You can use Random Number Table. - Computerized random number generator.

s. Béatrice Marianne Ewalds-Kvist I get using a random number generator for randomly getting participants but my concern is determining a specific number of participants per group that would total the sample size from groups with different number of members.

s. Béatrice Marianne Ewalds-Kvist

There are few methods pick your sample.

1. Using Randomized table using number generator - This will generate different results each time (You should not make any mistakes).

2. Using proportionate of the different groups to the population and randomly picking the members of each group (Using random table). Please refer the excel sheet [(Size of the groups / Population Size)* Required Sample size ]. Refer the excel sheet attached.

3. If you can classify the groups into different strata (Based on a unique characteristic - Strata Sampling ) and then use random table to pick the persons.

The basis, you formulate the sample is important, if you have viva presentation.

Good Luck

[email protected]

Rajeevan

Shanmugavel Rajeevan [(Size of the groups / Population Size)* Required Sample size ] - Is this statistically sound?

Alvin Carreon

It is a form of proportionate sample method that you can apply random sampling(Refer the Excel Sheet). Since you sample group does not have similar characteristics ( I am assuming). Proportions will be in line with the populations.

Shanmugavel Rajeevan I get it. Thank you for your suggestion. Does this mean that I have to assign a number to each member of a group and use a random number generator to choose who can participate or I can just opt to have the participants answer at their convenience provided that the total number of participants per group is in accordance to the formula you provided? I.e. population of 500 in a group so the participants should be 60 individuals who can spare the time to answer the questionnaires.

Alvin Carreon if you can assign a number to each individual and pick a randomly 60 persons (based on the random numbers generated by a software Eg-Excel Random Function ) that would be ideal (The plausibility of doing this is doubtable) or as indicated by convenience sampling methodolgy count in the sixty people who who can spare the time to answer the questionnaires. (Please Google the unknown / research jargon terms).

End of the day your data collection process for the sample should reflect the following; (Traps at VIVA questions)

1. Your sample should be representative of the population

2. Maintaining randomness

Good Luck

Dear Alvin Carreon how, at the first place, did you decide on the magic number of 528 as your total sample size?

Second, do you for sure have the full list of all the members of the sub-groups of your population of interest? Can you share about your population of interest with the reasons of such large differences between the sub-groups (i.e., from 50 to 1000)?

Finally, the issue here might not be regarding randomness of sampling but the appropriateness of the population of interest (with very very small to large sub-groups).

Best regards

Muhammad Zia Aslam I used this site https://www.surveymonkey.com/mp/sample-size-calculator/ and put the population size using 95% level of confidence and 4% margin of error. The numbers I gave were just an example. Shanmugavel Rajeevan thanks for pointing me to proportionate sampling and the formula you gave.

If you want all the groups of the population to be proportionally represented in the sample you will have to make a stratified sampling, and not a simple random sampling.

See in: https://www.thoughtco.com/stratified-sampling-3026731