1. ## Nominal vs. Ordinal

Say i'm using age groups which are not equal in size and interval (1-2,2-4,5-8years)
is it better when using a regression model to set this variable as nominal or ordinal
On one side, It's obviuos why it should be ordinal
But on the other hand, lets assume only the second age group is a risk factor, would i detect it also if it would be ordinal
Thanks,
Aviv

2. ## Re: Nominal vs. Ordinal

Any reason you can't break them into evenly spaced intervals?

3. ## Re: Nominal vs. Ordinal

Yes, that was just an example, in this case:
1. theres a different characteristics relvant to each of this groups in the model
Another example i have is:
2. when im dividing into three groups to create equal in size groups (bt the intervals are not equal)

4. ## Re: Nominal vs. Ordinal

Given your first post, if you can regroup I would say treat them as nominal and use the biggest or smallest group as the reference.

I would imagine there could be a way to treat them as ordinal and then weight them accordingly, but I have never done anything like that.

Thanks.

 Tweet

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts