Using the California Psychological Inventory to assess the Big Five personality domains: A hierarchical approach Christopher J. Soto a, * , Oliver P. John b, * a Department of Psychology, Colby College, 5550 Mayflower Hill, Waterville, ME 04901-8855, USA b Psychology Department, University of California, Berkeley, 3210 Tolman Hall, MC 1650, Berkeley, CA 94720-1650, USA article info Article history: Available online 5 November 2008 Keywords: California Psychological Inventory Facets Five factor personality model Hierarchical structure Personality measures Test construction abstract The present research developed a new, multi-step approach for hierarchically assessing the Big Five per- sonality domains from a large and diverse pool of existing questionnaire items: those of the California Psychological Inventory (CPI). First, the Abridged Big Five Circumplex (AB5C) structural model was used to organize the CPI item pool in Big Five factor space. Using the 10 resulting CPI-AB5C circumplexes as starting points, 16 facet scales were then developed to assess specific personality characteristics within the broad Big Five domains. Finally, principal components analysis with validimax rotation was used to score the five domains from the facet scales. In three independent samples, the resulting CPI-Big Five measure demonstrated strong reliability, convergence with self- and peer-reports, and discriminant validity. Availability of the new measure brings more than a half-century’s worth of archival CPI data to bear on contemporary research questions about the Big Five. Additionally, the process of developing the CPI-Big Five measure illustrates some of the challenges that may arise when attempting to assess new psychological constructs from existing measures, as well as methods for addressing such challenges. Ó 2008 Elsevier Inc. All rights reserved. Christopher J. Soto’s work on this project was supported in part by a Chancellor’s Opportunity Predoctoral Fellowship from the Univer- sity of California, Berkeley and by a Graduate Research Fellowship from the National Science Foundation; Christopher J. Soto and Oliver P. John also received support from a grant from the Retirement Research Foundation, a University of California Faculty Research Grant, and National Institute of Mental Health Grant MH-43948. The authors thank Lewis R. Goldberg for sharing access to the community validation sample data. They thank Joshua Eng, Carol J. Garvan, and Laura P. Naumann for serving as judges for the corre- late-matching task. Finally, they thank Lewis R. Goldberg, Ravenna Helson, Laura P. Naumann, and the Mills Lab at the University of California, Berkeley for their helpful feedback on this work. 1. Introduction A consensus has emerged among personality researchers that the most important individual differences in adults’ personality characteristics can be organized in terms of five broad trait domains collectively known as the Big Five: Extraversion, Agree- ableness, Conscientiousness, Neuroticism, and Openness. This consensus has led to the construction of several dedicated Big Five measures since the early 1990s, such as the NEO Personality Inventory (Costa & McCrae, 1985, 1992; McCrae, Costa, & Martin, 2005), the Big Five Inventory (John, Donahue, & Kentle, 1991; see John, Naumann, & Soto, 2008), and sets of trait-descriptive marker adjectives (Goldberg, 1992; Saucier, 1994). Using these measures, much new data has been collected to address research questions ranging from how the Big Five affect job performance (Hogan & Holland, 2003) to how they are expressed in people’s living spaces (Gosling, Ko, Mannarelli, & Morris, 2002). Addressing some fundamental questions about the Big Five requires the analysis of long-term longitudinal data. Such ques- tions include how and why the Big Five change across the life course, and how standing on the Big Five in adolescence and young adulthood predicts relationship, career, and health outcomes years or decades later. Unfortunately, long-term longitudinal data spanning the full life course are not yet available from studies using the many relatively new Big Five measures, nor will such data be available in the near future. This situation creates a dilemma for researchers who wish to examine lifespan personality development from a Big Five perspec- tive. Must such researchers twiddle their thumbs for the next few decades, waiting for recently begun longitudinal studies to run their course? Or are there other approaches that would help them address their research questions more immediately? The present research aims to address this dilemma by developing and validat- ing a method for assessing the Big Five trait domains from an exist- ing measure that has been widely used in personality research for more than a half century, namely the California Psychological 0092-6566/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jrp.2008.10.005 * Corresponding authors. E-mail addresses: cjsoto@colby.edu (C.J. Soto), o_johnx5@berkeley.edu (O.P. John). Journal of Research in Personality 43 (2009) 25–38 Contents lists available at ScienceDirect Journal of Research in Personality journal homepage: www.elsevier.com/locate/jrp