On Wed, Oct 31, 2012 at 5:17 PM, Anat (Manes) Tchetchik
<anatmanes@gmail.com> wrote:
> Jay
> one last question, I now realize that the dependent caterogical var.
> I'm using, "Tier", is distributed as follows: (with only four cases
> having the value 4).
> Shell I drop these cases?
>
> Tier | Freq. Percent Cum.
> ------------+-----------------------------------
> 1 | 56 37.84 37.84
> 2 | 58 39.19 77.03
> 3 | 30 20.27 97.30
> 4 | 4 2.70 100.00
> ------------+-----------------------------------
> Total | 148 100.00
I'd merge them into 3. It's situations like this category for which
the tests of threshold are most important. Chances are good that this
one wouldn't be much distinguished from that of 3 anyway.
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/
