拓展后的GGUM在LOT-R中的适用性分析

该研究对拓广等级展开模型（GGUM）进行了拓展,取消GGUM中关于主观反应类别阈限对称的假设,并将拓展之后的新模型和GGUM同时用于生活取向测验修订版（LOT-R）的被试反应数据分析,采用新编的单项目、两项目对和三项目组χ2/df计算程序计算和比较新模型和GGUM在该测验数据上的拟合差异。结果显示,新编程序与Stark等人开发的MODFIT程序具有同样的有效性,新模型在这些指标上的值显著小于GGUM,并且均小于3,表明新模型较GGUM更适合于分析LOT-R的反应数据,说明新模型更适用于分析具有多个评定等级的人格测验数据。根据以上结果,该研究认为,未来人格测验的数据分析应该使用没有对主观反应类别阈限进行对称限定的新拓展的模型更合理。     

展开模型和优势模型在职业兴趣测验中的比较研究

该研究调查了展开模型(GGUM)和优势模型(GRM)对职业兴趣测验反应数据的拟合情况,并对展开模型和优势模型两种测验编制方法在职业兴趣测验中进行了比较。结果发现:(1)展开模型的模型拟合情况和测量精度优于累积模型,两种模型对被试能力参数估计的差异主要体现在极端被试上,对兴趣水平极端高的被试,展开模型的估计值更精确;(2)采用展开模型编制的测验在信度上远远高于Likert方法编制的测验,中间区域题目的增加提高了测验的信度,但两种方法在测验的效标关联效度上没有差异。结果表明,在职业兴趣的测量上,展开模型更精确;在职业兴趣测验的编制上,GGUM和Likert法没有差异,反而Likert法具有简便、易懂的优势。     

IRT展开模型及对非累积反应机制的检测

被试回答人格测验题目时并不是特质水平越高其得分率越高,这称为非累积反应机制。广义等级展开模型GGUM就是针对这一机制提出来的。使用EPQ和五因素人格问卷发现GGUM比累积IRT模型有更好的模型拟合度和测量精度。研究结果表明GGUM有其合理性,且有助于反应心理过程机制的深入探讨。     

人格测验的项目反应过程与展开模型研究述评

一些研究者提出了人格测验的项目反应理想点过程假设, 并在此基础上开发了拓广等级展开模型(GGUM)用于现有人格测验的数据分析和新量表的建构, 显示出了较优势模型更优良的一些性质。不过, 现有项目反应过程的研究结果来自于大样本的调查, 缺乏实验证据的支持, GGUM也存在不适用于分析多类别人格测验数据的局限性。未来需要对GGUM进行拓展, 剔除其主观反应类别阈限对称的限定。此外, 还要重视配对格式人格测验数据分析模型的研究和开发。     

项目反应理论新进展：基于3PLM和GRM的混合模型

IRT中的计量模型较多,不同计量模型适合不同特点的数据资料,实际工作者应根据实际情况选择适当的IRT模型来分析数据。我国是个考试、测评大国,测评的题型丰富多样,在实际应用IRT时,一个模型往往很难反应所有数据资料本身的特点,这时可考虑应用多个IRT模型（即“混合模型”）来分析,以达到对数据的最佳拟合。本文对混合模型的思想方法及原理、参数估计的实现、以及模型性能进行了研究,发现：（1）本文自主开发的混合模型参数估计程序Mix_Tu具有较高的返真性,且与国际知名测量软件Parscale相当。（2）在“项目异常”情况下,Mix_Tu程序对参数b和c的估计受数据异常程度的影响要大于Parscale程序,而对参数a的估计受数据异常程度的影响要小于Parscale程序,而在参数theta上两个程序相当。（3）在“被试异常”情况下,Mix_Tu程序对所有参数的估计受数据异常程度的影响均要小于Parscale程序,Mix_Tu程序表现的更为稳健。     

认知诊断模型中项目水平模型比较统计量的健壮性

使用模拟研究方法比较了以往研究中提出的基于观察信息矩阵、三明治矩阵的Wald（分别表示为W_Obs、W_Sw）、似然比（Likelihood Ratio）统计量以及新提出的基于经验交叉相乘信息矩阵的Wald统计量（W_XPD）在模型——数据失拟条件下进行项目水平上模型比较时的表现。结果显示：（1）W_Sw的一类错误控制率有很强的健壮性。（2）W_XPD在Q矩阵错误设定的大多数条件下的表现优于W_Sw。结论：模型—数据拟合良好时可以使用W_Sw进行项目水平上的模型比较,当模型与数据失拟时W_XPD可能是更好的选择。     

