Two-level Confirmatory Factor Analysis Research Articles

Applying the Robust Chi-Square Goodness-of-Fit Test to Multilevel Multitrait-Multimethod Models: A Monte Carlo Simulation Study on Statistical Performance

As the robust maximum likelihood χ2 goodness-of-fit test had been found to yield inflated type-I error rates for certain two-level confirmatory factor analysis (CFA) models, a new correction for the test was implemented in Mplus version 8.7. In this simulation study, we inspected whether the corrected test statistics follow the expected χ2 distributions when applying more complex two-level models for multitrait-multimethod data with varying sample sizes and correlations within trait factors. Investigating rejection rates and probability-probability plots, we found that the new correction markedly and sufficiently reduced previously inflated rejection rates in conditions with within-trait correlations equal to 1, 100 between-level units, and 10 or 20 within-level units. In other conditions, rejection rates were hardly affected or not sufficiently reduced by the new correction. While in most conditions, 2 within-level units did not suffice, 5 within-level units and 250 between-level units were enough to yield correct rejection rates given within-trait correlations did not exceed 0.80. Correlations above 0.80 required larger sample sizes. In planning studies with multilevel CFA models, researchers should be aware that sample size requirements for likelihood-based model fit evaluations can depend on several different factors and might consider conducting Monte Carlo simulations tailored to their specific modeling conditions.

OKULDA KENDİNİ TOPARLAMA GÜCÜ: BİR ÖLÇEK UYARLAMA ÇALIŞMASI

This study aimed to adapt the School Resilience Scale (Caleon& King, 2020) to Turkish culture. The participants of this study were high school students. The study consisted of three stages. The first stage regarded the scale translation for linguistic equivalence. In the second stage, the structural validity and reliability of the scale were tested. In the third stage, a test-retest reliability of the scale was performed. Fifty-three students (52.8% girls) participated in the first group, 615 students (52.7% girls) in the second group, and 88 students (67% girls) in the third group. The confirmatory factor analysis (CFA) findings to test construct validity showed that the four-factor structure of the scale had an excellent fit index. Next, a two-level confirmatory factor analysis was conducted, which confirmed the two-level model. The Cronbach's Alpha reliability coefficient for the total scale was .89, while it was between .78 and .84 for the sub-dimensions of the scale. In sum, the findings evidenced that the school resilience scale was a valid and reliable measurement tool that can be used to assess Turkish high school students' school resilience.

Mengurai Faktor Terpenting Pada Kualitas Pelayanan Pembiayaan Murabahah Pada Bank Perkreditan Rakyat Syariah

The quality of service perceived by users has been generally reviewed and researched previously. This study aims to describe the important factors that reflect the quality of service in murabahah financing of a Islamic Rural Bank (BPRS) AM which is operating in Bandung Regency. A total of 140 respondents who are customers of BPRS AM in Bandung Regency are the samples in this study. Two-level confirmatory factor analysis (second order factor analysis) is used to answer the research question posed in this study, namely what are the most important aspects that can be reflected in murabahah services for customers. Factor analysis is used to test the factor weight values of each indicator of the variables observed in the study. The results of data analysis reveal that the quality of murabahah financing services can be well reflected in the aspects of guarantees of trust, responsiveness, price, physical evidence, and reliability provided to murabahah financing customers. In addition, the murabahah service model has a level of validity that can be well confirmed. Findings also reveals that respondents perceive the quality of murabahah services as being in the moderate category. Among five dimensions of murabahah service quality, the aspect of physical evidence was found to be the most important factor in reflecting the quality of murabahah services of BPRS AM operating in Bandung Regency. On the other hand, the respondent's assessment also shows that the aspect of guarantee or trust is the aspect that is rated the highest in murabahah services. In addition to the generally well-confirmed model, the findings of this study are useful as a basis for further exploration of the concept of service quality in Islamic Rural banks.

