Investigating the Structure of Intelligence Using Latent Variable and Psychometric Network Modeling: A Commentary and Reanalysis

In a recent publication in the Journal of Intelligence, Dennis McFarland mischaracterized previous research using latent variable and psychometric network modeling to investigate the structure of intelligence. Misconceptions presented by McFarland are identified and discussed. We reiterate and clarify the goal of our previous research on network models, which is to improve compatibility between psychological theories and statistical models of intelligence. WAIS-IV data provided by McFarland were reanalyzed using latent variable and psychometric network modeling. The results are consistent with our previous study and show that a latent variable model and a network model both provide an adequate fit to the WAIS-IV. We therefore argue that model preference should be determined by theory compatibility. Theories of intelligence that posit a general mental ability (general intelligence) are compatible with latent variable models. More recent approaches, such as mutualism and process overlap theory, reject the notion of general mental ability and are therefore more compatible with network models, which depict the structure of intelligence as an interconnected network of cognitive processes sampled by a battery of tests. We emphasize the importance of compatibility between theories and models in scientific research on intelligence.

Download Full-text

Psychometric Models of Individual Differences in Reading Comprehension: A Reanalysis of Freed, Hamilton, and Long (2017)

10.31234/osf.io/hvzyx ◽

2019 ◽

Author(s):

Sara Anne Goring ◽

Christopher J. Schmank ◽

Michael J. Kane ◽

Andrew R. A. Conway

Keyword(s):

Reading Comprehension ◽

Individual Differences ◽

Cognitive Abilities ◽

Latent Variable ◽

Convergent Validity ◽

Cognitive Task ◽

Network Modeling ◽

Network Models ◽

Measurement Model ◽

Common Cause

Individual differences in reading comprehension have often been explored using latent variable modeling (LVM), to assess the relative contribution of domain-general and domain-specific cognitive abilities. However, LVM is based on the assumption that the observed covariance among indicators of a construct is due to a common cause (i.e., a latent variable; Pearl, 2000). This is a questionable assumption when the indicator variables are measures of performance on complex cognitive tasks. According to Process Overlap Theory (POT; Kovacs & Conway, 2016), multiple processes are involved in cognitive task performance and the covariance among tasks is due to the overlap of processes across tasks. Instead of a single latent common cause, there are thought to be multiple dynamic manifest causes, consistent with an emerging view in psychometrics called network theory (Barabási, 2012; Borsboom & Cramer, 2013). In the current study, we reanalyzed data from Freed et al. (2017) and compared two modeling approaches: LVM (Study 1) and psychometric network modeling (Study 2). In Study 1, two exploratory LVMs demonstrated problems with the original measurement model proposed by Freed et al. Specifically, the model failed to achieve discriminant and convergent validity with respect to reading comprehension, language experience, and reasoning. In Study 2, two network models confirmed the problems found in Study 1, and also served as an example of how network modeling techniques can be used to study individual differences. In conclusion, more research, and a more informed approach to psychometric modeling, is needed to better understand individual differences in reading comprehension.

Download Full-text

Recent Integrations of Latent Variable Network Modeling With Psychometric Models

Frontiers in Psychology ◽

10.3389/fpsyg.2021.773289 ◽

2021 ◽

Vol 12 ◽

Author(s):

Selena Wang

Keyword(s):

Graphical Models ◽

Latent Variable ◽

Network Modeling ◽

Network Models ◽

Latent Variable Models ◽

Independent Learning ◽

Psychometric Models ◽

R Packages

The combination of network modeling and psychometric models has opened up exciting directions of research. However, there has been confusion surrounding differences among network models, graphic models, latent variable models and their applications in psychology. In this paper, I attempt to remedy this gap by briefly introducing latent variable network models and their recent integrations with psychometric models to psychometricians and applied psychologists. Following this introduction, I summarize developments under network psychometrics and show how graphical models under this framework can be distinguished from other network models. Every model is introduced using unified notations, and all methods are accompanied by available R packages inducive to further independent learning.

Download Full-text

Interpreting psychometric models

10.31237/osf.io/x6a7s ◽

2019 ◽

Author(s):

Riet van Bork

Keyword(s):

