I am doing a project comparing average GPAs in male and fema

I am doing a project comparing average GPAs in male and female students over the last 5 years. Here is a link to the data I am using: http://studentresearch.ucsd.edu/stats-data/mean-gpa.html

What level of significance should I use to do my hypothesis testing?

Which statistical testing should I use? (z or t) and why should I use that one?

And how do I find the critical value? Thanks!!! :)

Solution

level of significance is of your choice. If you wnat your results to be very precise!Near perfect and less error!

a level of 0.01 will be good!

z- test ! you have to use z-test! because , you are collecting data for all the males and all the females! so, both of them are populations! You are comparing 2 populations!and in 45 different time points, you have the data for mean gpa\'s! and population s.d\'s cna be calculated! you need to know the total no. of males and females! and then you cna standadrize using a 2-sample z-test!

critical value ,you can find using a Normal distribution — inverse cumulative distribution function table!

like here the critical value is = z ( 0.99) = 2.3263 ( for 1 tailed test)

and = z( 0.995) = 2.5758 ( for 2 tailed test , both sides)