The p-value of a two-sided t-test.

[Marco Stefano]

Dear all
I have two group based on dummy variable. I have the mean, standard deviation, minimum and maximum. Please, how can I get the average value for each group for each variable as well as the p-value of a two-sided t-test. Thanks a lot.

[Clyde Schechter]

Run -help tttest- and then look specifically at the instructions for -ttesti-, which is the command that will do this.

[Marco Stefano]

Thanks Clyde.
I did it as you have mentioned. Please , I have two following questions :
1-) I got this table, but I do not know the interpretations for the values below the tables - I mean the p-value and other things.
2-) because I have too many variables , and when i run the code stata give me this message -too many variables specified
r(103);-.

Two-sample t test with equal variances
------------------------------------------------------------------------------
Group | Obs Mean Std. Err. Std. Dev. [95% Conf. Interval]
---------+--------------------------------------------------------------------
0 | 468 .0426858 .0023686 .051241 .0380313 .0473403
1 | 51 .0705214 .0122765 .0876716 .0458633 .0951794
---------+--------------------------------------------------------------------
combined | 519 .0454211 .0024745 .0563724 .0405599 .0502823
---------+--------------------------------------------------------------------
diff | -.0278356 .0082302 -.0440043 -.0116668
------------------------------------------------------------------------------
diff = mean(0) - mean(1) t = -3.3821
Ho: diff = 0 degrees of freedom = 517

Ha: diff < 0 Ha: diff != 0 Ha: diff > 0
Pr(T < t) = 0.0004 Pr(|T| > |t|) = 0.0008 Pr(T > t) = 0.9996

[Bruce Weaver]

Hello Marco. I'm a bit confused. You say you have "the mean, standard deviation, minimum and maximum". Do you mean you have the mean and SD for each of two groups? Or do you just have the overall mean & SD?

In order to use the -ttesti- command Clyde directed you to, you need n, mean and SD for each group separately.

    Immediate form of two-sample t test

        ttesti #obs1 #mean1 #sd1 #obs2 #mean2 #sd2 [, options2]

HTH.

[Bruce Weaver]

2-) because I have too many variables , and when i run the code stata give me this message -too many variables specified
r(103);-.

Marco, please review the FAQ, and pay attention especially to item 12.

[Marco Stefano]

Hello Bruce. Thank you.
Please, I am not professional in stata, so please let me explain to you and Clyde What I mean and What I did.
Well,I have dummy variable equal 1 Financial firms and 0 Non financial firms. I have the "the mean, standard deviation, minimum and maximum" on the whole sample by using sim Varlist.
I need to have the "the mean, standard deviation, minimum and maximum" for each variable for each group and my advisor said the most important thing is to get the p-value of a two-sided t-test. So, I do not know how to do that in stata. Thanks a lot.

[Bruce Weaver]

Hi Marco. Unfortunately, I am still confused. See item 12.2 in the FAQ, and then use -dataex- to share part of your data file so that we can see what you are working with. And use code delimiters (described in item 12.3) to show us the code that you used to generate the output shown in post #3. Thanks.

[Marco Stefano]

Hi Bruce. Much appreciated. Thanks. I am attaching an example of my dateset and the Code that I have used in post #3.
input long FirmID int Year float GroupDum double(X1 X2 X3 X4)
16834 2019 1 .00187057616284 .03227519259037 . .
28881 2019 1 .01879157025818 .24420714913271 .08256095890757001 .
47718 2019 1 .004117736957 .07457386438763 .00152085132479 .06587418795537
44868 2019 1 .00206467586212 .0371636441709 . .00024022278439
19945 2019 1 .00543650574791 .3632171757673 .00152085132479 .00157275263716
43997 2019 1 .01133497425939 .1485511346097 .00861740721532 .
95684 2019 0 .00531338568905 .08329652934095 .01076756420894 .013895406372050001
30185 2019 0 .00760396995359 .12694083462574 .0048863140689 .009659249274450001
43311 2019 0 .00460111177885 .06538564184681 .004 .01273170731707
93365 2019 0 .0038733683775299997 .06452755556867 .00637488439329 .0124602662646
95327 2019 0 .00620271379122 .09930199086515 .00567095530305 .010790814636199998
43411 2019 0 .008620147571359999 .12674961884621 .00464114012391 .01930897129284
95358 2019 0 .00880273120363 .12493727683278999 .0062048617691 .008545996252080001
44316 2019 0 .00959606102855 .07505896135285 .01204197047347 .02151867987044
43887 2019 0 .010479624700479999 .07478326311673 .012166751430210001 .01999058994054
45107 2019 0 .01074533090408 .07396482914934001 .01422433450121 .02687423173089
95392 2019 0 .020411725999439998 .15811246997683 .05822342464209 .05878705895615
32560 2019 0 .00791720482413 .05360590573538 .03614740172242 .037201945046580004
19556 2019 1 .00971518215784 .03911845750316 .07802252250294 .14384789124989
95435 2019 1 .00496310616646 .052071462090229996 .05374644056271 .15551872157763
23598 2019 1 .00719372453085 .07868924918168001 .03617625275387 .05098395209962
95264 2019 1 -.01678346515395 -.15704737966964 .08062779760892999 .21120665232572
31687 2019 0 .023223787167449998 .14720761829183998 .03812014582343 .07592328419718
------------------.
ttest X1, by(GroupDum)

[Clyde Schechter]

OK, so the output of that command applied to the example data is:

. ttest X1, by(GroupDum)

Two-sample t test with equal variances
------------------------------------------------------------------------------
   Group |     Obs        Mean    Std. err.   Std. dev.   [95% conf. interval]
---------+--------------------------------------------------------------------
       0 |      13    .0097993    .0016016    .0057748    .0063096     .013289
       1 |      10    .0048705    .0028889    .0091355   -.0016647    .0114056
---------+--------------------------------------------------------------------
Combined |      23    .0076563    .0015958    .0076533    .0043468    .0109659
---------+--------------------------------------------------------------------
    diff |            .0049289    .0031144               -.0015479    .0114056
------------------------------------------------------------------------------
    diff = mean(0) - mean(1)                                      t =   1.5826
H0: diff = 0                                     Degrees of freedom =       21

    Ha: diff < 0                 Ha: diff != 0                 Ha: diff > 0
 Pr(T < t) = 0.9358         Pr(|T| > |t|) = 0.1285          Pr(T > t) = 0.0642

The output is actually pretty self explanatory. But for guidance, I have put the two group means in red, and the two-tailed t-test of group mean equality in blue. The standard errors in each group are to the right of the group means, and the standard deviations to the right of those.

To get the minimum and maximum of the outcome in each group, which are not part of the -ttest- output, you can do this:

by GroupDum, sort: summ X1

[Marco Stefano]

Thanks a lot Clyde. I have got it. Thanks