Barley, Stephen R. "Technology as an occaision for structuring: evidence from observations of CT scanners and the social order of radiology departments", Administrative Science Quarterly, 31 (1986) 78-108
In this paper, Barley demonstrates the effectiveness of an observational test methodology to study the structural impact of technology under dualistic assumptions. He finds that adoption of new technologies can have different effects on structure depending on the actions of the members.
Barley 's testing premise is that since structure is temporal and exists through action, one must "adopt a longitudinal view as well as a cross-sectional view" when observing actions to assess the evolution of structure. New technology, when introduced into a structure-action dualism, becomes a social object itself "whose meanings [are] defined by the context of [it's] use".
Based on a dualistic model, Barley develops an observation methodology that categories observed action into scripts and measures script frequency and distribution against observed structural phases. He uses this methodology to compare the impact of CT Scanner technology on structure (especially roles and norms) at radiology departments in two Boston hospitals.
Barley found that the initial staffing decisions and relative initial CT scanner experience between the radiologists and technicians had a strong effect on the development of roles and norms in the use of CT scanners. At one hospital, the high initial experience level of the technicians relative to the radiologists caused the radiologists to adopt a "hands-off" policy, partly due to anxiety over ignorance of CT scanner technology that potentially could weaken the dominant role of the radiologist in the organization.
At the other hospital, inexperienced technicians initially performed poorly due to conflictory and micro-managed training supervision by two radiologists experienced in CT technology. Only when the technicians began to work with inexperienced radiologists did they receive the support and consistency needed to learn and perform their tasks confidently.
In effect, sociologists have as much trouble predicting the behavior of a specific organization as psychologists have predicting the behavior of a specific individual.