采用理想点方法检验生活取向测验及其修订版的单维性

该文从项目反应过程出发,认为可能是由于原有的维度分析方法将生活取向测验（LOT）及其修订版（LOT-R）这两个测验的项目反应理想点过程错误地限定为优势过程,导致两个测验违背单维性的结论;进而采用主成分分析的未旋转成分负荷散点图和对应分析的维度分数散点图,并结合项目相关矩阵法验证LOT和LOT-R的维度。结果显示LOT-R的单维性得到了验证,LOT由于包含两个测量应对的项目而导致违背单维性。最后作者提出应该与理想点过程相一致的心理计量学模型对LOT-R的数据进行分析。     

认知诊断模型中项目水平模型比较统计量的健壮性

使用模拟研究方法比较了以往研究中提出的基于观察信息矩阵、三明治矩阵的Wald（分别表示为W_Obs、W_Sw）、似然比（Likelihood Ratio）统计量以及新提出的基于经验交叉相乘信息矩阵的Wald统计量（W_XPD）在模型——数据失拟条件下进行项目水平上模型比较时的表现。结果显示：（1）W_Sw的一类错误控制率有很强的健壮性。（2）W_XPD在Q矩阵错误设定的大多数条件下的表现优于W_Sw。结论：模型—数据拟合良好时可以使用W_Sw进行项目水平上的模型比较,当模型与数据失拟时W_XPD可能是更好的选择。     

基于GT和多面Rasch模型的结构化面试分析

该研究应用GT和多面Rasch模型对结构化面试数据进行分析,并提出一些建议针对某辅导员招聘面试数据,运用GT从宏观上分析应聘者、考官和项目所带来的总体误差大小,在此基础上,运用多面Rasch模型从微观上进一步探查考官严厉度、应聘者能力差异、项目难易度及侧面偏差．结果表明：1）GT分析表明应聘者产生的变异较大（90．65％）,说明面试可靠性较高,且当考官数为2时可靠性已较好.2）多面Rasch模型分析出了各侧面效应中的非拟合因素及交互效应中的偏差因素,表明面试误差主要来自考官间严厉度的差异及其自身一致性的不稳定。将GT与多面Rasch模型相结合分析面试数据不仅能测查出评价过程各方面的问题因素,并能更好地作整体把握。     

认知诊断资料-模型拟合检验统计量及应用

本文将IRT常用资料-模型拟合检验统计量χ^2和G^2引入认知诊断领域,具体讨论了这两个统计量在认知诊断资料-模型拟合检验的可行性及其侦查效果,并讨论了其在实际中的应用,为研究者及实际应用者在认知诊断资料模型拟合检验中提供借鉴及方法学支持。研究发现：（1）χ^2和G^2统计量在认知诊断资料-模型拟合检验中,犯Ⅰ类错误和Ⅱ错误概率均小于5%,表明χ^2和G^2统计量均能有效地侦查项目失拟情况,均可用于认知诊断中的资料-模型拟合检验。（2）测验长度、被试样本容量、认知属性个数等因素均会影响χ^2和G^2统计量的侦查效果。（3）就所犯两类错误率而言,χ^2统计量优于G^2统计量。（4）两统计量均能有效地侦查出项目因属性被错误标定而导致的失拟,因而它们在侦查属性错误标定中有一定的应用前景。     

GGUMLINK: A computer program to link parameter estimates of the generalized graded unfolding model from item response theory

The generalized graded unfolding model (GGUM) is an item response theory (IRT) model that implements symmetric, nonmonotonic, single-peaked item characteristic curves. The GGUM is appropriate for measuring individual differences for a variety of psychological constructs, especially attitudes. Like other IRT models, the location and scale (i.e., the metric) of parameter estimates from the GGUM are data dependent. Therefore, parameter estimates from alternative calibrations will generally not be comparable, even when responses to the same items are analyzed. GGUMLINK is a computer program developed to reexpress parameter estimates from two separate GGUM calibrations in a common metric. In this way, the results from separate calibrations of model parameters can be compared. GGUMLINK can secure a common metric by using one of five methods that have recently been generalized to the GGUM. The GGUMLINK executable program is available free and may be downloaded from http://www.education.umd.edu/EDMS/tutorials/index.html.     

