TY - JOUR
T1 - Towards understanding assessments of the big five
T2 - Multitrait-multimethod analyses of convergent and discriminant validity across measurement occasion and type of observer
AU - Biesanz, Jeremy C.
AU - West, Stephen
PY - 2004/8
Y1 - 2004/8
N2 - Multitrait-multimethod analyses were used to examine the degree of convergent and discriminant validity of the Big Five. Phase 1 examined self-reports of the Big Five across three measurement occasions. Self-reports of the Big Five traits were stable, but were moderately intercorrelated. Phase 2 examined assessments of the Big Five across different types of informants (self, peer, and parent). Assessments converged across types of informants and, importantly, there was no evidence of correlation between the Big Five traits across the perspectives of different types of informants. The present results suggest that the degree of orthogonality of the Big Five traits depends on the source of the data. A single informant produces Big Five traits that are intercorrelated, whereas diverse informants tend to produce a much more orthogonal structure. Discussion focuses on methodological considerations in examining levels of convergent and discriminant validity and the theoretical implications for understanding personality assessments and the relationship between three-and five-factor models of personality.
AB - Multitrait-multimethod analyses were used to examine the degree of convergent and discriminant validity of the Big Five. Phase 1 examined self-reports of the Big Five across three measurement occasions. Self-reports of the Big Five traits were stable, but were moderately intercorrelated. Phase 2 examined assessments of the Big Five across different types of informants (self, peer, and parent). Assessments converged across types of informants and, importantly, there was no evidence of correlation between the Big Five traits across the perspectives of different types of informants. The present results suggest that the degree of orthogonality of the Big Five traits depends on the source of the data. A single informant produces Big Five traits that are intercorrelated, whereas diverse informants tend to produce a much more orthogonal structure. Discussion focuses on methodological considerations in examining levels of convergent and discriminant validity and the theoretical implications for understanding personality assessments and the relationship between three-and five-factor models of personality.
UR - http://www.scopus.com/inward/record.url?scp=3142660149&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=3142660149&partnerID=8YFLogxK
U2 - 10.1111/j.0022-3506.2004.00282.x
DO - 10.1111/j.0022-3506.2004.00282.x
M3 - Review article
C2 - 15210019
AN - SCOPUS:3142660149
SN - 0022-3506
VL - 72
SP - 845
EP - 876
JO - Journal of personality
JF - Journal of personality
IS - 4
ER -