T-test for the difference in the means of two subgroups in my data

[FitzGerald Blindman]

Hi everyone, I would like for your help.

I have cross-sectional data which divided into treatment group and control group. In each group there are men and women. Now I want to check out that the groups are not different from each other in any parameter except to the treatment. So I want to calculate the means and to do t-test to see in they are statistically difference from each other, and than export the results to Excel and present them with asterisks. I attach screenshot from Lavy & Nussbaum (2023) for an example:

So I know of curse that for means of variables I can write:

sum var1 if control == 1
sum var1 if control == 0

And so on, but I really need your help in understand how to do t-test on the means of the difference groups and how to export the results to Excel.

I thank you already,

Fitzgerald.

[Andrew Musau]

You can use the mean command to obtain standard errors for the means. The following uses estout from SSC.

sysuse auto, clear
eststo domestic: mean mpg weight turn displacement if !foreign
eststo foreign: mean mpg weight turn displacement if foreign
eststo diff: quietly estpost ttest mpg weight turn displacement, by(foreign) unequal

esttab domestic foreign diff, label collab(none) ///
cells("b(pattern(1 1 0) fmt(2) label(Mean)) b(pattern(0 0 1) fmt(2) label(Diff.) star)" ///
se(par pattern(1 1 0) fmt(2)))  title("T-test of comparisons of means") ///
nonumbers mlab("Domestic" "Foreign" "Difference", lhs("Variables")) ///
starlevels(* 0.1 ** 0.05 *** 0.01) varwidth(25) ///
addnote("Standard errors in parentheses." "* p<0.1, ** p<0.05, *** p<0.01")

Res.:

. esttab domestic foreign diff, label collab(none) ///
> cells("b(pattern(1 1 0) fmt(2) label(Mean)) b(pattern(0 0 1) fmt(2) label(Diff.) star)" ///
> se(par pattern(1 1 0) fmt(2)))  title("T-test of comparisons of means") ///
> nonumbers mlab("Domestic" "Foreign" "Difference", lhs("Variables")) ///
> starlevels(* 0.1 ** 0.05 *** 0.01) varwidth(25) ///
> addnote("Standard errors in parentheses." "* p<0.1, ** p<0.05, *** p<0.01")

T-test of comparisons of means
-------------------------------------------------------------------
Variables                     Domestic      Foreign   Difference   
-------------------------------------------------------------------
Mileage (mpg)                    19.83        24.77        -4.95***
                                (0.66)       (1.41)                
Weight (lbs.)                  3317.12      2315.91      1001.21***
                               (96.43)      (92.32)                
Turn circle (ft.)                41.44        35.41         6.03***
                                (0.55)       (0.32)                
Displacement (cu. in.)          233.71       111.23       122.48***
                               (11.82)       (5.30)                
-------------------------------------------------------------------
Observations                        52           22           74   
-------------------------------------------------------------------
Standard errors in parentheses.
* p<0.1, ** p<0.05, *** p<0.01

Export in CSV format and then open this in Excel.

[Andrew Musau]

Also note that dtable was designed to make such descriptive statistics easy to obtain. Using the example in #2:

sysuse auto, clear
dtable mpg weight turn disp, by(foreign, tests nototal) nformat(%3.0f N)

Res.:

. dtable mpg weight turn disp, by(foreign, tests nototal) nformat(%3.0f N)
note: using test regress across levels of foreign for mpg, weight, turn, and displacement.

---------------------------------------------------------------------
                                         Car origin                  
                             Domestic            Foreign        Test 
---------------------------------------------------------------------
N                               52 (70.3%)          22 (29.7%)       
Mileage (mpg)               19.827 (4.743)      24.773 (6.611) <0.001
Weight (lbs.)          3,317.115 (695.364) 2,315.909 (433.003) <0.001
Turn circle (ft.)           41.442 (3.968)      35.409 (1.501) <0.001
Displacement (cu. in.)    233.712 (85.263)    111.227 (24.881) <0.001
---------------------------------------------------------------------

You could change the layout of the table either using available options or using collect.

help dtable
help collect