Qs on Final Project

Qs on Final Project

by Anjana Sharma -

Appreciate your humoring some random questions about final project:

1. Do we need to label variables if we know what they mean? For example, if I know that "OM5a" means "A1C value", should I be labeling that for the benefit of the reviewers? Same goes for if I know that 1=yes, 2=no, etc. for an entry?

2. Does table 1 need to have P value to show i can calculate it? i was just going to show the demographics of the two arms of an RCT. Or does it really not matter?

3. Any way we can see the grading criteria for the final project, so I know I'm not missing anything?

Re: Qs on Final Project

by Michael Peters -

Anjana,

1) Yes, you should label your variables as reasonable terms.  The objective for the final project is to create tables that are of "publishable quality"  as such all terms in your table and figures should be terms that are universally accepted in your field.  This also goes for labels of 1 or 2.  A simple code that I use to do this is.

label define var 1 "Yes" 2 "No", replace

label values var var

The first line of this code tells stata that 1 = "Yes" and 2 = "No"

The second line puts that label onto the variable of the same name.

2)  You do not NEED p-values but generally for a Table 1 in an RCT these are included.  If you feel that it is not needed, that is ok (we are going to be pretty easy here when it comes to grading), but I think it would be better for you if you just included the p-values.

3) As for grading criteria, the simplest answer here is to approach this as you would a submission to a journal or a grant.  Make the figure labels legible and make sure a reviewer will be able to understand the primary points you are trying to convey.