Rescaling coefficients and t stats from a regression

[Mike Kraft]

Hi
I have a regression using ivreg2. Is there a simple way (or an option) to rescale all coefficients and t statistics from the regression by multiplying by 100 when reporting the results?
Thanks

[Andrew Musau]

replace outcome= outcome*100

[Mike Kraft]

Dear Andrew Musau
It does not seem to work in my case. The error message is "variable outcome not found". Did you mean to estimate the betas and t stats and then replace each? I was wondering if there is another way to do so such as - rescale - in figures?

[Nick Cox]

For outcome substitute the name of your own outcome, response, target, regressand or dependent variable.

[Mike Kraft]

The problem would be that I want to rescale all numbers in the table and produce the table using say --esttab--

eststo: xi: qui reg profit var1 var2 var3 i.week*i.month, r
outreg2 using Table_profit.xls, append  sym(***,**,*) tstat dec(3) bdec(3) excel cttop(profit)

esttab, stats(N) b(a3) starlevels(*  0.10 ** 0.05 *** 0.010) label nogaps nonote    

Thus it would be less efficient to estimate the model parameters and t stats. and then multiply by 100.

In graphs, for example, one can use rescale(100). How can I do this in regressions?

[Nick Cox]

Mike Kraft I think the answer is still no. I suggest the main point is that just about any command producing statistical results based on the units of one or more original variables could also carry some rescaling option that might be useful. The data arrive as proportions but I want percents, or vice versa; the data arrive as cm but I need mm; the data arrive as inches but I need mm, and so on. In essence Stata puts the responsibility of choosing units of measurement on the user, and I don't think other software is typically different.