For your third example; if you exclude people because they don't fit the criteria of your population or study design then that should not be seen as non-response. It is still good to mention, especially if the criteria are very strict, but your response rate should be based on the total amount of respondents who were eligible.

Dear Muhammad,

Request you to provide some more information on what you are looking forward to find out. Here are some thoughts however.

Proportional z test is good if you want to compare the response rate at a given confidence level. However, for one sample proportinal z test you need population response rate and compare it with one group of sample at a given cinfidence level, to find out if the response rate of your group significantly differ from population response rate.

For 2 sample proportional z test, instead of with population, you may compare response rate between 2 groups at a given confidence level, to identify if there is significant difference between response rates of the two groups.

Thanks.

Dear @Jolien van Dijk, response rate in example 3, should it be reported 100% or 88.1%?

Dear @Praheli Dhar Chowdhuri, all three examples are from published studies. Which study calculated/reported response rate right/wrong? Please share your comments.

Muhammad Adnan Maybe you can clarify what the criteria you mentioned were in example 3? If it were regular criteria that defined who was eligible for the study then it should have been 100%. If the criteria was related to nonresponse then that 88,1% is correct.

Muhammad Adnan Study 2 I'm not sure what you mean in the difference between sample size and invited. Presuming the invited number was the drawn sample and 10834 individuals responded making it the realised sample of respondents. Then study 2 calculates the response rate correctly.

For study 3 as a few people have commented it depends why the 18 individual cases were excluded. If they were excluded due to not belong to your population of interest (say wrong age cohort) and then ineligible for your survey this is not non-response. If they were excluded due to non-contact (couldn't find them) or refusal to take part then this is non-response and (133/151)*100 would the appropriate way to calculate non-response.

Thanks for your comments, dear @Matthew Moore.

Dear Muhammad,

For option 3, excluding respondents based on researcher's set criteria, can not determine response rate. Response rate is how many responders has responded, given a chance to response.

For option 2, 'participated' is ok. But have little confusion in understanding what is meant by 'invited' and 'sample size'.

Thanks,

Praheli

@Praheli Dhar Chowdhuri

In study 2, authors calculated sample size was 10800. To meet this, they invited 12486 people to participate in the survey. Of them 10834 people participated. Hopefully you understand it.