Same technology, different outcome? Reinterpreting Barley's Technology as an Occasion for Structuring

In the last few decades, several studies have found the same technology implemented in highly similar organizational settings to be associated with very different consequences for structure and process. The seminal study in this stream of research is Barley's (1986) Technology as an Occasion for Structuring, which reported that two similarly composed radiology departments implemented the same technology (computerized tomography scanners), yet experienced very different structural outcomes. In this paper I re-analyze the original study's data under three different statistical assumptions. First, I performed an arcsine transformation on the dependent variable where the original study used the raw probabilities. Second, I specified a power regression model in which the original study employed a linear regression. Finally, I user fewer dummy variables in the ‘combined’ regression models to determine the distinct phases through which the two hospitals evolved. Taken together, these assumptions produce very different results from the original study. Specifically they indicate that the radiology departments did not decentralize at different rates and did not do so over a different number of distinct phases. From my analysis come three specific recommendations for research investigating the consequences of information technology in similarly constituted organizations: (1) exchange the default assumption of homogeneity of outcomes with one of heterogeneity; (2) explicitly account for both the observable properties of technology and the context of its use; and (3) state clearly and a priori the standard used to classify structural and organizational outcomes as ‘different’.