Subtask Analysis of Process Data Through a Predictive Model
This provides a solution for researchers analyzing behavioral patterns in process data, though it is incremental as it builds on existing segmentation and predictive modeling techniques.
The paper tackles the challenge of analyzing irregular and large-scale human-computer interaction process data by developing a computationally efficient method that segments lengthy processes into short subprocesses for complexity reduction, clustering, and interpretation, using sequential action predictability with a predictive model and Shannon entropy, and demonstrates its application on PIAAC 2012 data.
Response process data collected from human-computer interactive items contain rich information about respondents' behavioral patterns and cognitive processes. Their irregular formats as well as their large sizes make standard statistical tools difficult to apply. This paper develops a computationally efficient method for exploratory analysis of such process data. The new approach segments a lengthy individual process into a sequence of short subprocesses to achieve complexity reduction, easy clustering and meaningful interpretation. Each subprocess is considered a subtask. The segmentation is based on sequential action predictability using a parsimonious predictive model combined with the Shannon entropy. Simulation studies are conducted to assess performance of the new methods. We use the process data from PIAAC 2012 to demonstrate how exploratory analysis of process data can be done with the new approach.