Reading focus
Syllabus: How Do We Describe and Measure Individual Differences?
Follow the three studies from the choice of facets and a 110-item pool, through selection of the final 60 items, to testing in separate validation samples.
Why you’re reading it. See how conceptual criteria, balanced item wording, factor models, and peer reports combine in the development of an actual questionnaire.
A question to consider. The BFI-2 divides conscientiousness into Organization, Productiveness, and Responsibility. Which facet would you expect to matter most for completing a course, and what evidence in the paper bears on using facet scores rather than a single domain score?
Why this reading
This paper documents how a widely used personality questionnaire was rebuilt. Across three studies, the authors define fifteen facets within the Big Five domains, write and select items, and test the resulting 60-item inventory in new samples, making visible the judgments behind a finished scale.
Before you read
The Big Five are broad personality domains, here labeled Extraversion, Agreeableness, Conscientiousness, Negative Emotionality, and Open-Mindedness. Facets are narrower traits within a domain. The bandwidth–fidelity trade-off contrasts broad measures that summarize much behavior with narrow ones that describe and predict specific criteria more precisely. Acquiescence is a tendency to agree, or disagree, regardless of content; balancing positively and negatively worded items helps separate it from the trait. Principal components and confirmatory factor analyses test whether items and facets group as intended.
Read for
- Follow how facets were chosen, pairing one facet that is central to its domain and largely independent of the others with two that broaden it, and how that choice shapes what the domain scale measures.
- Trace the move from a 110-item pool to 60 items, noting which criteria were conceptual and which empirical, and why results from the development sample were not treated as the main evidence.
- Examine the evidence from the two validation samples—internal consistency, retest reliability, self–peer agreement, factor models that include acquiescence, and prediction of criteria—and what each kind of evidence can show.
Consider afterwards
- The authors find Organization to be the most distinct Conscientiousness facet but arguably the least conceptually interesting, and the other two facets predicted several criteria more strongly. What does this suggest about using internal structure alone to decide what a scale should contain?
- Balanced wording controls acquiescence at the scale level. What problems might remain when the inventory is translated or read aloud to respondents with little formal schooling?
Editorial context
The development and validation samples came from economically developed, English-speaking countries and were mostly young or middle-aged adults with at least secondary education. The authors call for more representative samples and population norms, and note that the original BFI is less reliable among younger and less educated respondents. Most criteria were self-reports, with some peer reports. The BFI-2 measures the Big Five and fifteen facets; it does not show that this taxonomy exhausts personality, and its brevity required excluding content the authors acknowledge is important. A 2017 correction revises reference citations and the attribution of the BFI-2 items; it does not change the reported results. Evidence gathered in one population does not establish the same measurement properties elsewhere; Laajaj and colleagues examine that problem for items from the original BFI.