Creating new variables by combining others on a Likert-scale

[Anna weel]

I'm wondering if I did this correct.
I have 26 variables that are on a likert scale but only score from 0-3. I want to make three different groups out of these 26 questions and then calculate the mean, and sd. This is what I did:
gen Bulimia = c016c + c016d + c016i + c016r + c016u + c016y
gen Anorexia = c016a+ c016f+ c016g+ c016j+ c016k + c016l+ c016n+ c016p+ c016q + c016v + c016w + c016x + c016z
gen Binge = c016b +c016e+ c016m+ c016o + c016s +c016t

Did I do this correct and is this going to give me the desired outcome, or did I miss something?

In addition I would like to scale the variables like but I'm not sure how to do this, so if someone can give me a hint that would be amazing.
0-7 normal
8-11 mild
12-20 significant tendency
21-39 severe tendency

Thank you for your kind answer.
Best regards
Anna

[Nick Cox]

You'll find just about every possible view on this kind of manipulation from it is indefensible to this is what people do in this field.

I just note that

Your three composite variables are the sum of (respectively) 6, 13 and 6 variables, but you refer to 26 variables.-- so presumably there is a c016h variable too.

Only the second variable can extend to 39 in sum; that's impossible for the other two. So, I don't know what people do, perhaps multiply up by 39/18.

I find it puzzling that people think there's fine detail that allows addition, and then want to coarsen the scale again.

[Anna weel]

Dear Nick,

Thank you for bringing to my attention that I had missed the c016h in the sequence. I have now added it and I hope this is the correct way to proceed, since there were no specific instructions provided.
Regarding the comment "I find it puzzling that people think there's fine detail that allows addition, and then want to coarsen the scale again," I am unsure of its meaning and implications.
As someone who is new to Stata and only trained in R-studio, I am currently struggling to become familiar with this program. I would appreciate any assistance someone can provide to help me become more proficient in using Stata. If you have any tips or resources that you can recommend, I would be very grateful.

Thank you for your help.

[Anna weel]

I did figure it out but I'm sure there is a better way to do this.
recode Bulimia (0/2=0) (3/5=1) (6/10=2) (11/18=3), generate(Bulimia_cat)

gen bulimia_cat = .

replace bulimia_cat = 0 if bulimia >= 0 & bulimia <= 2
replace bulimia_cat = 1 if bulimia >= 3 & bulimia <= 5
replace bulimia_cat = 2 if bulimia >= 6 & bulimia <= 10
replace bulimia_cat = 3 if bulimia >= 11 & bulimia <= 18
label define eatscale 0 "Normal" 1 "Mild" 2 "Significant tendency" 3 "Severe tendency"
label values Bulimia_cat eatscale

If someone has a suggestion that would be great.