T-test and ANOVA help

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

michael1986
Posts: 1
Joined: Mon Mar 07, 2016 8:46 pm

T-test and ANOVA help

Postby michael1986 » Mon Mar 07, 2016 9:05 pm

Hi,

I would just like some confirmation on the tests that I plan to run on my data.

I am looking to see if gender affects football game attendance (4 groupings e.g. one game, under 5 games, every 2-3 games, every game), I plan to run independent samples t-test.

I then want to find if there game attendance effects purchase intentions (measured on Likert scales), for this I plan to run ANOVA.

Am I on the right path with these tests?

Thank you very much for your help.
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: T-test and ANOVA help

Postby GerineL » Tue Mar 08, 2016 8:23 am

I am looking to see if gender affects football game attendance (4 groupings e.g. one game, under 5 games, every 2-3 games, every game), I plan to run independent samples t-test.
it sounds like your dependent variable is ordinal level and your independent variable is binary (nominal level).
That means you should use the nonparametric variant of the t-test (mann whitney u test) instead).
I then want to find if there game attendance effects purchase intentions (measured on Likert scales), for this I plan to run ANOVA.
here it depends a bit: Is your likert scale transformed to an actual scale? I assume it is.
Then you have:
ordinal -> scale, which is usually analyzed in a regression type of model with dummy variables.

The problem is that an ordinal variable is neither nominal (i.e., you do not randomly assign numbers to the variable: someone who scores 4 visits the game more often than someone who scores 2), nor scale (the difference between 1 and 2 is not the same as the difference between 3 and 4).

Who is online

Users browsing this forum: No registered users and 2 guests

cron