Abstract
This paper highlights the dynamics that led to the narrowing of the achievement gap between students in the Hebrew-speaking and Arabic-speaking schools in Israel. Instead of looking at one point of time (Zuzovsky, 2005), this paper examines changes between 1999 to 2003 in certain school/class-level variables and relates them to the uneven increase in achievement between these two populations.
Notes
1. Sephardi (also mizrachi): A Jewish person of Spanish or Portuguese origin, now used loosely to refer to any Jewish person who is not of American or European (Ashkenazi) descent.
2. This is the result of the sampling procedure of TIMSS that allows the sampling of at least one class per sampled school.
3. The regression coefficients of all first-level student variables employed in the analyses were found to have a nonsignificant random effect over schools, and so they were specified as having fixed effect.
4. This distinction has several implications in terms of socioeconomic and cultural differences between the two populations.
5. As a general principle, models should contain a reduced set of variables by at least an order of magnitude from the number of cases to the number of model parameters. With 245 schools participating in the 1999 and in 2003 studies, the second-level model parameter should not exceed more than 24 regression terms.
6. Proficiency score used in IEA studies. For more details, see the TIMSS-2003 Technical Report (Martin, Mullis, & Chrostowski, 2004).
7. Due to an increase by 1 SD above the mean value of the standardized selected school variable involved in the analysis.