Csaba Csíkos: Determining sample size in educational investigations

This study focuses on the factors that affect sample size in educational investigations. Two types of research are addressed: (1) large sample investigations and (2) educational experiments. In large sample investigations three factors are analyzed: i) the number of aspects from which a large sample can be considered representative; ii) the role of confidence intervals in the estimation of population mean; and iii) subjective factors concerning relevance and soundness of research results. Determining sample size in educational experiments requires taking account of i) the probability of type I error; ii) the power of the experiment; and iii) effect size. Beyond discussing theoretical relations between factors affecting sample size, detailed examples and illustrations are given in order to help researchers in designing investigations and determining sample size.

MAGYAR PEDAGÓGIA 104. Number 2. 183-201. (2004)

Levelezési cím / Address for correspondence: Csaba Csíkos, Department of Education University of Szeged

 
PDF