On the Influence of Sample Size to Precision of Percentages and Means in a Survey Research

Abstract

One of the most frequently asked questions in statistics is how to determine the sample size. For many educational research pieces, inappropriate, inadequate, or excessive sample sizes continue to influence research findings' quality and accuracy. The study demonstrates the influence of sample size on the precision of measurements like percentage and means in survey research. In the study, the researcher utilized a sample size formula based on the margin of error and confidence level to determine the different sample sizes (100, 150, 300, and 1200) based on the 1761 total population of a state university study was conducted. A census on the population was conducted to determine the students' profile concerning sex, age, year level, and course and seven benchmark statements on the students’ attitude towards the university using a validated instrument. Accordingly, a number was assigned to each of the respondents during the encoding in statistical software. After this, a random number generator was utilized to determine the respondents in different sample sizes from the census's encoded data. Based on the statistical analyses, the result shows the precision of percentages in the four profile variables. The means of the seven benchmark statements do not vary significantly in terms of sample size. On inferential statistics, only the sample sizes 300 and 1200 had the same decision with the population. To conclude, the proper calculation can determine the sample size necessary to have a specific percent chance that the result will not vary from the population.