Ant Colony Optimization and Local Weighted Structural Equation Modeling. A Tutorial on Novel Item and Person Sampling Procedures for Personality Research

Abstract

Measurement in personality development faces many psychometric problems. First, theory-based measurement models do not fit the empirical data in terms of traditional confirmatory factor analysis. Second, measurement invariance across age, which is necessary for a meaningful interpretation of age-associated personality differences, is rarely accomplished. Finally, continuous moderator variables, such as age, are often artificially categorized. This categorization leads to bias when interpreting differences in personality across age. In this tutorial, we introduce methods to remedy these problems. We illustrate how Ant Colony Optimization can be used to sample indicators that meet prespecified demands such as model fit. Further, we use Local Structural Equation Modeling to resample and weight subjects to study differences in the measurement model across age as a continuous moderator variable. We also provide a detailed illustration for both tools with the Neuroticism scale of the openly available International Personality Item Pool -NEO inventory using data from the UK sample (N = 15 827). Combined, both tools can remedy persistent problems in research on personality and its development. In addition to a step-by-step illustration, we provide commented syntax for both tools.

A (More) Behavioural Science of Personality in the Age of Multi–Modal Sensing, Big Data, Machine Learning, and Artificial Intelligence

“…data mining of websites and multi‐modal sensing) to sample a plethora of variables from many people and run some form of supervised or unsupervised machine learning over that big data to solve certain problems—often better or more accurately than we could do with simpler or piecemeal manual analyses (e.g. when creating new short forms of scales; Dörendahl & Greiff, 2020; Jankowsky, Olaru, & Schroeders, 2020; Olaru, Schroeders, Hartung, & Wilhelm, 2019). Harnessing these new opportunities should be encouraged; in fact, they probably should even be integrated into the common methodological, statistical, and quantitative training of psychologists.…”

Science Self-Concept – More Than the Sum of its Parts? A Comparison of Different Assessment and Modeling Approaches

“…These results should be interpreted with caution: at lower levels of the ability distribution, the robustness of the estimated SEM parameters is affected by a small effective sample size (Olaru et al, 2019). Nonetheless, the results indicate a significant drop in correlation among lesscompetent students.…”

“…After calculating the sample weights, the SEM model of interest will be fitted at each focal point within a defined moderator range, so that it is possible to explore trajectories of the model parameters along the continuous moderator. The method is explained in more detail in Hildebrandt et al (2016) or in a tutorial on LSEM (Olaru, Schroeders, Hartung, & Wilhelm, 2019). All LSEM were estimated with the R package sirt (Robitzsch, 2019).…”

A Reappraisal of the Threshold Hypothesis of Creativity and Intelligence

“…A recent extension of the structural equation models that ameliorate these drawbacks are local structural equation models (LSEM;Hildebrandt et al, 2016), which combines the advantages of confirmatory factor analysis and the treatment of intelligence as a continuous variable. In a nutshell, LSEM involves the fitting of several "conventional" structural equation models along the distribution of a continuous moderator with weighted observations (Olaru et al, 2019). The weight of each observation is based on the proximity of an observation to a specific value of the moderator, so that observations near this focal point provide more information to model estimation than more distant points.…”

“…In contrast, to MGCFA that rely on the categorization of intelligence as a moderator (e.g., Holling & Kuhn, 2008), LSEM allows for the investigation of the factor structure of creativity (Figure 2) across the intelligence continuum. LSEM is a person-sampling method that is applied to investigate deviations in the measurement model across observations (Olaru et al, 2019). Compared to MGCFA, which requires the grouping of participants, the observations in LSEM are weighted as a function of their proximity to a focal point of intelligence (Hildebrandt et al, 2009).…”