Identifying Mid-Range Patterns of Action: Tools for the Analysis of Organizational Routines

Capturing frequent action patterns can be very useful when trying to explain how individuals or organizations accommodate to different situations. Organizational routines have been defined as repetitive, recognizable patterns of interdependent action. However, there is a gap in our repertoire of sequence methods for identifying and analyzing the action patterns. We have tools for analyzing analyze pairs of actions, or whole sequences, but not the mid-range sub-sequences that can differentiate between performances of a routine and influence outcomes. Here, we introduce two classes of methods that address this gap: regular expressions and sequential pattern mining. Regular expressions are deductive; pattern mining is inductive. We demonstrate and compare the use of these methods using data from MindLab (a computer based, multi-player simulation game), and provide examples for each method.