Testing measurement invariance using multigroup CFA: differences between educational groups in human values measurement

This article applies the testing procedures for measurement invariance using multigroup confirmatory factor analysis (MGCFA). It illustrates these procedures by investigating the factorial structure and invariance of the Portraits Value Questionnaire (PVQ, Schwartz et al.: J. Cross Cult. Psychol. 32(5), 519–542 (2001)) across three education groups in a population sample (N  =  1,677). The PVQ measures 10 basic values that Schwartz postulates to comprehensively describe the human values recognized in all societies (achievement, hedonism, self-direction, benevolence, conformity, security, stimulation, power, tradition and universalism). We also estimate and compare the latent means of the three education groups. The analyses show partial invariance for most of the 10 values and parameters. As expected, the latent means show that less educated respondents attribute more importance to security, tradition, and conformity values.     

Rater source factors represent important subcomponents of the criterion construct space, not rater bias

We contrast normative accuracy and ecological perspectives on applications of the multitrait–multimethod methodology to multisource performance ratings and review existing research that provides critical tests of these perspectives. Existing research supports the ecological perspective which proposes that the rater source effects that are typically found in analysis of multisource performance ratings do not represent mere halo biases but alternative, perhaps equally valid perspectives on ratee performance. We suggest that future research view multifaceted research designs in the broader context of a prototype multidimensional data relational system such as that proposed by Lance, Baranik, Lau, and Scharlau (Lance, C. E., Baranik, L. E., Lau, A. R., & Scharlau, E. A. (in press). If it's not trait it must be method: (Mis)application of the multitrait–multimethod design in organizational research. In C. E. Lance & R. J. Vandenberg (Eds.), Statistical and methodological myths and urban legends: Received doctrine, verity, and fable in the organizational and social sciences. Mahwah, NJ: Erlbaum).     

Testing for measurement equivalence of human values across online and paper-and-pencil surveys

The following study investigates the measurement equivalence of an online and paper-and-pencil (PAP) survey of human values. For this purpose, a total of 250 respondents completed the 21-item version of the Portrait Value Questionnaire either online (n = 125) or by PAP (n = 125). This questionnaire was developed by Schwartz (Advances in experimental social psychology, Academic Press, New York, 1992) and has been included in the European Social Survey since 2002 to test his theory of basic human values. Measurement invariance was tested via a multiple group confirmatory factor analysis. The assessment of invariance included the three levels of configural, metric, and scalar invariance, and the latent means of the values between both samples were compared. Results of this study show that the measurements are invariant at the three levels (configural, metric, and scalar), but there are latent mean differences between the values across the surveys. These differences may be partly explained by age and level of education differences between the two samples. Based on these findings we conclude that the methods of measurement are essentially invariant.     

An assessment of equivalence between Internet and paper-based surveys: evidence from collectivistic cultures

Little research exists that addresses the equivalence in collectivistic cultures of paper- versus Internet-based surveys. This study addressed this gap and examined the measurement equivalence of individual innovativeness scales between Internet surveys and paper-based surveys within a collectivistic culture (with China serving as our example). The study analyzed and compared survey data from both paper and web-based surveys using confirmatory factor analysis. The assessment of invariance included the levels of configural, metric, scalar, and covariance invariance. The means and variance of latent variables were also compared. The results show that measurements are invariant at the two levels (configural and metric), and the covariances between latent variables are also equivalent, but the mean and variance differences of latent variables are apparent. The results indicate that when conducting research in collectivistic cultures and collecting data from distinct survey modes, researchers should concern themselves with the potential of extreme response patterns and the inclination of social desirability responding, as well as considering the measurement invariance across survey modes.     

Validation and measurement invariance of the multidimensional qualitative job insecurity scale

This contribution introduces the Multidimensional Qualitative Job Insecurity Scale (MQJIS). Drawing from the qualitative job insecurity literature and addressing some of other scales’ limitations, a multidimensional model is proposed and investigated by means of confirmatory factor analysis and multilevel confirmatory factor analysis. Study 1 aims to explore the psychometric properties and factorial structure of MQJIS in an Italian sample of blue-collar workers (N?=?583), showing that a model with one higher-order factor (i.e., qualitative job insecurity) and four dimensions (i.e., social relationships, employment conditions, working conditions, and work content) shows a good fit to the data and good reliability indices. Study 2 aims to investigate MQJIS measurement invariance across several groups, based on country of origin, age, and gender. Results on a sample of Belgian and Italian workers (N?=?710) show that MQJIS met the criteria for uniqueness invariance across genders and scalar invariance across countries and age groups. Significance, implications, and future directions stemming from the initial validation and the confirmed measurement invariance of this scale are discussed.

     

Confirmatory factor analyses of the children’s attitudes and behaviors towards animals scale in two eastern cultural contexts

The purpose of the present study was to investigate the factor stability of the Children’s Attitudes and Behaviors Towards Animals Scale (CABTA) across two eastern contexts: China and Malaysia. Confirmatory factor analyses (CFAs) were conducted on data collected from 700 Chinese mothers and 700 Chinese fathers, and 385 Malay parents of children aged 6–12 years. The CFA results suggested that the model was a good fit for both the Chinese and Malaysian responses. Furthermore, invariance testing showed that although there was factor loading invariance across cultures, the assumption of intercept invariance did not hold. This suggests that although the scale is suitable for use in both cultures, mean comparisons across contexts could be problematic. The internal reliabilities for the Typical and Malicious cruelty subscales were adequate across the samples. All findings show that the CABTA is a promising scale for international use.     

Instrumental values

This paper studies identification of partial differences of nonseparable structural functions. A model is defined which admits structural functions exhibiting a degree of monotonicity with respect to a latent variate. The model identifies partial differences when there are instrumental values of covariates over which the latent variate exhibits a local quantile invariance, and a local order condition holds. The result is useful when covariates exhibit discrete variation, as arises often in practice, and when restricting latent variates and covariates to be statistically independent is unpalatable. The results are illustrated with data from the returns-to-schooling study of Angrist and Krueger [1991. Does compulsory schooling attendance affect schooling and earnings? Quarterly Journal of Economics 106, 979–1014].     

Alternative measures of political efficacy: models and means

In the 1982 and 1984 American national election surveys the CPS deleted a subset of its longstanding measures of political efficacy. This paper employs covariance structure analysis to test a two-factor measurement model based on alternative indicators. The model has an excellent fit for data from the 1964–84 national surveys as well as for different educational, gender, and racial groups in the 1984 study. Consistent with previous theorizing, the internal efficacy factor is significantly more stable than the external factor. Also, the two efficacy factors correlate as expected with measures of general personal competence and political trust. The paper is the first to use LISREL structured means tests to test group differences in levels of efficacy or other important political attitudes. Since this technique is not well-known, a tutorial appendix describing its implementation in LISREL VI is included to facilitate future research.     

Interpreting the results of studies using latent variable models to assess data quality: an empirical example using confirmatory factor analysis

Latent variable models (LVMs) offer one route to examine the quality of data collected in surveys. The possibility exists that individuals equivalent in their true level of a construct or variable being measured are unlikely to have equivalent observed responses as a function of an extraneous variable, e.g., group membership. This potential is labeled here as differential item functioning (DIF). Survey methods generally considers measurement bias to be estimators that do no not accurately reflect true values. DIF may be thought of as differential measurement bias, i.e., measurement bias conditional on group membership. As a function of group membership, the degree, amount, or type of measurement bias changes. DIF has the potential to negatively affect the quality of data. LVMs, e.g., confirmatory factor analysis (CFA), offer one tool to assess DIF. However, few published examples exist in the survey research field and training in the interpretation of these models is lacking. The purpose of the current paper is to describe CFA sufficiently for interpretive purposes and demonstrate an empirical application of CFA to assess survey data quality to provide further interpretive guidance. References are provided for analysts wishing to conduct analyses of this type.     

Investigating the Portrait Values Questionnaire at two transitions in adulthood: Retirement and University

In this study, two issues regarding the use of the Portrait Value Questionnaire (PVQ; Schwartz et al.: J. Cross. Cult. Psychol. 32:519–542, 2001) with older adults were addressed: (1) whether the quasi-circumplex structure of values, developed on younger adult samples, also emerges among older adults; and (2) whether the PVQ demonstrates measurement invariance across age groups and equivalence over time. The 40-item version of the PVQ was completed by 433 retired adults and 173 university students in Montreal, Canada. In both retired and student samples, the quasi-circumplex structure of values emerged using exploratory techniques, but was not supported by constrained confirmatory factor analyses. A modified 26-item model was necessary to achieve adequate fit indices in both samples, suggesting problems of multi-collinearity and internal discriminant validity. However, using the revised 26-item model of the values, support was found for most types of measurement invariance across age groups and time. The PVQ should, therefore, be considered appropriate for use across the adult lifespan and in longitudinal research. Results are discussed with regard to the Schwartz (In: Zanna M (ed) Advances in experimental social psychology. New York: Academic Press, pp. 1–65, 1992) value theory and the conceptualizing of values across the lifespan.     

Does measurement invariance hold for the official Mexican multidimensional poverty measure? A state-level analysis 2012

One of the main goals in poverty measurement is making comparisons of prevalence and severity across geographical units. This is attained by merely disaggregating the index in question. The underlying assumption is that comparisons across units are tenable, inasmuch as the same indicators are utilised for constructing the index. Nonetheless, in practice, this assumption is very rarely tested. From the statistical perspective, measurement invariance (MI) must hold for comparisons to be valid, and violations thereof indicate that a given poverty index measures different things across different countries, states, counties, etc. Consequently, differentials in severity and prevalence cannot be attributed exclusively to the underlying construct (i.e. poverty) but to factors not considered in the measure. This article tests whether MI holds for two indexes: the Mexican official multidimensional measure (MPM) and an adjusted multidimensional measure (MPM-A) that uses less severe thresholds. The analysis is conducted using a novel method called the ‘alignment method’. It uses these two measures and the method as an illustration of why it is vital to introduce MI tests into poverty measurement. The results suggest that partial strong MI holds for the official measure and MI is violated when the thresholds are adjusted. Partial strong MI guarantees making valid comparisons across the 32 states. Should the official measure requires to be updated with other thresholds, it would be necessary to adjust the threshold or drop the indicator for water deprivation.     

How do high-involvement work processes influence employee outcomes? An examination of the mediating roles of skill utilisation and intrinsic motivation

How do high-involvement work processes influence employee outcomes? In this paper, we use Vandenberg et al.'s [Vandenberg, R. J., Richardson, H. A., & Eastman, L. J. (1999). The impact of high involvement work processes on organisational effectiveness: A second order latent variable approach. Group and Organisational Management, 24, 300–339] elaboration of Lawler's [Lawler, E. E. (1986). High-involvement management: Participative strategies for improving organizational performance. San Francisco, CA: Jossey-Bass] model to test direct links and to explore skill utilisation and intrinsic motivation as mediators. Survey data were collected from a large New Zealand organisation providing distribution services and results were analysed through structural equation modelling. While there are important direct effects, the results demonstrate that both skill utilisation and intrinsic motivation transmit high-involvement processes into valuable outcomes for employees. The benefits to employees of such processes are direct but also lie in the way they make the work itself more motivating and enable them to deploy and grow their skills. This implies that forms of work organisation and supervision that offer workers greater opportunity for discretion, and involvement in the decisions that concern them, create the conditions for greater learning and, in turn, contribute to their well-being.     

Confirmatory and Structural Categorical Latent Variables Models

The paper proposes a general framework for modeling multiple categorical latent variables (MCLV). The MCLV models extend latent class analysis or latent transition analysis to allow flexible measurement and structural components between endogenous categorical latent variables and exogenous covariates. Therefore, modeling frameworks in conventional structural equation models, for example, CFA and MIMIC models are feasible in the MCLV circumstances. Parameter estimations for the MCLV models are performed by using generalized expectation–maximization (E–M) algorithm. In addition, the adjusted Bayesian information criterion provides help for model selections. A substantive study of reading development is analyzed to illustrate the feasibility of MCLV models.     

Maternal full-time employment and overweight children: Parametric, semi-parametric, and non-parametric assessment

We use the matched mother-child data from the 2000 wave of the US National Longitudinal Survey of Youth 79 (NLSY79) to assess the impact of full-time working mothers on children’s body mass index (BMI) and the likelihood of becoming overweight. Parametric, semi-parametric and non-parametric methods are employed to correct the bias of selection on observables and unobservables. Pros and cons of various methods are discussed and specification tests are conducted. In general, we find that a mother’s full-time employment does have some impact on her children’s BMI and likelihood of becoming overweight across models and inference procedures.     

Comparing levels of anti-Muslim attitudes across Western countries

In this study we describe differences of the ethnic majorities?? attitudes toward Muslims across Western countries. Using data from the Pew Global Attitudes Survey (2005), we were able to increase and test cross-cultural comparability of anti-Muslim attitudes. We constructed a single factor CFA model with three indicators, which was tested for scalar equivalence. Our results indicate that anti-Muslim attitudes differ significantly across the countries in our analysis. Germany and the Netherlands turned out to display relatively high levels of anti-Muslim attitudes, whereas these levels were relatively low in Great-Britain and the USA. We conclude our study with giving some post-hoc explanations for the differences found across countries.     

