Organizational configurations and performance: the role of statistical
power in extant research
Strat. Mgmt. J., 20 (4):
385--395 (April 1999)DOI: 10.1002/(SICI)1097-0266(199904)20:4<385::AID-SMJ24>3.0.CO;2-X
Abstract
The relationship between organizational configurations and performance
has been a frequent albeit controversial subject of research in the
field of strategic management. Many studies have failed to find a
link between configurations and performance, leading prominent researchers
to question the value of the concept of organizational configurations.
Before the concept can be discarded, however, other plausible explanations
for the lack of findings should be examined. This paper examines
the possible role of statistical power. Specifically, it may be that
the sample sizes in many studies are too small to detect relationships
between configurations and performance when such links are, in fact,
present. Analysis of 24 tests of the configurations–performance link
revealed that only 8 percent had samples large enough to detect all
important relationships. Thus, there is reason to suspect that insufficient
statistical power may help to explain extant results. Given these
findings, suggestions are presented for improving the ability of
configurational studies to detect relationships.