Non-parametric test? Or something else?

[Ondin]

Hi there, I was wondering about something. Here's my problem:

If I have three choices and they're chosen as follows:
Choice 1: 40%
Choice 2: 30%
Choice 3: 30%

Number of people who chose: 1.000


Then I have the sam three choices at some other time and they're chosen as follows:
Choice 1: 38%
Choice 2: 33%
Choice 3: 29%

Number of people who chose: 1.500


Isn't there a non parametric test I can use to test the hypothesis that those percentages are the same between those two observations?

[JohnM]

Chi-square goodness of fit test.

[Ondin]

Al right, thanks alot, I'll try it when I get home. Your help is much appreciated.

Regards,
Ondin.