Leadership competency inventory: A systematic process of developing and validating a leadership competency scale

This article reports psychometric evaluation of the Penn State Leadership Competency Inventory (LCI). The 32‐item LCI was validated on a sample of 323 managers in the health care industry. Preliminary validity and reliability evidence of the LCI was established through exploratory factor analysis (EFA), item‐total correlations, Cronbach's alpha coefficients, and confirmatory factor analysis (CFA). The four‐factor leadership competency scale, comprising supervisory and managerial competencies, organizational leadership, personal mastery, and resource leadership, accounted for 58% of variance. According to CFA results, the model fit of the four latent factors of the LCI was confirmed to be appropriate. Cross‐validation with other populations is needed to confirm the factor structure. Limitations and further research recommendations are discussed.     

Semiparametric tests of conditional moment restrictions under weak or partial identification

We propose two new semiparametric specification tests which test whether a vector of conditional moment conditions is satisfied for any vector of parameter values θ₀. Unlike most existing tests, our tests are asymptotically valid under weak and/or partial identification and can accommodate discontinuities in the conditional moment functions. Our tests are moreover consistent provided that identification is not too weak. We do not require the availability of a consistent first step estimator. Like Robinson [Robinson, Peter M., 1987. Asymptotically efficient estimation in the presence of heteroskedasticity of unknown form. Econometrica 55, 875–891] and many others in similar problems subsequently, we use k-nearest neighbor (knn) weights instead of kernel weights. The advantage of using knn weights is that local power is invariant to transformations of the instruments and that under strong point identification computation of the test statistic yields an efficient estimator of θ₀ as a byproduct.     

Risk-based classification of supplier relationships

The management and exploitation of external resources has increased and become a new source of business success. This has given rise to various new risks and therefore increased the need for collaborative risk management and learning. The objective of this paper is to explore the differences in risk management and learning across supplier relationships classified through network risks. The paper provides a theoretical review of supplier relationships and risk management, as well as a survey-based empirical study conducted in one case company's supply network. A set of network risks and risk-management measures is developed by means of factor analysis, and a supplier classification by means of cluster analysis. According to the results, the exploitation of collaborative risk management and learning is highest among the most strategic supplier relationships. A classification typology based on suppliers’ opinions helps both parties in the relationship and enables the efficient exploitation of mutual risk management with collaborative learning as one of its keystones.     

Income convergence within the MENA countries: A panel unit root approach

This article aims at testing the convergence hypothesis in MENA region using new tests of a unit root in panel data. Evans and Karras [Evans P., & Karras G. (1996). Convergence revisited. Journal of Monetary Economics, 37, 249–265] and Bernard and Jones [Bernard A., & Jones C. I. (1996). Productivity across industries and countries: Time series theory and evidence. The Review of Economics and Statistics, 135–146] recommend this technique to evaluate the income convergence hypothesis. According to them it avoids econometric problems of the cross-countries growth regressions testing convergence and sample bias of the multivariate cointegration techniques. We test for both absolute and the conditional convergence with panel unit root tests using the Summers and Heston's data 5.6 and 6.1 on the periods of 1960 to 1990 and from 1960 to 2000. The absolute convergence hypothesis use panel unit roots test with no fixed individual effects. The catching-up hypothesis is not rejected for most groups of countries of the region during both periods. If we allow a break in the unit root tests, the hypothesis is not rejected for more groups. The conditional convergence requires panel unit root tests with fixed individual effects. Again, during the whole periods, the conditional convergence is not rejected for the major part of the remaining groups of MENA countries.     

Multivariate residual‐based finite‐sample tests for serial dependence and ARCH effects with applications to asset pricing models

In this paper, we propose several finite‐sample specification tests for multivariate linear regressions (MLR). We focus on tests for serial dependence and ARCH effects with possibly non‐Gaussian errors. The tests are based on properly standardized multivariate residuals to ensure invariance to error covariances. The procedures proposed provide: (i) exact variants of standard multivariate portmanteau tests for serial correlation as well as ARCH effects, and (ii) exact versions of the diagnostics presented by Shanken ( 1990 ) which are based on combining univariate specification tests. Specifically, we combine tests across equations using a Monte Carlo (MC) test method so that Bonferroni‐type bounds can be avoided. The procedures considered are evaluated in a simulation experiment: the latter shows that standard asymptotic procedures suffer from serious size problems, while the MC tests suggested display excellent size and power properties, even when the sample size is small relative to the number of equations, with normal or Student‐t errors. The tests proposed are applied to the Fama–French three‐factor model. Our findings suggest that the i.i.d. error assumption provides an acceptable working framework once we allow for non‐Gaussian errors within 5‐year sub‐periods, whereas temporal instabilities clearly plague the full‐sample dataset. Copyright © 2009 John Wiley & Sons, Ltd.