Performing Univariate Tests and Creating a Summary Table in Stata

[Johanna Wald]

Hello Statalist Community,

I am currently working on a project where I need to perform univariate tests for several variables over a specific time period. The variables in question are dacc_noturnover, dacc_cfoturnover, dacc_ceoturnover, and dacc_ceocfoturnover. My goal is to analyze these variables for the years 2017, 2018, and 2019.

Specifically, I am looking to calculate both the median and mean for each of these variables for each year. I would like to present this data in a table format, where each variable is listed with the mean and median values for 2017, 2018, and 2019 side by side.

Could anyone guide me on how to:

1. Perform univariate tests for these variables in Stata, focusing on calculating the median and mean for each year.
2. Create a table in Stata that aligns with the following structure:

Variable	2017		2018		2019
	Mean	Median	Mean	Median	Mean	Median
dacc_noturnover	[X]	[x]	[x]	[x]	[x]	[x]

Any assistance or guidance on how to approach this in Stata would be greatly appreciated.

Thank you in advance for your help!

Best regards

Johanna

My code:

---

scalar minObs= 1
drop if fyear !=2016 & fyear!= 2017 & fyear != 2018 & fyear != 2019
 
tostring sic, replace

drop sic_2
gen sic_2 = substr(sic,1,2)
destring sic_2, replace

summ sic_2
scalar a= 0 /*r(min)*/
scalar b= 99 /*r(max)*/
di sic_2

summ fyear
scalar c= 2017 /*r(min)*/
scalar d= 2019 /*r(max) */

summ dummy_T_CEOCFO
scalar e= 0
scalar f= 1

summ dummy_T_CEO
scalar g = 0
scalar h = 1

summ dummy_T_CFO
scalar i = 0
scalar k = 1

summ No_Turnover
scalar m = 1
scalar n = 1

drop uhat
gen uhat=.

xtset gvkey fyear

/*drop obs
gen obs= [_n]
summ obs
scalar e= r(min)
scalar f= r(max)*/

/*Erstellen der notwendigen Variablen*/
drop tac_scaled
gen tac_scaled= (ib-oancf)/L.at

drop inverse_lagassets
gen inverse_lagassets= 1/L.at

drop lagrevrec
gen lagrevrec= (d.revt - d.rect)/L.at

drop ppeg_scaled
gen ppeg_scaled = cond(ppegt == ., ppent/L.at, ppegt/L.at)

forvalues j = 1/`=_N' {
 if fyear[`j']>= c & fyear[`j']<=d & sic_2[`j'] >= a & sic_2[`j'] <= b  {
   
   capture noisily {
    reg tac_scaled inverse_lagassets lagrevrec ppeg_scaled if sic_2 == sic_2[`j'] & fyear == fyear[`j'] & _n != `j' & No_Turnover == 1, nocons
        di e(N)
        if e(N) >= minObs {
            predict uhat_2 in `j', resid
            replace uhat = uhat_2 in `j'
            drop uhat_2
        }
    }
foreach fyear in 2017 2018 2019 {
    drop dacc_turnover
    gen dacc_turnover = .
    replace dacc_turnover = uhat if dummy_T_CFO == 1 | dummy_T_CEO == 1 | dummy_T_CEOCFO == 1
}
  }
}
drop dacc_noturnover
gen dacc_noturnover = .
replace dacc_noturnover = uhat if No_Turnover == 1

drop dacc_cfoturnover
gen dacc_cfoturnover = .
replace dacc_cfoturnover = uhat if dummy_T_CFO == 1 & dummy_T_CEO == 0 & dummy_T_CEOCFO == 0

drop dacc_ceoturnover
gen dacc_ceoturnover = .
replace dacc_ceoturnover = uhat if dummy_T_CEO == 1 & dummy_T_CFO == 0 & dummy_T_CEOCFO == 0

drop dacc_ceocfoturnover
gen dacc_ceocfoturnover = .
replace dacc_ceocfoturnover = uhat if dummy_T_CEOCFO == 1

 
tabstat dacc_turnover , statistics(median, mean, min, max)
 tabstat dacc_noturnover, statistics(median, mean, min, max)
  tabstat dacc_ceoturnover , statistics(median, mean, min, max)
   tabstat dacc_cfoturnover , statistics(median, mean, min, max)
    tabstat dacc_ceocfoturnover , statistics(median, mean, min, max)