Meta-analysis of the impact of research methods on findings of first-mover advantage

A long-standing hypothesis is that firms that enter a market early (‘first movers’) tend to have higher performance than their followers (‘first-mover advantage’). Recently, researchers have begun to argue that the statistical tests that support this relationship are limited in their applicability. That is, it is suggested that because of the methods used, these tests show the relationship only for certain subsets of firms, markets, and types of performance. We performed a meta-analysis to determine whether the findings are in fact sensitive to the methods used. We discovered that tests using market share as their performance measure were sharply and significantly more likely to find a first-mover advantage than tests using other measures (such as profitability or survival). Also significantly more likely to find an advantage were tests that sample from individually selected industries and those that include no measures of the entrants' competitive strength. Conversely, we found little evidence that ‘survivor bias’ (the exclusion of nonsurviving entrants from the sample) affects a test's findings. The data further suggest that tests that use none of the questioned research practices will find a first-mover advantage no more often than can be accounted for by random statistical error alone.