How was this collected?<p>The twitter thread seems to suggest that current couples were asked what year they met and how they met, and the chart was filled out from there. If so, isn’t there some strong survivorship bias to these data, i.e., there’s a preference for longer lasting couples having met certain ways?
What's really interesting to me is the temporary flattening in "met online" around 2000 - 2005. What caused that?<p>My theory: it's the transitional period between online dating as a novelty, and online dating as an obvious default for youth who grew up online.