I'm planning to do a research based on questionnaire/ survey and population are clinical researchers from the whole country. The count of researchers in the country in unknown (I don't know how to get this info). My aim is to publish the results.
1- How to calculate the sample size?
2- Should IRB approve the research?
3- What is the least number of questions they should answer to be included in the stats? or percentage
4- Can I cite the same article(s) in more than on research? The information found in these articles are relevant and I can use for literature.
Thanks in advance.
You need to identify the hypotheses, power, or beta (usually 0.80), p-value threshold, or alpha (usually 0.05), the statistical method you want to use and hopefully find an approximate effect you want to find (smaller effect needs larger sample). G*power is great and free. Power analysis assumes perfect distributions, so, you can consider increasing the sample size by perhaps 10-20%, to compensate for skewness and outliers in a continuous variable.
1. There're numerous sample size calculators online, you can choose one depending on the outcome (prevalence, odds ratio etc).
2. IRB is likely to approve it if your research does not acquire sensitive information or doing any harm to the participants.
3. There is no minimum number of questions, depends on the scope of your research.
4. Yes you can.
You need to identify the hypotheses, power, or beta (usually 0.80), p-value threshold, or alpha (usually 0.05), the statistical method you want to use and hopefully find an approximate effect you want to find (smaller effect needs larger sample). G*power is great and free. Power analysis assumes perfect distributions, so, you can consider increasing the sample size by perhaps 10-20%, to compensate for skewness and outliers in a continuous variable.
Sample size calculation is a science by itself. In simple way, it depends on the type of sample you are targeting, the research design, the variables, the statistics that you will use..etc.
for example, when you have a hypothesis and you will apply inferential statistics, you may use the power sampling technique. This requires from you to specify the alpha level {mostly acceptable to be .05}, the power (standard to be .80), and the effect size. The effect size vary depend on the statistics that you will use [it is possible to choose between small, moderate, and large] which depends on the literature to estimate. There is a widely used program online for sample calculation based on power called G. power, it is free to download and use.
For survey design, there are online links that calculate the sample size based on confidence interval and confidence level. Here is one link
https://www.surveysystem.com/sscalc.htm
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size