Estimating optimal decision trees for treatment assignment: The case of K > 2 treatment alternatives.

For many problems in clinical practice, multiple treatment alternatives are available. Given data from a randomized controlled trial or an observational study, an important challenge is to estimate an optimal decision rule that specifies for each client the most effective treatment alternative, given his or her pattern of pretreatment characteristics. In the present paper we will look for such a rule within the insightful family of classification trees. Unfortunately, however, there is dearth of readily accessible software tools for optimal decision tree estimation in the case of more than two treatment alternatives. Moreover, this primary tree estimation problem is also cursed with two secondary problems: a structural missingness in typical studies on treatment evaluation (because every individual is assigned to a single treatment alternative only), and a major issue of replicability. In this paper we propose solutions for both the primary and the secondary problems at stake. We evaluate the proposed solution in a simulation study, and illustrate with an application on the search for an optimal tree-based treatment regime in a randomized controlled trial on K = 3 different types of aftercare for younger women with early-stage breast cancer. We conclude by arguing that the proposed solutions may have relevance for several other classification problems inside and outside the domain of optimal treatment assignment.

Ready for the worst? Negative affect in anticipation of a stressor does not protect against affective reactivity.

INTRODUCTION: Lay wisdom suggests feeling negative while awaiting an upcoming stressor-anticipatory negative affect-shields against the blow of the subsequent stressor. However, evidence is mixed, with different lines of research and theory indirectly suggesting that anticipatory negative affect is helpful, harmful, or has no effect on emotional outcomes. In two studies, we aimed to reconcile these competing views by examining the affective trajectory across hours, days, and months, separating affective reactivity and recovery. METHODS: In Study 1, first-year students (N = 101) completed 9 days of experience sampling (10 surveys/day) as they received their first-semester exam grades, and a follow-up survey 5 months later. In Study 2, participants (N = 73) completed 2 days of experience sampling (60 surveys/day) before and after a Trier Social Stress Test. We investigated the association between anticipatory negative affect and the subsequent affective trajectory, investigating (1) reactivity immediately after the stressor, (2) recovery across hours (Study 2) and days (Study 1), and (3) recovery after 5 months (Study 1). RESULTS: Across the two studies, feeling more negative in anticipation of a stressor was either associated with increased negative affective reactivity, or unassociated with affective outcomes. CONCLUSION: These results run counter to the idea that being affectively ready for the worst has psychological benefits, suggesting that instead, anticipatory negative affect can come with affective costs.

Efficient estimation of bounded gradient-drift diffusion models for affect on CPU and GPU.

Computational modeling plays an important role in a gamut of research fields. In affect research, continuous-time stochastic models are becoming increasingly popular. Recently, a non-linear, continuous-time, stochastic model has been introduced for affect dynamics, called the Affective Ising Model (AIM). The drawback of non-linear models like the AIM is that they generally come with serious computational challenges for parameter estimation and related statistical analyses. The likelihood function of the AIM does not have a closed form expression. Consequently, simulation based or numerical methods have to be considered in order to evaluate the likelihood function. Additionally, the likelihood function can have multiple local minima. Consequently, a global optimization heuristic is required and such heuristics generally require a large number of likelihood function evaluations. In this paper, a Julia software package is introduced that is dedicated to fitting the AIM. The package includes an implementation of a numeric algorithm for fast computations of the likelihood function, which can be run both on graphics processing units (GPU) and central processing units (CPU). The numerical method introduced in this paper is compared to the more traditional Euler-Maruyama method for solving stochastic differential equations. Furthermore, the estimation software is tested by means of a recovery study and estimation times are reported for benchmarks that were run on several computing devices (two different GPUs and three different CPUs). According to these results, a single parameter estimation can be obtained in less than thirty seconds using a mainstream NVIDIA GPU.

: An R Package for performing kernel change point detection on the running statistics of multivariate time series.

In many scientific disciplines, researchers are interested in discovering when complex systems such as stock markets, the weather or the human body display abrupt changes. Essentially, this often comes down to detecting whether a multivariate time series contains abrupt changes in one or more statistics, such as means, variances or pairwise correlations. To assist researchers in this endeavor, this paper presents the package for performing kernel change point (KCP) detection on user-selected running statistics of multivariate time series. The running statistics are extracted by sliding a window across the time series and computing the value of the statistic(s) of interest in each window. Next, the similarities of the running values are assessed using a Gaussian kernel, and change points that segment the time series into maximally homogeneous phases are located by minimizing a within-phase variance criterion. To decide on the number of change points, a combination of a permutation-based significance test and a grid search is provided. stands out among the variety of change point detection packages available in because it can be easily adapted to uncover changes in any user-selected statistic without imposing any distribution on the data. To exhibit the usefulness of the package, two empirical examples are provided pertaining to two types of physiological data.

