I'm conducting a cross-sectional KAP survey exploring a specific population's knowledge attitude and practice in a particular area of healthcare. There is an informational intervention in the survey after answering the KAP questions. After the KAP questions and the intervention, the knowledge and attitude questions will be asked again to study the short-term benefit of the intervention.
My questions are:
1. The general population of the study group is around 3000. What's the best way to calculate the sample size?
2. Should I include a control group for the intervention (they will be asked the knowledge and attitude questions again at the end of the survey without seeing the intervention.?
3. If I include a control group, do I have to double the sample size?
Thanks
Depends on the circumstances. However, sample size can be estimated using formula-based approaches, power analysis etc.
Yes, including a control group can help control for potential confounding variables that may affect the outcome of your intervention. A control group can help you determine the effect of the intervention by comparing the post-intervention knowledge and attitude scores of the intervention group with those of the control group. By including a control group, you can increase the validity and reliability of your study results.
Including a control group does not necessarily mean that you have to double the sample size. The sample size for the control group can be smaller than that for the intervention group, depending on the desired level of statistical power and the expected effect size. The sample size for the control group can be calculated separately using the same methods as for the intervention group. However, if you have limited resources or a small population, it may not be feasible to include a separate control group. In this case, you may consider using a pre-post design, where the same group of participants is assessed both before and after the intervention. However, this design has some limitations, such as the lack of a control group and the potential for testing effects
I suggest using a power analysis with the G*power program. Having a control group is no different from any other two-way comparison (e.g., male and female), so you can assess the power you need for a t-Test.
Hello!
Here are some suggestions for your questions:
n = (Z^2 * p * (1-p)) / E^2
where n is the sample size, Z is the z-score corresponding to the desired level of confidence (e.g., 1.96 for a 95% confidence level), p is the estimated proportion of the population with the outcome of interest, and E is the desired margin of error.
To estimate p, you may need to conduct a pilot study or use data from previous studies. Alternatively, you can assume a conservative estimate of p (e.g., 0.5) to get a larger sample size.
Keep in mind that this formula assumes simple random sampling, so if you plan to use a more complex sampling method, such as stratified or cluster sampling, you may need to adjust the formula accordingly.
I hope this helps! Good luck with your survey.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression