# Thread: Sample Size Determination

1. ## Sample Size Determination

Let’s say I have a population of 70000. The goal is to get a “statistically significant” sample (respondents ) for a survey.

The sample size is 383 @ 95% confidence level and 5% margin of error. This means that if you sent surveys to 70000, then you would need to receive at least 383 surveys in order to have a statiscially respresentative sample size.

Case 1:

Let’s say the expected return rate is 19%. So if you need 383 respondents and you expect 19% of the people to respond, then you need to send surveys’ to 383/0.19=2016 people.

Suppose, we want a return rate to be 20%, but historically the return rate has been only 19%. So, thinkng of adjusment, 0.20/0.19×2016=2122

Question, is this adjustment meaningful?

Case 2:

Let’s say we need to stratify the sample members. We have information on return rates in each of this group from the past. So, I first proportionally stratify my sample (383). Next, I divided the strata sample sizes by return rates to get number of surveys to sent out to get the same response rate as last year. Finally I adjust the sample sizes by multiply it by 0.2/0.19 as in case 1.

Please see attached for computations.

The final sample size here is 2027.

Can anyone tell me if my methodology is correct? If not, what is an more appropriate way to do this?

 Tweet

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts