Person-centered data analyses: Observation Oriented Modeling as an alternative and rational data analytics approach
Publisher
University of Tennessee at Chattanooga
Place of Publication
Chattanooga (Tenn.)
Abstract
“How many people in my study/intervention behaved or responded in a manner consistent with theoretical expectation?" This is arguably one of the most important questions we can ask about research and practice efforts as psychologists. Most traditional statistical analyses, however, are designed to answer questions about average group differences, variable associations, and relative risk. Alternative methods of thinking about and analyzing data are therefore needed. In this presentation I demonstrate and discuss novel techniques for computing a simple percentage for a wide variety of study designs that quantifies the number of individuals in a given study whose pattern of behavior or responses match theoretical expectation. This percentage essentially treats persons as an effect size, and it can easily be understood by scientists, professionals, and laypersons alike because its meaning requires no statistical training to understand. The methods for computing this person-centered effect size are moreover rooted in traditional nonparametric procedures of hypothesis testing and can, therefore, be used to draw inferences. As will be discussed, the inferences drawn from these techniques are explanatory in nature rather than focused on parameter estimation. I will demonstrate how these novel, straightforward, and intuitively compelling techniques work by presenting and discussing several re-analyses of recently published studies.
Date
October 2019
Subject
Industrial and organizational psychology
Document Type
presentations
Language
English
Rights
http://rightsstatements.org/vocab/InC/1.0/
License
http://creativecommons.org/licenses/by/4.0/
Person-centered data analyses: Observation Oriented Modeling as an alternative and rational data analytics approach
“How many people in my study/intervention behaved or responded in a manner consistent with theoretical expectation?" This is arguably one of the most important questions we can ask about research and practice efforts as psychologists. Most traditional statistical analyses, however, are designed to answer questions about average group differences, variable associations, and relative risk. Alternative methods of thinking about and analyzing data are therefore needed. In this presentation I demonstrate and discuss novel techniques for computing a simple percentage for a wide variety of study designs that quantifies the number of individuals in a given study whose pattern of behavior or responses match theoretical expectation. This percentage essentially treats persons as an effect size, and it can easily be understood by scientists, professionals, and laypersons alike because its meaning requires no statistical training to understand. The methods for computing this person-centered effect size are moreover rooted in traditional nonparametric procedures of hypothesis testing and can, therefore, be used to draw inferences. As will be discussed, the inferences drawn from these techniques are explanatory in nature rather than focused on parameter estimation. I will demonstrate how these novel, straightforward, and intuitively compelling techniques work by presenting and discussing several re-analyses of recently published studies.
Department
University of Tennessee at Chattanooga. Dept. of Psychology