Uclue has now closed (more details). So long and thanks so much for everything.
Ask a Question | Browse Questions

EXPIRED

Question: Sample size for multiple choice survey

Please carefully read the Disclaimer and Terms & conditionsT&Cs.
Priced at $50.00

Actions: Add Comment

dankirk 

Customer

 13 Apr 2017 12:39 UTCThu 13 Apr 2017 - 12:39 pm UTC 

I have a question about minimum sample sizes for a multiple choice survey.

We plan to run a survey with a multiple choice question, where respondents are allowed to choose only one choice. Here's a hypothetical but representative survey question:

Which feature do you want us to build next:
A) Feature 1
B) Feature 2
C) Feature 3
D) Feature 4
E) None of the above

How does one determine how many responses are necessary to achieve 85% confidence? (This is a business question, I'm happy with a lower level of confidence than a medical test).

How would sample size change if we assume that likely distribution of responses is something like this:
A) 0%
B) 10%
C) 3%
D) 0%#
E) 87%

I can find online calculators for sample size, like this:
http://www.raosoft.com/samplesize.html

But I can't figure out how to use it for a multiple choice question. Explaining how to use it for a case like this so I can answer my own question next time would be ideal, if that's possible.

 

David Sarokin 

Researcher

 14 Apr 2017 17:06 UTCFri 14 Apr 2017 - 5:06 pm UTC 

dankirk,

Wish I could help, but I'm just a mite too far removed from the days when I actively used statistics. Hopefully, someone here will have some useful input.

David

 

myoarin 

User

 15 Apr 2017 22:41 UTCSat 15 Apr 2017 - 10:41 pm UTC 

Dankirk,

You certainly bring interesting questions to Uclue.  As usual, my comment is just that, a couple of possibly pertinent remarks. 

Including "None of the above" may provide interesting results, but it is also an invitation for respondents to avoid the decision for one of the other items, inflating that percentage and reducing the others.  You may have reflected that in your second example.  

The hypothetical but representative survey question:  "Which features do you want us to build next?" suggests the problem of phrasing those questions to attract answers to the products or services the organization can provide, probably the ones already being explored, plus perhaps a less specific one that isn't already on the table, but which later could be, if respondents liked it. If they did, that would detract from responses to the other questions, but could suggest future direction better than "none of the above."  

But whom am I suggesting that to?   I'll shut up and hope a statistician can help.

Myo

 

Actions: Add Comment

 

Frequently Asked Questions | Terms & Conditions | Disclaimer | Privacy Policy | Contact Us | Spread the word!

Sat 16 Dec 2017 - 1:14 am UTC - © 2017 Uclue Ltd