>Any study into rebellion is going to be difficult to gauge because what constitutes 'rebellion' in the context of family and upbringing is going to vary wildly between families and even individuals.

There probably are some straightforward ways of getting at this. I wonder why you can't just ask the participant to rank on a Likert Scale how much they agree with statements like "my personal values contrast with my parents" or "my actions contrast with those of my nuclear family". As long as you don't code it in charged language, people, in Western countries at least, are open about this stuff.

Real issue I take is, I can't help but feel like birth order effects would vary, and be moderated by things like sibling closeness, age, or perceived sibling social status. Like, I can see an opposite effect to tje rebel hypothesis for people particularly close to their siblings, while those who are less close would be more likely to strive to be unique.