Thursday, October 25, 2012

Stats Nerd post

I was poking through some of my files yesterday and found my paper from my advanced statistics course I took in my last semester of college. It was nice to review it, and I got a laugh out of a typo. It was also useful to see what specific statistical tests I had run, as that may come in handy when I discuss my statistical experience while job hunting. (I wonder how many other undergraduate students got to run multivariate analysis of variance. My college had that advanced stats elective course taught by the psychology department).

Not that I'm going to post the whole paper, these are just the specific tests I had run. The # x # refers to the number of groups an independent variable has.

  • The first design was a 2 × 10 × 4 between-subjects MANOVA with 4 dependent variables. The effect of gender by subject by level was analyzed.
  • For the second design I used a 10 × 2 between-subjects ANOVA, analyzing the effect of subject by gender on learned a lot.       
  • For the third design I used a 10 × 2 between-subject ANCOVA, analyzing the effects of subject by gender on learned a lot while controlling for level.   
  • For the fourth design I used a 2 × 10 between-subjects MANCOVA. The effects of gender by subject on learned a lot, satisfied with instruction, recommend instructor, and grades reflect performance was studied while controlling for level was analyzed.
  • For the fifth design I used a one way between subjects ANOVA. I tested the effect of gender on learned a lot. 

I had analyzed course evaluations for my project, looking at levels of classes (100 v 200 ...), gender, and course subject. Frame of reference for the stats. Now back to cover letters, job postings, and such.

No comments: