基于认知理论的测验编制技术—项目生成

传统测验重视统计技术,不重视测量结构的心理学意义,使得它的功能局限于筛选,不能提供更多的信息。由于认知心理学理论、心理计量学以及计算机技术的发展,使得基于认知理论指导下的项目生成技术逐渐成熟。该文从项目生成的概念、方法,及研究的意义和难点等方面对项目生成技术作了一个简要述评,以期推进我国认知心理学与心理测量学相结合的研究工作。     

Person-organization fit: Testing socialization and attraction-selection-attrition hypotheses

Using the socialization and attraction-selection-attrition (ASA) frameworks, this study examined the relation between employees’ work values and their organization’s values (person-organization fit). With a two year time interval, 140 respondents participated in a longitudinal study. After entry, socialization served to enhance homogeneity. The work values underwent small changes and the perception of fit with the organization grew. Despite high retention rates results confirmed the attrition effect. Results indicated that the lower the perceived match between own and organizational values at entry, the more likely it was that someone left the organization over time. We concluded that socialization as well as attrition mechanisms were present at the same time.     

Affective Synchrony in Dual‐ and Single‐Smoker Couples: Further Evidence of “Symptom‐System Fit”?

Couples in which one or both partners smoked despite one of them having a heart or lung problem discussed a health-related disagreement before and during a period of laboratory smoking. Immediately afterwards, the partners in these 25 couples used independent joysticks to recall their continuous emotional experience during the interaction while watching themselves on video. A couple-level index of affective synchrony, reflecting correlated moment-to-moment change in the two partners' joystick ratings, tended to increase from baseline to smoking for 9 dual-smoker couples but decrease for 16 single-smoker couples. Results suggest that coregulation of shared emotional experience could be a factor in smoking persistence, particularly when both partners in a couple smoke. Relationship-focused interventions addressing this fit between symptom and system may help smokers achieve stable cessation.     

Identifiability and Equivalence of GLLIRM Models

The generalized logit–linear item response model (GLLIRM) is a linearly constrained nominal categories model (NCM) that computes the scale and intercept parameters for categories as a weighted sum of basic parameters. This paper addresses the problems of the identifiability of the basic parameters and the equivalence between different GLLIRM models. It is shown that the identifiability of the basic parameters depends on the size and rank of the coefficient matrix of the linear functions. Moreover, two models are observationally equivalent if the product of the respective coefficient matrices has full column rank. Finally, the paper also explores the relations between the parameters of nested models. I would like to express my gratitude to the editor and three anonymous reviewers for their helpful suggestions on earlier versions of the paper. This work was supported by the Comunidad de Madrid (Spain) grant: CCG07-UAM/ESP-1615.     

Stressed out over possible failure: The role of regulatory fit on claimed self-handicapping

The self-protective mechanism of self-handicapping appears to be motivated by the need to protect ability attributions in the face of concern about possible failure. Indeed, the present research finds a correlation between trait self-handicapping and chronic prevention focus. Moreover, the present research examines the role of “regulatory fit” on the use of claimed self-handicapping by exposing high and low trait self-handicappers to performance situations framed in prevention or promotion terms. Consistent with our regulatory fit hypothesis, high self-handicappers (HSHs) handicapped significantly more (by reporting higher levels of stress) when the task was framed in prevention focus rather than promotion focus terms, and did so even when the viability of the handicap was dubious. Self-handicapping in the prevention focus condition was mediated by elevated feelings of evaluative concern. The findings suggest that conditions of regulatory fit (i.e., HSHs under prevention focus) can lead to increased use of self-handicapping.     

多级评分题计算机自适应测验选题策略比较

研究比较了多级评分题计算机化自适应测验五种选题策略的优劣。应用的IRT模型是Samejima的等级反应模型。参加比较的选题策略有难度均值与能力匹配法、难度中值与能力匹配法、信息量最大法和两种A分层法。比较指标采用了能力估计值返回真值偏差、能力估计标准差、人均用题数和试题调用次数标准差四个。研究采用蒙特卡罗模拟法,结果显示每种方法各有优劣,在分层得当情况下,A分层法(中)的综合效果最佳     

应用项目反应理论创建图形推理测验题库

自编235个图形推理测验题目。采用铆测验等值设计,以72个联合型瑞文测验题目为铆题,对初中到大学各能力层次的1733名男性进行了测验。使用BILOG MG3.0（边际极大似然估计）对实测数据进行了分析,采用Logsitic 3参数模型。剔除数据与模型拟合不好的题目以及信息函数最大值小于0.3的题目,最终建立一个包含181道题目的题库。该题库可以用于淘汰智力较低的应征青年     

计算机自动化项目生成概述

近年来,计算机自动化项目生成作为应对测验项目曝光问题的很具前景的一项技术,逐渐得到越来越多研究者的重视。这种技术是以认知心理学和心理测量学为基础,在测试过程中借助计算机自动生成目标难度水平的项目。该文首先简要介绍了这种技术的理论基础和两种具体方法,然后以项目设计系统法为例说明自动化项目生成研究的具体步骤与优点,最后评述了这种技术的局限性及未来的发展趋势     

项目组合在结构方程模型中的应用

项目组合（itemparceling）是对同一量表中的若干项目进行整合并形成新的观测指标的过程。虽然一直以来它都是一个有争议的议题,但随着其在结构方程模型中的应用日益广泛,它越来越受到研究者重视。文章从项目组合的基本逻辑、优缺点以及具体方法等方面对项目组合的研究进行了概括,并在此基础上提出了使用的具体建议：（1）根据研究的目的与具体情境选择是否需要组合;（2）组合之前必须首先确定概念的维度性;（3）项目组合最好建立在一定的理论基础上等等。未来的研究可以深入探讨各种组合方法对模型拟合以及参数估计的影响以及项目组合在一些SEM高阶应用中的效果,并进一步与项目反应理论等测量理论相结合     

A Unified Nonparametric IRT Model for <Emphasis Type="Italic">d</Emphasis>-Dimensional Psychological Test Data (<Emphasis Type="Italic">d</Emphasis>-ISOP)

The (univariate) isotonic psychometric (ISOP) model (Scheiblechner, 1995) is a nonparametric IRT model for dichotomous and polytomous (rating scale) psychological test data. A weak subject independence axiom W1 postulates that the subjects are ordered in the same way except for ties (i.e., similarly or isotonically) by all items of a psychological test. A weak item independence axiom W2 postulates that the order of the items is similar for all subjects. Local independence (LI or W3) is assumed in all models. With these axioms, sample-free unidimensional ordinal measurements of items and subjects become feasible. A cancellation axiom (Co) gives, as a result, the additive isotonic psychometric (ADISOP) model and interval scales for subjects and items, and an independence axiom (W4) gives the completely additive isotonic psychometric (CADISOP) model with an interval scale for the response variable (Scheiblechner, 1999). The d-ISOP, d-ADISOP, and d-CADISOP models are generalizations to d-dimensional dependent variables (e.g., speed and accuracy of response). The author would like to thank an Associate Editor and two anonymous referees and also Professor H.H. Schulze for their very valuable suggestions and corrections.