Numbers 1.4 - first t-test analysis ever!

Here is my published shot on screencast and here is my image.  You can also access my document on my Google Drive.

For this exercise, my null hypothesis would be: There is no relationship between ethnicity and scores on the 2013 National Assessment of Educational Progress.  

My research hypothesis (or test hypothesis) would be:  There is a relationship between ethnicity and scores on the 2013 National Assessment of Educational Progress.

I ran a t-test for all of the scores, comparing White students' scores with Hispanic students' scores, then White students' scores with Black students' scores, and finally, Black students' scores with Hispanic students' scores.  I used the Data Analysis Toolpak add on in Excel and computed t-tests with the assumption that there were equal variances.  Each P value was 0.00 which indicates that these score differences are not simply due to chance.  The null hypothesis can be rejected.  There clearly is a relationship between ethnicity and scores on the 2013 NAEP.

If I were to represent this data in an academic paper using APA style I would first have to find and reference the data itself.  I believe the National Center for Education Statistics web page would be an excellent source as it allows you opportunities to select and access the exact data you wish to focus on as you can see here.


I would also have to include some references to P values, how they are calculated, what they mean, and how they are used in research.  Having just read two books by Neil J. Salkind that touched on this topic, I would probably reference at least one of those or look for others to explain this topic.

I would also do some additional research to support my findings, to see how results on this test compare to other similar tests and what experts believe causes these ethnic differences as well as what educators and communities are doing in an effort to limit these differences.