用结构方程建模（SEM）估计概论理论（GT）中的评分者信度

评分者的信度分析,已从经典测量理论的方法（Kendall和谐系数）发展为基于现代测量理论（如GT）的方法,但由于GT中方差分量估计的工具和途径有限,使该方法的推广受到限制。本文介绍了结构方程建模（SEM）估计GT中方差分量的必要性及原理,并通过对某省直国家机关公务员录用考试结构化面试的评分者信度分析,阐述用SEM估计概化理论中不同设计下的评分者信度研究。     

概化理论方差分量估计的跨分布分析

方差分量估计是进行概化理论分析的关键。采用MonteCarlo模拟技术,探讨心理与教育测量数据分布对概化理论各种方法估计方差分量的影响。数据分布包括正态、二项和多项分布,估计方法包括Traditional、Jackknife、Bootstrap和MCMC方法。结果表明:(1)Traditional方法估计正态分布和多项分布数据的方差分量相对较好,估计二项分布数据需要校正,Jackknife方法准确地估计了三种分布数据的方差分量,校正的Bootstrap方法和有先验信息的MCMC方法(MCMCinf)估计三种分布数据的方差分量结果较好;(2)心理与教育测量数据分布对四种方法估计概化理论方差分量有影响,数据分布制约着各种方差分量估计方法性能的发挥,需要加以区分地使用。     

在GLMM框架下统一GT与IRT

本文首次提出使用广义线性混合模型(Generalized Linear Mixed Model, GLMM)对概化理论(GT)和项目反应理论(IRT)进行统合,即在一次统计中就能同时获得GT和IRT所需要的估计结果。模拟研究结果显示：相比于传统的GT方差分量估计方法——期望均值平方(Expected Mean Squares, EMS),GLMM可以获得更准确的方差分量、G系数和Φ系数,而且GLMM获得的题目难度参数估计精度优于传统Rasch模型。实证研究展示GLMM在实际心理测量数据分析中的应用。     

概化理论中的模型选择、数据解释和指标比较——评刘远我等的两篇论文

对刘远我等在1998年《心理学报》和1999年《心理科学》上发表的两篇慨化理论(GT)的应用研究论文．提出了对GT的一些主要问题的不同理解。这些问题包括模型选择、方差分量解释、一致性指标比较等。     

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。     

GT与IRT的比较: 北京奥运会男子10米跳台跳水分析

概化理论(GT)和项目反应理论(IRT)从两个不同的方向发展了经典测量理论, GT和IRT中的多面Rasch测量模型(MFRM)在主观评分中都可以用来估计评分中各变异来源对变异的贡献, 对测评的信度进行估计, 提出测评改进意见。12名运动员参加了2008北京奥运会男子10米跳台跳水决赛, 比赛共6个回合, 7名裁判独立对他们在各个回合的表现进行打分。GT和MFRM比较一致地认为运动员自身、回合、运动员与回合的交互效应是运动员得分的重要变异来源, 而裁判员对运动员得分差异的贡献不显著。MFRM同时还估计出难度系数是影响男子跳台跳水成绩的重要变异来源, 在评分等级6.5附近存在步校准错乱, 得出的运动员成绩排序与2008奥运实际排序有所不同。在GT中难度系数作为隐藏侧面, 其效应未能分离出来。GT和MFRM从两个不同的方面给测量提供改进意见: GT发现可以通过增加回合数来提高g系数, 而增加裁判数对其影响不大。MFRM给出各侧面的要素(如某裁判、运动员等)的估计值及其标准误, 它给出的诊断性拟合统计也有助于甄别异常得分或评分模式。     

基于概化理论的方差分量变异量估计

概化理论广泛应用于心理与教育测量实践中, 方差分量估计是进行概化理论分析的关键。方差分量估计受限于抽样, 需要对其变异量进行探讨。采用蒙特卡洛(Monte Carlo)数据模拟技术, 在正态分布下讨论不同方法对基于概化理论的方差分量变异量估计的影响。结果表明: Jackknife方法在方差分量变异量估计上不足取; 不采取Bootstrap方法的“分而治之”策略, 从总体上看, Traditional方法和有先验信息的MCMC方法在标准误及置信区间这两个变异量估计上优势明显。     