Generalized graded unfolding model with structural equation for subject parameters

The generalized graded unfolding model (GGUM) is capable of analyzing polytomous scored, unfolding data such as agree‐disagree responses to attitude statements. In the present study, we proposed a GGUM with structural equation for subject parameters, which enabled us to evaluate the relation between subject parameters and covariates and/or latent variables simultaneously, in order to avoid the influence of attenuation. Additionally, an algorithm for parameter estimation is newly implemented via the Markov Chain Monte Carlo (MCMC) method, based on Bayesian statistics. In the simulation, we compared the accuracy of estimates of regression coefficients between the proposed model and a conventional method using a GGUM (where regression coefficients are estimated using estimates of θ). As a result, the proposed model performed much better than the conventional method in terms of bias and root mean squared errors of estimates of regression coefficients. The study concluded by verifying the efficacy of the proposed model, using an actual data example of attitude measurement.     

Middle Response Scale Options are Inappropriate for Ideal Point Scales

Purpose

The purpose of this study is to provide a theoretical rationale for the inappropriateness of middle response options on the response scales offered on ideal point scales, and to provide empirical support for this argument to assist ideal point scale development.

Design/Methodology/Approach

The same ideal point scale was administered in three quasi-experimental groups varying only in the response scale offered: the three groups received either a four-, five-, or six-option response scale. An ideal point Item Response Theory model, the Generalized Graded Unfolding Model (GGUM), was fit to the response data, and model-data fit was compared across conditions.

Findings

Responses from the four- and six-option conditions were fit well by the GGUM, but responses from the five-option condition were not fit well.

Implications

Despite the scale being constructed to follow the tenets of ideal point responding, the GGUM was unable to provide a reasonable probabilistic account of responding when the response scale contained a middle option. The authors find support for the argument that an odd-numbered response scale does not match the principles of ideal point responding, and can actually result in misspecifying the underlying response process.

Originality/Value

Although a growing body of research has suggested that attitude and personality measurement is best conceptualized under the assumptions of ideal point responding, little practical advice has been given to researchers or practitioners regarding scale creation. This was the first study to theoretically and empirically assess the response scale on ideal point scales, and offer guidance for constructing ideal point scales.     

DIMENSIONS OF COMPLIANCE-GAINING SITUATIONS

A seven-factor model of situation perception, including “personal benefits,”“intimacy,”“rights,”“resistance,”“dominance,”“situation apprehension,” and “relational consequences” was proposed. Confirmatory factor analysis was employed to test the fit of the model to the data. One sample (N=450) was employed to test a six-factor sub model (excluding relational consequences) and a separate sample (N= 270), employing a different situation, was utilized to test the seven-factor model. The proposed model proved to fit the data well (x²/df ratios of 3.14 and 2.97, respectively) and was superior to alternative models. Future research recommendations were provided.     

Personality similarity in twins reared apart and together

We administered the Multidimensional Personality Questionnaire (MPQ) to 217 monozygotic and 114 dizygotic reared-together adult twin pairs and 44 monozygotic and 27 dizygotic reared-apart adult twin pairs. A four-parameter biometric model (incorporating genetic, additive versus nonadditive, shared family-environment, and unshared environment components) and five reduced models were fitted through maximum-likelihood techniques to data obtained with the 11 primary MPQ scales and its 3 higher order scales. Solely environmental models did not fit any of the scales. Although the other reduced models, including the simple additive model, did fit many of the scales, only the full model provided a satisfactory fit for all scales. Heritabilities estimated by the full model ranged from .39 to .58. Consistent with previous reports, but contrary to widely held beliefs, the overall contribution of a common family-environment component was small and negligible for all but 2 of the 14 personality measures. Evidence of significant nonadditive genetic effects, possibly emergenic (epistatic) in nature, was obtained for 3 of the measures.     

物质主义价值观量表在大学生群体中的修订

以449名大学本科生为被试,对物质主义价值观量表（MVS）进行中文版的修订,考察其信效度指标。探索性因素分析发现,修订后的量表具有与原量表基本相同的三个维度;验证性因素分析结果显示,MVS的三因素结构拟合较好;MVS的内部一致性信度为0．792,重测信度为0．830;MVS的效标关联效度良好。表明修订后的MVS具有较好的心理测量学属性,可作为测量我国大学生物质主义价值观的工具。     

Measuring intention in language learning: a confirmatory factor analysis

Intention in language learning has not been studied effectively in research on second language (L2) learning. The goal is to fill this gap by designing and testing a measure of L2 learning intention. The scale was differentiated into two distinct but correlated components, goal intention and implementation intention, within the L2 context. The two intention scales were examined for reliability and validity using a series of standard psychometric procedures. A confirmatory factor model was then constructed and tested with a sample of 333 senior high school and college students. The results showed that a modified model had good psychometric characteristics and reasonable fit to the data.