using Chi-Square for two samples
Message posted by Sherry (via 142.165.107.205) on November 20, 2001 at 8:26 PM (ET)

I have 2 samples of data (totally independent samples). Sample 2 data is from the literature so I dont have much more information for that sample.

for both data sets, I have the percentage of times something occurs (e.g., seeing someone getting killed) in a person's lifetime. I want to compare sample 1 and sample 2 to see if their percentages are significantly different. i dont have means and stand dev for Sample 2 so I can't do T-tests.

e.g., Sample 1 => 16 out of 30 saw someone getting killed
for sample 2 => 18 out of 80 saw someone getting killed.

How do I do a Chi-Square test to test if they are different?
That is what my advisor wants to use.

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

Re: using Chi-Square for two samples
Message posted by Darius (via 200.23.217.10) on November 21, 2001 at 4:42 PM (ET)

Does the proportion of defectives meet requirements?

read

http://www.itl.nist.gov/div898/handbook/prc/section2/prc24.htm

Re: using Chi-Square for two samples
Message posted by Phil (via 165.247.228.113) on November 24, 2001 at 10:54 AM (ET)