I don't think it is representative enough. As I said, it is a self selected sample. Just with that you are skewing data. Not just that, but you are excluding precisely the people who have it easier to met other people and get dates in other ways.

If you look at studies about marriage or long term partnership, most would say that people know others at work, school or by being introduced by common friends... That is, you are excluding the most common methods to met people in that sample.

Online data is also biased because they only react to what you show them. And I hope you aren't so naive to think that the algorythm that decides which profiles are shown to you doesn't affect the reactions. People are manipulated in those apps by being shown only a small fraction of the profiles. For example: Tinder was know to show profiles of younger and less educated women to men, so we don't really know how they would react to older, more educated women. And the other way around, if as a woman I was show men 10 or more years my senior, it is no wonder I swiped left on most. Yet, when faced men my age things worked differently.

Also, you really think that people doesn't lie in apps? That is absurd. Many people use photoshop, pictures from someone else, filters, angles, wear hats or something to hide baldness, lie about how much money they made, pretend to want to date but only wanting attention, try to lure a posible cheating boyfriend... The situations are countless for people lying in the apps.

I don't dislike the results, but I dislike how people missrepresents studies by appliying them outside the scope of the specific study because not only is dishonest, it is bad science.

In fact, it would be interesting to see if the results change if you apply different algorythms to different sample sets. Or if you add new fields to filter the profiles. For example: what would happen if women filters would affect to which men they would show up and the other way around. So if I decided on only a 5 years age difference, my profile didn't even show up for the men 10, 15 years my seniors. Because it is quite possible that just by doing that the proportion of "swipe right" increased. But OK cupid and other apps would never do that, because otherwise, they might show less profiles to the men who usually are willing to pay premium. And it is also the reason that they add features as not allowing you to block men with premium accounts or allowing them to text you even if you haven't swiped right on them in some apps.

If you want to make a complete analysis, you should take into account that OK cupid is a company that aims to make money and that their settings are created with that in mind, not to facilitate long term relationships, but to "fidelize" customers (who would leave the app once they find someone long term).

So I encourage you to take that data bu add variants into the analysis and apply it only to the population it represents.