Corporate Website Job Recruitment Platform: A Moderated Mediation Model of Organisational Image

Corporate companies’ websites recruitment in advanced countries is well established. However, lesser is known about the Malaysian students and employed jobseekers’ intention to apply for jobs on the corporate companies’ websites. Therefore, this study aims to assess the subjective perceptions of these two groups’ intentions to apply for jobs on the websites. This study also aims to analyse the mediating effect of organisational image on perceived person-job fit (P-J fit) and perceived website usability (PWSU) which may be moderated by these two groups of jobseekers’ intentions. The conceptual model was first analysed using two-level confirmatory factor analysis. Adopting Hayes’ Model 7 template, AMOS software was used to differentiate the two groups’ perceptions. The results indicate that the perceived P-J fit and PWSU are the main factors influencing jobseekers’ intentions to apply for jobs on corporate websites. The moderated mediation technique has identified an indirect effect between perceived P-J fit and the intention to apply for jobs indicating the importance of established organisational image that it arouses applicants’ pursuit intention. Thus, the corporate human resource managers need to ensure the sites are designed for easy navigation and job advertisements that contain updated job information and realistic job preview. Using this platform, companies can reduce costs and time for the recruitment of talents. This study provides the subjective perceptions of jobseekers’ intention study and it contributes to the first component of Planning Theory of Intention. Lastly, this research also contributed to an alternate research method of analysing the moderated mediation using AMOS.

JOB SEEKERS’ ACCEPTANCE TOWARDS PRODUCING VIDEO RESUMES FOR COMPANY’S RECRUITMENT

The purpose of this study is to explore the job seekers’ acceptance towards producing video resumes for the company’s video-based recruitment. The study aims to analyse the moderating effect of self-video resumes by experienced and non-experienced applicants’ in terms of their perceived stress and coping strategies in creating the videos when applying for jobs. The quota sampling method was used to collect data from four faculties in a private university. The conceptual model was developed with constructs consisting of perceived stress, problem-focused coping, self-efficacy, self-presentation and acceptance of video resumes. The model was validated by two-level confirmatory factor analysis (CFA). Subsequently, the moderation effects between the two groups of applicants were computed to uncover different group behaviours. The perceived stress and problem-focused coping responses were found to be significantly related to self-presentations for video resumes. A direct positive relationship was also found between the applicants’ self-efficacy and their acceptance of video resumes. The results of the moderation analysis indicate a significant difference in perceived stress between groups but not in the problem-focused coping strategies.

Multilevel Structural Equation Modeling for Cross-National Comparative Research

This contribution focuses on a model that is gaining currency in cross-national research, namely multilevel structural equation modelling (MSEM). Similarly to standard multilevel modelling (MLM), this model distinguishes between various levels of analysis (e. g., individuals nested within countries) and, in doing so, takes the hierarchical structure of cross-national data into account. However, MSEM incorporates a latent-variable approach into the multilevel framework, making it possible to assess the measurement quality of latent constructs. As such, MSEM is a synthesis of structural equation modeling (SEM) and MLM that combines the best of both worlds. The MSEM approach makes it possible to model multilevel mediations and group-level outcomes, and therefore provides a more complete representation of Coleman’s bathtub model. This contribution presents the statistical and conceptual background of MSEM in a formal but accessible manner. The paper discusses applications of MSEM that are particularly useful for cross-national comparative research (CNCR), namely two-level confirmatory factor analysis (CFA), multilevel mediation models, and models for group-level outcomes. A practical step-by-step strategy on how MSEM can be used for applied research is provided and illustrated by means of a didactical example.

Attitudes to School Teacher and Peers – Mixed Ethnicity Groups Compared to Majority and Minority Student Networks

Aims: The general aim of this study was to investigate students’ attitudes to school, teacher and peers in voluntarily-chosen networks with respect to ethnicity and gender.
 Study Design: Self-reports on attitudes were collected from 12-year old students in 77 grade 6 classes in the city of Göteborg, Sweden. Social networks were made up of students who voluntarily and reciprocally chose each other to be with during breaks.
 Place and Duration of Study: Sampling of all data autumn 2003. Work up of sociometric data 2016.
 Methodology: Based on a previous two-level confirmatory factor analysis, three school factors and four relational factors were identified. Social network analysis of the choices was carried out using a Matlab program identifying reciprocal (bilateral) choices. Three categories of network were identified using the names of their members. Mixed ethnicity networks were compared to majority (Scandinavian), and minority networks (Non-Scandinavian). Members outside networks were labelled “Outsiders”.
 Results: One out of five students was classified non-Scandinavian. Non-Scandinavian networks had high scoring for the categories “Interest in School” and “Working Atmosphere”. In Scandinavian networks, girls were more anxious at school and scored relationships to their classmates and the view of their peers significantly lower than the male students. Girls in this category also felt that disruption in the classroom was more common while the boys were, apparently, more tolerant of it. Within Mixed networks, gender differences were exclusively dependent on Non-Scandinavian students’ attitudes. Non-Scandinavian girls in the Mixed networks were surprisingly similar to their female Scandinavian peers. Individuals outside networks were overrepresented among Non-Scandinavians. They showed weaker relationships with classmates, had more problems with peers and were more anxious at school. Furthermore, they considered disruption during lessons to be common.
 Conclusion: Promotion of a mixture of ethnicities appeared best for improving attitudes to school. Immigrant children were overrepresented in the Outsider group, and at risk of rejection.

