Stata新手求助：基于现有虚拟变量生成新偏好变量的语法问题

Alright, let's tackle this Stata variable creation problem together. First, I need to lay out some reasonable assumptions about your existing variables since you didn't specify their exact coding—just tweak these if they don't align with your data:

Let’s assume:

• Sex_at_birth = 1 = Assigned male at birth, Sex_at_birth = 0 = Assigned female at birth
• isHE = 1 = Heterosexual (异性恋), isBI = 1 = Bisexual (双性恋), isHO = 1 = Homosexual (同性恋)
• These orientation variables are mutually exclusive (only one is 1 per observation; if not, you’ll want to clean that first to avoid conflicting logic)

Based on standard understandings of sexual orientation and attraction, here’s the logic for each new variable:

• likesMEN: Equals 1 if the person is attracted to men
  • Assigned male + homosexual → attracted to men
  • Assigned female + heterosexual → attracted to men
  • Bisexual (regardless of assigned sex) → attracted to men
• likesWOMEN: Equals 1 if the person is attracted to women
  • Assigned male + heterosexual → attracted to women
  • Assigned female + homosexual → attracted to women
  • Bisexual (regardless of assigned sex) → attracted to women
• likesBOTH: Equals 1 only if the person is bisexual (since bisexuality implies attraction to both genders)

You can use either explicit replace commands (great for readability) or concise conditional expressions. Here’s both options:

Option 1: Readable Step-by-Step Code #

* Initialize all new variables to 0 first
gen likesMEN = 0
gen likesWOMEN = 0
gen likesBOTH = 0

* Update likesMEN based on attraction logic
replace likesMEN = 1 if (Sex_at_birth == 1 & isHO == 1) | (Sex_at_birth == 0 & isHE == 1) | (isBI == 1)
* Set to missing if any key variable is missing
replace likesMEN = . if missing(Sex_at_birth, isHE, isBI, isHO)

* Update likesWOMEN based on attraction logic
replace likesWOMEN = 1 if (Sex_at_birth == 1 & isHE == 1) | (Sex_at_birth == 0 & isHO == 1) | (isBI == 1)
replace likesWOMEN = . if missing(Sex_at_birth, isHE, isBI, isHO)

* Update likesBOTH (only bisexuals)
replace likesBOTH = 1 if isBI == 1
replace likesBOTH = . if missing(isBI)

Option 2: Concise Conditional Expression Code #

* Generate likesMEN directly with conditionals
gen likesMEN = ( (Sex_at_birth == 1 & isHO == 1) | (Sex_at_birth == 0 & isHE == 1) | isBI == 1 )
replace likesMEN = . if missing(Sex_at_birth, isHE, isBI, isHO)

* Generate likesWOMEN directly with conditionals
gen likesWOMEN = ( (Sex_at_birth == 1 & isHE == 1) | (Sex_at_birth == 0 & isHO == 1) | isBI == 1 )
replace likesWOMEN = . if missing(Sex_at_birth, isHE, isBI, isHO)

* Generate likesBOTH directly
gen likesBOTH = (isBI == 1)
replace likesBOTH = . if missing(isBI)

Pro Tip #

If your variable coding is different (e.g., Sex_at_birth = 1 is female instead of male), just flip the corresponding conditions. For example, if Sex_at_birth = 1 = female, the likesMEN logic would become (Sex_at_birth == 1 & isHE == 1) | (Sex_at_birth == 0 & isHO == 1) | isBI == 1.

内容的提问来源于stack exchange，提问作者Mandarc