Stepwise variable selection in factor analysis

It is very important to choose appropriate variables to be analyzed in multivariate analysis when there are many observed variables such as those in a questionnaire. What is actually done in scale construction with factor analysis is nothing but variable selection.In this paper, we take several goodness-of-fit statistics as measures of variable selection and develop backward elimination and forward selection procedures in exploratory factor analysis. Once factor analysis is done for a certain numberp of observed variables (thep-variable model is labeled the current model), simple formulas for predicted fit measures such as chi-square, GFI, CFI, IFI and RMSEA, developed in the field of the structural equation modeling, are provided for all models obtained by adding an external variable (so that the number of variables isp + 1) and for those by deleting an internal variable (so that the number isp – 1), provided that the number of factors is held constant.A programSEFA (Stepwise variable selection in Exploratory Factor Analysis) is developed to actually obtain a list of the fit measures for all such models. The list is very useful in determining which variable should be dropped from the current model to improve the fit of the current model. It is also useful in finding a suitable variable that may be added to the current model. A model with more appropriate variables makes more stable inference in general.The criteria traditionally often used for variable selection is magnitude of communalities. This criteria gives a different choice of variables and does not improve fit of the model in most cases.The URL of the programSEFA is http://koko15.hus.osaka-u.ac.jp/~harada/factor/stepwise/.     

考虑题目选项信息的非参数认知诊断计算机自适应测验

选择题中的作答选项能提供额外诊断信息, 为充分利用选项信息, 研究提出认知诊断计算机自适应测验(CD-CAT)中两种处理选择题选项信息的非参数选题策略和变长终止规则。模拟研究的结果发现：(1)定长条件下两种非参数选题策略的分类准确性整体要高于参数选题策略; (2)两种非参数选题策略较参数选题策略具有更加均衡的题库使用情况; (3)非参数选题策略在两种新的变长终止规则下具有更高的分类准确率; (4)两种非参数选题策略均适用于选择题CD-CAT情境, 使用者可任选其一进行测验分析。     

Variable Neighborhood Search Heuristics for Selecting a Subset of Variables in Principal Component Analysis

The selection of a subset of variables from a pool of candidates is an important problem in several areas of multivariate statistics. Within the context of principal component analysis (PCA), a number of authors have argued that subset selection is crucial for identifying those variables that are required for correct interpretation of the components. In this paper, we adapt the variable neighborhood search (VNS) paradigm to develop two heuristics for variable selection in PCA. The performances of these heuristics were compared to those obtained by a branch-and-bound algorithm, as well as forward stepwise, backward stepwise, and tabu search heuristics. In the first experiment, which considered candidate pools of 18 to 30 variables, the VNS heuristics matched the optimal subset obtained by the branch-and-bound algorithm more frequently than their competitors. In the second experiment, which considered candidate pools of 54 to 90 variables, the VNS heuristics provided better solutions than their competitors for a large percentage of the test problems. An application to a real-world data set is provided to demonstrate the importance of variable selection in the context of PCA.     

A novel machine learning approach for early detection of hepatocellular carcinoma patients

Liver cancer is quite common type of cancer among individuals worldwide. Hepatocellular carcinoma (HCC) is the malignancy of liver cancer. It has high impact on individual’s life and investigating it early can decline the number of annual deaths. This study proposes a new machine learning approach to detect HCC using 165 patients. Ten well-known machine learning algorithms are employed. In the preprocessing step, the normalization approach is used. The genetic algorithm coupled with stratified 5-fold cross-validation method is applied twice, first for parameter optimization and then for feature selection. In this work, support vector machine (SVM) (type C-SVC) with new 2level genetic optimizer (genetic training) and feature selection yielded the highest accuracy and F₁-Score of 0.8849 and 0.8762 respectively. Our proposed model can be used to test the performance with huge database and aid the clinicians.     

Sensorimotor Variables and Forced Expiratory Volume as Correlates of Speed,Accuracy, and Variability in Reaction Time Performance in Late Adulthood

The relationship between performance on measures of sensorimotor and physiological function and performance on measures of reaction time (RT) was investigated using an individual differences approach. Older participants showed greater intra-individual variability on the RT measures. They were also slower on measures of four choice visual RT but not on measures of three choice auditory RT. Sensorimotor and physiological variables explained nearly all the age differences in performance on the RT tasks and showed similar patterns of relationships with measures of speed, accuracy, and intra-individual variability. The results support a common factor interpretation of the relationships among these variables in old age.     

