Hi everyone,
(Please excuse the ambiguity within the question - protecting my idea)
I recently discovered some indexes which I and colleagues would like to see if scores on these indexes across different countries is related to a particular public health concern.
In saying that, the total sample size we could achieve is 195+ (depending on how which number you take as countries in the world) - however we likely only have DV data (on the public health concern) in around 37 countries - maybe slightly more - so 20% of the 'total possible population'.
Is this sufficient for regression analyses? Would rule of thumbs regarding sample size be different if the total population is so low? Are there different thoughts from different areas of research demographers etc?
Thanks.
Dear Russell
Determining sample size is a very important issue because samples that are too large may waste time, resources and money, while samples that are too small may lead to inaccurate results. In many cases, we can easily determine the minimum sample size needed to estimate a process parameter, such as the population mean.
If the population is known (sample size calculation with finite population) then you can use any of the following calculators to know the sample size.
1. https://www.surveysystem.com/sscalc.htm
2. https://www.calculator.net/sample-size-calculator.html
3. https://apps.who.int/iris/handle/10665/40062
I mostly used the following online calculator:
4. https://select-statistics.co.uk/calculators/sample-size-calculator-population-proportion/
Some related research paper available on researchgate:
1.Article Sample Size Determination in Health Studies: A Practical Manual.
2. Article Sample size calculation
When You have not any idea about the population ( here in your question i.e. the prevalence of any disease) so you need to use the sample size calculation with a non-finite population.
And in this case,(sample size calculation with non-finite population) you go with the Cochran formula, for this please follow the link:
https://www.researchgate.net/deref/https%3A%2F%2Fwww.statisticshowto.com%2Fprobability-and-statistics%2Ffind-sample-size%2F
or you can visit any other link related to Cochran formula (Sorry I do not know how to type the formula here, so please go to any of the link or just search):
1. https://www.researchgate.net/post/Is_that_possible_to_use_Cochran_formula_sample_size_determination_for_heterogeneous_groups_populations
2. https://www.researchgate.net/post/How_do_I_determine_sample_size_for_each_strata_in_survey2
As per the design of your study determine the sample size for the study.
Please find some of the files (related to sample size calculation) as attached files.
Hope you find all the information useful in the way to determine sample size as the case you have.
Happy researching..
Dear Russell
Determining sample size is a very important issue because samples that are too large may waste time, resources and money, while samples that are too small may lead to inaccurate results. In many cases, we can easily determine the minimum sample size needed to estimate a process parameter, such as the population mean.
If the population is known (sample size calculation with finite population) then you can use any of the following calculators to know the sample size.
1. https://www.surveysystem.com/sscalc.htm
2. https://www.calculator.net/sample-size-calculator.html
3. https://apps.who.int/iris/handle/10665/40062
I mostly used the following online calculator:
4. https://select-statistics.co.uk/calculators/sample-size-calculator-population-proportion/
Some related research paper available on researchgate:
1.Article Sample Size Determination in Health Studies: A Practical Manual.
2. Article Sample size calculation
When You have not any idea about the population ( here in your question i.e. the prevalence of any disease) so you need to use the sample size calculation with a non-finite population.
And in this case,(sample size calculation with non-finite population) you go with the Cochran formula, for this please follow the link:
https://www.researchgate.net/deref/https%3A%2F%2Fwww.statisticshowto.com%2Fprobability-and-statistics%2Ffind-sample-size%2F
or you can visit any other link related to Cochran formula (Sorry I do not know how to type the formula here, so please go to any of the link or just search):
1. https://www.researchgate.net/post/Is_that_possible_to_use_Cochran_formula_sample_size_determination_for_heterogeneous_groups_populations
2. https://www.researchgate.net/post/How_do_I_determine_sample_size_for_each_strata_in_survey2
As per the design of your study determine the sample size for the study.
Please find some of the files (related to sample size calculation) as attached files.
Hope you find all the information useful in the way to determine sample size as the case you have.
Happy researching..
An n of 37 is small to be conducting regression. My main concern would be stability of results, especially if you include several correlated predictors. You might find that your results will not replicate. There is also concern about statistical power--if your effect size is not very big, it is unlikely you will find an effect. I would start with correlations and see if any of your predictors relate to your DV.
The data is likely to be better in better developed countries, so this is not really a question of power and sample size to make a whole world inference.
rather you should restrict the inference to the 37 you have.
provided you don't try to look at too many independent regression variables (say less than 6 or so), you should be able to get something meaningful
Data from 37 countries are from a random sample. So statistically speaking you are in the trouble zone. A hypotheses test, confidence interval, .... are not valid. I would start with correlation. Regression can be done with 10n = sample size rule. n= number of predictors. This would be an absolute minimum. It would help if existing evidence underpins the selection of predictors. Otherwise false positives can obscure results. Also there is the confounding. Good luck
Cross sectional descriptive study based on current census of each countries, countries with larger census have similar challenges , hence sample few, countries with less census have different challenge thus sample many countries from the target population.
Sample Size Rule
Sekaran (2013) wrote:
"Roscoe (1975) proposes the following rules of thumb for determining sample size:
1. Sample sizes larger than 30 and less than 500 are appropriate for most research.
2. Where samples are to be broken into sub-samples;(male/females, juniors/seniors, etc.), a minimum sample size of 30 for each category is necessary.
3. In multivariate research (including multiple regression analyses),the sample size should be several times (preferably 10 times or more) as large as the number of variables in the study.
4. For simple experimental research with tight experimental controls (matched pairs, etc.), successful research is possible with samples as small as 10 to 20 in size."
Reference
Sekaran, U., 2003. Research methods for business: A skill building approach. John Wiley & Sons.
I think, you need to consider calculated individual samples from each countries or you should group similar countries in cluster to calculate optimum amount of samples. Otherwise it's difficult to select one country to represent another countries without assigning solid reasoning.
It seems you may also need to do multi-level modeling since there may be clustering effects within countries. Then sample size at the higher levels are also important for consideration - not just the first level
- Badges
- Science topic
- Similar topics
- Medicine
- Public Health