The Student Room Group
Competition - post a photo you've taken of nature >>
Competition - post a photo you've taken of nature >>
Competition - post a photo you've taken of nature >>

What type of statistical test to use ?

Hey i am a currently working on my university dissertation and am not strong when it comes to statistics, i would just like to ask whether i am using the correct statistical tests to analysis data in my experiment.

My experiment is looking at the effects of music ( violent vs non violent ) on stress scores. Participants will compete a stress questionnaire , listen to either a violent or non violent piece of music and then complete the questionnaire again.

So to analysis this data i will use 2 paired samples t tests ? (One looking at the difference in the two test scores in the violent music condition and one looking at the difference in the two test scores in the non violent music condition).

The second part of my experiment involves the participants completing a cognitive face recognition task designed to measure aggressive cognition. For this i will carry out an independent samples t test to look at the difference in aggression scores across the two groups of participants?

Thank You in advance
Andrew
Reply 1
Avatar for Jam'
Jam'
15
What numerical values for the second test will you collect?
That seems about right

As a general rule, t-test when comparing means.
Original post by andrew1992
Hey i am a currently working on my university dissertation and am not strong when it comes to statistics, i would just like to ask whether i am using the correct statistical tests to analysis data in my experiment.

My experiment is looking at the effects of music ( violent vs non violent ) on stress scores. Participants will compete a stress questionnaire , listen to either a violent or non violent piece of music and then complete the questionnaire again.

So to analysis this data i will use 2 paired samples t tests ? (One looking at the difference in the two test scores in the violent music condition and one looking at the difference in the two test scores in the non violent music condition).

For the first part, I think it would be better to use a 2 x 2 independent groups ANOVA, rather than t-tests.

Your two factors would be 1) Music, which has two levels - violent and non violent, and 2) Exposure, which has two levels - before and after. Then once you've run your analysis, you would record main effects and interaction effects, saying whether they're significant or not.

That's how I'd do it anyway, but if I were you I'd double check with your project supervisor if you can.

You need to be more clear about what you're doing for the second part.
(edited 11 years ago)

Quick Reply

Related discussions

Latest

Trending

Trending

Articles for you