Using the Triad Task as a Measure of Thought Style: A Validation Study

Initially developed to measure object categorization, the triad task has been adapted for use in different research fields and exhibits much variability in how it is implemented but limited validity evidence in our review of the literature. We report on a validation study of a common variant of the task as a measure of thought style in cultural psychology, following the three phases of validity (Loevinger, 1957). With the use of item response theory, we found support for a unidimensional construct in one American and one Chinese sample, but the measure showed other suboptimal psychometric properties. We subsequently improved the measure via item development, instruction modification, and invariance testing in three additional American samples. As a result, we found the expected within-culture differences. Buttressing the validity evidence of the triad measure reduces the threat of measurement-related alternative explanations and thus facilitates more valid group comparisons.