And on all dating apps there a survivorship bias where people who have dating difficulties stick around on dating platforms while people who do not have difficulties quickly leave.
The real problem here might be that
A) men overrate attractiveness.
Or
B) men’s attractiveness is selected for on OKcupid but women’s attractiveness is not. Thus there is a survivorship skew.
But with this explanation, you run into a chicken and egg problem.
To what extent is the male survivorship problem/effect merely exposed or reflected by the data from a dating app… vs … intentionally caused by, exacerbated by the dating app?
The dating apps all know the answer to that question in extreme detail.
They don’t let us know though.
And they also have precise data on the same kind of thing for women.
They have this data, yes. Its not chicken and egg as much as it is a feedback loop. The way you’d probably control for it is to look at attractiveness polling only from new users. If it looks markedly different from overall attractiveness data then what you have is a bias that is related to time spent on the site.
This IS OKcupid users.
And on all dating apps there a survivorship bias where people who have dating difficulties stick around on dating platforms while people who do not have difficulties quickly leave.
The real problem here might be that
A) men overrate attractiveness.
Or
B) men’s attractiveness is selected for on OKcupid but women’s attractiveness is not. Thus there is a survivorship skew.
But with this explanation, you run into a chicken and egg problem.
To what extent is the male survivorship problem/effect merely exposed or reflected by the data from a dating app… vs … intentionally caused by, exacerbated by the dating app?
The dating apps all know the answer to that question in extreme detail.
They don’t let us know though.
And they also have precise data on the same kind of thing for women.
They have this data, yes. Its not chicken and egg as much as it is a feedback loop. The way you’d probably control for it is to look at attractiveness polling only from new users. If it looks markedly different from overall attractiveness data then what you have is a bias that is related to time spent on the site.