项目表述效应的统计控制：以中文版生活定向测验为例

本研究以中文版生活定向测验（LOT-R）为例,使用CTCM和CTCU方法进行建模,旨在探讨中文表述的量表中是否存在项目表述效应以及影响项目表述效应的特质因素。采用中文版生活定向测验等量表组成的问卷对334名大学生进行测试。结果表明,中文版LOT-R存在项目表述效应,是一种反向题的项目表述效应,也可称为语言标签效应;在分离了反向题项目表述效应后,发现中文版LOT-R只存在乐观因子,而不再有悲观因子。这表明乐观-悲观同属于一种人格特质,而非两种不同的人格特质;反向题项目表述效应是一种结构性误差,会导致量表结构的扭曲。研究还发现担心错误和父母期望与反向题项目表述效应存在显著的负相关,表明越是担心错误或父母期望水平越高的被试,他们对反向题的项目表述效应（语言标签效应）越小。社会赞许性对反向题项目表述效应没有影响。     

双因子模型：多维构念测量的新视角

双因子模型是一种既有全局因子又有局部因子的模型,近年来有了许多应用。本文讨论了双因子模型和高阶因子模型在数学模型、参数之间的关系,概念上和应用上的差异;概述了双因子模型在信度研究、平衡量表、探索性因子分析和项目反应理论中的应用。作为例子,在Rosenberg自尊量表结构的研究中,通过双因子模型分析了自尊特质效应与项目表述方法效应。     

项目表述方法对中文Rosenberg自尊量表因子结构的影响

该研究探索项目表述方法对量表因子结构的影响,在多特质-多方法(MTMM)矩阵框架下对关联特质-相关模型(The correlated trait-correlated method,CTCM)和关联特质-相关独特性模型(The correlated trait-correlated uniqueness,CTCU)等8个相关方法效应模型进行比较.结果发现,导致中文Rosenberg自尊量表两因子结构的主要是由项目表述的方法效应引起的.因此,在编制量表过程中应考虑方法效应对因子结构的影响.     

项目表述效应对人格测验信度和效标效度的影响——以核心自我评价量表(CSES)为例

项目表述效应是指由项目表述方式的差异引起的与测量内容无关的系统变异,项目表述效应模型的统计本质是一种双因子模型。本研究以核心自我评价量表(CSES)为例,探讨项目表述效应对人格测验信效度的影响。采用核心自我评价量表、生活满意度量表和积极情感消极情感量表对340名“蚁族”进行测查。结果表明,CSES在核心自我评价特质以外,还存在一个反向题项目表述效应因子;忽视项目表述效应对CSES的同质性信度和效标关联效度有重要影响：高估CSES的同质性信度,低估核心自我评价与生活满意度、积极情感的正相关,高估核心自我评价与消极情感的负相关。     

大学生媒体影响、体重指数和身体自尊的关系

采用问卷测量法,考察了432名大学生的媒体影响、体重指数与身体自尊的关系。结果表明:(1)媒体影响、现实体重指数、理想体重指数和身体自尊各个分量表之间存在明显的相关关系;(2)根据变量的相关关系构建的结构方程模型有着较为理想的拟合指数,表明媒体影响、体重指数对身体自尊有着直接或间接的影响作用。     

编制与评价职业兴趣测验中值得明确的几个问题

本文纵观国内外职业兴趣测验的历史和现状,影响职业兴趣测验准确性的因素如受测者的年龄、受测者的性别、测验量表的多少、职业决策过程、职业兴趣测验与其它相关测验的整合作了分析和探讨,并在此基础上提出了编制与评价职业兴趣测验时的一些建议,如受测者年龄不低于１４岁,要按职业的文化要求考虑性别常模分不分开,要客观评价兴趣测验的实证效度等．     

自恋与自尊：社会赞许性的中介作用

目的：考察社会赞许性在自恋与自尊关系中所起的中介作用。方法：对614名大学生进行自恋人格问卷（NPI）、自尊问卷（SES）、马洛一克罗恩社会赞许量表（MCSD）测评。结果：高自恋者的自尊水平显著高于低自恋者;自恋总分及权欲、优越感、自我钦羡三个因子与自尊显著性正相关,特权感因子与自尊负相关;自恋总分及权欲、优越感、自我钦羡因子对自尊有较好的直接预测效应,以社会赞许性为中介能更好地解释特权感因子与自尊的关系。结论：自恋与自尊关系稳定,自恋的权欲、优越感、自我钦羡因子预测健康的自尊,特权感因子预测防御性的自尊。     