Latent variable selection in multidimensional item response theory models using the expectation model selection algorithm

The aim of latent variable selection in multidimensional item response theory (MIRT) models is to identify latent traits probed by test items of a multidimensional test. In this paper the expectation model selection (EMS) algorithm proposed by Jiang et al. (2015) is applied to minimize the Bayesian information criterion (BIC) for latent variable selection in MIRT models with a known number of latent traits. Under mild assumptions, we prove the numerical convergence of the EMS algorithm for model selection by minimizing the BIC of observed data in the presence of missing data. For the identification of MIRT models, we assume that the variances of all latent traits are unity and each latent trait has an item that is only related to it. Under this identifiability assumption, the convergence of the EMS algorithm for latent variable selection in the multidimensional two-parameter logistic (M2PL) models can be verified. We give an efficient implementation of the EMS for the M2PL models. Simulation studies show that the EMS outperforms the EM-based L₁ regularization in terms of correctly selected latent variables and computation time. The EMS algorithm is applied to a real data set related to the Eysenck Personality Questionnaire.     

An examination of employee perceptions of a subjective performance appraisal system

This study sought to identify factors related to employee perceptions of the accuracy of performance ratings they received using a subjective rating system. It also sought to determine if the relationship between these correlates and perceived fairness and accuracy was moderated by employee sex and/or race. The results were based on the analysis of items from questionnaires completed by 234 government employees whose job performance was rated on a graphic rating scale. A factor analysis of the predictor items yielded three factors. The first factor dealt with employees' confidence in their supervisors' qualifications to accurately rate their performance, the second with matters related to a discussion of appraisal outcomes (i.e., rewards and career progress), and the third with the relevance of the appraisal instrument. The measure of perceived fairness and accuracy was regressed onto scores derived for each of the three factors, plus a set of contrast variables representing the various race/sex groupings and a set of demographic variables (viz., job tenure, education level, and supervisory status). Factors 1 and 2, race, and supervisory status were found to be significantly related to perceived fairness and accuracy. Moreover, the relationship of the criterion with job tenure was moderated by race, that with factor 2 was moderated by sex. Management implications of these findings are discussed.     

An efficient cervical disease diagnosis approach using segmented images and cytology reporting

Cervical cancer is the second most common cancer in women globally. A computer aided cervical disease diagnosis system that can relieve pressure on medical experts and save the cost is proposed. To implement our approach in the reality of cervical diseases diagnosis, a multi-modal framework is designed for three kinds of cervical diseases diagnosis that integrates uterine cervix images, Thinprep Cytology Test, human papillomavirus test, and patients’ age. However, too many features increase memory storage costs and computational costs, and it affects the spread of this system in poor areas. Feature selection not only eliminates redundant or irrelevant features but also finds the factors that influence the disease most first is performed in multi-modal frameworks for cervical diseases diagnosis. The detailed process of the method is as follows: first, according the representative color, an efficient image segmentation algorithm is developed; then from three different types of segmented images, we extract color features and texture features for interpreting uterine cervix images; next, Boruta algorithm is applied to feature selection; finally, the performance of Random Forests that utilizes selected features for cervical disease diagnosis is investigated. In the experiment, the proposed multi-modal diagnostic approach gives the final diagnosis for three different kinds of cervical diseases with 83.1% accuracy, which significantly outperforms methods using any single source of information alone. The validation cohort is applied to validate the efficiency of our method, and the performance of random forest obtained by using only 1.2% of features is like or even better than using 100% of features.     

Stochastic search item selection for factor analytic models

In this paper we implement a Markov chain Monte Carlo algorithm based on the stochastic search variable selection method of George and McCulloch (1993) for identifying promising subsets of manifest variables (items) for factor analysis models. The suggested algorithm is constructed by embedding in the usual factor analysis model a normal mixture prior for the model loadings with latent indicators used to identify not only which manifest variables should be included in the model but also how each manifest variable is associated with each factor. We further extend the suggested algorithm to allow for factor selection. We also develop a detailed procedure for the specification of the prior parameters values based on the practical significance of factor loadings using ideas from the original work of George and McCulloch (1993). A straightforward Gibbs sampler is used to simulate from the joint posterior distribution of all unknown parameters and the subset of variables with the highest posterior probability is selected. The proposed method is illustrated using real and simulated data sets.     

Understanding subgroup differences with general mental ability tests in employment selection: Exploring socio-cultural factors across inter-generational groups

In employment selection, general mental ability (GMA) tests predict training and job performance but also lead to subgroup differences which in turn can produce adverse impact against minority groups. Although researchers have explored genetic, developmental, and environmental explanations for ethnic group differences, few studies have explored socio-cultural factors comparing immigrant and non-immigrant job applicants. Given that many ethnic job applicants may also be immigrants, understanding these factors can provide insight into GMA test score differences. The purpose of this paper is to explore the impact of individual and socio-cultural factors on GMA test scores with immigrant and non-immigrant bus driver applicants. This is the first study of its kind to our knowledge that has attempted to disentangle the effects of socio-cultural factors from race/ethnicity in the study of subgroup differences. Incorporating these variables between non-visible minority and minority groups accounted for considerable variance in GMA test scores across groups. The implications of focusing on socio-cultural variables to enhance our understanding of subgroup differences are discussed. Our results specifically suggest that practitioners attend to the issue of the intersecting grounds of potential discrimination when using GMA tests in personnel selection.     

Factor copula models for mixed data

We develop factor copula models to analyse the dependence among mixed continuous and discrete responses. Factor copula models are canonical vine copulas that involve both observed and latent variables, hence they allow tail, asymmetric and nonlinear dependence. They can be explained as conditional independence models with latent variables that do not necessarily have an additive latent structure. We focus on important issues of interest to the social data analyst, such as model selection and goodness of fit. Our general methodology is demonstrated with an extensive simulation study and illustrated by reanalysing three mixed response data sets. Our studies suggest that there can be a substantial improvement over the standard factor model for mixed data and make the argument for moving to factor copula models.     

Latent Structure of the Wisconsin Card Sorting Test after Pediatric Traumatic Head Injury

The performance of 80 pediatric patients with traumatic head injury (THI) on the Wisconsin Card Sorting Test (WCST; Heaton, Chelune, Talley, Kay, & Curtiss, 1993) was examined to determine the underlying latent structure. Exploratory factor analysis with oblique rotation identified three factors: response accuracy, failure to self-monitor, and learning. The response accuracy factor was directly affected by both age and length of coma. Level of performance on this factor also covaried with post-injury psychometric intelligence. It is concluded that interpretation of the WCST results of children with THI should consider the multifactorial nature of the instrument in combination with injury severity characteristics and demographic variables.     

Retention performance as a function of the distinctiveness of memory traces and retention task

Summary In the acquisition phase of an incidental learning paradigm, words that are exemplars of categories were judged as to whether or not they belonged to a specified category. This was followed by a forced choice recognition task. The effects of three variables were studied: a) whether the acquisition task called for a positive or a negative response; b) the number of exemplar words per category; and c) whether distractor items in the recognition test came from the same category as the target words or from other categories. The results showed that recognition errors were influenced by all three variables. Unlike the variables a) and c) which showed significant main effects, the effect of number of exemplars per category was significant only for the same category distractor condition. In this condition more exemplars per category led to more recognition errors. The results are explained by assuming that retention performance is determined by two factors. The first factor is the distinctiveness of the memory traces. This is supposed to be a joint product of task-specific encoding and stimulus characteristics. The second factor is the distinctive similarity between the retrieval information and the stored memory traces. Although the results show some sublte effects that are not easy to explain, it is argued that this model offers a good account of the general findings.     

Latent structure of the Wisconsin Card Sorting Test after pediatric traumatic head injury.

The performance of 80 pediatric patients with traumatic head injury (THI) on the Wisconsin Card Sorting Test (WCST; Heaton, Chelune, Talley, Kay, & Curtiss, 1993) was examined to determine the underlying latent structure. Exploratory factor analysis with oblique rotation identified three factors: response accuracy, failure to self-monitor, and learning. The response accuracy factor was directly affected by both age and length of coma. Level of performance on this factor also covaried with post-injury psychometric intelligence. It is concluded that interpretation of the WCST results of children with THI should consider the multifactorial nature of the instrument in combination with injury severity characteristics and demographic variables.     

A key role for experimental task performance: effects of math talent, gender and performance on the neural correlates of mental rotation

The neurophysiological mechanisms underlying superior cognitive performance are a research area of high interest. The majority of studies on the brain-performance relationship assessed the effects of capability-related group factors (e.g. talent, gender) on task-related brain activations while only few studies examined the effect of the inherent experimental task performance factor. In this functional MRI study, we combined both approaches and simultaneously assessed the effects of three relatively independent factors on the neurofunctional correlates of mental rotation in same-aged adolescents: math talent (gifted/controls: 17/17), gender (male/female: 16/18) and experimental task performance (median split on accuracy; high/low: 17/17). Better experimental task performance of mathematically gifted vs. control subjects and male vs. female subjects validated the selected paradigm. Activation of the inferior parietal lobule (IPL) was identified as a common effect of mathematical giftedness, gender and experimental task performance. However, multiple linear regression analyses (stepwise) indicated experimental task performance as the only predictor of parietal activations. In conclusion, increased activation of the IPL represents a positive neural correlate of mental rotation performance, irrespective of but consistent with the obtained neurocognitive and behavioral effects of math talent and gender. As experimental performance may strongly affect task-related activations this factor needs to be considered in capability-related group comparison studies on the brain-performance relationship.     

CHANGES IN STUDENT AND TEACHER RESPONSES IN OBSERVED AND GENERALIZED SETTINGS AS A FUNCTION OF SUPERVISOR OBSERVATIONS

Two studies are reported in which the effects of supervisor observations of teachers' performance rate and accuracy on both teachers' and severely handicapped students' behaviors were compared with baseline supervisor observations that did not specify rate and accuracy feedback. The latter observation procedure (nonspecific feedback) was more typical of the existing practices of school supervisors. The dependent variables for the teachers were rate and accuracy of teacher behaviors to student responses in individualized instructional settings involving discrete trials or task-analysis steps. The dependent variables for students were the rates of correct and incorrect responding to teacher presentations during the supervisor observation periods. Both studies used a multiple baseline design. The results of Study 1 showed that there were educationally significant changes in teacher and, in turn, student performance as a result of the use of the rate and accuracy procedure. The results of Study 2 replicated those of Study 1 while correcting limitations of Study 1. This latter study also demonstrated that the observation effects of the procedure generalized to teacher performance throughout the day. The results are discussed in terms of the efficiency and cost effectiveness of the procedure to train, monitor, and reinforce teacher effectiveness.     

Global self-esteem and method effects: Competing factor structures,longitudinal invariance,and response styles in adolescents

The Rosenberg Self-Esteem Scale (RSES) is a widely used measure for assessing self-esteem, but its factor structure is debated. Our goals were to compare 10 alternative models for the RSES and to quantify and predict the method effects. This sample involves two waves (N =2,513 9th-grade and 2,370 10th-grade students) from five waves of a school-based longitudinal study. The RSES was administered in each wave. The global self-esteem factor with two latent method factors yielded the best fit to the data. The global factor explained a large amount of the common variance (61% and 46%); however, a relatively large proportion of the common variance was attributed to the negative method factor (34 % and 41%), and a small proportion of the common variance was explained by the positive method factor (5% and 13%). We conceptualized the method effect as a response style and found that being a girl and having a higher number of depressive symptoms were associated with both low self-esteem and negative response style, as measured by the negative method factor. Our study supported the one global self-esteem construct and quantified the method effects in adolescents.     

基于基尼指数的双目标CD-CAT选题策略

双目标CD-CAT的测验结果既可用于形成性评估也可用于终结性评估。基尼指数可度量随机变量的不确定性程度, 值越小则随机变量的不确定程度越低。本文用基尼指数度量被试知识状态类别以及能力估计置信区间后验概率的变化, 提出基于基尼指数的选题策略。Monte Carlo实验表明与已有的选题策略相比, 新策略的知识状态分类精度和能力估计精度都较高, 同时能有效兼顾题库利用均匀性, 并能快速实时响应, 且受认知诊断模型和被试知识状态分布的影响较小, 可用于实际测验中含多种认知诊断模型的混合题库。     

Personality attributes as predictors of psychological well-being for NCOs

The authors examined the nature of the relationships between job-specific personality dimensions and psychological well-being for noncommissioned officers (NCOs) in the Turkish Armed Forces (TAF). A job-specific personality inventory, comprising measures of 11 personality dimensions was developed for selection purposes. The inventory was administered to a representative sample of 1,428 NCOs along with a general mental health inventory developed by the authors, which consisted of 6 dimensions of psychological well-being. Exploratory and confirmatory factor analyses suggested existence of a single factor underlying the 6 psychological well-being dimensions, Mental Health, and 2 latent factors underlying the 11 personality dimensions, Military Demeanor and Military Efficacy. The 2 personality constructs explained 91% of the variance in the Mental Health construct. A stepwise regression indicated that beta weights of the personality measures were significant except for military bearing, orderliness, and dependability. Results suggest that job-specific personality attributes were predictive of mental health. Implications of the findings for the selection of NCOs are discussed.