Situational Optimization Function Analysis: An Ideal Performance Analysis Inspired on Lewin's Equation

Víthor Rosa Franco

doi:10.1037/met0000319

Situational Optimization Function Analysis: An Ideal Performance Analysis Inspired on Lewin's Equation

2021-08-19

0

OA

AI

DOI：10.1037/met0000319

原文链接

原文求助

摘要

En 中文

This study presents the situational optimization function analysis (SOFA) and has three aims. First, to develop a Bayesian implementation of SOFA. Second, to compare this implementation with three other maximum likelihood-based models in their accuracy to estimate true scores. The third aim is to show how joint modeling can be used for validity research. A simulation study was used to examine the second aim, while an empirical example was used to illustrate the third aim. The simulation study used three data generating processes, with varying degrees of deviation from linear models and with different sample sizes. Results of the simulation study showed that the Bayesian implementation supersedes the other models. In the empirical example, data collected from 66 participants using an iterated prisoner dilemma and a scale measuring cooperation-competition attitudes were used. Results showed that joint modeling is the best fitting model, also increasing the correlation between the true scores of both measures (deviations from the iterated prisoner dilemma and the scale). Finally, implications, limitations and future studies are discussed. Translational Abstract Most psychometrical models are applied to correlational data based only on response patterns. However, it is possible to develop different models if the condition to which the study is subjected to is taken into account. Based on both Lewin's equation and ideal performance analysis, we propose a framework for developing measurement models in experimental settings where dispositional traits are measured after accounting for the effects of environmental, or situational, variables. We named our framework situational optimization function analysis (SOFA), as the estimates of individual differences are derived from a function that optimizes the effect of the experimentally manipulated situational variables. For illustrative purposes, we also developed a semiparametric Bayesian model that conforms to the SOFA framework and applied it in two different studies. The first study is a Monte Carlo simulation, aimed at testing the quality of our Bayesian model. The second study is an empirical study using data from a social dilemma setting, with the purpose of testing whether cooperative-competitive attitudes affect the behavior of participants in a game whose optimal strategy is to compete, as well as showing how joint modeling can be used in validity studies in the SOFA framework. We finish this article with some discussions on main theoretical and methodological implications, limitations of this study, and some guidelines for future studies.

Keyword:

measurement

SOFA framework

Bayesian modeling

joint modeling

Situational Optimization Function Analysis: An Ideal Performance Analysis Inspired on Lewin's Equation

摘要

期刊

机构

引用论文