Abstract
Influential individuals who shape the attitudes and behaviors of their peers are termed opinion leaders. The interpretability of research on opinion leadership, however, has frequently been hampered by the use of divergent instruments that follow either a domain-specific view of opinion leadership or a more domain-independent trait conceptualization. In two studies, multitrait-multimethod matrices are analyzed regarding the convergent and discriminant validity of both approaches. Study I ( N = 407) demonstrates the stability of the opinion leadership scales over time and their discriminant validity to established measures in personality (e.g., extraversion) and attitudinal research (e.g., involvement). Study II ( N = 185) replicates these results in the form of a multitrait-multiinformant design, demonstrating high convergence of self- and peer-assessments. However, different operationalizations of domain-independent opinion leadership displayed limited convergent validity, indicating that they capture different, albeit related, traits.
Original language | English |
---|---|
Pages (from-to) | 94-102 |
Number of pages | 6 |
Journal | Journal of Individual Differences |
Volume | 32 |
Issue number | 2 |
Publication status | Published - 2011 |
Fields of science
- 501001 General psychology
- 501003 Occupational psychology
- 501004 Differential psychology
- 501 Psychology
- 501016 Educational psychology
- 501021 Social psychology
- 501020 Legal psychology
- 508007 Communication science
- 508009 Media research
JKU Focus areas
- Management and Innovation
- Social Systems, Markets and Welfare States
- Social and Economic Sciences (in general)