Dynamic factor analysis of nonstationary multivariate time series

A dynamic factor model is proposed for the analysis of multivariate nonstationary time series in the time domain. The nonstationarity in the series is represented by a linear time dependent mean function. This mild form of nonstationarity is often relevant in analyzing socio-economic time series met in practice. Through the use of an extended version of Molenaar's stationary dynamic factor analysis method, the effect of nonstationarity on the latent factor series is incorporated in the dynamic nonstationary factor model (DNFM). It is shown that the estimation of the unknown parameters in this model can be easily carried out by reformulating the DNFM as a covariance structure model and adopting the ML algorithm proposed by Jöreskog. Furthermore, an empirical example is given to demonstrate the usefulness of the proposed DNFM and the analysis.     

A dynamic factor model for the analysis of multivariate time series

As a method to ascertain the structure of intra-individual variation,P-technique has met difficulties in the handling of a lagged covariance structure. A new statistical technique, coined dynamic factor analysis, is proposed, which accounts for the entire lagged covariance function of an arbitrary second order stationary time series. Moreover, dynamic factor analysis is shown to be applicable to a relatively short stretch of observations and therefore is considered worthwhile for psychological research. At several places the argumentation is clarified through the use of examples.I would like to thank WM. van der Molen, G. J. Mellenbergh and L. H. M. Oppenheimer, who provided valuable ideas that led to this formulation.     

A cognitive multivariate time series and its analysis

The method of analyzing a multivariate time series, by factoring a matrix polynomial of linear transfer function coefficients, is briefly outlined and shown to be potentially applicable to cognitive experiments where the inputs are numerically predictable and the outputs suitably quantified. An experiment on sequences of 90 figures, judged in terms of similarity and of the probability of exceeding or not exceeding some size, was analyzed individually for 12 subjects and their dynamic characteristics expressed in pole zero diagrams. Some dynamic components neglected by classical time-independent psychological methods are identified.     

Classical conditioning model of panic disorder: response to Wolpe and Rowan

We refute Wolpe and Rowan's (1989) classical conditioning theory of panic disorder based upon several lines of evidence presented above. While there is no doubt that hyperventilation and conditioning are important factors in the production of panic attacks, Wolpe and Rowan's (1989) hypothesis is inconsistent with recent data concerning the limited role of hyperventilation and the importance of cognitive mediation in the production of panic attacks.     

Classical fear conditioning in the anxiety disorders: a meta-analysis

Fear conditioning represents the process by which a neutral stimulus comes to evoke fear following its repeated pairing with an aversive stimulus. Although fear conditioning has long been considered a central pathogenic mechanism in anxiety disorders, studies employing lab-based conditioning paradigms provide inconsistent support for this idea. A quantitative review of 20 such studies, representing fear-learning scores for 453 anxiety patients and 455 healthy controls, was conducted to verify the aggregated result of this literature and to assess the moderating influences of study characteristics. Results point to modest increases in both acquisition of fear learning and conditioned responding during extinction among anxiety patients. Importantly, these patient-control differences are not apparent when looking at discrimination studies alone and primarily emerge from studies employing simple, single-cue paradigms where only danger cues are presented and no inhibition of fear to safety cues is required.     

Four simultaneous component models for the analysis of multivariate time series from more than one subject to model intraindividual and interindividual differences

A class of four simultaneous component models for the exploratory analysis of multivariate time series collected from more than one subject simultaneously is discussed. In each of the models, the multivariate time series of each subject is decomposed into a few series of component scores and a loading matrix. The component scores series reveal the latent data structure in the course of time. The interpretation of the components is based on the loading matrix. The simultaneous component models model not only intraindividual variability, but interindividual variability as well. The four models can be ordered hierarchically from weakly to severely constrained, thus allowing for big to small interindividual differences in the model. The use of the models is illustrated by an empirical example.This research has been made possible by funding from the Netherlands Organization of Scientific Research (NWO) to the first author. The authors are obliged to Tom A.B. Snijders, Jos M.F. ten Berge and three anonymous reviewers for comments on an earlier version of this paper, and to Kim Shifren for providing us with her data set, which was collected at Syracuse University.     

Detecting correlation changes in multivariate time series: A comparison of four non-parametric change point detection methods

Change point detection in multivariate time series is a complex task since next to the mean, the correlation structure of the monitored variables may also alter when change occurs. DeCon was recently developed to detect such changes in mean and\or correlation by combining a moving windows approach and robust PCA. However, in the literature, several other methods have been proposed that employ other non-parametric tools: E-divisive, Multirank, and KCP. Since these methods use different statistical approaches, two issues need to be tackled. First, applied researchers may find it hard to appraise the differences between the methods. Second, a direct comparison of the relative performance of all these methods for capturing change points signaling correlation changes is still lacking. Therefore, we present the basic principles behind DeCon, E-divisive, Multirank, and KCP and the corresponding algorithms, to make them more accessible to readers. We further compared their performance through extensive simulations using the settings of Bulteel et al. (Biological Psychology, 98 (1), 29-42, 2014) implying changes in mean and in correlation structure and those of Matteson and James (Journal of the American Statistical Association, 109 (505), 334-345, 2014) implying different numbers of (noise) variables. KCP emerged as the best method in almost all settings. However, in case of more than two noise variables, only DeCon performed adequately in detecting correlation changes.     

