A framework is presented for building and testing models of dynamic regulation by categorizing sources of differences between theories of dynamics. A distinction is made between the dynamics of change, i.e., how a system self–regulates on a short time scale, and change in dynamics, i.e., how those dynamics may themselves change over a longer time scale. In order to clarify the categories, models are first built to estimate individual differences in equilibrium value and equilibrium change. Next, models are presented in which there are individual differences in parameters of dynamics such as frequency of fluctuations, damping of fluctuations, and amplitude of fluctuations. Finally, models for within–person change in dynamics over time are proposed. Simulations demonstrating feasibility of these models are presented and OpenMx scripts for fitting these models have been made available in a downloadable archive along with scripts to simulate data so that a researcher may test selected models’ feasibility within a chosen experimental design.

The article accepted for publication can be downloaded as a PDF.

]]>The ten year anniversary of the COGITO Study provides an opportunity to revisit the ideas behind the Cattell data box. Three dimensions of the persons × variables × time data box are discussed in the context of three categories of researchers each wanting to answer their own categorically different question. The example of the well-known speed-accuracy tradeoff is used to illustrate why these are three different categories of statistical question. The 200 persons by 100 variables by 100 occasions of measurement COGITO data cube presents a challenge to integrate theories and methods across the dimensions of the data box. A conceptual model is presented for the speed-accuracy tradeoff example that could account for cross-sectional between persons effects, short term dynamics, and long term learning effects. Thus, two fundamental differences between the time axis and the other two axes of the data box include ordering and time scaling. In addition, nonstationarity in human systems is a pervasive problem along the time dimension of the data box. To illustrate, the difference in nonstationarity between dancing and conversation is discussed in the context of the interaction between theory, methods, and data. An information theoretic argument is presented that the theory-methods-data interaction is better understood when viewed as a conversation than as a dance. Entropy changes in the development of a theory-methods-data conversation provide one metric for evaluating scientific progress.

The article accepted for publication can be downloaded as a PDF.

]]>Damped Linear Oscillators estimated by 2nd-order Latent Differential Equation (LDE) have assumed a constant equilibrium and one oscillatory component. Lower-frequency oscillations may come from seasonal background processes, which non-randomly contribute to deviation from equilibrium at each occasion and confound estimation of dynamics over shorter timescales. Boker (2015) proposed a model of individual change on multiple timescales, but implementation, simulation, and applications to data have not been demonstrated. This study implemented a generalization of the proposed model; examined robustness to varied timescale ratios, measurement error, and occasions-per-person in simulated data; and tested for dynamics at multiple timescales in experience sampling affect data. Results show small standard errors and low bias to dynamic estimates at timescale ratios greater than 3:1. Below 3:1, estimate error was sensitive to noise and total occasions; rates of non-convergence increased. For affect data, model comparisons showed statistically significant dynamics at both timescales for both participants.

The article accepted for publication can be downloaded as a PDF.

]]>Maintained Individual Data Distributed Likelihood Estimation (MIDDLE) is a novel paradigm for research in the behavioral, social, and health sciences. The MIDDLE approach is based on the seemingly-impossible idea that data can be privately maintained by participants and never revealed to researchers, while still enabling statistical models to be fit and scientific hypotheses tested. MIDDLE rests on the assumption that participant data should belong to, be controlled by, and remain in the possession of the participants themselves. Distributed likelihood estimation refers to fitting statistical models by sending an objective function and vector of parameters to each participants’ personal device (e.g., smartphone, tablet, computer), where the likelihood of that individual’s data is calculated locally. Only the likelihood value is returned to the central optimizer. The optimizer aggregates likelihood values from responding participants and chooses new vectors of parameters until the model converges. A MIDDLE study provides significantly greater privacy for participants, automatic management of opt- in and opt-out consent, lower cost for the researcher and funding institute, and faster determination of results. Furthermore, if a participant opts into several studies simultaneously and opts into data sharing, these studies automatically have access to individual-level longitudinal data linked across all studies.

The final draft of the article accepted for publication can be downloaded as a PDF.

]]>An equilibrium involves a balancing of forces. Just as one maintains upright posture in standing or walking, many self-regulatory and interpersonal behaviors can be framed as a balancing act between an ever changing environment and within-person processes. The emerging balance between person and environment, the equilibria, are dynamic and adaptive in response to development and learning. A distinction is made between equilibrium achieved solely due to a short timescale balancing of forces and a longer timescale *preferred equilibrium* which we define as a state towards which the system slowly adapts. Together, these are developed into a framework that this article calls Adaptive Equilibrium Regulation (AER), which separates a regulatory process into two timescales: a faster regulation that automatically balances forces and a slower timescale adaptation process that reconfigures the fast regulation so as to move the system towards its preferred equilibrium when an environmental force persists over the longer timescale. This way of thinking leads to novel models for the interplay between multiple timescales of behavior, learning, and development.

The article accepted for publication can be downloaded as a PDF.

]]>The new software package OpenMx 2.0 for structural equation and other statistical modeling is introduced and its features are described. OpenMx is evolving in a modular direction and now allows a mix-and-match computational approach that separates model expectations from fit functions and optimizers. Major backend architectural improvements include a move to swappable open-source optimizers such as the newly-written CSOLNP. Entire new methodologies such as Item Factor analysis (IRT) and State-space modeling have been implemented. New model expectation functions including support for the expression of models in LISREL syntax and a simplified multigroup expectation function are available. Ease-of-use improvements include helper functions to standardize model parameters and compute their Jacobian-based standard errors, access to model components through standard R $ mechanisms, and improved tab completion from within the R Graphical User Interface.

The manuscript of this article accepted for publication can be downloaded as a PDF. This article may not exactly replicate the ﬁnal version published in Psychometrika. It is not the copy of record.

]]>This paper introduces an Item Factor Analysis (IFA) module for OpenMx, a free, open-source, and modular statistical modeling package that runs within the R programming environment on GNU/Linux, Mac OS X, and Microsoft Windows. The IFA module offers a novel model specification language that is well suited to programmatic generation and manipulation of models. Modular organization of the source code facilitates the easy addition of item models, item parameter estimation algorithms, optimizers, test scoring algorithms, and fit diagnostics all within an integrated framework. Three short example scripts are pre- sented for fitting item parameters, latent distribution parameters, and a multiple group model. The availability of both IFA and structural equation modeling in the same software is a step toward the unification of these two methodologies.

The manuscript of this article accepted for publication can be downloaded as a PDF. This article may not exactly replicate the ﬁnal version published in Educational and Psychological Measurement. It is not the copy of record.

]]>This is an application of contemporary multilevel regression modeling to the prediction of academic performances of 1st-year college students. At a first level of analysis, the data come from N > 16,000 students who were college freshman in 1994–1995 and who were also participants in high-level college athletics. At a second level of analysis, the student data were related to the different characteristics of the C = 267 colleges in Division I of the NCAA. The analyses presented here initially focus on the prediction of freshman GPA from a variety of high school academic variables. The models used are standard multilevel regression models, but we examine nonlinear prediction within these multilevel models, and additional outcome variables are considered. The multilevel results show that (a) high school grades are the best available predictors of freshman college grades, (b) the ACT and SAT test scores are the next best predictors available, (c) the number of high school core units taken does not add to this prediction but does predict credits attained, (d) college graduation rate has a second-level effect of a small negative outcome on the average grades, and (e) nonlinear models indicate stronger effects for students at higher levels of the academic variables. These results show that standard multilevel models are practically useful for standard validation studies. Some difficulties were found with more advanced uses and interpretations of these techniques, and these problems lead to suggestions for further research.

The manuscript of this article accepted for publication can be requested as a pdf file from Steve Boker.

]]>One of the major theoretic frameworks through which human development is studied is a process-oriented model involving selection, optimization, and compensation. These three processes each provide accounts for methods by which gains are maximized and losses minimized throughout the lifespan, and in particular during later life. These processes can be cast within the framework of dynamical systems theory and then modeled using differential equations. The current article will review basic tenets of selection, optimization, and compensation whilst introducing language and concepts from dynamical systems. Four categories of interindividual differences and intraindividual variability in dynamics are then described and discussed in the context of selection, optimization, and compensation.

The manuscript of this article accepted for publication can be downloaded as a PDF. This article may not exactly replicate the ﬁnal version published in The Journal of Gerontopsychology and Geriatric Psychiatry. It is not the copy of record.

]]>Studies suggest that within-person changes in estrogen and progesterone predict changes in binge eating across the menstrual cycle. However, samples have been extremely small (maximum N = 9), and analyses have not examined the interactive effects of hormones that are critical for changes in food intake in animals. The aims of the current study were to examine ovarian hormone interactions in the prediction of within-subject changes in emotional eating in the largest sample of women to date (N = 196). Participants provided daily ratings of emotional eating and saliva samples for hormone measurement for 45 consecutive days. Results confirmed that changes in ovarian hormones predict changes in emotional eating across the menstrual cycle, with a significant estradiol x progesterone interaction. Emotional eating scores were highest during the midluteal phase, when progesterone peaks and estradiol demonstrates a secondary peak. Findings extend previous work by highlighting significant interactions between estrogen and progesterone that explain midluteal increases in emotional eating. Future work should explore mechanisms (e.g., gene–hormone interactions) that contribute to both within- and between- subjects differences in emotional eating.

The full text of this article can be downloaded from APA Psycnet as a PDF.

]]>