Author ORCID Identifier
https://orcid.org/0000-0002-8671-0518
Date of Award
5-2-2022
Degree Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Psychology
First Advisor
Dr. John Horgan
Second Advisor
Dr. Anthony Lemieux
Third Advisor
Dr. Dominic Parrott
Fourth Advisor
Dr. Elizabeth Tighe
Fifth Advisor
Dr. Therese Pigott
Abstract
Radicalism, extremism, and related phenomena have been measured myriad ways, with little standardization. The most widely used metric—the Activism and Radicalism Intentions Scales (the ARIS: Moskalenko & McCauley, 2009)—has been translated, rescaled, reworded, reorganized, and used with populations never originally tested or necessarily intended for, with little scrutiny. To support the ARIS’s use across the past decade of research, Fodeman tested for Measurement Equivalence/ Invariance (ME/I), or potential bias in the measurement of radicalism across the past decade's use of the ARIS. Fodeman did so via Integrated Data Analysis (IDA)--an approach similar to, but more powerful than, meta-analysis. Among the comparable studies, Fodeman did not detect any significant bias. Indeed, the only significant difference between comparable studies was a significantly higher radicalism score for Decker and Pyrooz’s (2019) sample--the only criminal sample across all bodies of research with the ARIS. This was not unexpected, as radicalism is commonly associated with criminality. While this work gives some credence to the use of the ARIS across multiple contexts, more in-depth analyses with larger sample sizes will have to test for ME/I between cross-classified cohorts (e.g., by translation, country, age group, general vs. specific vs. at risk populations, etc.), activism, and other radicalism items. When advanced statistical techniques such as Moderated Non-Linear Factor Analysis (MNLFA) are further developed, future studies will also have to test for ME/I across rescaling of ARIS items, likely requiring a bridging study in which multiple scales are given to the same participants. It is this type of intensive, rigorous data collection and statistical analysis found in medicine, education, and other content areas to which we radicalism researchers can aspire.
DOI
https://doi.org/10.57709/28914817
Recommended Citation
Fodeman, Ari, "(Statistical) Power to the People: Testing for Measurement Invariance Across Integrated Radicalism Research." Dissertation, Georgia State University, 2022.
doi: https://doi.org/10.57709/28914817
File Upload Confirmation
1