Classical conditioning in the leechMacrobdella Ditetra as a function of CS and UCS intensity

Classical conditioning in the leech was obtained using two levels of shock as the UCS, two levels of light as the CS; for each condition there was a CS-UCS paired and unpaired group. Analysis showed significant effects of pairing and a UCS intensity by trials interaction during acquisition. Spontaneous recovery was evident during the second day of extinction.     

Dismal: a spreadsheet for sequential data analysis and HCI experimentation

Dismal is a spreadsheet that works within GNU Emacs, a widely available programmable editor. Dismal has three features of particular interest to those who study behavior: (1) the ability to manipulate and align sequential data, (2) an open architecture that allows users to expand it to meet their particular needs, and (3) an instrumented and accessible interface for studies of human-computer interaction (HCI). Example uses of each of these capabilities are provided, including cognitive models that have had their sequential behavior aligned with subject’s protocols, extensions useful for teaching and doing HCI design, and studies in which keystroke logs from the timing package in Dismal have been used.     

Simple reaction time as a race between signal detection and time estimation: A paradigm and model

A new paradigm for simple reaction time is described in which signal detection and time production trials are interleaved. A model is proposed which views the signal detection and time production processes as independent and engaged in a race whose outcome determines the observed response latencies. The model also allows the distribution of detection latencies to be extracted from the data.     

Classical discrimination conditioning of the rabbit's eyelid response using pontine stimulation as a conditioned stimulus

Classical discrimination conditioning of the nictitating membrane/eyelid response was performed on seven rabbits using stimulation of the pontine nuclei or middle cerebellar peduncle as the conditioned stimulus (CS) and an air puff as the unconditioned stimulus (US). The rabbits learned to discriminate between a CS paired with a US and delivered to one pontine nucleus (the CS+) and a CS presented alone and delivered to the contralateral pontine nucleus (the CS-). Subsequent reversal of the discrimination was also achieved when the CS+ and CS- stimulation sites were interchanged. The results are interpreted as support for the idea that essential plasticity for classical eyelid conditioning occurs efferent to the pontine nuclei, possibly in regions of the cerebellum.     

The conditional emotional response as a model of pavlovian conditioning

Integrative Psychological and Behavioral Science - Many investigators have assumed that the conditional emotional response (CER) with its attendant response measure, response suppression, can be...     

A lengthy look at the daily grind: time series analysis of events, mood, stress, and satisfaction

The present study investigated processes by which job stress and satisfaction unfold over time by examining the relations between daily stressful events, mood, and these variables. Using a Web-based daily survey of stressor events, perceived strain, mood, and job satisfaction completed by 14 university workers, 1,060 occasions of data were collected. Transfer function analysis, a multivariate version of time series analysis, was used to examine the data for relationships among the measured variables after factoring out the contaminating influences of serial dependency. Results revealed a contrast effect in which a stressful event associated positively with higher strain on the same day and associated negatively with strain on the following day. Perceived strain increased over the course of a semester for a majority of participants, suggesting that effects of stress build over time. Finally, the data were consistent with the notion that job satisfaction is a distal outcome that is mediated by perceived strain.     

Chemical aversion conditioning as a treatment for alcoholism: A re-analysis

The literature on chemical aversion conditioning is characterized by the lack of controlled clinical research. The existing data derive primarily from methodologically inadequate studies. Although short-term conditioned aversion reactions to alcohol have been demonstrated, the independent efficacy of this technique in clinical treatment remains to be shown. Beyond the failure to demonstrate the value of adding chemical aversion conditioning to more standard treatments for alcoholism, evaluation in terms of broader outcome criteria (e.g. safety, intrusiveness, acceptability, the availability of alternative methods, and cost-effectiveness) indicates that chemical aversion conditioning cannot be recommended as a standard form of treatment for alcoholism.     

Manipulation of response-contingent unconditioned-stimulus intensity in human eyelid conditioning: A two-phase model analysis

Four groups of human subjects were given 360 classical eyeblink conditioning trials. All groups received the same UCS (unconditioned stimulus) intensity on[Formula: see text] (nonconditioned response) trials but differed in the intensity presented on CR trials. Response probability increased as a positive function of UCS intensity on CR trials. Phase 1 of the two-phase model was longer when no UCS was presented on CR trials, but did not differ in duration among the remaining three groups. Most subjects could be described with a single operator in Phase 2, the operator limit increasing as a positive function of CR-contingent UCS intensity. For subjects requiring different operator limits on CR and[Formula: see text] trials, the latter was lower with high CR-trial intensities but higher with low CR-trialintensities. The results were interpreted to be more consistent with drive theory than with "law-of-effect" or two-factor theories.     

Fear conditioning, meaning, and belongingness: a selective association analysis

Twenty-three subjects rated the belongingness of pairs of conditionable (photographic slides) and unconditioned (e.g., shock, tone, human scream) stimuli. Forty new subjects were then classically conditioned, using rating-defined high (angry face/scream) and low (landscape/scream) belongingness pairs. Finger-pulse responses to the high-belongingness pairs showed superior acquisition and resistance to extinction. Another 40 subjects were conditioned to compound stimuli: a slide (either landscape or angry face) that was the same over trials, and a yellow or blue background that was the discriminant cue for the unconditioned stimulus (scream). When the angry face (the high-belongingness slide) was the invariant part of the compound, relatively poorer differential pulse-volume and skin-conductance conditioning was observed. Thus, depending on the task, a priori belongingness rendered stimuli selectively conditionable, either enhancing or inhibiting visceral response associations.