经典真分数理论与概化理论信度观评析

经典真分数理论信度观和概化理论信度观是当代心理测量学中的两大信度理论。该简要地介绍了两种信度观的基本思想,并在此基础上,对曲种信度理论的特征进行系统的分析和比较。章认为：概化理论信度观在理论假设、测量误差源的综合分析技术、测量条件的推广研究等方面,比经典真分数理论具有史多的优越性,但概化理论信度观沣重所测心埋特质的单维性和方差分量估计上的负值现象等局限性,大火限制了概化理论在实际测量中的应用。章指出,就目前的发展现状而言,没有哪一种信度理论能够包容或取代另一种信度理论,由此提出了当代信度理论合理使用的几点建议。     

经典真分数理论与概化理论信度观评析

经典真分数理论信度观和概化理论信度观是当代心理测量学中的两大信度理论。该文简要地介绍了两种信度观的基本思想,并在此基础上,对两种信度理论的特征进行系统的分析和比较。文章认为：概化理论信度观在理论假设、测量误差源的综合分析技术、测量条件的推广研究等方面,比经典真分数理论具有更多的优越性,但概化理论信度观注重所测心理特质的单维性和方差分量估计上的负值现象等局限性,大大限制了概化理论在实际测量中的应用。文章指出,就目前的发展现状而言,没有哪一种信度理论能够包容或取代另一种信度理论,由此提出了当代信度理论合理使用的几点建议。     

考试评分缺失数据的概化理论分析

考试评分缺失数据较为常见,如何有效利用现有数据进行统计分析是个关键性问题。在考试评分中,题目与评分者对试卷得分的影响不容忽视。根据概化理论原理,按考试评分规则推导出含有缺失数据双侧面交叉设计（p×i×r）方差分量估计公式,用Matlab7.0软件模拟多组缺失数据,验证此公式的有效性。结果发现：（1）推导出的公式较为可靠,估计缺失数据的方差分量偏差相对较小,即便数据缺失率达到50％以上,公式仍能对方差分量进行较为准确地估计;（2）题目数量对概化理论缺失数据方差分量的估计影响最大,评分者次之,当题目和评价者数量分别为6和5时,公式能够趋于稳定地估计;（3）学生数量对各方差分量的估计影响较小,无论是小规模考试还是大规模考试,概化理论估计缺失数据的多个方差分量结果相差不大。     

Prediction of Simulated Infantry-Combat Performance From a General Measure of Individual Aptitude

The relationships between individual soldier aptitudes and the potential for successful combat performance have not been systematically studied. The purpose of this research was to determine the relationships between one measure of individual aptitude, the General Technical (GT) composite from the Armed Services Vocational Aptitude Battery (ASVAB), and performance during simulated infantry combat. Eleven 3-man teams of infantry soldiers defended against enemy rifle squads performing a Movement to Contact, Hasty Attack mission. The correlation between GT and casualties inflicted was r(30) = .37, p < .05; the correlation between GT and casualties sustained was r(30) = -.25, p > .05. Marksmanship scores did not correlate with aptitude or the two criterion measures, partially due to limited variance. Given further validation, these results could provide important information for Army policy and management decisions.     

校正的Bootstrap方法对概化理论方差分量及其变异量估计的改善

Bootstrap方法是一种有放回的再抽样方法, 可用于概化理论的方差分量及其变异量估计。用Monte Carlo技术模拟四种分布数据, 分别是正态分布、二项分布、多项分布和偏态分布数据。基于p×i设计, 探讨校正的Bootstrap方法相对于未校正的Bootstrap方法, 是否改善了概化理论估计四种模拟分布数据的方差分量及其变异量。结果表明：跨越四种分布数据, 从整体到局部, 不论是“点估计”还是“变异量”估计, 校正的Bootstrap方法都要优于未校正的Bootstrap方法, 校正的Bootstrap方法改善了概化理论方差分量及其变异量估计。     

