Actually, your question is unanswerable.

What do you want to do with your sample? Perform a test for the mean or a proportion? Carry out a regression analysis? Carry out some sophisticated hierarchical model? Etc.

All these methods require different approaches to computing the desired sample size.

In statistics, population implies an indefinite number of data. to determine sample size, you should have an estimate of the population variance. In next step, you may determine your error (type I and II) critical values. This article may help you step forward.

Actually, your question is unanswerable.

What do you want to do with your sample? Perform a test for the mean or a proportion? Carry out a regression analysis? Carry out some sophisticated hierarchical model? Etc.

All these methods require different approaches to computing the desired sample size.

Total population is unknown! Do you mean you only part of the target population or do you mean you do not know the population size?

I should note that when we say a population is unknown, this means that we do not have information about 2 things:

1- We do not know population parameters (Mu and Sigma)

2- We do not know population probability mass (for discreet variables) or density (for continuous variables of population) function.

Thats all.

For both of these problems we can have some ideas that can help to reach a good sample size.

Aissa, try these

Sample size for estimating the population mean

(with infinite or unknown population), margin of error,

variance and confidence level given

n = z2 δ2

e2

Sample size for estimating the population proportion

(with infinite or unknown population), margin of error,

proportion and confidence level given

n = z2PQ

e2

2 is squared

Dear Scholars

Thank you for your answers and sorry for the delay in responding you.

Actually, i know that saturation is the criterion for qualitative research however i got confused about the determination of the sample size in my current study. the problem is that respondents of the questionnaire will be a certain ethnic entrepreneurs of the businesses who are both owners and managers. The figure i got so far about the total population is 1883 businesses that includes entrepreneurs for the target ethnicity but this number is raw and includes also those shareholder with very low participation which doesn't give right for management.

Note that i am studying the determinants of performance among immigrant entrepreneurs and i want to carry some cor-relational and regression analysis.

It is really a nice question in practical sense. Generally sampling in forests or even sampling for Household for larger areas also require this information, as generally researchers are not able to get this information due to cost and time. However, on the other hand, the sample size formula requires the size of population.

You may use this as reference

https://www.researchgate.net/publication/260082130_R_D_Management_in_State_Universities_and_Colleges_in_the_Philippines_Sampling_in_Business_and_Management_Research?ev=prf_pub

Data R& D Management in State Universities and Colleges in the Ph...

it is quite interesting the responses i read and now seems to realised that my study sample population is dependant on intended usage.

My concern relative to the question ask is;

what will be my sample size for stakeholders (experts, local community as heritage bearers and development partners) who are my respondents in a selected settlement?. It is also important to note that the population of all the three stakeholders in the study area is unknown.

Iam working on "assessment of architectural heritage conservation for tourism development amongst indigenous communities".