STRATEGIC ENTREPRENEURIALISM IN ANALYSIS: GLOBAL PROBLEMS IN RESEARCH

A decade ago, a series of published articles debated the amount of method variance in organizational measures from all over' the world (Bagozzi and Yi, 1990; Spector, 1987; Williams, Cote, and Buckley, 1989). Interestingly, while all three studies analyzed the same data sets, they reached different conclusions concerning the presence of method variance. Spector (1987) reported that method effects were relatively minor, Bagozzi and Yi (1990) found them to be of moderate importance, and Williams et al. (1989) found method effects to be quite pervasive. Up to this time, no one has offered reasonable explanations as to why these results were found. While some of the differences in results may be due to the use of different analytic methods; even the use of the same technique (Confirmatory Factor Analysis - CFA) resulted in different conclusions. The differences found by these researchers (Bagozzi and Yi, 1990; Williams et al., 1989) are likely due to the multiple solutions problem caused by factor indeterminacy which is common with factor analytic techniques which is likely to exist with domestic and international data (Cattell and Gorsuch, 1963; Gorsuch, 1974; 1983; Rozeboom, 1992). A series of simulations were performed which indicate that there are, in fact, many reasonable solutions which can be fit to MTMM data using CFA and that therefore it appears that the multiple solutions problem is also significant for CFA as it is for other factor analytic methods. Some possible remedies for the problem of fitting MTMM models are proposed. We suggest that researchers need to be adaptable in the analytic methods which they employ and the topics that they explore (entrepreneurial researchers) and to carefully examine their data so that the research questions match the capabilities of the techniques utilized.