Paper Type
full
Description
The widespread availability of digital trace data provides new opportunities for researchers to understand human behaviors at a large scale. Sequences of behavior, captured when individuals interface with an information system, can be analyzed to uncover behavioral trends and tendencies. Rather than assume homogeneity among actors, in this study we introduce a method for identifying subsets of the population which demonstrate similar behavioral trends. The objective of this analysis would be to identify a finite set of behavioral archetypes, which we define as distinct patterns of action displayed by unique subsets of a population. This study makes a contribution to the literature by introducing a novel methodology for analyzing sequences of digital traces. We apply our technique to data from a lab experiment featuring thirty twenty-person teams communicating over Skype.
Recommended Citation
Schecter, Aaron and Contractor, Noshir, "Uncovering Latent Archetypes from Digital Trace Sequences: An Analytical Method and Empirical Example" (2019). ICIS 2019 Proceedings. 7.
https://aisel.aisnet.org/icis2019/data_science/data_science/7
Uncovering Latent Archetypes from Digital Trace Sequences: An Analytical Method and Empirical Example
The widespread availability of digital trace data provides new opportunities for researchers to understand human behaviors at a large scale. Sequences of behavior, captured when individuals interface with an information system, can be analyzed to uncover behavioral trends and tendencies. Rather than assume homogeneity among actors, in this study we introduce a method for identifying subsets of the population which demonstrate similar behavioral trends. The objective of this analysis would be to identify a finite set of behavioral archetypes, which we define as distinct patterns of action displayed by unique subsets of a population. This study makes a contribution to the literature by introducing a novel methodology for analyzing sequences of digital traces. We apply our technique to data from a lab experiment featuring thirty twenty-person teams communicating over Skype.