心理科学领域内的客观测量——Rasch模型之特点及发展趋势

Rasch模型是在国外学术界受到广泛关注和深入研究的一个潜在特质模型。该模型为解决心理科学领域内测量的客观性问题提供了一个可行性很高的解决方案。而国内关于Rasch模型的理论探讨和应用研究却并不多见。不同于一般项目反应理论, Rasch模型要求所收集的数据必须符合模型的先验要求, 而不是使用不同的参数去适应数据的特点。Rasch模型的主要特点(包括个体与题目共用标尺、线性数据、参数分离)确保了客观测量的实现。未来关于Rasch模型的研究方向包括多维度Rasch模型、测验的等值与链接、计算机自适应性考试, 大型应用测量系统(比如Lexile系统)等等。     

华人人际关系的概念化--针对中国香港地区大学生的实证研究

以杨中芳所构建的华人人际关系概念化理论模型为蓝本,并采用以往的实证研究方法为基础,揭示了当代华人对人际关系如何进行概念化及其潜在的认知维度。杨氏模型认为华人人际关系主要由既定成分、情感成分和工具成分组成。该研究以中国香港地区的大学生为被试进行调查,首先采用访谈的形式,分家族、学校及日常生活情境建立人际关系清单。然后采用多维量法(Multi-dimensional scaling)处理被试对关系进行分类的数据。结果发现,在不同的生活情境中,被试均依据情感程度对关系进行分类。在等级差异明显的家族和学校情境中,被试依据辈份和等级差异的突显性进行分类;而在等级差异小的生活情境中,则采用功利程度为标准。本文对被试在不同情境中所采用的不同维度进行了讨论,并提出华人人际关系的义务具有工具性质的实证结果。     

Stevens’ Direct Scaling Methods and the Uniqueness Problem

Stevens postulated that we can use the responses of a participant in a ratio scaling experiment directly to construct a psychophysical function representing the participant's sensations. Although Stevens' methods of constructing measurement scales are widely used in the behavioral sciences, the problem of which scale type is appropriate to describe ratio scaling data is still unresolved. To deal with this problem, we develop a theoretical framework to specify the scale type attained by Stevens' direct scaling methods. It is shown, under fairly mild background assumptions, that the behavioral axioms presented in this paper are necessary and sufficient for the psychophysical functions to be ordinal-, interval-, log-interval-, or ratio-scales. Furthermore, suggestions on how to test these behavioral axioms are provided. Requests for reprints should be sent to thomas.     

Nonparametric Estimation of Item and Respondent Locations from Unfolding-type Items

Unlike their monotone counterparts, nonparametric unfolding response models, which assume the item response function is unimodal, have seen little attention in the psychometric literature. This paper studies the nonparametric behavior of unfolding models by building on the work of Post (1992). The paper provides rigorous justification for a class of nonparametric estimators of respondents’ latent attitudes by proving that the estimators consistently rank order the respondents. The paper also suggests an algorithm for the rank ordering of items along the attitudes scale. Finally, the methods are evaluated using simulated data. This research was supported in part by an Educational Testing Service Gulliksen Fellowship, and by the National Science Foundation, Grant DMS-97.05032. The author would like to thank Brian Junker for his help and support on this paper and Paul Holland, Steve Fienberg, and Jay Kadane for their helpful comments.     

采用自由分类的方法探析汉族的人格结构

采用自由分类和多维标度的方法,分析了582名汉族被试的人格结构。结果表明,汉族拥有“积极/消极”、“性格/气质”和“外倾/内倾”这三种人格结构维度。     

Analysis of a two-alternative force-choice signal detection theory model

A basic problem in psychophysics is recovering the mean internal response and noise amplitude from sensory discrimination data. Since these components cannot be estimated independently, several indirect methods were suggested to resolve this issue. Here we analyze the two-alternative force-choice method (2AFC), using a signal detection theory approach, and show analytically that the 2AFC data are not always suitable for a reliable estimation of the mean internal responses and noise amplitudes. Specifically, we show that there is a subspace of internal parameters that are highly sensitive to sampling errors (singularities), which results in a large range of estimated parameters with a finite number of experimental trials. Four types of singular models were identified, including the models where the noise amplitude is independent of the stimulus intensity, a situation often encountered in visual contrast discrimination. Finally, we consider two ways to avoid singularities: (1) inserting external noise to the stimuli, and (2) using one-interval forced-choice scaling methods (such as the Thurstonian scaling method for successive intervals).     

Mapping the Structure of Human Values through Conceptual Representations

The present research provides the first direct examination of human values through concept categorization tasks that entail judging the meaning of values. Seven studies containing data from nine samples (N = 1086) in two countries (the UK and Brazil) asked participants to compare the meaning of different values found within influential quasi‐circumplex model of values. Different methods were used across experiments, including direct similarity judgment tasks, pile sorting, and spatial arrangement. The results of these diverse conceptual assessments corresponded to spatial configurations that are broadly convergent with Schwartz's model, both between and within participants. © 2018 The Authors. European Journal of Personality published by John Wiley & Sons Ltd on behalf of European Association of Personality Psychology     

计算机化多维测验中作答时间和作答精度数据的联合分析

随着心理与教育测量研究的发展和科技的进步,计算机化(大规模)测验逐渐受到人们的关注。为探究在计算机化多维测验中如何利用作答时间数据来辅助评估多维潜在能力,以及为我国义务教育阶段教育质量监测提供数据分析方法上的理论支持。本研究以2012年和2015年国际学生能力评估(PISA)计算机化数学测验数据为例,提出了一种可同时利用作答时间和作答精度数据的联合作答与时间的多维Rasch模型。根据新模型对PISA数据的分析结果,表明引入作答时间数据,不仅有助于提高模型参数的估计精度,还有助于数据分析者利用被试的作答时间信息来做进一步的决策和干预(e.g., 对异常作答行为或预备知识的诊断)。     

Middle Response Functioning in Likert-responses to Personality Items

Two studies examined whether the middle response option in graphic rating scales indicates a moderate standing on a trait/item, or rather a “dumping ground” for unsure or non-applicable (N/A) responses. Study One identified middle response-option dysfunction. Study Two indicated that respondents use the middle response option as an N/A proxy, even under implicit ‘skip if you do not know’ instructional sets. Although middle response category ‘misuse’ did not adversely affect reliability and validity in these studies, it is recommended that assessment developers (especially in on-line administration contexts) regularly include an N/A response option when administering graphic rating scales.

An Efficient MCMC Algorithm to Sample Binary Matrices with Fixed Marginals

Uniform sampling of binary matrices with fixed margins is known as a difficult problem. Two classes of algorithms to sample from a distribution not too different from the uniform are studied in the literature: importance sampling and Markov chain Monte Carlo (MCMC). Existing MCMC algorithms converge slowly, require a long burn-in period and yield highly dependent samples. Chen et al. developed an importance sampling algorithm that is highly efficient for relatively small tables. For larger but still moderate sized tables (300×30) Chen et al.’s algorithm is less efficient. This article develops a new MCMC algorithm that converges much faster than the existing ones and that is more efficient than Chen’s algorithm for large problems. Its stationary distribution is uniform. The algorithm is extended to the case of square matrices with fixed diagonal for applications in social network theory. I am indebted to my colleague Gunter Maris for his suggestion to add a Metropolis–Hastings step as the finishing touch of the algorithm.