Prepaid parameter estimation without likelihoods.

In various fields, statistical models of interest are analytically intractable and inference is usually performed using a simulation-based method. However elegant these methods are, they are often painstakingly slow and convergence is difficult to assess. As a result, statistical inference is greatly hampered by computational constraints. However, for a given statistical model, different users, even with different data, are likely to perform similar computations. Computations done by one user are potentially useful for other users with different data sets. We propose a pooling of resources across researchers to capitalize on this. More specifically, we preemptively chart out the entire space of possible model outcomes in a prepaid database. Using advanced interpolation techniques, any individual estimation problem can now be solved on the spot. The prepaid method can easily accommodate different priors as well as constraints on the parameters. We created prepaid databases for three challenging models and demonstrate how they can be distributed through an online parameter estimation service. Our method outperforms state-of-the-art estimation techniques in both speed (with a 23,000 to 100,000-fold speed up) and accuracy, and is able to handle previously quasi inestimable models.

mobileQ: A free user-friendly application for collecting experience sampling data.

In this article we introduce mobileQ, which is a free, open-source platform that our lab has developed to use in experience sampling studies. Experience sampling has several strengths and is becoming more widely conducted, but there are few free software options. To address this gap, mobileQ has freely available servers, a web interface, and an Android app. To reduce the barrier to entry, it requires no high-level programming and uses an easy, point-and-click interface. It is designed to be used on dedicated research phones, allowing for experimenter control and eliminating selection bias. In this article, we introduce setting up a study in mobileQ, outline the set of help resources available for new users, and highlight the success with which mobileQ has been used in our lab.

Efficient simulation of diffusion-based choice RT models on CPU and GPU.

In this paper, we present software for the efficient simulation of a broad class of linear and nonlinear diffusion models for choice RT, using either CPU or graphical processing unit (GPU) technology. The software is readily accessible from the popular scripting languages MATLAB and R (both 64-bit). The speed obtained on a single high-end GPU is comparable to that of a small CPU cluster, bringing standard statistical inference of complex diffusion models to the desktop platform.

CHull: a generic convex-hull-based model selection method.

When analyzing data, researchers are often confronted with a model selection problem (e.g., determining the number of components/factors in principal components analysis [PCA]/factor analysis or identifying the most important predictors in a regression analysis). To tackle such a problem, researchers may apply some objective procedure, like parallel analysis in PCA/factor analysis or stepwise selection methods in regression analysis. A drawback of these procedures is that they can only be applied to the model selection problem at hand. An interesting alternative is the CHull model selection procedure, which was originally developed for multiway analysis (e.g., multimode partitioning). However, the key idea behind the CHull procedure--identifying a model that optimally balances model goodness of fit/misfit and model complexity--is quite generic. Therefore, the procedure may also be used when applying many other analysis techniques. The aim of this article is twofold. First, we demonstrate the wide applicability of the CHull method by showing how it can be used to solve various model selection problems in the context of PCA, reduced K-means, best-subset regression, and partial least squares regression. Moreover, a comparison of CHull with standard model selection methods for these problems is performed. Second, we present the CHULL software, which may be downloaded from http://ppw.kuleuven.be/okp/software/CHULL/, to assist the user in applying the CHull procedure.

Combining Experience Sampling and Mobile Sensing for Digital Phenotyping With m-Path Sense: Performance Study.

BACKGROUND: The experience sampling methodology (ESM) has long been considered as the gold standard for gathering data in everyday life. In contrast, current smartphone technology enables us to acquire data that are much richer, more continuous, and unobtrusive than is possible via ESM. Although data obtained from smartphones, known as mobile sensing, can provide useful information, its stand-alone usefulness is limited when not combined with other sources of information such as data from ESM studies. Currently, there are few mobile apps available that allow researchers to combine the simultaneous collection of ESM and mobile sensing data. Furthermore, such apps focus mostly on passive data collection with only limited functionality for ESM data collection. OBJECTIVE: In this paper, we presented and evaluated the performance of m-Path Sense, a novel, full-fledged, and secure ESM platform with background mobile sensing capabilities. METHODS: To create an app with both ESM and mobile sensing capabilities, we combined m-Path, a versatile and user-friendly platform for ESM, with the Copenhagen Research Platform Mobile Sensing framework, a reactive cross-platform framework for digital phenotyping. We also developed an R package, named mpathsenser, which extracts raw data to an SQLite database and allows the user to link and inspect data from both sources. We conducted a 3-week pilot study in which we delivered ESM questionnaires while collecting mobile sensing data to evaluate the app's sampling reliability and perceived user experience. As m-Path is already widely used, the ease of use of the ESM system was not investigated. RESULTS: Data from m-Path Sense were submitted by 104 participants, totaling 69.51 GB (430.43 GB after decompression) or approximately 37.50 files or 31.10 MB per participant per day. After binning accelerometer and gyroscope data to 1 value per second using summary statistics, the entire SQLite database contained 84,299,462 observations and was 18.30 GB in size. The reliability of sampling frequency in the pilot study was satisfactory for most sensors, based on the absolute number of collected observations. However, the relative coverage rate-the ratio between the actual and expected number of measurements-was below its target value. This could mostly be ascribed to gaps in the data caused by the operating system pushing away apps running in the background, which is a well-known issue in mobile sensing. Finally, some participants reported mild battery drain, which was not considered problematic for the assessed participants' perceived user experience. CONCLUSIONS: To better study behavior in everyday life, we developed m-Path Sense, a fusion of both m-Path for ESM and Copenhagen Research Platform Mobile Sensing. Although reliable passive data collection with mobile phones remains challenging, it is a promising approach toward digital phenotyping when combined with ESM.

