Descriptive Statistic for Categorical Variable

[Caleb Hall-Paterson]

Hi all,

I am trying to create a table of descriptive statistics for my explanatory variables, one of which is Industry Classification. I want the summary statistic to break it down by industry but my current code, as shown below, is failing.

sum BICSLevel1_n=="Communications" BICSLevel1_n="Consumer Discretionary" BICSLevel1_n="Consumer Staples" BICSLevel1_n="Energy "BICSLevel1_n="Financials" BICSLevel1_n="Health Care" BICSLevel1_n="Industrials" BICSLevel1_n="Materials" BICSLevel1_n="Technology" BICSLevel1_n="Utilities" LoanTrancheSizeBM LoanTrancheSizeAM OriginalMaturityBM OriginalMaturityAM AverageLifeBM AverageLifeAM CouponBM CouponAM InterestAccruedBM InterestAccruedAM CreditSpreadBM CreditSpreadAM FlexStatus_n=="No Flex" FlexStatus_n=="Flexed Up" FlexStatus_n=="Flexed Down" CallProvision_n="No Call" CallProvision_n="Soft Call" CallProvision_n="Hard Call" InterestFrequency_n=="MONTHLY" InterestFrequency_n=="QUARTERLY" Guaranteed_n=="No" Guaranteed_n=="Yes" SponsorLed_n=="No" SponsorLed_n=="Yes" Amended_n=="No" Amended_n=="Yes" OriginalIssueDiscountSecurity_n=="No" OriginalIssueDiscountSecurity_n=="Yes"

I also want the summary to cover mean, median, standard deviation and number of observations. The traditional

sum

does not seem to include median, is there an option somewhere to change this? Some guidance for both these issues would be much appreciated.

Thank you in advance,
Caleb

[Carlo Lazzaro]

Caleb:
do you mean something along the following lines?

use "https://www.stata-press.com/data/r16/nlswork.dta"

. bysort union: tabstat tenure, stat(count mean sd p50 min max)

---------------------------------------------------------------------------------------------------------------------------------------
-> union = 0

    variable |         N      mean        sd       p50       min       max
-------------+------------------------------------------------------------
      tenure |     14539  3.602758  3.872048      2.25         0     24.75
--------------------------------------------------------------------------

---------------------------------------------------------------------------------------------------------------------------------------
-> union = 1

    variable |         N      mean        sd       p50       min       max
-------------+------------------------------------------------------------
      tenure |      4471  5.160982  4.661994  3.583333         0  25.91667
--------------------------------------------------------------------------

---------------------------------------------------------------------------------------------------------------------------------------
-> union = .

    variable |         N      mean        sd       p50       min       max
-------------+------------------------------------------------------------
      tenure |      9091   1.35603  1.817795       .75         0      20.5
--------------------------------------------------------------------------

.

[Caleb Hall-Paterson]

Hi Carlo,

Thank you, the above code you have given does in fact work well. However, I want to export the data to Microsoft Word and using the following code;

bysort BICSLevel1_n: tabstat AmtRecovered, stat(mean sd count)
outreg2 using myreg.doc, replace ctitle(Model 1)

I receive this error message - matrix e(b) not found; run/post a regression, or specify varlist for non-regression outputs

Any advice as to what to do from here?

Kind regards,
Caleb

[Marcos Almeida]

Try - putdocx - instead.

[Caleb Hall-Paterson]

Hi Marcos,

I tried - putdocx - but I am still getting no success. For clarification, where should it come in the code?

Thanks,
Caleb

[Marcos Almeida]

Maybe this text of the Stata Manual will be helpful.