Can political inequality be reduced in the classroom? Testing the compensation hypothesis and the BFLPE on youth civic competence

ABSTRACT

This study examines the general question of how classroom characteristics are linked with differences among students in civic competence, which is seen to be an important basis for political inequality. A resource-mobilisation account of youth civic competence is presented, and this is tested using hierarchical linear modelling and International Civic and Citizenship Study 2009 data. The determinants of youth civic competence are explored at the individual, family, and classroom levels, where resource and mobilisation factors at each level are examined. Evidence for classroom effects are tested using Campbell’s compensation hypothesis and insights derived from Marsh’s big-fish-little-pond effect (BFLPE). This case study of the Czech Republic shows limited evidence for an open classroom climate reducing civic competence differences between low- and high-SES students, and no evidence of BFLPE increasing such differences among youths.

KEYWORDS:

• Civic knowledge
• political attitudes
• open classroom climate
• classroom composition (BFLPE)
• ICCS

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. BFLPE is explained using social comparison and reference group theory from social psychology. While not directly connected with resource and mobilisation literatures of political science discussed in this study, the notion that BFLPE critically depends on a student’s frame of reference does reflect resource differences and the differential impact of student mobilisation factors in different contexts.

2. The student weight (TOTWGTC) prepared by the IEA was not used because non-response adjustments made to the school data may make these values biased. Consequently, the data were weighted using a procedure outlined by Brese, Jung, Mirazchiyski, Schulz, and Zuehlke (2011, p. 31). At the student level, a within-school weight was applied (i.e., WGTFAC2S x WGTADJ2S x WGTADJ3S). For the classroom level, the weight used was a product of the WGTFAC1 and WGTADJ1S weighting variables. All models were estimated using STATA v.14.2 (SE).

3. The ICC variance components statistics for the three-level model are (1) civic knowledge ICC (school) ≤ .01, ICC (classroom) = .07; (2) voting ICC (school) ≤ .01, ICC (classroom) = .03; (3) political participation ICC (school) ≤ .01, ICC (classroom) = .03. The ICC estimates at the school level are not significant at the p ≤ .05 level. With a two-level model, the results are (1) civic knowledge ICC (classroom) = .08, (2) voting ICC (classroom) = .03; (3) political participation ICC (classroom) = .04. All ICC estimates are significant (p ≤ .05).

4. Table A2 reveals that almost all of the independent variables are scales estimated using confirmatory factor analysis (CFA). The model fit statistics (root mean square error of approximation [RMSEA] and comparative fit index [CFI]) for these scales show good model fit. Moreover, these scales are reasonably internally consistent (as indicated by the Cronbach’s alpha statistic estimates ≥ .66). However, Table A2 shows that the factor loadings for some indicators on the explanatory variable scales presented later in regression models of Table 1 are not high, although when modelled as summated rating scales the inter-correlation between scale variables is always adequate. The justification for including indicators in regression factor scales with low loadings (≤ .50), especially for SES, is theoretical.

5. Cohen (1988, pp. 477–478) advised that the f² statistic be interpreted as follows: f² ≥ .02, f² ≥ .15, and f² ≥ .35 represent small, medium, and large effect sizes, respectively. Basically, Cohen’s f² statistic shows the marginal effect size of including an additional covariate to a set of explanatory variables already in the model.

Additional information

Funding

Research for this article was kindly provided by the Czech Science Foundation within the framework of a Centre of Excellence project entitled “Dynamics of Change in Czech Society” [project number GA ČR GB14-36154G].

Notes on contributors

Aleš Kudrnáč

Aleš Kudrnáč is a post-doc researcher at the Department of Political Sociology, Institute of Sociology of the Czech Academy of Sciences. His research focuses mainly on political attitudes, values, and behaviour of youth. He is an author of several academic articles on the topic of youth political attitudes (e.g., Political Studies, Journal of Youth Studies).

Pat Lyons

Pat Lyons is a senior researcher at the Department of Political Sociology, Institute of Sociology of the Czech Academy of Sciences. He has a PhD in political science from Trinity College, Dublin. His main areas of research are public opinion, political attitudes, and behaviour.