Regression Assumptions

Monday nights is my stats/research methods class. 

It's pretty much a doozy. 

Every time I go to this class I start to get irritated, because last I checked, I'm not a math major and by the end of class my notes look like this:

  What the heck?! 

Every Monday night I ask myself, why is this important for a degree in public administration??...don't answer that, I know the real answer, but that's not the answer, nor the sympathy I'm looking for.  

BUT on the bright side of things, we totally learned something fun today:

HOMOSKEDASTICITY

I know right?! My spell check was like, "whoa what the effff??"

Now, sitting in a statistics/research methods class for 3 hours is terribly long, boring, confusing, and...confusing. So you can imagine my delight when our professor busted out this bad boy.

 I learned that Homoskedasticity occurs when all random variables in a sequence or vector have the same finite variance, or when the variance of error term is equal to all observations, blah blah blah... which isn't nearly as fun and exciting as just saying the word!

 But wait! It gets better!

HOMOSKEDASTICITY has a counterpart.

 wait for it...

HETEROSKEDASTICITY

Which would be the opposite, or absence of homoskedasticity...don't ask me what that looks like or implies, I haven't really thought about it yet...homework isn't due until next week and I'm having too much fun just saying it the word. 

See, regression assumptions have a sexual preference too!