Multivariate Behavioral Research最新文献

This special issue is a collection of papers inspired by Dr. Molenaar's work and innovations - a tribute to his passion for advancing science and his ability to ignite a spark of creativity and innovation in multiple generations of scientists. Following Dr. Molenaar's creative breadth, the papers address a wide variety of topics - sharing of new methodological developments, ideas, and findings in idiographic science, study of intraindividual variation, behavioral genetics, model inference/identification/selection, and more.

With the increased use of time series data in human research, ranging from ecological momentary assessments to data passively obtained, researchers can explore dynamic processes more than ever before. An important question researchers must ask themselves is, do I think all individuals have similar processes? If not, how different, and in what ways? Dr. Peter Molenaar's work set the foundation to answer these questions by providing insight into individual-level analysis for processes that are assumed to differ across individuals in at least some aspects. Currently, such assumptions do not have a clear taxonomy regarding the degree of homogeneity in the patterns of relations among variables and the corresponding parameter values. This paper provides the language with which researchers can discuss assumptions inherent in their analyses. We define strict homogeneity as the assumption that all individuals have an identical pattern of relations as well as parameter values; pattern homogeneity assumes the same pattern of relations but parameter values can differ; weak homogeneity assumes there are some (but not all) generalizable aspects of the process; and no homogeneity explicitly assumes no population-level similarities in dynamic processes across individuals. We demonstrate these assumptions with an empirical data set of daily emotions in couples.

Inspired by Peter Molenaar's Houdini transformation, we consider the idea of touchstones between different models. Touchstones represent instances where models that appear different on the surface can have equivalent characteristics. Touchstones can appear as identical tests of model parameters. They can exist in the mean structure, in the covariance structure, or in both. In the latter case, the models will generate identical mean and covariance structures and will fit the data equally well. After showing some examples of touchstones and how they result from constraints on a general model, we show how that idea can suggest Molenaar's Houdini transformation. This transformation allows one to take a latent variable model and derive an equivalent model comprised solely of manifest (observed) variables. As equivalent models, the parameters of one can be transformed into the parameters of the other.

Increasingly, behavioral scientists encounter data where several individuals were measured on multiple variables over numerous occasions. Many current methods combine these data, assuming all individuals are randomly equivalent. An extreme alternative assumes no one is randomly equivalent. We propose state space mixture modeling as one possible compromise. State space mixture modeling assumes that unknown groups of people exist who share the same parameters of a state space model, and simultaneously estimates both the state space parameters and group membership. The goal is to find people that are undergoing similar change processes over time. The present work demonstrates state space mixture modeling on a simulated data set, and summarizes the results from a large simulation study. The illustration shows how the analysis is conducted, whereas the simulation provides evidence of its general validity and applicability. In the simulation study, sample size had the greatest influence on parameter estimation and the dimension of the change process had the greatest impact on correctly grouping people together, likely due to the distinctiveness of their patterns of change. State space mixture modeling offers one of the best-performing methods for simultaneously drawing conclusions about individual change processes while also analyzing multiple people.

Psychological science is divided into two distinct methodological traditions. One tradition seeks to understand how people function at the individual level, while the other seeks to understand how people differ from each other. Methodologies that have grown out of these traditions typically rely on different sources of data. While both use statistical models to understand the structure of the data, and these models are often similar, Molenaar (2004) showed that results from one type of analysis rarely transfer to the other, unless unrealistic assumptions hold. This raises the question how we may integrate these approaches. In this paper, we argue that formalized theories can be used to connect intra- and interindividual levels of analysis. This connection is indirect, in the sense that the relationship between theory and data is best understood through the intermediate level of phenomena: robust statistical patterns in empirical data. To illustrate this, we introduce a distinction between intra- and interindividual phenomena, and argue that many psychological theories will have implications for both types of phenomena. Formalization provides us with a methodological tool for investigating what kinds of intra- and interindividual phenomena we should expect to find if the theory under consideration were true.

Advances in ability to comprehensively record individuals' digital lives and in AI modeling of those data facilitate new possibilities for describing, predicting, and generating a wide variety of behavioral processes. In this paper, we consider these advances from a person-specific perspective, including whether the pervasive concerns about generalizability of results might be productively reframed with respect to transferability of models, and how self-supervision and new deep neural network architectures that facilitate transfer learning can be applied in a person-specific way to the super-intensive longitudinal data arriving in the Human Screenome Project. In developing the possibilities, we suggest Molenaar add a statement to the person-specific Manifesto - "In short, the concerns about generalizability commonly leveled at the person-specific paradigm are unfounded and can be fully and completely replaced with discussion and demonstrations of transferability."

Gender is person-specific, and it influences and is influenced by a breadth of multidimensional psychological factors, including cognition. Directionality is important for research on gender and cognition, as debate surrounds, for instance, whether masculine self-concepts precede spatial skills, or whether the reverse is true. In order to provide novel insights into the individualized nature of these relations, a person-specific network approach devised by Peter Molenaar and the first author - group iterative multiple model estimation for multiple solutions (GIMME-MS) - was applied to 75-day intensive longitudinal data on gender self-concept (i.e., femininity-masculinity, instrumentality, and expressivity) and cognition (i.e., mental rotations and verbal recall) from 103 young adults. GIMME-MS estimates individualized networks that contain same-day and next-day directed relations, prioritizing relations common across participants. It is ideal for analyzing behavioral time series with unclear directionality, as it generates multiple solutions from which an optimal one is selected. GIMME-MS revealed notable heterogeneity in the presence, direction, and nature of relations from gender self-concept to cognition (∼26% of participants) and vice versa (∼21% of participants). Findings are wholly novel in revealing the person-specific nature of gender and its cognitive dynamics, yet somehow, unsurprising given the revolutionary corpus of Peter Molenaar.

Testing for Granger causality relies on estimating the capacity of dynamics in one time series to forecast dynamics in another. The canonical test for such temporal predictive causality is based on fitting multivariate time series models and is cast in the classical null hypothesis testing framework. In this framework, we are limited to rejecting the null hypothesis or failing to reject the null - we can never validly accept the null hypothesis of no Granger causality. This is poorly suited for many common purposes, including evidence integration, feature selection, and other cases where it is useful to express evidence against, rather than for, the existence of an association. Here we derive and implement the Bayes factor for Granger causality in a multilevel modeling framework. This Bayes factor summarizes information in the data in terms of a continuously scaled evidence ratio between the presence of Granger causality and its absence. We also introduce this procedure for the multilevel generalization of Granger causality testing. This facilitates inference when information is scarce or noisy or if we are interested primarily in population-level trends. We illustrate our approach with an application on exploring causal relationships in affect using a daily life study.

Following Kelderman and Molenaar's demonstration that a factor model with person specific factor loadings is almost indistinguishable from the standard factor model in terms of overall fit, we examined person specific measurement models in Item Response Theory, person specific discrimination and difficulty parameters were created by adding random variation at the item by person level. Using standard fitting algorithms for the 2PL IRT there was modest evidence of person- or item-level misfit using common diagnostic tools. The item difficulties were well-estimated, but the item discriminations were noticeably underestimated. As found by Kelderman and Molenaar, factor scores were estimated with less than expected reliability due to the underlying heterogeneity. The person specific models considered here are basically limiting cases of IRT models with multilevel, mixture, or differential item functioning structure. We conclude with some thoughts regarding real-world sources of heterogeneity that might go unacknowledged in common testing applications.

Rapid developments over the last several decades have brought increased focus and attention to the role of time scales and heterogeneity in the modeling of human processes. To address these emerging questions, subgrouping methods developed in the discrete-time framework-such as the vector autoregression (VAR)-have undergone widespread development to identify shared nomothetic trends from idiographic modeling results. Given the dependence of VAR-based parameters on the measurement intervals of the data, we sought to clarify the strengths and limitations of these methods in recovering subgroup dynamics under different measurement intervals. Building on the work of Molenaar and collaborators for subgrouping individual time-series by means of the subgrouped chain graphical VAR (scgVAR) and the subgrouping option in the group iterative multiple model estimation (S-GIMME), we present results from a Monte Carlo study aimed at addressing the implications of identifying subgroups using these discrete-time methods when applied to continuous-time data. Results indicate that discrete-time subgrouping methods perform well at recovering true subgroups when the measurement intervals are large enough to capture the full range of a system's dynamics, either via lagged or contemporaneous effects. Further implications and limitations are discussed therein.