Στατιστική ανάλυση της μαθητικής επίδοσης σε σχέση με σχολικά και κοινωνικά χαρακτηριστικά

The aim of this dissertation is twofold. Initially, a methodology for the development of the optimal sample size is developed employing the Cronbach’s alpha coefficient in a Bayesian statistical setup in the case of cluster sampling. The determination of the optimal sample size in empirical studies in social sciences is useful not only in testing theoretical research questions but it also constitutes an important practical issue since the sample must be adequately informative without unnecessarily compromising available resources. The suggested approach estimates the probability that the reliability coefficient exceeds a predetermined value, imposed by the researcher. Furthermore, since this is a pre-experimental procedure, the calculation of the optimal sample size takes into account parameters of the prior distribution. Then, utilizing the new technique, a country-wide study was conducted to analyze student performance in the examinations for entrance in the tertiary education in Greece with respect to social and school characteristics. The analysis was based on a representative sample from the country’s public lyceums. The data contained information about social characteristics of the students at their last year of study as well as information about the schools. Finally, student performance in tertiary education entrance examination was recorded. The statistical techniques made use of descriptive statistics, structural equation modeling and hierarchical linear models. A general conclusion is that students with good grades in previous years have increased chances of entering tertiary education. This is also affected by additional help at home whereas, at least at this stage, the effect of parent education and the use of cultural items are less important factors. Other variables affecting student grades are gender and the area of residence in particularly with respect to the presence of students with immigrant background and teachers’ training. These findings should be evaluated within the specific timeframe of the students’ educational paths. The educational system of Greece at the last grade of secondary education is particularly exam-oriented thus stifling any creative-based seeking of knowledge where social and cultural differences surface and encouraging standardized techniques to obtain higher grades. Hence, further research is necessary to analyze student performance in previous grades while it would also be interesting to examine the students’ educational paths especially after entering higher education where cultural and social capital are more important factors.

Grantor dc:publisher

University of Peloponesse

Year dc:date

2009

Authors dc:creator

• Limakopoulou, Aristea
• Λιμακοπούλου, Αριστέα

• Εκπαιδευτική έρευνα
• Μαθητική επίδοση
• Σχολικά χαρακτηριστικά
• Κοινωνικά χαρακτηριστικά
• Μέγεθος δείγματος
• Στατιστική κατά Bayes
• Θεωρία δομικών εξισώσεων
• Ιεραρχικά γραμμικά μοντέλα
• Educational research
• Student performance
• School characteristics
• Social characteristics
• Sample size
• Bayesian statistics
• Structural equations modeling
• Hierarchical linear models

Language dc:language

gre

Identifier

10.12681/eadd/1543

OAI identifier oai:identifier

oai:10442/1543