The Multidimensionality of Firm Size
Posted: 26 Oct 1998
"Firm size" is frequently used in accounting research as a proxy for theoretical constructs or as a control variable to mitigate missing variables bias. We provide empirical evidence that "firm size" is not uniquely related to a single theoretical construct and hence lacks construct validity. The lack of construct validity means that the interpretation of any empirical result with "firm size" as a variable is theoretically ambiguous. We apply cluster analysis and confirmatory factor analysis to data generated by a sample of accounting academics and identify five dimensions of the theoretical domain that have been proxied by measures of "firm size" in the literature. If empirical research is to generate theoretically interpretable results, refined proxies must be developed for each of these dimensions.
JEL Classification: M41, C31
Suggested Citation: Suggested Citation