How many subjects required?

Home > Statistics Every Writer Should Know > The Stats Board > Discusssion

How many subjects required?
Message posted by Janice lloyd on November 2, 1999 at 12:00 AM (ET)

I am undertaking a research project that involves asking guide dog users around 300 questions about life with their guide dogs. I have 10 participants. How many will I need in order to analyse results?

READERS RESPOND:
(In chronological order. Most recent at the bottom.)

Re: How many subjects required?
Message posted by jg on November 3, 1999 at 12:00 AM (ET)

You have too many questions and too few participants. First, have a two part questionnair. A dozen or so carefully prepared questions that you will try to analyze statistically and the rest that may give you interesting ideas for future questionnairs but can't be statistically meaningful.

The carefully prepared questionnair should cover key issues, have at least two questions that basically ask the same thing to estimate reliability and at least one question that will tell you if they are lying or very careless in their answers to estimate validity. There are many other factors to consider but this should get you started.

Also questinnairs are not always designed to get valid and reliable answers. The questioner may be simply trying to put ideas in peoples heads - push questionnairs. Or the questioner may wish to add a venier of scientific validity to ideas that have already been decided or that the questioner wants to be true.

Your $5 contribution helps cover part the $500 annual cost of keeping this site online.