[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
RE: st: RE: Changing -correlate- results
I share your view, Nick. Thank you all.
From: "Nick Cox" <email@example.com>
Well, yes it is. Here, and almost always
elsewhere, there is presumably no substantive
difference, but I'd argue that equal values
should get equal ranks.
> I've tried both the -egen- and -sort, stable- approaches;
> each gives me
> identical results after each execution of the do file. With
> -egen- the
> correlation coefficient is always 0.7498, and with -sort,
> stable-, the
> result is always 0.7608. I suppose the difference in the two
> results lies in how tied ranks are treated.
* For searches and help try: