การขนาดตัวอย่างที่เหมาะสมสาหรับการวิจัยทางสังคมศาสตร์

Abstract

In survey research sampling, the issue that researchers are faced with is the problem of determining an appropriate sample size. So, the researchers must realise that determining the sample size should concern at least two reasons; 1) to estimate the proportion of the population and 2) to estimate the population mean. The author hypothesised that both the formula used to calculate sample size by estimating the proportion of the population such as Taro Yamane's formula, and the formula to calculate sample size by estimating the average population such as Krejcie and Morgan’s formula for quantitative researches in the past sometimes went wrong. The results showed that, in determining the sample size using Taro Yamane’s formula and R. V. Krejcie and R. W. Morgan’s formula, the results of the sample size calculations were similar since the two formulae were alike, and, in some cases, the calculation of sample size by using Taro Yamane’s formula may not be appropriate if the error of estimates and the population size were too small, since that could cause mistakes or inaccuracies. Accordingly, the appropriate formula to correct the problems, namely “Wonlop’s formula”, is thus suitable for such circumstances, either the error of estimates or the population size is small, that can consequently produce errors and alter the reliability of the research.