I got it.
And I raised my hand and said (hee hee!), "So if you want to explore just the questions that interest you, choose your own contrasts and deal with the overlap. But if you want to explain all the variance, choose an orthogonal set of contrasts and deal with the complex contrasts."
And the prof smiled and gave me a big thumbs up and went on with the lecture.
And I sat back in my seat with a huge satisfied grin on my face because - FINALLY! - something about some of the choices we might make about what statistical procedures to use when made sense to me.
Awesome. And by "awesome" I mean, "Mwahahahahahahahaaa!" :) :)
5 comments:
So would you recommend principal component factor extraction, or principal axis?
Don't forget that your decision will affect the percentage of total variance explained by the results.
The NY State 8th grade math test was tridimensional last year, but all three factors only explain one-third of the variance. Oops.
我完全看不懂.
Amy: I couldn't agree more.
Jeremy: Um, what are you talking about??
You know, I've taken a good deal of statistics, and I have no idea what you guys are talking about. I mean, I am familiar (in a vague sense) with factor analysis/principal components analysis, but what's all this about contrasts? Please explain (feel free to do so via email).
Amy, on the other hand, is absolutely right.
smarty-pants.
and I wish I could understand what Amy said. but I can't.
Post a Comment