大学新生自尊与成就动机的关系及外显自尊的中介效应

采用自尊自评量表（SES）,内隐联想测验（IAT）和成就动机量表（AMS）,对100A大学新生外显自尊,内隐自尊以及成就动机进行测量,探测大学新生外显自尊,内隐自尊与成就动机的关系及外显自尊的中介效应。结果发现,外显自尊、内隐自尊对成就动机有中等程度的相关关系;IAT2对成就动机有直接预测作用,而IAT1却没有直接预测作用,但是外显自尊在IATI和成就动机之间具有中介作用。     

自尊与亲社会行为的关系:来自元分析的证据

已有关于自尊与亲社会行为关系的研究结论有所不同,采用元分析的方法对自尊与亲社会行为之间的关系进行探讨.通过文献检索,共纳入原始文献28篇,独立效应值35个,包含15715名被试.通过随机效应模型分析,主效应检验发现自尊与亲社会行为呈中等程度正相关(r=0.32);调节效应检验表明年龄、自尊测量工具及亲社会行为测量工具可...     

羞耻感对男性暴力犯自我攻击的影响：自尊和愤怒内部表达的中介作用

为探索暴力犯的羞耻感与自我攻击的关系及自尊和愤怒内部表达的中介作用，采用羞耻感量表、自尊量表、愤怒内部表达和自我攻击分量表对533名成年男性暴力犯进行调查。结果显示：（1）羞耻感、自尊、愤怒内部表达和自我攻击之间相关均显著；（2）羞耻感显著正向预测自我攻击，自尊和愤怒内部表达在其中起并列中介作用，分别占总效应的17.91%和34.33%。本研究揭示了羞耻感是导致暴力犯自我攻击的重要因素，且通过降低自尊和增加愤怒内部表达产生影响。     

Factor structure of the Penn State Worry Questionnaire: examination of a method factor

The Penn State Worry Questionnaire (PSWQ) was originally designed as a unifactorial measure of pathological trait worry. However, recent studies supported a two-factor solution with positively worded items loading on the first factor and reverse-scored items loading on a second factor. The current study compared this two-factor model to a negative wording method factor solution among college students. A method factor model with all PSWQ items loading on a single worry factor and reverse-scored items loading on a negative wording method factor provided as good a fit as the two-factor model. This method factor alone did not predict a generalized anxiety disorder diagnosis. Finally, the psychometric properties of an abbreviated scale containing only positively worded items were examined. The PSWQ appears to measure a single unitary construct, but response patterns differ between positively worded and reverse-scored items. Theoretical implications for pathological worry and assessment-related issues are discussed.     

Measuring shyness: analysis of the Revised Cheek and Buss Shyness scale

A sample of 741 university students completed the Revised Cheek and Buss Shyness scale. Item analysis showed that the scale has sound psychometric properties. There was no gender difference in total shyness scores although there were differences on two of the 14 items. A series of confirmatory factor analyses tested the fit of several models of the factor structure of the scale and showed that a two-factor model taking into account differences in the direction of item wording provided a satisfactory and parsimonious fit to the data. The implications of the factor structure for understanding the nature of trait shyness are discussed.     

Comparison of classical and modern methods for measuring and correcting for acquiescence

Likert-type self-report scales are frequently used in large-scale educational assessment of social-emotional skills. Self-report scales rely on the assumption that their items elicit information only about the trait they are supposed to measure. However, different response biases may threaten this assumption. Specifically, in children, the response style of acquiescence is an important source of systematic error. Balanced scales, including an equal number of positively and negatively keyed items, have been proposed as a solution to control for acquiescence, but the reasons why this design feature worked from the perspective of modern psychometric models have been underexplored. Three methods for controlling for acquiescence are compared: classical method by partialling out the mean; an item response theory method to measure differential person functioning (DPF); and multidimensional item response theory (MIRT) with random intercept. Comparative analyses are conducted on simulated ratings and on self-ratings provided by 40,649 students (aged 11–18) on a fully balanced 30-item scale assessing conscientious self-management. Acquiescence bias was explained as DPF and it was demonstrated that: the acquiescence index is highly related to DPF; balanced scales produce scores controlled for DPF; and MIRT factor scores are highly related to scores controlled for DPF and the random intercept is highly related to DPF.     

Mind your words: positive and negative items create method effects on the Five Facet Mindfulness Questionnaire

Mindfulness, a construct that entails moment-to-moment effort to be aware of present experiences and positive attitudinal features, has become integrated into the sciences. The Five Facet Mindfulness Questionnaire (FFMQ), one popular measure of mindfulness, exhibits different responses to positively and negatively worded items in nonmeditating groups. The current study employed confirmatory factor analysis with a large undergraduate sample to examine the validity of a hierarchical mindfulness model and whether response patterns related to item wording arose from method effects. Results indicated that a correlated facets model better explained the data and that negative and positive wording constituted substantive method effects. This study suggests that the FFMQ measures components that may relate to, but do not seem to directly reflect, a latent variable of mindfulness. The authors recommend against the use of an FFMQ total score, favoring individual scale scores, and further examination of method effects in mindfulness scales.     

Effects of estimation methods on making trait-level inferences from ordered categorical items for assessing psychopathology

In assessments of attitudes, personality, and psychopathology, unidimensional scale scores are commonly obtained from Likert scale items to make inferences about individuals' trait levels. This study approached the issue of how best to combine Likert scale items to estimate test scores from the practitioner's perspective: Does it really matter which method is used to estimate a trait? Analyses of 3 data sets indicated that commonly used methods could be classified into 2 groups: methods that explicitly take account of the ordered categorical item distributions (i.e., partial credit and graded response models of item response theory, factor analysis using an asymptotically distribution-free estimator) and methods that do not distinguish Likert-type items from continuously distributed items (i.e., total score, principal component analysis, maximum-likelihood factor analysis). Differences in trait estimates were found to be trivial within each group. Yet the results suggested that inferences about individuals' trait levels differ considerably between the 2 groups. One should therefore choose a method that explicitly takes account of item distributions in estimating unidimensional traits from ordered categorical response formats. Consequences of violating distributional assumptions were discussed.     

Method Effects on an Adaptation of the Rosenberg Self-Esteem Scale in Greek and the Role of Personality Traits

Rosenberg's Self-Esteem Scale is a balanced, 10-item scale designed to be unidimensional; however, research has repeatedly shown that its factorial structure is contaminated by method effects due to item wording. Beyond the substantive self-esteem factor, 2 additional factors linked to the positive and negative wording of items have been theoretically specified and empirically supported. Initial evidence has revealed systematic relations of the 2 method factors with variables expressing approach and avoidance motivation. This study assessed the fit of competing confirmatory factor analytic models for the Rosenberg Self-Esteem Scale using data from 2 samples of adult participants in Cyprus. Models that accounted for both positive and negative wording effects via 2 latent method factors had better fit compared to alternative models. Measures of experiential avoidance, social anxiety, and private self-consciousness were associated with the method factors in structural equation models. The findings highlight the need to specify models with wording effects for a more accurate representation of the scale's structure and support the hypothesis of method factors as response styles, which are associated with individual characteristics related to avoidance motivation, behavioral inhibition, and anxiety.     

Latent growth curve analysis with dichotomous items: Comparing four approaches

A Monte Carlo study was used to compare four approaches to growth curve analysis of subjects assessed repeatedly with the same set of dichotomous items: A two‐step procedure first estimating latent trait measures using MULTILOG and then using a hierarchical linear model to examine the changing trajectories with the estimated abilities as the outcome variable; a structural equation model using modified weighted least squares (WLSMV) estimation; and two approaches in the framework of multilevel item response models, including a hierarchical generalized linear model using Laplace estimation, and Bayesian analysis using Markov chain Monte Carlo (MCMC). These four methods have similar power in detecting the average linear slope across time. MCMC and Laplace estimates perform relatively better on the bias of the average linear slope and corresponding standard error, as well as the item location parameters. For the variance of the random intercept, and the covariance between the random intercept and slope, all estimates are biased in most conditions. For the random slope variance, only Laplace estimates are unbiased when there are eight time points.