Home > Statistics Every Writer Should Know > The Stats Board > Discusssion
Testing the differences between 2 means
Message posted by Shaun (via 193.132.229.110) on December 11, 2001 at 11:40 AM (ET)
Hi
This is my problem.
300 people were asked their strength of agreement to the statement 'I love watching TV'.
The respondents answered on a ten point scale.
The results were
Sample 1 Sample 2
0 - 0 0 - 0
1 - 0 1 - 0
2 - 0 2 - 0
3 - 0 3 - 3
4 - 2 4 - 2
5 - 6 5 - 4
6 - 16 6 - 19
7 - 22 7 - 31
8 - 32 8 - 64
9 - 67 9 - 91
10 - 80 10 - 36
Mean1 = 8.65 Mean2 = 8.24
I am familiar with t and z tests for comparing means but these tests assume the distributions are Normal. The data I have is skewed heavily, it is not Normally distributed.
Can anyone tell me if these tests still apply and also is it possible to calculate confidence limits?
If the answer is very technical can someone point me to the topic which covers this.
thank you
READERS RESPOND:
(In chronological order. Most recent at the bottom.)
Re: Testing the differences between 2 means
Message posted by Tomi (via 154.32.142.114) on December 11, 2001 at 6:39 PM (ET)
Your sample sizes are large enough to make the t test OK.
Your $5 contribution helps cover part the $500 annual cost of keeping this site online.