![Sample our Education journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5932/TOC-Subject-Sample-2021-Education.jpg"})
Abstract
The Czech education system is characterised by early tracking and high educational inequalities. The most controversial element of the system is the academic track, which is attended by students after the completion of primary school. The paper focuses on fifth graders’ application process and their success in entrance examinations. The process is examined by binary logistic regression and descriptive analyses of the data collected in the Trends in International Mathematics and Science Study conducted in 2011 in grade 4, and in a survey of students attending grade 5 implemented on the same sample within the Czech Longitudinal Study in Education. The analyses show that students’ family background has a significant effect on both application and success during admission procedures. The paper discusses the role of the family in the application process and in preparation for the entrance examinations, and compares the situation in the Czech Republic with that in other countries.
Notes
1. The aspirations were defined as an intention to stay in full-time education after the age of 16.
2. More detailed information on the proportion of students at eight-year gymnasia and their development in the Czech Republic and abroad can be found in Greger and Holubová (Citation2011).
3. Available from http://toiler.uiv.cz/rocenka/rocenka.asp. The data were provided for the school year in which the students included in this study started their studies in the eight-year gymnasium.
4. A detailed discussion of the history and post-1989 re-introduction of multi-year gymnasia is provided in another work (Greger, Citation2005).
6. The CLoSE survey is funded by the Czech Science Foundation. The grant also includes the monitoring of adult education and the transition from nursery to basic school. The grant is used by three organisations: the Economics Institute of the Academy of Sciences of the Czech Republic, the National Training Fund and the Institute for Research and Development of Education of the Faculty of Education, Charles University. The Institute for Research and Development of Education is responsible for the research on the transition to basic school and for the monitoring of students within compulsory education.
7. Binary logistic regression is a regression model where the dependent variable is dichotomous. In our analyses, we explain two dichotomous variables (application to study and admission to study) by a set of independent variables that characterise student aptitude and student family background.
8. We followed the recommendations of IEA and OECD experts not to average plausible values but to use the first plausible value in the analyses in a situation where special routines are not available for calculations with all five plausible values. No special routines for logistic regression where plausible values are used as independent variables were available when this analysis was carried out.
9. The answer to the question: How much time have you devoted to preparation for the entrance examinations in the last semester? I prepared (a) every day or almost every day, (b) once or two times per week, (c) once or two times per month, (d) never or almost never.
10. The Czech Republic used the opportunity provided by the fact that PIRLS and TIMSS were run in the same year and combined the studies. That means that there is one sample for both studies that allows achievement in all the domains that were tested to be compared. In this study, we use only mathematics achievement.
11. The TIMSS and PIRLS samples were drawn using the multistage cluster-sampling technique: first, schools were sampled randomly from the database of all schools with fourth-grade students, then one or two classes were selected randomly from these schools. The general idea behind the jackknife procedure is to split a single sample into multiple subsamples and use the fluctuation among the subsamples to obtain an estimate of the overall sampling variability. The subsamples were generated using the variables JKZONE and JKREP, which are included in the TIMSS & PIRLS database (for more information see Martin & Mullis, Citation2013).
12. Both the application and the success model were also tested with the interactions between achievement and parental education. The interactions proved to be significant and eliminated the effect of both mathematical achievement and parental education. However, the R 2 of the model with the interaction term was lower in both cases than the R 2 of the original model. We thus decided to interpret the model without interactions.