Date of Award
August 2020
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Industrial-Organizational Psychology
Committee Member
Marissa L Shuffler
Committee Member
Gerald F Goodwin
Committee Member
Patrick J Rosopa
Committee Member
Fred S Switzer
Abstract
Teams have been an integral part of organizational success for several decades and as such, researchers have sought to better understand all aspects of work teams. To better inform research and practice, Marks, Mathieu and Zaccaro (2001) advanced a theory and framework of team processes that has become a seminal piece in our field. Their theory proposed that ten team processes could be mapped on to three second order constructs (transition, action, and interpersonal phases). Mathieu and colleagues (2019) developed and validated a measure designed specifically to align with Marks et al. (2001) framework. While much needed, this measure is not without limitations, namely its self-report nature and associated subjectivity.
The current study proposes a means for overcoming those limitations by using machine learning to automate the Mathieu et al. (2019) measure. This study used traditional human coding methods to code data from three different sources to include teams across various contexts. Data was used from NASA HERA teams, medical teams, and student engineering teams. Then, the researcher trained various models using Natural Language Classifier software (provided through IBM Watson) to create an automated coding scheme. The results of this study are mixed. Using Natural Language Classifier, various models were trained and tested according to the Marks et al. (2001) framework. However, once tested, the accuracy of the model was not up to standard. This study provides a fruitful avenue for future research; the models can be refined by collecting further data and then retraining the models.
Recommended Citation
Flynn, Michelle Leigh, "Towards More Efficient Behavioral Coding of Teamwork via Machine Learning" (2020). All Dissertations. 2685.
https://open.clemson.edu/all_dissertations/2685