An assessment of the construct validity of the Belbin Self-Perception Inventory and Observer's Assessment from the perspective of the five-factor model

Abstract

This paper examines the construct equivalency of the two measures used in Belbin's (1994) Interplace system - the self-report Self-Perception Inventory MK2 (SPI) and the observer-rating Observers Assessment (OA) - with regard to their use as parallel forms and their construct validity relative to the five-factor method (FFM) of personality assessment. A total of 152 managers (101 men, 51 women) completed the SPI and were assessed by four knowledgeable co-workers each (608 raters) using the OA. Results showed that the convergent and discriminant construct validities of the two measures did not support the contention that they were parallel forms. A second study further examined the equivalency of the measures when 123 participants self-rated both a normative SPI and the OA. Despite attenuating the effects of ipsativity and self vs. observer discrepancies the two measures were not supported as parallel forms. Finally, the construct validity of ipsative and normative versions of the SPI and a self-rater OA was examined by correlating their respective role scales with a standard measure of the five basic personality dimensions, the NEO-PI-R. Though the ipsative, forced-choice format of the SPI was identified as a source of invalidity, results showed that neither the SPI, either in ipsative or normative version, nor the OA evidenced convergent and discriminant construct validity relative to the NEO-PI-R. Implications of these findings are discussed. © 1996 The British Psychological Society.