Descriptive statistics for numerical and categorical variables

[Taiba Chau]

Hi, I am trying to create a descriptive statistics table for numerical and categorical variables in Stata. I have read that for categorical variables you would instead report percentage or the proportion of the observations in each group. I am struggling to do this though. An example of what I am trying to replicate is below, which came from a paper

For my project, my categorical variables are mental wellbeing, age, education, employment, marital status. My numerical variables are gender, age and income. At the same time I wanted to split this between ethnic and non-ethnic people. Like in this example they split it by G1, G2, G2.5 and natives. Any help?

[Jared Greathouse]

Please give your example data using the dataex command. For us to provide meaningful feedback, you must provide your example data using the dataex command, the real data from an easily importable source (i.e., Github), or the equivalent of a toy example.
Otherwise, anything we say is simply a waste of time. Note, that I'm not trying to be mean in saying this, I'm saying this because if we can't see your dataset as you do with a minimal worked example, anything we suggest is just guesswork. The reason that I'm emphasizing this is because questions like this one likely have a relatively simple fix, but even simple fixes can be wildly overcomplicated without a minimal worked example of a dataset and code that you've tried to accomplish your task.

So please, provide us with your example data that encapsulates the problem and I'm more than okay with helping you solve this Taiba Chau

[Carlo Lazzaro]

Taiba:
as an aside to Jared's helpful advice, I would point you out to the brand new section on tables included in the Stata.pdf manual, 17th release.

[Jared Greathouse]

Additionally, if the question is just about calculation

sysuse auto, clear
mean foreign

Gives us the percentage of rides that're foreign in this dataset. Note that I'm not at my computer to test the table command, but the way the math works is like any other average. If we have 10 cars and 5 are foreign, we have 5/10 or (1+1+1+1+1)/(10)

[Taiba Chau]