Latent Variable ◽

Latent Trait ◽

Network Models ◽

Latent Variable Models ◽

Causal Mechanism ◽

Variable Model ◽

Psychological Construct ◽

Psychological Constructs ◽

Common Causes ◽

Psychometric Models

The field of psychometrics aims to develop theories on how to measure psychological constructs through observable behavior. This dissertation focuses on two psychometric theories that differ in how the psychological construct is related to observable behaviors. Latent trait theory understands psychological constructs as underlying common causes of observed behavior that explain the associations between certain behaviors. Alternatively, in the psychological network theory, behaviors correlate because they mutually reinforce each other and the psychological construct refers to the resulting cluster of associated behaviors. These different theories about how to conceptualize psychological constructs and how to relate these constructs to observable behavior can be formally defined in a set of equations and assumptions that make up a psychometric model. The chapters in this dissertation focus on two types of psychometric models: Latent variable models and network models. Part I of the dissertation focuses on the interpretation of the latent variable model. Part II of the dissertation makes a comparison between latent variable models and network models. While psychometric models can be interpreted as representations of a theory about the data-generating mechanism, this is not necessary. Psychometric models are often viewed as mere descriptions of data. This dissertation shows the importance of thinking through the choice of interpreting psychometric models either as a representation of a causal mechanism or as a description of the data and provides insights in the implications of that choice.

Download Full-text

Differential Effects of Executive Processes on Working Memory

Journal of Individual Differences ◽

10.1027/1614-0001/a000211 ◽

2016 ◽

Vol 37 (4) ◽

pp. 239-249

Author(s):

Xuezhu Ren ◽

Tengfei Wang ◽

Karl Schweizer ◽

Jing Guo

Keyword(s):

Working Memory ◽

Cognitive Processes ◽

Latent Variable ◽

Working Memory Capacity ◽

Memory Capacity ◽

Attention Control ◽

Executive Processes ◽

Differential Effects ◽

Prepotent Response ◽

Two Measures

Abstract. Although attention control accounts for a unique portion of the variance in working memory capacity (WMC), the way in which attention control contributes to WMC has not been thoroughly specified. The current work focused on fractionating attention control into distinctly different executive processes and examined to what extent key processes of attention control including updating, shifting, and prepotent response inhibition were related to WMC and whether these relations were different. A number of 216 university students completed experimental tasks of attention control and two measures of WMC. Latent variable analyses were employed for separating and modeling each process and their effects on WMC. The results showed that both the accuracy of updating and shifting were substantially related to WMC while the link from the accuracy of inhibition to WMC was insignificant; on the other hand, only the speed of shifting had a moderate effect on WMC while neither the speed of updating nor the speed of inhibition showed significant effect on WMC. The results suggest that these key processes of attention control exhibit differential effects on individual differences in WMC. The approach that combined experimental manipulations and statistical modeling constitutes a promising way of investigating cognitive processes.

Download Full-text

The Validity of Specific Cognitive Abilities for the Prediction of Training Success in Germany

Journal of Personnel Psychology ◽

10.1027/1866-5888/a000110 ◽

2014 ◽

Vol 13 (3) ◽

pp. 123-133 ◽

Cited By ~ 11

Author(s):

Wiebke Goertz ◽

Ute R. Hülsheger ◽

Günter W. Maier

Keyword(s):

Cognitive Abilities ◽

Personnel Selection ◽

Meta Analysis ◽

Mental Ability ◽

Occupational Category ◽

General Mental Ability ◽

Training Success ◽

Specific Cognitive Abilities ◽

Better Than

General mental ability (GMA) has long been considered one of the best predictors of training success and considerably better than specific cognitive abilities (SCAs). Recently, however, researchers have provided evidence that SCAs may be of similar importance for training success, a finding supporting personnel selection based on job-related requirements. The present meta-analysis therefore seeks to assess validities of SCAs for training success in various occupations in a sample of German primary studies. Our meta-analysis (k = 72) revealed operational validities between ρ = .18 and ρ = .26 for different SCAs. Furthermore, results varied by occupational category, supporting a job-specific benefit of SCAs.

Download Full-text

Is Retest Bias Biased?

Journal of Personnel Psychology ◽

10.1027/1866-5888/a000149 ◽

2016 ◽

Vol 15 (2) ◽

pp. 45-54 ◽

Cited By ~ 4

Author(s):

Jason G. Randall ◽

Anton J. Villado ◽

Christina U. Zimmer

Keyword(s):

Sex Differences ◽

Emotional Stability ◽

Adverse Impact ◽

Mental Ability ◽

Ability Test ◽

General Mental Ability ◽

Psychological Mechanisms

Abstract. The purpose of this study was to test for race and sex differences in general mental ability (GMA) retest performance and to identify the psychological mechanisms underlying these differences. An initial and retest administration of a GMA assessment separated by a six-week span was completed by 318 participants. Contrary to our predictions, we found that race, sex, and emotional stability failed to moderate GMA retest performance. However, GMA assessed via another ability test and conscientiousness both partially explained retest performance. Additionally, we found that retesting may reduce adverse impact ratios by lowering the hiring threshold. Ultimately, our findings reinforce the need for organizations to consider race, sex, ability, and personality when implementing retesting procedures.

Download Full-text