Douglas Rugh (D.Rugh) wrote:
I have a question about the levels of measurement in a multi-level regression model. The sampling frame is a representative cross section of students throughout the United States at each grade level. The three selection stages are: (1) geographic areas or primary sampling units (PSUs), (2) schools within PSUs, and (3) students within sampled schools.
As for my model, I have an individual level measure of alcohol consumption which is ordinal.
ordinal = each level is ranked, but the increment between ranks is not constant. You cannot use this as a response variable in multiple regression unless you convert it to an interval scale (increment between levels is constant). I'll bet this would not be too hard. (famous naive words :)
I also have a school level measure of alcohol consumption, which is an aggregate of the individuals within a school--also ordinal.
Same situation here.
Then I have a Primary Sampling Unit variable on a scale from 1 to 72. This third level measure is the one I am confused about. It is not an ordinal measure, it is a nominal measure.
nominal = there is no relationship between each 'level' - more like separate categories actually. The word 'scale' is forced a bit, when you have a nominal scale variable.
You have a variable with 72 separate categories? are you sure that the group marked 64 is not simply a little more of the same thing as the group marked 62?
How do I handle that in the regression? It is unreasonable to dummy code this variable.
If you can't 'dummy code' this one, then it really is a nominal variable. My first response is to suggest that you consider the 72 as separate 'groups' and do some kind of one-way AoV, with 72 separate groups. Ouch! You would still need an interval or ratio scale variable to measure, that would be analyzed via the AoV.
If you are missing that, then I think the next fallback position would be some kind of chi-sq analysis of count/frequency for the 72 groups.
Nah, that doesn't tell you much in any case. But it is something.
So what else can be done with nominal & ordinal data?
Cheers & good luck, Jay
-- Jay Warner Principal Scientist Warner Consulting, Inc. 4444 North Green Bay Road Racine, WI 53404-1216 USA
Ph: (262) 634-9100 FAX: (262) 681-1133 email: [EMAIL PROTECTED] web: http://www.a2q.com
The A2Q Method (tm) -- What do you want to improve today?
. . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
