Harmonizing Long and Short Form Personality Inventory Scores: a parsimonious linking approach for aging research

Introduction: Prior research on cognitive aging has suggested that personality traits—especially the Big Five from the Five Factor Model (FFM)—are potential contributors to cognitive resilience in later life. However, the existence of different personality inventories, particularly different versions across instruments, presents a methodological challenge for facilitating further investigation across cohorts of older adults. To address this issue, the current study investigates whether a parsimonious harmonization of two personality inventories would be feasible, and if the combined datasets would reveal patterns of association between personality and cognitive outcomes. Methods: Data was drawn from the SuperAging Research Initiative, which includes SuperAgers and Cognitively Average Older Adults who completed two versions of the NEO Personality Inventory, the Revised NEO Personality Inventory (NEO-PI-R; N = 141) and a shortened version, the NEO Five-Factor Inventory-3 (NEO-FFI-3; N = 212). Their shared items were evaluated to assess harmonization feasibility and psychometric equivalence. Results: A confirmatory factor analysis (CFA) of 55 identical items between the two instruments indicated acceptable model fit, supporting harmonization. Measurement invariance testing comparing the models demonstrated moderate psychometric equivalence through configural, partial metric, and partial scalar invariance. A logistic regression of the combined data revealed that lower Neuroticism was associated with increased odds of SuperAger classification. Conclusion: The simplified harmonization approach of using shared items across forms was supported, enabling the integration of the two datasets for subsequent analyses.