Date of Award
May 2021
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Industrial-Organizational Psychology
Committee Member
Fred Switzer
Committee Member
Patrick Rosopa
Committee Member
Grigori Yourganov
Committee Member
Claudio Cantalupo
Abstract
This study replicates and then refutes portions of an article published in Nature by Gerlach, Farb, Revelle, & Nunes Amaral (2018) on personality clusters. The central claim of the current study is that the clusters were actually biases in the data, based on central tendency and social desirability biases. We find that with proper preprocessing of our data, that all personality clusters found in the Gerlach et al. (2018) study cease to exist as anything but random noise. The interpretation of these findings is that careless responding, response styles, and characteristics of Likert scale style data can lead to artificial clustering, leading to improper interpretation of the frequency of occurrence of certain arrangements of personality traits. The implications of these findings are that unsupervised machine learning approaches can be especially useful in personality research, but misuse of these approaches can lead to misleading results.
Recommended Citation
Ligato, Joseph, "Personality Style Clusters Using Unsupervised Machine Learning" (2021). All Dissertations. 2769.
https://tigerprints.clemson.edu/all_dissertations/2769