Tell me briefly,who do you want to be? Psychometric properties of long- and short-measures of traits and change goals

Previous research has found that most people want to change their big five personality traits. Moreover, these change goals predict actual growth in corresponding traits. However, one major concern that often arises in intensive longitudinal research is keeping measures as short as possible, to minimize survey length—and consequently reduce participant reactivity and fatigue. This paper describes the results of a 16-wave, intensive/weekly longitudinal study (N = 662) in which participants self-reported their traits and change goals using [1] the 60-item (C-)BFI2, [2] the 30-item (C-)BFI2S, [3] the 15-item (C-)BFI2XS, and [4] the 10-item (C-)TIPI. Results indicated that shorter measures of traits and change goals predicted cross-sectional, concurrent outcomes with relatively equal validity, as compared to longer measures. In contrast, longer measures of both traits and change goals were superior to shorter ones when examining longitudinal/predictive effects (e.g., test–retest reliability, growth models, or within-person correlations). Thus, this research elucidates that measure selection should depend on the goals and design of particular studies (e.g., cross-sectional research can likely use short measures without consequence, whereas longitudinal studies likely need longer measures to reliably detect effects).

Plain Language Summary

This study used a relatively large sample (for psychological studies) and found that asking more questions (e.g., 6–12 questions per trait) about people's personalities (as opposed to fewer questions, e.g., 2–4 questions per trait) allowed researchers to identify complex time-based relationships between the variables. Thus, studies that ask only 2–4 questions about personality traits may not be able to detect complex, time-ordered phenomena. In contrast, studies that include 6–12 questions per trait may be able to illuminate more complex dynamics in traits.