Dear Abi Ranjan,

Not the sample but the proportion of the sample to the general population is important. To calculate the necessary size of the sample please look at http://www.raosoft.com/samplesize.html

Regarding the response rate -- good journals accept articles with the results of large surveys with response rate below 10%. Formally, if you follow the suggestions of Raosoft Sample Size Calculator and do have a sufficient number of responses, it is acceptable. However, for a student project you should have some information on both respondents and non-respondents and to look if they differ in some characteristics. This can be itself a valuable source of information (inclination or avoidance of responses can be related with some important factors).

Success!

Igor Gurkov

Dear Abi Ranjan,

Not the sample but the proportion of the sample to the general population is important. To calculate the necessary size of the sample please look at http://www.raosoft.com/samplesize.html

Regarding the response rate -- good journals accept articles with the results of large surveys with response rate below 10%. Formally, if you follow the suggestions of Raosoft Sample Size Calculator and do have a sufficient number of responses, it is acceptable. However, for a student project you should have some information on both respondents and non-respondents and to look if they differ in some characteristics. This can be itself a valuable source of information (inclination or avoidance of responses can be related with some important factors).

Success!

Igor Gurkov

I wozld say for some complex statistcal analysis your sa.ple should be over 100. But I totally agree with Igor