Measurement invariance,factor analysis and factorial invariance

Several concepts are introduced and defined: measurement invariance, structural bias, weak measurement invariance, strong factorial invariance, and strict factorial invariance. It is shown that factorial invariance has implications for (weak) measurement invariance. Definitions of fairness in employment/admissions testing and salary equity are provided and it is argued that strict factorial invariance is required for fairness/equity to exist. Implications for item and test bias are developed and it is argued that item or test bias probably depends on the existence of latent variables that are irrelevant to the primary goal of test constructers.Presidential address delivered at the Annual Meeting of the Psychometric Society in Berkeley, California, June 18–20, 1993.     

A validational note on the revised Socialization Scale of the California Psychological Inventory

The revised (1987) Socialization Scale of the California Psychological Inventory was studied in a sample of 121 delinquent and nondelinquent males to determine if the revised scale, compared to the original version of the scale (1957), could differentiate among a group of nonoffenders and a comprehensive typology of offenders. Results of an ANOVA and product-moment correlations provided significant and considerable support for the discriminability of the revised So Scale. Obtained results indicate that the revised instrument can differentiate offenders from nonoffenders to about the same, or a greater, degree than the original scale.     

An extension of the rating scale model with an application to the measurement of change

The polytomous unidimensional Rasch model with equidistant scoring, also known as the rating scale model, is extended in such a way that the item parameters are linearly decomposed into certain basic parameters. The extended model is denoted as the linear rating scale model (LRSM). A conditional maximum likelihood estimation procedure and a likelihood-ratio test of hypotheses within the framework of the LRSM are presented. Since the LRSM is a generalization of both the dichotomous Rasch model and the rating scale model, the present algorithm is suited for conditional maximum likelihood estimation in these submodels as well. The practicality of the conditional method is demonstrated by means of a dichotomous Rasch example with 100 items, of a rating scale example with 30 items and 5 categories, and in the light of an empirical application to the measurement of treatment effects in a clinical study.Work supported in part by the Fonds zur Förderung der Wissenschaftlichen Forschung under Grant No. P6414.     

“Celebrities to lean on!” Perceived celebrity support across two cultures—Iran and the United States

Though many people loathe celebrities and their lifestyles, celebrities can also be perceived as being available to provide community support. The purpose of this study was to establish the psychometric properties of the Perceived Celebrity Support Inventory. The scale was developed to capture three aspects of perceived celebrity support to the community: instrumental (i.e., providing tangible and concrete assistance), informational (i.e., advice, explanation, and information), and emotional (i.e., empathizing, encouraging, listening, and comforting) support. Factor and item analyses have revealed robust psychometric properties in samples drawn from the United States (US) and Iran. The US sample perceived higher community support from celebrities than the Iranian sample. In the Iranian sample, perceived celebrity support for the community was associated with more positive attitudes and less hatred of celebrity culture. In the US, perceived celebrity support was associated with negative attitudes and elevated hatred towards celebrities. The findings advance the literature by providing an instrument with which to measure positive aspects of celebrity influence across cultures.     

A hot new way to measure aggression: Hot sauce allocation

Laboratory experiments investigating aggressive behavior have operationalized and assessed aggression in a variety of ways; however, these measures are often problematic because they do not create a situation in which participants perceive potential for real harm to come to the target, there is a risk of actual harm to the target, or they are too familiar to participants. To overcome these limitations, we developed a new method for measuring aggression, specifically, the amount of hot sauce administered to a target known to dislike spicy foods. We summarize a series of experiments assessing theory‐based hypotheses regarding aggression in which this measure is employed. We then briefly consider the strengths and limitations of this new measure. Aggr. Behav. 25:331–348, 1999. © 1999 Wiley‐Liss, Inc.     

A three-factor scale of attitudes toward guns

Guns figure prominently in American culture. There is, however, considerable variability in attitudes toward guns. The beliefs and feelings that comprise attitudes toward guns may well be important moderators of numerous social behaviors. For this reason, a three-dimensional attitudes scale was constructed. The first factor consists of an abstract set of beliefs concerning the right of the American public to own or not own guns. The second and third factors tap more concrete beliefs about the consequences of gun ownership. Specifically, the second factor assesses the degree to which people believe gun ownership affords protection against crime. The third factor measures the degree to which people believe guns stimulate crime. Discriminant validity of the subscales is shown by the pattern of correlations obtained with the three dimensions and various other social and personality measures. Regression analyses indicate that the subscales are uniquely predicted by different social variables. Discussion centers on the usefulness of such a measure for future research on aggression, social policy, and attitude change.     

Legitimized aggression and sexual coercion: Individual differences in cultural spillover

Correlations between aggressive attitudes and sexual coercion have consistently been found in numerous empirical studies across social science disciplines. Extrapolating from sociological research linking indices of legitimate aggression by state to statewide frequencies of rape, we have extended the investigation of legitimate aggression and coercive sexuality to the individual level. The specific purposes of this research were to index the breadth and level of endorsement of legitimized aggression at the individual level and to measure the association between such an index and coercive sexual behavior. These purposes were achieved across the course of three studies in which we created a new dispositional measure, the Proclivity for Legitimized Aggression Questionnaire (PLAQ), and replicated a sociocultural level correlation with coercive sexual behavior (Study 1); assessed the individual differences level construct validity of the PLAQ (Study 2); and tested whether endorsement of items on the PLAQ were related to content‐relevant behaviors (Study 3). The PLAQ was internally consistent, modestly but significantly correlated with a measure of self‐reported coercive sexual behavior, and characterized by promising construct validity. Aggr. Behav. 27:26–43, 2001. © 2001 Wiley‐Liss, Inc.     

等效性检验——结构方程模型评价和测量不变性分析的新视角

常用的结构方程模型拟合指数存在一定局限, 如χ ²以传统零假设为目标假设, 无法验证模型, 而RMSEA和CFI等描述性的拟合指数不具备推断统计性质, 等效性检验有效弥补了这些问题。首先说明等效性检验如何评价单个模型的拟合, 并解释其与零假设检验的不同, 然后介绍等效性检验如何分析测量不变性, 接着用实证数据展示了等效性检验在单个模型评价和测量不变性检验中的效果, 并与传统模型评价方法比较。