There’s some evidence that it’s inherited, under the umbrella of other inherited OCD behaviors.
That’s not about rhinotillexis (and what it is about, provides zero evidence because it’s one of the worst genetics papers I’ve ever read).
Will these experiments work? Could I find any subjects willing to spend their time advancing science in this direction? Will I find this utterly devoid of any interest once the comedic novelty of talking to people about it wears off? That I don’t know, but I hope I’ll be able to post a follow-up a year from now.
In all these cases you have a statistical power issue. How large an effect on infections or environmental awareness or whatever does a functional (evolutionary) explanation need? Well, the benefit has to exceed the cost to even potentially evolve, so you can look at the cost for a lower bound. How costly is picking your nose? Not very: it is a hand gesture done a few times a day for a few seconds in idle moments of downtime. You probably burn more calories taking a walk for an hour or three than you do picking your nose over a lifetime. Thus, the benefit need only be vanishingly small and ever so slightly different from zero to pencil out. And that’s not an effect you would ever expect to detect. So a null result here would be uninformative.
Already at https://www.gwern.net/docs/biology/boogers/index
That’s not about rhinotillexis (and what it is about, provides zero evidence because it’s one of the worst genetics papers I’ve ever read).
In all these cases you have a statistical power issue. How large an effect on infections or environmental awareness or whatever does a functional (evolutionary) explanation need? Well, the benefit has to exceed the cost to even potentially evolve, so you can look at the cost for a lower bound. How costly is picking your nose? Not very: it is a hand gesture done a few times a day for a few seconds in idle moments of downtime. You probably burn more calories taking a walk for an hour or three than you do picking your nose over a lifetime. Thus, the benefit need only be vanishingly small and ever so slightly different from zero to pencil out. And that’s not an effect you would ever expect to detect. So a null result here would be uninformative.