling expectation, where someone changes how they treat another person and thereby helps change that person's actual behavior. Biased observation and a changed outcome are different processes.
For more specific definitions, the APA Dictionary describes observer bias as expectations influencing an observational record, and the experimenter expectancy effect as researchers conveying expectations that influence participants’ responses.
The site's entries on observer bias and experimenter bias cover these research forms in more detail. When an expectation changes how someone is treated and their performance changes as a result, see the Pygmalion effect.
Expectation bias vs. confirmation bias
The ideas overlap: expectations can shape how evidence is interpreted, while confirmation bias concerns favoring information that supports an existing belief. The useful distinction is the process involved, rather than assuming every expectation changes the outcome itself.
What does the research show?
Fingerprint judgments
Dror, Charlton and Péron (2006) presented fingerprint experts with previously examined prints in a different context. Several judgments changed. It was a small preliminary study, but it showed that expertise did not make those judgments immune to contextual information.
Clinical outcome assessments
In clinical research, Hróbjartsson and colleagues (2012) compared blinded and nonblinded outcome assessments across 21 trials. Nonblinded assessments produced more favorable treatment estimates on average.
Wine and color
Morrot, Brochet and Dubourdieu (2001) asked 54 oenology students to describe wines, including a white wine colored red with an odorless dye. The colored white wine received descriptions associated with red wine. The result shows that visual information influenced how participants described the odor; the authors discussed verbalization as a possible stage at which this influence occurred.
What is an example of expectation bias in aviation?
Imagine a pilot expecting a particular instruction and interpreting an ambiguous cue in line with it. The expectation supplies an answer before the information is clear. This illustrates the possibility; it is not a report of a specific aviation incident.
What is expectation bias in product and listening tests?
When people know which product they are trying, what they expect of it can shape what they report. A listener who knows that one speaker costs ten times more than another may hear the expensive one as clearer. Blind testing removes that information: the listener does not know which product is playing. To investigate the influence of labels or price information, compare blinded and sighted assessments under otherwise comparable conditions. Differences can provide evidence of an information effect, but other differences between the tests must also be controlled. The same logic applies to taste tests, software trials and any comparison where the assessor can see the label.
What is expectation bias in hiring?
A hiring manager might hear an ambiguous answer more favorably because the candidate attended a prestigious school. The same answer from another candidate might receive less credit. The problem is the standard of interpretation changing with the prior expectation.
How can expectation bias be addressed?
Set assessment criteria before examining the result, limit access to irrelevant context where feasible, and check ambiguous observations against independent evidence. Simply telling people to be open-minded is a weak substitute for examining how the judgment is made. Even agreement among observers can be misleading if they share the same expectations.
