Sample Size Calculation for Continuous and Discrete Data
Paul I. Louangrath · Zenodo (CERN European Organization for Nuclear Research) · 2019
The purpose of this paper is to provide a practical guidance to researcher in social science on sample size determination. Sample size calculation is a basic and indispensable requisite for applied research in social science. Most research in social science is about population studies. In population studies, researchers could only study the sample of the population because detailed examination of the population is not feasible. In order for the sample to represent the population, a minimum sample must be obtained. Thus, minimum sample determination becomes a critical requisite in survey collection, interviews, or data collection. In this paper, we present minimum sample calculation methods for continuous and discrete data in non-time series scenarios. The data came from randomly generated values by using Excel command: rand()*100 for test sample sizes of n = 5, 10, 20, 30, 50, 100, 200, 300, 400, 500, and 1,000. We proposed a new minimum sample size method that consistently produces n = 30.