Hi, I have never used the command but I tried doing it and got the following:
clear
input byte(Mental_Health Gender) int Age byte(Marital Educ Employ) float(People_IN_House Inc ethnic)
1 1 58 1 3 5 1 1.72575 0
1 2 42 1 3 1 3 .7030833 0
1 1 23 6 4 7 1 .3068 0
1 1 21 6 3 1 2 .671125 0
1 1 30 6 3 3 1 .95875 0
1 2 18 6 2 1 2 .1534 0
1 2 24 6 3 2 1 1.72575 0
1 2 26 4 2 1 2 .287625 0
1 2 32 1 3 6 3 .57525 0
1 2 44 1 2 1 2 1.534 0
1 2 74 1 2 5 1 2.10925 0
1 1 57 1 3 1 1 1.34225 0
1 1 68 1 3 5 1 .95875 0
1 1 71 1 6 5 1 .95875 1
1 2 23 6 4 1 1 .95875 0
1 2 44 3 5 2 3 .57525 0
1 1 21 6 3 1 1 .95875 1
1 1 32 6 3 1 1 3.068 0
1 1 45 6 3 1 1 1.72575 0
1 1 58 4 3 8 2 .671125 0
1 1 68 1 3 5 1 1.72575 0
1 1 69 1 3 5 2 .671125 0
1 2 20 6 3 1 3 .57525 0
1 2 26 1 3 2 2 .862875 1
1 2 27 6 5 3 1 1.72575 0
1 2 30 6 3 1 1 1.34225 0
1 2 36 5 4 2 2 .287625 0
1 2 37 1 4 10 2 .862875 0
1 2 38 1 3 1 2 1.054625 0
1 2 46 4 2 1 3 .19175 0
1 2 47 1 3 3 2 1.534 0
1 2 53 1 1 1 4 .4314375 1
1 2 68 3 3 5 1 1.34225 0
1 2 69 3 2 5 1 .57525 0
1 1 22 6 3 7 4 .767 1
1 1 23 6 3 1 1 2.10925 0
1 1 30 6 5 7 1 .57525 0
1 1 31 1 3 1 3 .57525 0
1 1 31 6 3 1 1 2.10925 0
1 1 35 1 2 1 6 .51133335 0
1 1 38 1 3 1 2 .671125 0
1 1 40 1 3 1 1 1.72575 0
1 1 41 3 3 3 1 3.068 0
1 1 41 6 3 3 1 1.34225 0
1 1 42 1 3 1 3 .7030833 0
1 1 47 1 5 2 1 3.068 0
1 1 47 1 5 3 1 .95875 1
1 1 50 1 3 3 3 1.5979167 0
1 1 52 1 3 1 2 1.534 0
1 1 54 6 3 1 2 1.054625 0
1 1 55 1 2 1 4 .2396875 0
1 1 57 1 3 1 1 3.068 0
1 1 59 1 5 1 2 2.396875 0
1 1 60 1 5 1 1 2.10925 0
1 1 61 1 3 5 1 2.10925 0
1 1 62 1 3 5 2 .862875 0
1 1 62 1 3 5 1 2.10925 0
1 1 62 1 5 1 2 .862875 0
1 1 67 1 2 5 1 1.34225 0
1 1 68 6 1 5 1 .57525 0
1 1 69 1 3 5 1 3.068 0
1 1 70 1 4 5 2 1.054625 0
1 1 79 1 1 5 1 .57525 0
1 1 79 3 2 5 2 1.054625 0
1 2 18 6 2 2 1 .3068 0
1 2 19 6 4 7 1 .95875 0
1 2 20 1 3 1 1 2.10925 0
1 2 20 6 3 1 1 .95875 0
1 2 20 6 3 1 3 1.0226667 0
1 2 20 6 3 1 3 1.0226667 0
1 2 20 6 3 8 3 .7030833 0
1 2 21 6 3 7 3 1.5979167 0
1 2 22 6 1 7 2 .671125 1
1 2 23 6 3 1 5 .6136 0
1 2 25 6 3 5 1 .95875 0
1 2 25 6 3 7 2 .1534 0
1 2 26 6 3 7 1 2.10925 0
1 2 29 6 3 10 3 .57525 0
1 2 29 6 5 1 1 .95875 0
1 2 30 1 3 10 4 .4314375 0
1 2 30 1 5 10 2 1.054625 1
1 2 32 1 3 2 3 .7030833 0
1 2 32 6 3 2 2 1.534 0
1 2 32 6 4 1 1 .95875 0
1 2 32 6 3 7 1 .57525 0
1 2 35 1 3 6 3 .57525 0
1 2 35 6 3 1 2 1.534 0
1 2 36 1 3 2 3 .57525 0
1 2 36 1 3 2 3 1.0226667 0
1 2 38 1 2 6 3 .3195833 1
1 2 39 1 3 2 3 .7030833 0
1 2 39 1 3 6 3 .57525 0
1 2 41 1 3 1 3 .57525 0
1 2 42 1 3 6 4 .3355625 0
1 2 43 1 4 1 3 .57525 0
1 2 44 1 3 1 6 .51133335 0
1 2 44 4 4 1 2 .287625 1
1 2 45 1 5 1 3 1.0226667 0
1 2 47 1 2 2 1 1.72575 0
1 2 49 1 3 6 2 1.054625 0
end
label values Mental_Health V8
label def V8 1 "very happy", modify
label values Gender V302
label def V302 1 "male", modify
label def V302 2 "female", modify
label values Age AGE
label values Marital V313
label def V313 1 "married", modify
label def V313 3 "widowed", modify
label def V313 4 "divorced", modify
label def V313 5 "separated", modify
label def V313 6 "never married or registered partnership", modify
label values Educ V336
label def V336 1 " 1 : Primary education or first stage of basic education", modify
label def V336 2 " 2 : Lower secondary or second stage of basic education", modify
label def V336 3 " 3 : (Upper) secondary education", modify
label def V336 4 " 4 : Post-secondary non-tertiary education", modify
label def V336 5 " 5 : First stage of tertiary education", modify
label def V336 6 " 6 : Second stage of tertiary education", modify
label values Employ V337
label def V337 1 "30h a week or more", modify
label def V337 2 "less then 30h a week", modify
label def V337 3 "self employed", modify
label def V337 5 "retired/pensioned", modify
label def V337 6 "housewife not otherwise employed", modify
label def V337 7 "student", modify
label def V337 8 "unemployed", modify
label def V337 10 "other", modify
[/CODE]

[Taiba Chau]

Any thoughts on how to do it correctly?

[Carlo Lazzaro]

Taiba:
looks OK; probably -ethnic- should be -label-led too.

[Taiba Chau]

Okay, so how would I then create the table? I am bit confused

[Carlo Lazzaro]

Taiba:
see -table- entries in Stata 17 .pdf manual.

[Taiba Chau]

For some reason, it won't allow me to open this. Any other support in trying to create the tables?