Multilevel analysis of student civics knowledge scores

ABSTRACT

Compositional effects of scholarly culture classroom/school climate on civic knowledge scores of 9th graders in the United States were examined using the International Association for the Evaluation of Educational Achievement (IEA) 1999 Civic Education Study data. Following Evans et al. (2010, 2014), we conceived that the number of books at home, referred to as the home literacy score by IEA, can be an index of scholarly culture of the student's home, and its aggregated average constitutes scholarly culture of the classrooms/schools. The results obtained through multilevel analysis indicated that there were indeed large unique compositional effects and its effect size was comparable to that of mean parent education, individual level scholarly culture, and parent education. Implications of the results in terms of educational policy were discussed.

KEYWORDS:

• Compositional effects
• IEA Civic Education Study 1999
• multilevel model

Note

Notes

1. The half-standardized coefficient was obtained by multiplying raw coefficients by respective standard deviation of each independent variable. This means that the half-standardized coefficient for the level 1 independent variable (x) will be obtained by multiplying the raw coefficient by its standard deviation, sd(x), and that the half-standardized coefficient for level 2 aggregated independent variable () will be obtained by multiplying the raw coefficient by its standard deviation (i.e., sd()), all of which are available in Table 1. Specifically, we obtained the half-standardized coefficient as follows: 3.710 × 1.47 = 5.45 for parent education, 3.226 × 1.32 = 4.26 for home literacy, 7.946 × 0.74 = 5.88 for mean parent education, and 7.824 × 0.61 = 4.77 for mean home literacy.