An overview of applied robust methods

Data in social sciences are typically non-normally distributed and characterized by heavy tails. However, most widely used methods in social sciences are still based on the analyses of sample means and sample covariances. While these conventional methods continue to be used to address new substantive issues, conclusions reached can be inaccurate or misleading. Although there is no ‘best method’ in practice, robust methods that consider the distribution of the data can perform substantially better than the conventional methods. This article gives an overview of robust procedures, emphasizing a few that have been repeatedly shown to work well for models that are widely used in social and behavioural sciences. Real data examples show how to use the robust methods for latent variable models and for moderated mediation analysis when a regression model contains categorical covariates and product terms. Results and logical analyses indicate that robust methods yield more efficient parameter estimates, more reliable model evaluation, more reliable model/data diagnostics, and more trustworthy conclusions when conducting replication studies. R and SAS programs are provided for routine applications of the recommended robust method.     

Conceptual framework from the Paris Psychosomatic School: A clinical psychoanalytic approach to oncology1

This article presents further clinical material from the Paris Psychosomatic School ( Aisenstein, 2006 ). The Freudian foundations of psychosomatics are detailed and post-Freudian developments focusing on the contribution of the Paris Psychosomatic School are outlined, in particular, the somatizing process as a result of regression and the somatizing process as a result of drive unbinding. The authors argue that the latter possibly gives rise to progressive and serious illness leading to death. The relationship of classical psychoanalysis to psychotherapeutic treatment from the angle of the Paris school is commented on. The authors then turn to two clinical presentations of women suffering from breast cancer. The method of evaluating the patients’ capacities for undergoing psychotherapeutic treatment and their mental capacity for healing is discussed. The face-to-face psychoanalytic treatment undertaken with the second case is discussed. Finally, the authors recall Freud’s insistence after 1920 on the opposition of the life drives and the death drives, which placed self-destruction at the centre of psychic functioning. They conclude that current research in biology and medicine, notably research concerning programmed cell death, will converge with psychoanalytic psychosomatics to illuminate somatizing processes and demonstrate the relevance of psychoanalytic treatment to patients who are capable of mental reorganization in the course of their illness and medical treatment.     

Applying linear systems analysis to dynamic behavior

In this paper we present an abbreviated discussion of the linear systems analysis in the time domain. We then consider the qualitative character of the behavioral dynamics predicted using the linear form of the analysis. The analysis is then extended to a second-order form. We illustrate some relevant new features introduced by the second-order form with a special case example.     

Order and chaos in fixed-interval schedules of reinforcement

Fixed-interval schedule performance is characterized by high levels of variability. Responding is absent at the onset of the interval and gradually increases in frequency until reinforcer delivery. Measures of behavior also vary drastically and unpredictably between successive intervals. Recent advances in the study of nonlinear dynamics have allowed researchers to study irregular and unpredictable behavior in a number of fields. This paper reviews several concepts and techniques from nonlinear dynamics and examines their utility in predicting the behavior of pigeons responding to a fixed-interval schedule of reinforcement. The analysis provided fairly accurate a priori accounts of response rates, accounting for 92.8% of the variance when predicting response rate 1 second in the future and 64% of the variance when predicting response rates for each second over the entire next interreinforcer interval. The nonlinear dynamics account suggests that even the “noisiest” behavior might be the product of purely deterministic mechanisms.     

The integration of multidimensional scaling and multivariate analysis with optimal transformations

The recent history of multidimensional data analysis suggests two distinct traditions that have developed along quite different lines. In multidimensional scaling (MDS), the available data typically describe the relationships among a set of objects in terms of similarity/dissimilarity (or (pseudo-)distances). In multivariate analysis (MVA), data usually result from observation on a collection of variables over a common set of objects. This paper starts from a very general multidimensional scaling task, defined on distances between objects derived from one or more sets of multivariate data. Particular special cases of the general problem, following familiar notions from MVA, will be discussed that encompass a variety of analysis techniques, including the possible use of optimal variable transformation. Throughout, it will be noted how certain data analysis approaches are equivalent to familiar MVA solutions when particular problem specifications are combined with particular distance approximations.This research was supported by the Royal Netherlands Academy of Arts and Sciences (KNAW). An earlier version of this paper was written during a stay at McGill University in Montréal; this visit was supported by a travel grant from the Netherlands Organization for Scientific Research (NWO). I am grateful to Jim Ramsay and Willem Heiser for their encouragement and helpful suggestions, and to the Editor and referees for their constructive comments.     

Resurgence of derived stimulus relations.

Resurgence has been shown in human and nonhuman operant behavior, but not in derived relational responses. The present study examined this issue. Twenty-three undergraduates were trained to make conditional discriminations in a three-choice matching-to-sample paradigm. The training resulted in three equivalence classes, each consisting of four arbitrarily configured visual stimuli. The same 12 stimuli were then reorganized, and the conditional discrimination training was repeated such that three new classes were possible. In a subsequent test of derived relations, most subjects showed response patterns that were consistent with the altered conditional discriminations. Subjects were then exposed to conditional discrimination trials under extinction. Most subjects continued to respond consistently with the most recently reinforced conditional discrimination trials. During the next phase, subjects were exposed to symmetry and equivalence trials. Responses consistent with the most recent training produced feedback saying that the responses were incorrect, whereas other responses produced no feedback. Most subjects showed a resurgence of responding that was consistent with their earlier training. Finally, subjects were exposed to conditional discrimination trials carried out in extinction. Most subjects continued to show a resurgence of responding that was consistent with their early training.     

Statistical inference based on latent ability estimates

The quality of approximations to first and second order moments (e.g., statistics like means, variances, regression coefficients) based on latent ability estimates is being discussed. The ability estimates are obtained using either the Rasch, or the two-parameter logistic model. Straightforward use of such statistics to make inferences with respect to true latent ability is not recommended, unless we account for the fact that the basic quantities are estimates. In this paper true score theory is used to account for the latter; the counterpart of observed/true score being estimated/true latent ability. It is shown that statistics based on the true score theory are virtually unbiased if the number of items presented to each examinee is larger than fifteen. Three types of estimators are compared: maximum likelihood, weighted maximum likelihood, and Bayes modal. Furthermore, the (dis)advantages of the true score method and direct modeling of latent ability is discussed.     

从黑夜走向黎明：动脉粥样硬化从早期发现到消退

年１０年来,动态粥样硬化的早期诊断已成为可能,影像学能无创直接检测血管功能和形态学变化。在此基础上,动态粥样硬化的防治也进入了一级预防的时代。一级预防使早期粥样硬化消退,特效降脂药也可使中、晚期粥样硬化患者明显受益。随着对动态粥样硬化发病机制认识的不断深入,到下世纪初,人类有可能终止动态粥样硬化在全球的肆虐。     

The Problem of local optimality with OSMOD

The possibility of obtaining locally optimal solutions with categorical data is pointed out for the original version of OSMOD development by Saito and Otsu. A revision of the initialization strategy in OSMOD is suggested, and its effectiveness in diminishing this possibility is demonstrated.     

A note on monotone polygons fitted to bivariate data

The monotone regression function of Kruskal and the rank image function of Guttman and Lingoes were fitted to bivariate normal samples and their statistical properties contrasted.This paper was written while the author was a Visiting Professor in the Department of Psychology, University College, London.