The sample size has to be adequate for statistical purposes such as adequately representing your population. The degree you are pursuing has nothing to do with it.
It depends on the method. If I were to guess and the issue relates to a structure questionnaire I would suggest at least 1000 observations/participants. This number would give you a +/- % margin of error in percentage estimation
The greater the sample size the better the results. You must also consider explanotory variables and Response Variables which is also known as independent variables and dependent variables.
When planning a quantitative study—like one for a Master’s thesis in social sciences or business—the required sample size depends on several factors including your expected effect size, the statistical power you want (usually 0.80), and the significance level (commonly α = 0.05).
Literature and Practical Considerations
Cohen’s Guidelines: Cohen’s work on statistical power analysis provides a detailed foundation for these estimates. His classic text, Statistical Power Analysis for the Behavioral Sciences (1988), remains a cornerstone for understanding how effect size influences required sample size. His work is also complemented by subsequent guidelines (e.g., Cohen, 1992) that include tabulated values for different effect sizes.
Practical Application in Social Science/Business Research: In many Master’s theses, researchers in social sciences or business operate within practical constraints such as limited resources and participant availability. As a result, studies often target medium effect sizes, which means aiming for a total sample size in the range of 150 to 300 may be both realistic and statistically adequate.
Practical Application in Social Science/Business Research: In many Master’s theses, researchers in social sciences or business operate within practical constraints such as limited resources and participant availability. As a result, studies often target medium effect sizes, which means aiming for a total sample size in the range of 150 to 300 may be both realistic and statistically adequate.
Software Tools: Tools like G*Power are frequently used to fine-tune these estimates. With G*Power, you can input the number of groups, effect size, power, and α level to obtain a more precise calculation for your specific study setup. It is a free-to use software used to calculate statistical power.
Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences, 2nd Edn. Hillsdale, NJ: Erlbaum.
Cohen, J. (1992). A power primer. Psychol. Bull. 112, 155–159.
The required sample size for a Master's thesis in social science or business using quantitative methods depends on factors like statistical power (usually 80%), effect size (small: 0.2, medium: 0.5, large: 0.8), and the complexity of your analysis. For simple surveys or descriptive statistics, 100–300 participants are typically sufficient. For more complex analyses like regression or SEM, a sample size of 200–500 is recommended. You can use tools like G*Power to estimate the exact sample size based on your specific research design and statistical requirements.
I agree with Dr. David L Morgan in saying that the degree you are pursuing has nothing to do with the sample size. The sample size should consider confidence level, margin of error and variability within the population. You might be referring to the discipline where the study will be conducted that affects the sample size. Yes, the discipline or field of study can influence the sample size.
For a Master’s thesis using quantitative methods in social sciences or business, a sample size of 100–300 respondents is generally acceptable, though the exact number depends on the research design, statistical techniques used, expected effect size, and population characteristics; power analysis can help determine the optimal sample size for reliable results.
While the ideal sample size in a quantitative Master’s thesis in social science or business depends on the research design and analysis method, a general guideline suggests a minimum of 100–150 responses for basic statistical analysis, with larger samples (200–400) preferred for more robust and generalizable results.