Published online by Cambridge University Press: 16 April 2020
Psychiatric and psychological testing are used extensively in both clinical and research contexts, with the goal of providing an objective indication of the construct being measured. At the same time, it has long been recognized that there is no such thing as a culture free or even fair test, to the extent that cultural bias-the constant and systematic statistical error due not to chance but to ethnic group membership-can seriously undermine the validity of a psychological or psychiatric test. The very construct being measured may not have a culturally compatible equivalent, or, relatedly, the behavior sampled in a given item or combination of items may not be associated with the construct in question in the same way across all cultures. In addition the means by which the behavior is sampled-the instrument construction, the sorts of questions or their structure, for example-can vary in familiarity across cultures. Finally, the individual items themselves may be poorly or ambiguously translated, may be predicated on specific cutlural norms, or, indeed, the meaning of the item can vary cross-culturally. Finally, sampling bias pertains to the representativeness of the sample, however, this is all the more complex given that constructs such as “race”, “ethnic group” and so forth are often poorly defined and ambiguous, and thus who is and who is not included in a given group may further undermine the validity of the test. Despite these difficulties, concrete steps can be taken to increase the cultural validity of the test used.