On the Use of Mixed Markov Models for Intensive Longitudinal Data.

Markov modeling presents an attractive analytical framework for researchers who are interested in state-switching processes occurring within a person, dyad, family, group, or other system over time. Markov modeling is flexible and can be used with various types of data to study observed or latent state-switching processes, and can include subject-specific random effects to account for heterogeneity. We focus on the application of mixed Markov models to intensive longitudinal data sets in psychology, which are becoming ever more common and provide a rich description of each subject's process. We examine how specifications of a Markov model change when continuous random effect distributions are included, and how mixed Markov models can be used in the intensive longitudinal research context. Advantages of Bayesian estimation are discussed and the approach is illustrated by two empirical applications.

Analyzing subcomponents of affective dysregulation in borderline personality disorder in comparison to other clinical groups using multiple e-diary datasets.

BACKGROUND: Affective dysregulation is widely regarded as being the core problem in patients with borderline personality disorder (BPD). Moreover, BPD is the disorder mainly associated with affective dysregulation. However, the empirical confirmation of the specificity of affective dysregulation for BPD is still pending. We used a validated approach from basic affective science that allows for simultaneously analyzing three interdependent components of affective dysregulation that are disturbed in patients with BPD: homebase, variability, and attractor strength (return to baseline). METHODS: We applied two types of multilevel models on two e-diary datasets to investigate group differences regarding three subcomponents between BPD patients (n = 43; n = 51) and patients with posttraumatic stress disorder (PTSD; n = 28) and those with bulimia nervosa (BN; n = 20) as clinical control groups in dataset 1, and patients with panic disorder (PD; n = 26) and those with major depression (MD; n = 25) as clinical control groups in dataset 2. In addition, healthy controls (n = 28; n = 40) were included in the analyses. In both studies, e-diaries were used to repeatedly collect data about affective experiences during participants' daily lives. In study 1 a high-frequency sampling strategy with assessments in 15 min-intervals over 24 h was applied, whereas the assessments occurred every waking hour over 48 h in study 2. The local ethics committees approved both studies, and all participants provided written informed consent. RESULTS: In contradiction to our hypotheses, BPD patients did not consistently show altered affective dysregulation compared to the clinical patient groups. The only differences in affective dynamics in BPD patients emerged with regard to one of three subcomponents, affective homebase. However, these results were not even consistent. Conversely, comparing the patients to healthy controls revealed a pattern of more negative affective homebases, higher levels of affective variability, and (partially) reduced returns to baseline in the patient groups. CONCLUSIONS: Our results indicate that affective dysregulation constitutes a transdiagnostic mechanism that manifests in similar ways in several different mental disorders. We point out promising prospects that might help to elucidate the common and distinctive mechanisms that underlie several different disorders and that should be addressed in future studies.

Categories versus dimensions in personality and psychopathology: a quantitative review of taxometric research.

Taxometric research methods were developed by Paul Meehl and colleagues to distinguish between categorical and dimensional models of latent variables. We have conducted a comprehensive review of published taxometric research that included 177 articles, 311 distinct findings and a combined sample of 533 377 participants. Multilevel logistic regression analyses have examined the methodological and substantive variables associated with taxonic (categorical) findings. Although 38.9% of findings were taxonic, these findings were much less frequent in more recent and methodologically stronger studies, and in those reporting comparative fit indices based on simulated comparison data. When these and other possible confounds were statistically controlled, the true prevalence of taxonic findings was estimated at 14%. The domains of normal personality, mood disorders, anxiety disorders, eating disorders, externalizing disorders, and personality disorders (PDs) other than schizotypal yielded little persuasive evidence of taxa. Promising but still not definitive evidence of psychological taxa was confined to the domains of schizotypy, substance use disorders and autism. This review indicates that most latent variables of interest to psychiatrists and personality and clinical psychologists are dimensional, and that many influential taxonic findings of early taxometric research are likely to be spurious.

Clusterwise HICLAS: a generic modeling strategy to trace similarities and differences in multiblock binary data.

In many areas of the behavioral sciences, different groups of objects are measured on the same set of binary variables, resulting in coupled binary object × variable data blocks. Take, as an example, success/failure scores for different samples of testees, with each sample belonging to a different country, regarding a set of test items. When dealing with such data, a key challenge consists of uncovering the differences and similarities between the structural mechanisms that underlie the different blocks. To tackle this challenge for the case of a single data block, one may rely on HICLAS, in which the variables are reduced to a limited set of binary bundles that represent the underlying structural mechanisms, and the objects are given scores for these bundles. In the case of multiple binary data blocks, one may perform HICLAS on each data block separately. However, such an analysis strategy obscures the similarities and, in the case of many data blocks, also the differences between the blocks. To resolve this problem, we proposed the new Clusterwise HICLAS generic modeling strategy. In this strategy, the different data blocks are assumed to form a set of mutually exclusive clusters. For each cluster, different bundles are derived. As such, blocks belonging to the same cluster have the same bundles, whereas blocks of different clusters are modeled with different bundles. Furthermore, we evaluated the performance of Clusterwise HICLAS by means of an extensive simulation study and by applying the strategy to coupled binary data regarding emotion differentiation and regulation.

Determination of n-dimensional reference ellipsoids using patient data.

A method is described for the calculation of n-dimensional reference ellipsoids, using patient data. The advantages and drawbacks of the use of reference ellipsoids for a set of different parameters, in contrast with the use of a reference range for every single parameter, are discussed. The use of reference ellipsoids in practice is illustrated with an example.

The use of patient data for the calculation of reference values for some haematological parameters.

We have investigated the use of patient data for the calculation of reference values for the parameters which are determined by the Hemalog. For this purpose we used the Bhattacharya plot. All the parameters, with the exception of leukocytes, appear to meet the main underlying assumption of this plot, namely that the frequency distribution is Gaussian. In the case of leukocytes, however, the frequency distribution could be resolved into two overlapping Gaussian curves, thus making it possible to calculate reference values for this parameter also. The reference values as calculated from 14,500 unselected data (excluding children) are in general agreement with the literature. Significant differences were however detected between a group of patients and a group of blood donors. When a Bhattacharya plot has to be constructed with relatively few data, smoothing of the observed frequencies is very helpful in deciding which part of the plot is linear. Smoothing was carried out using the least squares method with a quadratic equation. Since the classes are equally spaced, this involves only a simple numerical transformation of the frequencies.

Calculation of radioimmunoassay standard curves.

A method for calculating radioimmunoassay standard curves, based on the theory of Ekins et al., is described. Because a four-parameter model is used, nonlinear standard curves are the result. The calibration curve is fitted to the measured standard points by means of a weighted least-squares method. The program based on this model can be easily processed on a desk-top calculator. For all 250 runs of six different assays, very good standard curves could be obtained. The mean deviation between the concentrations of the standard points and the corresponding calculated values was about 6%. In 26% of the cases it could be shown that the model we describe gave significantly better results than did two simpler ones.