Significant predictors of mathematical literacy for top-tiered countries/economies, Canada, and the United States on PISA 2012: Case for the sparse regression model.

National ranking from the triennial Programme of International Student Assessment (PISA) often serves as a barometer of national performance and human capital. Though excessive student- and school-level covariates (n>700) may prove intractable for traditional least-squares estimate procedures, shrinkage methods may be more suitable for subset selection. With a focus on the United States, this paper proposes sparse regression for PISA 2012 to discover salient student- and school-level predictor variables for mathematical literacy achievement. The sparse regression analysis was conducted on 10 top-tiered OECD countries/economies, Canada, and the United States in mathematical literacy on the 2012 PISA. Two- and three-level hierarchical regression analyses were performed on Canadian and US students (N=26,522) along with five of the ten top-tiered countries/economies (N=58,385). Using the 'least absolute shrinkage and selection operator' (LASSO) technique, the study (1) identified salient predictor variables of mathematical literacy performance for the top-tiered countries/economies, Canada, and the United States and (2) used these salient variables to perform two- and three-level hierarchical regression on data from Canada and the United States along with five top-tiered countries/economies. Weights and replicates were used to account for complex sample design. A weighted, two-level confirmatory factor analysis was performed to identify latent constructs. Missing data were handled through multiple imputation. Separate two-level hierarchical models accounted for 32-35% student-level and 58-70% school-level variance in Canada and the United States, respectively; three-level models accounted for 33% of level-one variance, 62-65% level-two variance, and 13-44% of level-three variance for the US/Canada and US/Canada/top-tiered students, respectively. Following top-tiered countries/economies, Canadian students had high levels of self-efficacy, were more likely to encounter advanced concepts in class, were less activity/small group-centred, and were more likely to consider truancy a learning hindrance. Factor analyses revealed a positive relation with rigour and class organization (teacher-centred) for top-tiered countries and Canada, though not for the United States. For all countries, there was a strong relation between rigour and self-beliefs. Compared to top performers, a less rigorous curriculum, coupled with class and school factors, may explain lag in US performance.

Societal development, social stratification and power- and achievement-values Inglehart’s scarcity hypothesis and the theory of the social production functions in the comparative study of values

ABSTRACTPower- and achievement-values are seen to be important for productivity and societal stability. But what exact factors drive the forces to adhere to power- and achievement-values? In this study, the focus is on societal development and social stratification. Research on values in general often is not very careful in the distinction between different levels of analysis although the necessity has been outlined in several contributions to theoretical thinking in sociology. That is why this article comes from the position of methodological individualism and highlights explicitly the different levels of analysis and their assumed interrelations. It is shown that Inglehart’s scarcity hypothesis as an explaining mechanism has some weaknesses. As an alternative, the theory of social production functions is suggested that is capable to provide a framework of compatible societal- and individual-level mechanisms. The empirical analysis is based on data of the fourth round of the European Social Survey containing about 50,000 respondents in 31 countries. The comparability of power- and achievement-values across countries and across the societal and the individual level is tested by using two-level confirmatory factor analysis. The results show that comparability across countries and across both levels can be confirmed. Furthermore, opposing relations for societal development and social stratification are found: The higher the societal development, the lower the priority of power- and achievement-values; the higher the social stratification position, the more important the power- and achievement-values. Therefore, the scarcity hypothesis is rejected and the need for an alternative explanation is reinforced.

Collinear Latent Variables in Multilevel Confirmatory Factor Analysis: A Comparison of Maximum Likelihood and Bayesian Estimations.

Because variables may be correlated in the social and behavioral sciences, multicollinearity might be problematic. This study investigates the effect of collinearity manipulated in within and between levels of a two-level confirmatory factor analysis by Monte Carlo simulation. Furthermore, the influence of the size of the intraclass correlation coefficient (ICC) and estimation method; maximum likelihood estimation with robust chi-squares and standard errors and Bayesian estimation, on the convergence rate are investigated. The other variables of interest were rate of inadmissible solutions and the relative parameter and standard error bias on the between level. The results showed that inadmissible solutions were obtained when there was between level collinearity and the estimation method was maximum likelihood. In the within level multicollinearity condition, all of the solutions were admissible but the bias values were higher compared with the between level collinearity condition. Bayesian estimation appeared to be robust in obtaining admissible parameters but the relative bias was higher than for maximum likelihood estimation. Finally, as expected, high ICC produced less biased results compared to medium ICC conditions.

Comparison of Popular and Lonely Students’ Attitude, and Their Relationship to the Classroom Situation as Reported by Teachers

Aims: To investigate the attitudinal profiles of Popular and Lonely students in school classes and their similarities and differences, relationships to gender and cultural backgrounds, and the classroom characteristics reported by teachers. Study Design: Self-reports on attitudes were collected from 1531 students representing 77 grade 6 classes from comprehensive schools in the city of Goteborg, Sweden. The students were asked to answer 40 questions, each with five possible responses. Each student was asked to rank the three peers they would most prefer to work with in the classroom and to play with during breaks. In addition, data was gathered from the teacher of each class. Place and Duration of Study: Sampling of all data autumn 2003. Work up of sociometric data 2013. Methodology: Based on a previous two-level confirmatory factor analysis, three school factors and four relational factors were identified. Weighted adjacency matrices were used to quantify social relationships. The eigenvector of the largest eigenvalue of each adjacency matrix yielded individual indices drawn from the eigenvector components and a group index from the eigenvalue. Symmetrized matrices were used to reveal the most popular individuals in each network. Lonely individuals were outliers who were not nominated by other students and were identified based on bilateral choices using semi-symmetrized matrices. Results: Popular (11.5%) and Lonely (8.5%) students have similar attitudes to school and teachers. Around 10% of both groups had very positive attitudes for all factors. However, lonely students were three times likelier than popular students to have negative attitudes and also expressed more anxiety, considered disruption to be more common, and sometimes reported problematic relationships with their classmates and peers. Half of the popular students had negative or neutral attitudes to school factors and 40% had negative or neutral scores for relational factors. Conclusion: The working conditions were most positive in classrooms with only one large peer group and a minimum of Lonely students.

Construct validity of student perceptions of instructional quality is high, but not perfect: Dimensionality and generalizability of domain-independent assessments

In educational research, student ratings are frequently used to assess aspects of instructional quality. The present study investigates two key aspects of construct validity in student ratings in domain independent assessments: (1) the dimensionality of ratings and (2) their generalizability across classes and two subjects (English and German lessons). A large, representative sample of N = 6909 ninth grade students from 280 classes was used and a structural model postulating five central dimensions of instructional quality (structure, classroom management, understandableness, motivation, student involvement) was tested by means of a two-level confirmatory factor analysis. The five-factor model adequately described the structure of student ratings. In terms of generalizability, the assumption of equal measurement properties of student ratings across classes held for some of the investigated dimensions (structure and classroom management), but not for others (understandableness, motivation, and student involvement). Possible explanations for these differences are finally discussed.

A Two-Level Confirmatory Factor Analysis of a Modified Rosenberg Self-Esteem Scale

Classical factor analysis assumes independent and identically distributed observations. Educational data, however, are often hierarchically structured, with, for example, students being nested within classes. In this study, data on self-esteem gathered in a sample of 1,107 students within 72 school classes in Switzerland were analyzed using two-level confirmatory factor analysis. Considering a sequence of two-level confirmatory factor models, the results indicate that a one-factor model of self-esteem with an additional orthogonal method or response-style factor of negatively worded items adequately described within-class (individual) differences in self-esteem. By contrast, at the between-class level, a general factor of self-esteem was sufficient to capture school class differences in self-esteem. Thus, apart from other influences, for students, the social context (school class) seems to matter in forming their self-esteem. At the same time, the findings imply that studies examining self-esteem using samples of clustered observations should account for multiple levels of analysis (i.e., separate within- and between-class variances).

Estimation of factor scores in a two-level confirmatory factor analysis model

In this paper, two approaches for estimating the between-group and the within-group factor scores in a two-level confirmatory factor analysis model will be investigated. In each approach, the regression method and the generalized least-squares method are studied. These methods will be compared with each other theoretically and empirically with a simulation study.