An intervention to increase walking requires both motivational and volitional components: a replication and extension

An intervention to increase walking has previously been developed, consisting of three motivational techniques, designed to increase self-efficacy, and three volitional techniques, designed to help translate intentions into action. Previous research found large effects (d?=?0.90) on the objectively measured walking behaviour of 130 English adult volunteers, mediated by self-efficacy. The present study aimed to replicate this intervention, and decompose the intervention to assess whether both motivational and volitional intervention components are necessary. A three-group experimental design was employed, with n?=?35 adult volunteers randomly allocated to receive one of three interventions: (a) a "combined" intervention, containing motivational and volitional components in session at T1 and a filler task at T2, (b) a "motivation first" intervention, where the motivational components were received at T1 and the volitional components at T2, or (c) a "volition first" intervention, where the volitional components were received at T1 followed by motivational components at T2. At T2, there was a significant main effect of time, such that there was an increase in walking, but this did not differ between groups. At T3, the "combined" intervention group showed a large (d?=?1.06) and significant (p?=?0.036) increase in walking behaviour, in contrast to both other interventions (time?×?groups interaction, p?=?0.003). The "combined" intervention also produced a significant increase in self-efficacy, relative to the two other interventions. This study demonstrates generalisability of previous large intervention effects and suggests that use of both motivational and volitional components is optimal in producing change in walking behaviour. Future research should explore the mechanisms by which techniques to increase self-efficacy and planning interact.     

Analysis of covariance structures

A general method is presented for estimating variance components when the experimental design has one random way of classification and a possibly unbalanced fixed classification. The procedure operates on a sample covariance matrix in which the fixed classes play the role of variables and the random classes correspond to observations. Cases are considered which assume (i) homogeneous and (ii) nonhomogeneous error variance, and (iii) arbitrary scale factors in the measurements and homogeneous error variance. The results include maximum-likelihood estimations of the variance components and scale factors, likelihood-ratio tests of the goodness-of-fit of the model assumed for the design, and large-sample variances and covariances of the estimates. Applications to mental test data are presented. In these applications the subjects constitute the random dimension of the design, and a classification of the mental tests according to objective features of format or content constitute the fixed dimensions.Preparation of this paper has been supported in part by NSF Grant GB-939 and U. S. P. H. Grant GM-1286-01. Computer time was donated by the Computation Center, University of Chicago.Now at the University of Chicago.Now at the University of Georgia.     

Generalizability theory: A practical guide to study design,implementation, and interpretation

Generalizability Theory (GT) offers increased utility for assessment research given the ability to concurrently examine multiple sources of variance, inform both relative and absolute decision making, and determine both the consistency and generalizability of results. Despite these strengths, assessment researchers within the fields of education and psychology have been slow to adopt and utilize a GT approach. This underutilization may be due to an incomplete understanding of the conceptual underpinnings of GT, the actual steps involved in designing and implementing generalizability studies, or some combination of both issues. The goal of the current article is therefore two-fold: (a) to provide readers with the conceptual background and terminology related to the use of GT and (b) to facilitate understanding of the range of issues that need to be considered in the design, implementation, and interpretation of generalizability and dependability studies. Given the relevance of this analytic approach to applied assessment contexts, there exists a need to ensure that GT is both accessible to, and understood by, researchers in education and psychology. Important methodological and analytical considerations are presented and implications for applied use are described.     

STATISTICAL AND PSYCHOMETRIC INFERENCE IN PRINCIPAL COMPONENTS ANALYSIS

Statistical inference applied to principal components analysis deals with estimating the parameters of the correlation matrix, R, found in the population, from the characteristics of the sample matrix, R*. On the other hand, psychometric inference refers to estimating the internal consistency of the components themselves, so that the decisions about retaining a component for further analysis can be based upon psychometric criteria. A slightly modified approach to statistical inference, which focuses upon the variance of the Components in the population, has been suggested. This viewpoint can be extended to estimating the true score variance and the reliabilities of the components in the population of subjects. Psychometric tests of significance can then be made statistical in nature.