How to Determine Sample Size for Survey Research?

Рет қаралды 3,543

Күн бұрын

Selecting a suitable sample size is essential for deriving reasonable inferences from study results. While there are a number of commonly used guidelines for calculating sample size, researchers are still unsure which guideline to use for their individual investigations. ‘How large should the sample be?’ is one the most frequently asked questions in survey research.
The focus of this tutorial is on: First, discuss the factors that influence sample size decisions. Second, review existing rules of thumb related to the calculation of sample size and also perform power analysis using the G*Power program.
In this insightful video, we delve into the complexities of determining sample size for survey research, a perennial challenge for social science researchers. I use the paper titled "Sample size for survey research: Review and recommendations" published in the Journal of Applied Structural Equation Modeling," that succinctly encapsulates crucial considerations for estimating sample size. Exploring factors such as the nature of research, statistical analysis types, and the influence of research supervisors, the video provides valuable insights for young scholars and those working on papers or theses.
The speaker emphasizes the significance of tailoring sample size decisions based on the research's nature and the type of statistical analysis employed. Key considerations include the impact of analysis programs, the influence of research supervisors, and the role of multivariate analysis techniques.
Contents
00:00 - Introduction
00:13 - Challenges in Sample Size Calculation
01:42 - Factors Influencing Sample Size Decision
03:44 - Research Supervisor's Influence on Sample Size
05:04 - Consideration of Sample Size in Research
06:21 - Sample to Item Ratio for Exploratory Factor Analysis
06:59 - Sample to Variable Ratio
08:16 - Recommended Online Calculators
08:59 - Sample Size for Structural Equation Models
12:12 - Sample Size Guidelines for PLS-SEM
13:27 - Criticism of the 10 Times Rule
14:04 - Alternative Methods: Inverse Square Root and Gamma Exponential
14:42 - Power Table by Hair et al.
17:29 - G*Power Analysis for Sample Size Calculation
Reference Paper
Memon, M. A., Ting, H., Cheah, J. H., Thurasamy, R., Chuah, F., & Cham, T. H. (2020). Sample size for survey research: Review and recommendations. Journal of Applied Structural Equation Modeling, 4(2), 1-20.
Daniel Soper Sample Size Calculator
• A-priori Sample Size C...
Do you have any questions, tips, or ideas? Let me know in the comments section below!
About:
Research With Fawad is a Research initiative to help young and early career research learn the research tools and techniques. It is an initiative to share research knowledge and help researchers learn to research the easy way.
Learn more about Research With Fawad at www.researchwithfawad.com.
#Research #Sampling #SampleSize #SurveyResearch

Пікірлер: 14

@aftabhussain5858 5 ай бұрын

I appreciated your clarity of concept

@researchwithfawad 5 ай бұрын

Thanks. I am glad you liked it.

@chefberrypassionateresearcher 3 ай бұрын

I assume that population size is unknown/ infinite, therefore I plan to use Cochrane Formula- which gives a sample size of 384, for this study, at 95% confidence interval, 50% degree of variability, 5% as margin of error. All my questions are on Likert 5 point scale (Interval data). Is that appropriate to be used?.

@researchwithfawad 3 ай бұрын

Yes, it is appropriate.

@anitalow3628 5 ай бұрын

Could you produce two staged of PLS-ANN tutorial, particularly how the R-squared for ANN is calculated

@researchwithfawad 5 ай бұрын

Sure, when i work on it.

@chefberrypassionateresearcher Ай бұрын

SIR, If i want to compare three groups, then Daniel soper calculator gives me the minimum sample size "required for each group" OR "minimum sample size required altogether"?

@researchwithfawad Ай бұрын

Please refer to Multigroup Analysis using SmartPLS: Step-by-Step Guidelines for Business Research

@chefberrypassionateresearcher Ай бұрын

@@researchwithfawad I have read this paper. It also does not specify whether sample size given by g power is for complete data required, or for the groups. And this study has taken effect size of .25, but medium effect size is .15. This study also does not mention the number of predictors taken for the model.