Generating New Variable

[Tess Gecha]

I want to generate a new dichotomous variable where 1 denotes the presence of one or more types of abuse - verbal abuse, sexual abuse, and physical abuse - and 0 denotes the absence of any abuse. Verbal abuse, sexual abuse, and physical abuse variables already exist. People could have experienced multiple forms of abuse. Would the code below work?

gen anyabuse=1 if verbal==1 | sexual==1 | physical==1

replace anyabuse=0 if anyabuse!=1

[Jared Greathouse]

Great question, please provide us your dataset using dataex.

[Nick Cox]

Assuming the original values are 0, 1 or just possibly missing

gen anyabuse = verbal==1 | sexual==1 | physical==1

would do what you want, as would

gen anyabuse = inlist(1, verbal, sexual, physical)

as would (but this time missing values are not allowed)

gen anyabuse = verbal | sexual | physical

For more background, see https://www.stata-journal.com/articl...article=dm0099 or https://www.stata-journal.com/articl...article=dm0087 or https://www.stata.com/support/faqs/d...rue-and-false/

[Tess Gecha]

Unfortunately, I don't have access to it and am merely trying to help a colleague troubleshoot. Are you able to answer without it?

[Nick Cox]

I think #4 was an answer to #2.