What is a general linear model?

noetsi

noetsi

Fortran must die
#1
#1
Not sure of the terminology here although lots of dummy variables were used as predictors. This comes from a disertation I am reading.

"To explore the relationship between personal characteristics and employment outcome rates, a multiple regression was used. Since many of the personal characteristic variables were coded categorically, a general linear model was used to run the multiple regressions."

Is this just another way of saying multiple linear regression with dummy predictors?
 
K

katxt

Active Member
#2
#2
A simple view is a combination of regression (continuous predictors) and anova (categorical predictors) although the internal workings use just what you have said - regression with dummy predictors for the categories.
 
F

fed2

Active Member
#5
#5
Statistics terminology drives me crazy. Different people use differ terms for the same thing.
Click to expand...
I feel like there should be some sort of convention for naming stats procedures, like IUPAC does for chemicals.


I think multiple regression is really just a subset of general linear models, but software often gives different procedures for the two, which is where this language is coming from.
 
Dason

Dason

Ambassador to the humans
#7
#7
A general linear model is like a model but it's also linear so it's like a linear model but it's also like general so it's like a general linear model. You know?

But just because it's a general linear model doesn't mean the predictions will fall on a line. You know?

As if.

Totally.

Rolling with my homies.

Sorry - we put on Clueless a while ago.
 
noetsi

noetsi

Fortran must die
#8
#8
lol dason its too early to be drunk :p

I think what raised my concern is I have been reading about models like general additive models recently and I had not heard linear regression called general linear models before. I wondered if they had different assumptions.
 
Dason

Dason

Ambassador to the humans
#9
#9
noetsi said:
lol dason its too early to be drunk :p
Click to expand...
Lies. If you're gonna drink then drink early and end early. Drink lots of water after. You avoid the hangover that way. It's not like anybody should be going to bars now.

Plus it sounds like you've never watched Clueless. You're clueless about Clueless.

I think what raised my concern is I have been reading about models like general additive models recently and I had not heard linear regression called general linear models before. I wondered if they had different assumptions.
Click to expand...
General additive models are different but if all anybody says about a model is that it's a "linear regression" then it almost surely falls under the umbrella of general linear model.
 
noetsi

noetsi

Fortran must die
#10
#10
I watched clueless once. It seemed pretty clueless. I am not sure what the point was, unless having lots of attractive young people is the point.

Although I don't drink I know there are drinks, vodka is one I think, that do not generate traditional hangovers. They hurt you other ways. I have never figured out the value of getting drunk, even assuming I could afford to do so (and of course my faith teaches not to get drunk).

I hope you have a peaceful holidays Dason.
 
You must log in or register to reply here.