Doubly-Latent Models of School Contextual Effects: Integrating Multilevel and Structural Equation Approaches to Control Measurement and Sampling Error

We assessed evidence for a contextual effect of positive intergroup contact, whereby the effect of intergroup contact between social contexts (the between-level effect) on outgroup prejudice is greater than the effect of individual-level contact within contexts (the within-level effect). Across seven large-scale surveys (five cross-sectional and two longitudinal), using multilevel analyses, we found a reliable contextual effect. This effect was found in multiple countries, operationalizing context at multiple levels (regions, districts, and neighborhoods), and with and without controlling for a range of demographic and context variables. In four studies (three cross-sectional and one longitudinal) we showed that the association between context-level contact and prejudice was largely mediated by more tolerant norms. In social contexts where positive contact with outgroups was more commonplace, norms supported such positive interactions between members of different groups. Thus, positive contact reduces prejudice on a macrolevel, whereby people are influenced by the behavior of others in their social context, not merely on a microscale, via individuals' direct experience of positive contact with outgroup members. These findings reinforce the view that contact has a significant role to play in prejudice reduction, and has great policy potential as a means to improve intergroup relations, because it can simultaneously impact large numbers of people. diversity | trust | social norms | multilevel analysis

show abstract

“…1). We assessed the magnitude of the contextual effect by calculating an effect size measure (ES2; see SI Text) (26).…”

Section: Studies 1a To 1ementioning

confidence: 99%

Contextual effect of positive intergroup contact on outgroup prejudice

Christ

Schmid

Lolliot

et al. 2014

Proc. Natl. Acad. Sci. U.S.A.

296

250

View full text Add to dashboard Cite

show abstract

“…In this doubly latent model, the dependent variable (here: students' performance measured in the post-test) is added as a latent trait on both the individual and the class level. This model controls for measurement error (sampling of items on both levels) and sampling error (sampling of individuals in the aggregation from the individual to the class level) (Marsh et al, 2009). We chose this method because of the hierarchical structure in our data set.…”

Section: Discussionmentioning

confidence: 99%

“…Doubly latent model. To investigate whether biology teachers' CK, PCK, and CuK are predictive for students' science performance, we specified a doubly latent model (Marsh et al, 2009), which integrates a structural equation model and a multilevel model. In this doubly latent model, the dependent variable (here: students' performance measured in the post-test) is added as a latent trait on both the individual and the class level.…”

Section: Discussionmentioning

confidence: 99%

Using doubly latent multilevel analysis to elucidate relationships between science teachers’ professional knowledge and students’ performance

Mahler

Großschedl

Harms

2017

International Journal of Science Education

View full text Add to dashboard Cite

Teachers make a difference for the outcome of their students in science classrooms. One focus in this context lies on teachers' professional knowledge. We describe this knowledge according to three domains, namely (1) content knowledge (CK), (2) pedagogical content knowledge (PCK), and (3) curricular knowledge (CuK). We hypothesise a positive relationship between these three domains and students' performance in science. Students' science performance was conceptualised by system thinking performance in the context of biology teaching. In order to test our hypothesis, we examined the relationship between the knowledge triplet CK, PCK, and CuK and students' performance. 48 biology teachers and their students (N = 1036) participated in this study. Teachers' content-related professional knowledge and students' performance were measured by paper-and-pencil tests. Moreover, we used concept maps to further assess students' performance. By specifying doubly latent models, we found a significant positive relationship between biology teachers' PCK and students' performance. On the contrary, the results reveal no relationship between CK and CuK and students' performance. These findings are discussed in respect to modelling the interrelationship of teachers' content-related professional knowledge and students' learning in science, as well as concerning their relevance for further research and teacher education programmes. ARTICLE HISTORY

show abstract

“…One problematic aspect of the context effect analysis is that the observed group average obtained by aggregating individual observations may not be a very reliable measure of the unobserved group average if only a small number of L1 individuals is sampled from each L2 group [1,4]. A few researchers explored the integration of structural equation modeling (SEM) and multilevel modeling (MLM) to the issue of contextual analysis with the consideration of measurement error and sampling error [1,4,5].…”

Section: Introductionmentioning

confidence: 99%

“…When data are multilevel and a predictor variable varies both within clusters and between clusters (such as individual social economic status, SES) scores varying within schools and average SES scores varying between schools), researchers are frequently interested in estimating within-cluster and between-cluster relationships of the predictor to the criterion. Often people are interested in estimating the context coefficient [1,2,3], that is, the difference in the regression coefficients for the between-and within-cluster relationship. Contextual analysis evaluates whether the aggregated group characteristic (L2) has an effect on the outcome variable after controlling for individual level characters (L1).…”

Section: Introductionmentioning

confidence: 99%

Estimating the Context Effect in a Multilevel Latent Model with Small Sample Sizes: A Monte Carlo Simulation Study

Gao¹

2017

AJTAS

View full text Add to dashboard Cite

Abstract:In multilevel modeling, the relationships between the criterion and predictors are investigated at different levels. Often, the cluster-level predictors are measured by aggregating the individual-level measures. However, the aggregated clusterlevel predictors do not always reliably measure the cluster-level regression coefficient, and therefore the context coefficient. This study investigates an alternative approach: estimating cluster-level predictor on the latent cluster mean by using multilevel latent. A comparison is made of the accuracy of the context coefficient and standard error under a wide range of conditions. Results reveal that bias for context effect is small in multilevel latent model. Maximum likelihood (ML) estimator yields more accurate standard error estimation than robust maximum likelihood (MLR) when cluster number is small (less than 50). Very small cluster sample sizes (less than 10) should be avoided because they lack power and empirical sampling variance.

show abstract

Doubly-Latent Models of School Contextual Effects: Integrating Multilevel and Structural Equation Approaches to Control Measurement and Sampling Error

Cited by 426 publications

References 61 publications

Contextual effect of positive intergroup contact on outgroup prejudice

Contextual effect of positive intergroup contact on outgroup prejudice

Using doubly latent multilevel analysis to elucidate relationships between science teachers’ professional knowledge and students’ performance

Estimating the Context Effect in a Multilevel Latent Model with Small Sample Sizes: A Monte Carlo Simulation Study

Contact Info

Product

Resources

About