Content area
Full Text
Received 4 November 1996
Final revision received 7 April 1998
Key words: causal modeling; measurement issues; partial least squares; strategic management research issues
Advances in causal modeling techniques have made it possible for researchers to simultaneously examine theory and measures. However, researchers must use these new techniques appropriately. In addition to dealing with the methodological concerns associated with more traditional methods of analysis, researchers using causal modeling approaches must understand their underlying assumptions and limitations.
Most researchers are well equipped with a basic understanding of LISREL-type models. In contrast, current familiarity with PLS in the strategic management area is low. The current paper reviews four recent studies in the strategic management area which use PLS. The review notes that the technique has been applied inconsistently, and at times inappropriately, and suggests standards for evaluating future PLS applications. Copyright (C) 1999 John Wiley & Sons, Ltd.
Advances in causal modelling techniques have made it possible for researchers to simultaneously examine theory and measures. Such techniques can be thought of as superior to more traditional techniques (e.g., multidimensional scaling, factor analysis) in that they permit: (1) the explicit inclusion of measurement error, and (2) an ability to incorporate abstract and unobservable constructs (Fornell, 1982). Bagozzi (1980) suggests that causal models provide researchers with four key benefits: (1) they make the assumptions, constructs, and hypothesized relationships in a theory explicit; (2) they add a degree of precision to a theory, since they require clear definitions of constructs, operationalizations, and functional relationships; (3) they permit a more complete representation of complex theories; and (4) they provide a formal framework for constructing and testing both theories and measures.
The best-known causal modeling technique is LISREL (Joreskog and Sorbom, 1989; Hagedoorn and Schakenraad, 1994). However, LISREL is poorly suited to deal with small data samples (Fornell, 1982), and can yield nonunique or otherwise improper solutions in some cases (Fornell and Bookstein, 1982). An alternative causal modeling approach known as Partial Least Squares (PLS) has been developed to avoid some of these limitations (Wold, 1974, 1985), although use of PLS requires its own set of assumptions. PLS has been used both in other business disciplines (e.g., Duxbury and Higgins, 1991; Hulland and Kleinmuntz, 1994; Smith and Barclay, 1997; Zinkhan,...