A Comparison of Imputation Strategies for Ordinal Missing Data on Likert Scale Variables

This article compares a variety of imputation strategies for ordinal missing data on Likert scale variables (number of categories = 2, 3, 5, or 7) in recovering reliability coefficients, mean scale scores, and regression coefficients of predicting one scale score from another. The examined strategies include imputing using normal data models with naïve rounding/without rounding, using latent variable models, and using categorical data models such as discriminant analysis and binary logistic regression (for dichotomous data only), multinomial and proportional odds logistic regression (for polytomous data only). The result suggests that both the normal model approach without rounding and the latent variable model approach perform well for either dichotomous or polytomous data regardless of sample size, missing data proportion, and asymmetry of item distributions. The discriminant analysis approach also performs well for dichotomous data. Naïvely rounding normal imputations or using logistic regression models to impute ordinal data are not recommended as they can potentially lead to substantial bias in all or some of the parameters.     

ROY SCHAFER'S CONTRIBUTIONS TO PSYCHOLOGICAL TESTING: FROM CLINICAL SENSIBILITY TO THE ANALYTIC ATTITUDE

The author reviews Schafer's contributions to psychological testing, emphasizing his development of the test battery, his significant contributions to psychoanalytically oriented Rorschach interpretation, and his understanding of the complex interpersonal dynamics involved in psychological test interpretation. The author also discusses his use of Schafer's writing in his own teaching and academic work, noting that Schafer's contributions have not only provided innovative methods for examining test data, but have also promoted a respectful, humanistic, and individualized approach to the patient in testing and treatment. The author asserts that Schafer's later seminal contributions to psychoanalysis had their origins in his early career as a psychologist applying psychoanalytic ideas to testing.     

题目因素分析：基于SEM和基于IRT的两类方法

心理和教育测量一般只能达到顺序量表的水平,其测量数据与被测因子间并非简单线性关系。题目因素分析是用来描述测量题目与因子间非线性关系的统计模型。题目因素分析主要有基于结构方程模型和基于项目反应理论两类方法,两类方法之间存在紧密的联系,甚至可以看作是同一模型的两种表现形式。本文详细阐述了该关系,同时对两类方法在参数估计、模型拟合指标、测量一致性检验和支撑软件等方面的特点进行了分析和比较,以便研究者选择最为适合其研究的方法。     

Perfectionistic concerns predict increases in adolescents’ anxiety symptoms: a three-wave longitudinal study

Background and Objectives: Although perfectionism has been proposed to be a risk factor for the development of anxiety, research on perfectionism and anxiety symptoms in adolescents is scarce and inconclusive. The aim of the present study was to investigate whether the two higher-order dimensions of perfectionism – perfectionistic strivings and perfectionistic concerns – predict the development and maintenance of anxiety symptoms. An additional aim of the present study was to examine potential reciprocal effects of anxiety symptoms predicting increases in perfectionism.

Design: The study used a longitudinal design with three waves spaced 4–5 months apart.

Methods: A non-clinical sample of 489 adolescents aged 12–19 years completed a paper-and-pencil questionnaire.

Results: As expected, results showed a positive effect from perfectionistic concerns to anxiety symptoms, but the effect was restricted to middle-to-late adolescents (16–19 years old): Perfectionistic concerns predicted longitudinal increases in adolescents’ anxiety symptoms, whereas perfectionistic strivings did not. Furthermore, anxiety symptoms did not predict increases in perfectionism.

Conclusions: Implications for the understanding of the relationship between perfectionism and anxiety symptoms are discussed.     

概化理论缺失数据方差分量估计

各种心理调查、心理实验中, 数据的缺失随处可见。由于数据缺失, 给概化理论分析非平衡数据的方差分量带来一系列问题。基于概化理论框架下, 运用Matlab 7.0软件, 自编程序模拟产生随机双面交叉设计p×i×r缺失数据, 比较和探讨公式法、REML法、拆分法和MCMC法在估计各个方差分量上的性能优劣。结果表明：(1) MCMC方法估计随机双面交叉设计p×i×r缺失数据方差分量, 较其它3种方法表现出更强的优势; (2) 题目和评分者是缺失数据方差分量估计重要的影响因素。     

The Œdipus complex, crystallizer of the debate between psychoanalysis and anthropology

The way that anthropologists understand the Oedipus complex, in particular, is a good example of how they understand psychoanalysis in general. Indeed, it has crystallized a set of reactions marked by ignorance, misunderstanding, distortion and screening out and at the same time has provoked suspicion among anthropologists as to psychoanalysis, according to the preconceptions of the various schools of thought and authors implied, and this from the very first contacts up to nowadays. In what way did the psychoanalysts contribute to this and what representation did they, in turn, elaborate of anthropology? The purpose of this paper is to expose the epistemological and historical conditions of the emergence of this debate, and then to develop it by following chronology up to the 1950s and 1960s, while differentiating three major cultural areas, Great Britain, the USA and France, in order to get a clearer picture. From that point on, we will try to diversify our inquiry and to formulate some interpretative hypotheses. In particular, we think that a traumatic event may have inaugurated and organized the history of the relationship between the two disciplines, producing a situation of acculturation with multiple impacts, if we identify them with two cultures coming into contact: what is at stake here is Totem and Taboo in which Freud carries through the first major psychoanalytical approach of the interpretation of ethnographic facts, that leads him to transplant the universality of the Oedipus complex to the very root of the first social institutions and to pinpoint the presence of unconscious processes in their genesis.     

Data presentation options to manage variability in physical activity research

This paper presents seven tactics for managing the variability evident in some physical activity data. High levels of variability in daily step‐count data from pedometers or accelerometers can make typical visual inspection difficult. Therefore, the purpose of the current paper is to discuss several strategies that might facilitate the visual interpretation of highly variable data. The seven strategies discussed in this paper are phase mean and median lines, daily average per week, weekly cumulative, proportion of baseline, 7‐day moving average, change point detection, and confidence intervals. We apply each strategy to a data set and discuss the advantages and disadvantages.