Why it is important to distribute questionnaires double in amount than calculated sample size? e.g. If the calculated sample size is 300, then why it it is necessary to double up the sample e.g. 600? Would it be exposed to over sampling bias?? Why and how researchers can set the estimated percentage of return rate e.g. 50%? any reference?
I am not sure where you get this percentage. Generally speaking, not all the people who receive a questionnaire will be willing to respond. The response rate, i.e., the percentage of people who return a completed questionnaire, will vary. A well-designed, focussed questionnaire might get response rates of 90% or more; a poorly designed one might get no respondents at all.
This will depend on a number of factors, including the length and structure of the questionnaire; the workload of potential respondents, timing, literacy levels, ease of returning the questionnaire, and many more. Because of the number of factors involved, it's hard to give an estimate. However, you can make a rough estimate by looking at the published response rates of similar studies in the context you are investigating.
Thank you Dr. Kostoulas for your contribution.
I came to know that to get a desired responses (as per our sample size), it is recommended to distribute the questionnaires in double amount as per our calculated sample size. I am not sure about its reference and searching for this. If you have any idea..
Thank you
Hi Mehwish,
Further to Dr Kostoulas' excellent reply, a key factor is also the method of distribution. For example mailed questionnaires tend to have a lower reply rate than phone-ins.
Regards,
Dominic
I think that the double in amount because of the nature of data collected by questionnaires, almost be descriptive data which in quality and accuracy lower than scale data which prefer in scientific research , that is my idea , no references .
Good Luck
If the sample size was calculated through robust method, and 10 to 20% non-response expected added to the calculated sample size is enough. But no evidence base to distribute more than required number.
As the answers above suggest, there are several pieces involved in estimating your needed distribution relative to your required sample size. In no particular order, these include:
1) Method of distribution -- in order of typical response rate, the lowest is mail, then email, then phone, then handing out surveys and standing there until the participant fills it in. The latest figure I saw about mailed surveys was a 5% return rate, so if you need 300, you would have to mail out 6,000. Email seems to have about a 20% return rate, phone about 50%, and in person is about 80%. To find the return rate, you need to search on your distribution type.
2) Quality of the survey -- this can be tested through piloting to improve participants' ability to answer the questions. The issue here is that someone will typically stop taking a survey if they cannot understand or cannot answer an item.
3) What's in it for me? - this issue is very simple: participants respond more if there is something in it for them. This can be a reward or even just the chance for a reward (entering a person in a pool to win a prize), or just reinforcing the importance of a topic to the person responding. For example, as a type 1 diabetic, I am much more likely to respond to a survey that targets improving health than I am a survey that targets lumber prices.
I suspect that whoever told you 50% was taking a wild stab in the dark based on his/her personal experiences getting survey responses. You can increase your confidence in getting a good return rate if you do some piloting and some investigating about response rate of different delivery systems.
Good luck!!
Thank you all for your valuable contribution.
Here i want to ask Dr. Smith that can you please give any reference for your point 1. Where i can find its detail?
Here is a sampling of what I was able to find just searching on "Response Rate" and "Surveys." You get better information if you add your own area as a keyword (mine is education).
You should find this reference contains information useful to you:
University of Texas at Austin (2011). Response Rates. Retrieved 2014. Instructional Assessment Resource Database.
You will find it with a Google search.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size