I would like to be able to display the number of non-missing values for each combination of factor variable and group allocation in the body of a descriptive table. I'm most of the way to creating the table I'm after and need help with the last steps. Alternatively, there may be a better, maybe even obvious, approach that I haven't seen.
Firstly, I'm creating a dataset containing some missing data.
The tables below show that overall n is 16 and 21 for the Intervention and Control group, respectively. When counting for the cases where blinding is not missing, n is 14 and 20.
Note that I'm restricting to day=0 for my example but ultimately will be doing this for multiple time points (and also multiple variables).
There doesn't seem to be an easy way to count the number of non-missing observations for factor variables, but I can use the count statistic for continuous variables. Here I'm generating a copy of my variable "binding" that I will treat as continous.
Here's my first attempt and the resulting table. The table is close to what I want, but I would like to indent the categories of my factor variable.
Here's my second attempt. The categories are indented but I have an extra row that I don't want and the (n=#)s are on the wrong row - I want them on the vargrp header row.
How can I create a table with this structure?
Kind regards,
Suzanna
Firstly, I'm creating a dataset containing some missing data.
Code:
*Reading in the dataset.
use https://www.stata-press.com/data/r18/exercise.dta,clear
keep if inlist(day, 0, 4, 6, 8)
*Replacing the variable program with group_allocation, where 1=intervention and 2=control.
gen group_allocation=program
lab var group_allocation "Group allocation"
lab def grplbl 1 Intervention 2 Control
lab val group_allocation grplbl
drop program
gen blinding=2
replace blinding=3 if _n/3==int(_n/3)
bysort day: replace blinding=1 if _n/4==int(_n/4) & day!=0
lab def blindlbl 1 "Change or lack of change from baseline" ///
2 "Participant informed assessor" ///
3 "Guess"
lab val blinding blindlbl
*Setting a few values of the variable blinding to missing.
replace blinding=. if inlist(id,1,2,18)
Note that I'm restricting to day=0 for my example but ultimately will be doing this for multiple time points (and also multiple variables).
Code:
. tabulate group_allocation if day==0
Group |
allocation | Freq. Percent Cum.
-------------+-----------------------------------
Intervention | 16 43.24 43.24
Control | 21 56.76 100.00
-------------+-----------------------------------
Total | 37 100.00
. tabulate blinding group_allocation if day==0
| Group allocation
blinding | Intervent Control | Total
----------------------+----------------------+----------
Participant informed | 9 14 | 23
Guess | 5 6 | 11
----------------------+----------------------+----------
Total | 14 20 | 34
Code:
*Generating a copy of the variable blinding, which will be treated as continuous
*in order to count the number of nonmissing values.
gen blinding_c=blinding
Here's my first attempt and the resulting table. The table is close to what I want, but I would like to indent the categories of my factor variable.
Code:
*Clearing any collections that may be in Stata's memory.
collect clear
*Creating a separate descriptive table for each level of time.
*For day=0, including the sample frequency in the column header.
dtable blinding_c i.blinding if day==0, nolistwise ///
by(group_allocation,nototal missing) ///
sample(, statistic(frequency) place(seplabels)) ///
sformat("n=%s" frequency) ///
continuous(, statistic(count)) ///
title(Table 5: Blinding of assessor) ///
titlestyles(font(Calibri, size(12) bold)) ///
name(Day0)
collect style cell result[count], sformat((n=%s))
collect style header blinding, title(hide)
collect label levels var blinding_c "Blinding",modify
collect style cell result, halign(center)
collect preview
Table 5: Blinding of assessor
-----------------------------------------------------
Group allocation
Intervention Control
n=16 n=21
-----------------------------------------------------
Blinding (n=14) (n=20)
Participant informed assessor 9 (64.3%) 14 (70.0%)
Guess 5 (35.7%) 6 (30.0%)
-----------------------------------------------------
Here's my second attempt. The categories are indented but I have an extra row that I don't want and the (n=#)s are on the wrong row - I want them on the vargrp header row.
Code:
collect addtags vargrp[Blinding], fortags(var[blinding_c blinding])
collect layout (vargrp#var) (group_allocation#result)
Table 5: Blinding of assessor
-------------------------------------------------------
Group allocation
Intervention Control
-------------------------------------------------------
Blinding
Blinding (n=14) (n=20)
Participant informed assessor 9 (64.3%) 14 (70.0%)
Guess 5 (35.7%) 6 (30.0%)
-------------------------------------------------------
How can I create a table with this structure?
Code:
Table 5: Blinding of assessor
-------------------------------------------------------
Group allocation
Intervention Control
-------------------------------------------------------
Blinding (n=14) (n=20)
Participant informed assessor 9 (64.3%) 14 (70.0%)
Guess 5 (35.7%) 6 (30.0%)
-------------------------------------------------------
Suzanna