A Systematic Study into the Factors that Affect the Predictive Accuracy of Multilevel VAR(1) Models.

The use of multilevel VAR(1) models to unravel within-individual process dynamics is gaining momentum in psychological research. These models accommodate the structure of intensive longitudinal datasets in which repeated measurements are nested within individuals. They estimate within-individual auto- and cross-regressive relationships while incorporating and using information about the distributions of these effects across individuals. An important quality feature of the obtained estimates pertains to how well they generalize to unseen data. Bulteel and colleagues (Psychol Methods 23(4):740-756, 2018a) showed that this feature can be assessed through a cross-validation approach, yielding a predictive accuracy measure. In this article, we follow up on their results, by performing three simulation studies that allow to systematically study five factors that likely affect the predictive accuracy of multilevel VAR(1) models: (i) the number of measurement occasions per person, (ii) the number of persons, (iii) the number of variables, (iv) the contemporaneous collinearity between the variables, and (v) the distributional shape of the individual differences in the VAR(1) parameters (i.e., normal versus multimodal distributions). Simulation results show that pooling information across individuals and using multilevel techniques prevent overfitting. Also, we show that when variables are expected to show strong contemporaneous correlations, performing multilevel VAR(1) in a reduced variable space can be useful. Furthermore, results reveal that multilevel VAR(1) models with random effects have a better predictive performance than person-specific VAR(1) models when the sample includes groups of individuals that share similar dynamics.

The LMPCA program: a graphical user interface for fitting the linked-mode PARAFAC-PCA model to coupled real-valued data.

In behavioral research, PARAFAC analysis, a three-mode generalization of standard principal component analysis (PCA), is often used to disclose the structure of three-way three-mode data. To get insight into the underlying mechanisms, one often wants to relate the component matrices resulting from such a PARAFAC analysis to external (two-way two-mode) information, regarding one of the modes of the three-way data. To this end, linked-mode PARAFAC-PCA analysis can be used, in which the three-way and the two-way data set, which have one mode in common, are simultaneously analyzed. More specifically, a PARAFAC and a PCA model are fitted to the three-way and the two-way data, respectively, restricting the component matrix for the common mode to be equal in both models. Until now, however, no software program has been publicly available to perform such an analysis. Therefore, in this article, the LMPCA program, a free and easy-to-use MATLAB graphical user interface, is presented to perform a linked-mode PARAFAC-PCA analysis. The LMPCA software can be obtained from the authors at http://ppw.kuleuven.be/okp/software/LMPCA. For users who do not have access to MATLAB, a stand-alone version is provided.

Affect dynamics in relation to depressive symptoms: variable, unstable or inert?

Depression not only involves disturbances in prevailing affect, but also in how affect fluctuates over time. Yet, precisely which patterns of affect dynamics are associated with depressive symptoms remains unclear; depression has been linked with increased affective variability and instability, but also with greater resistance to affective change (inertia). In this paper, we argue that these paradoxical findings stem from a number of neglected methodological/analytical factors, which we address using a novel paradigm and analytic approach. Participants (N = 99), preselected to represent a wide range of depressive symptoms, watched a series of emotional film clips and rated their affect at baseline and following each film clip. We also assessed participants' affect in daily life over 1 week using experience sampling. When controlling for overlap between different measures of affect dynamics, depressive symptoms were independently associated with higher inertia of negative affect in the lab, and with greater negative affect variability both in the lab and in daily life. In contrast, depressive symptoms were not independently related to higher affective instability either in daily life or in the lab.