[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: RE: Pearson correlation with check
You can get a matrix of correlations using -correlate-
or with -pwcorr-. Given many missing data, the results
might be very different.
You can get a printout of significance levels using -pwcorr-.
Why you failed to find these commands, which are only a
-search- away from you, is not clear.
Getting significance results that mean anything much is quite
another matter. As panel data, these data will not be mutually
independent, so the standard P-values associated with Pearson
correlations do not apply. (The same issue would arise with any
other kind of correlation.) I doubt that you can calculate
meaningful P-values without some explicit model for the data
generating process. That's probably irrelevant given the large
notional "sample size".
> I need an help: I have a panel dataset of 1000 firms over ten
> years (1980-1990), a lot of variables (but the panel is
> incomplete, because some data are missing), and I must
> realize a table of Pearson correlations among measures of
> var1, var2, var3.
> The following should be the plot:
> var1 var2 var3
> Further, for each value I need to check if every correlation
> is significantly different from zero at the 0.05 level or
> higher. I don't know how to do the table and to check it...
> Any help will be greatly appreciated.
* For searches and help try: