Home Student Life & Education
At The Mix, we want to make our services as helpful as we can. To do this, we’d love to ask you a few questions about you, your visit to The Mix and its impact. It should take only about 5-10 minutes to complete. Take this survey and get a chance at winning a £200 Amazon voucher​.
Free Wednesday evening? Come along to our safeguarding co-design session from 6-7:30PM. sign up here

Statistics - t-tests

Former MemberFormer Member Posts: 1,876,324 The Mix Honorary Guru
Ok, I'm stuck on the simplest part of my statistical analysis...

I've done some t-tests in the SPSS statistics programme comparing some ratings between two groups.

I've got 2 sets of ratings for 3 questions, and the difference between the two groups is significant for one of the questions.

However, the difference between the means for the two groups is actually larger for one of the other questions, however it's not saying it's significant. Is this because of the standard deviation or something? I just don't really get how that works out because I thought t-tests literally compared the means between the groups.

Can anyone enlighten me on this? I'm also unsure which bits are important to report in my appendices. Obviously the t-value, df i guess and the sig value but what about SD, standard error mean and then mean diff and standard error diff? aghh so confusing!

Thanks to anyone that helps! :)


  • Former MemberFormer Member Posts: 1,876,324 The Mix Honorary Guru
    t = b-b^/ the square root of the standard deviation. dont think mean matters.
Sign In or Register to comment.