I actually am very curious to the responses, but whether the results are surprising or not, I think another value it would have is as a place to point people to. For example, Norm New-Guy or Felicity the Feminist says “I think X is a problem,” you can point them to the narratives, and say “8/10 women disagree.” (or vice versa)
Also, even if there are no “surprises” per se, it could still be enough to redefine your hypothesis space. For example, maybe before the results, I would guess that any one of six issues might be occurring to effect the gender ratio. After reading the samples, I notice that most the results focus on only 3 of my original issues, and maybe there is a new one women were discussing, that although it wasn’t in my original hypothesis space, wasn’t necessarily “surprising” (depending on your definition- perhaps you hadn’t thought of it yourself, but when someone said it, it made sense. I can see how you’d consider this a “surprise” though.)
In general, when making an example that could equally be made in either direction, I think it’s best to go the direction against what you think—or what others think you think.
So in the same way I think Yvain’s post would have been improved if his examples had been against positions he held, so too you might in future want your examples to be phrased more like
For example, Norm New-Guy or Felicity the Feminist says “I think X is a problem,” you can point eir to the narratives, and say “8/10 women disagree.” (or vice versa).
Obveously this is a purely about reducing system 1 negative reactions to posts and demonstrating an ability to visualise the other side’s hypothesis, and not a content issue at all. It’s much like the motivation behind Politics is the Mind Killer.
I actually am very curious to the responses, but whether the results are surprising or not, I think another value it would have is as a place to point people to. For example, Norm New-Guy or Felicity the Feminist says “I think X is a problem,” you can point them to the narratives, and say “8/10 women disagree.” (or vice versa)
Also, even if there are no “surprises” per se, it could still be enough to redefine your hypothesis space. For example, maybe before the results, I would guess that any one of six issues might be occurring to effect the gender ratio. After reading the samples, I notice that most the results focus on only 3 of my original issues, and maybe there is a new one women were discussing, that although it wasn’t in my original hypothesis space, wasn’t necessarily “surprising” (depending on your definition- perhaps you hadn’t thought of it yourself, but when someone said it, it made sense. I can see how you’d consider this a “surprise” though.)
In general, when making an example that could equally be made in either direction, I think it’s best to go the direction against what you think—or what others think you think.
So in the same way I think Yvain’s post would have been improved if his examples had been against positions he held, so too you might in future want your examples to be phrased more like
Obveously this is a purely about reducing system 1 negative reactions to posts and demonstrating an ability to visualise the other side’s hypothesis, and not a content issue at all. It’s much like the motivation behind Politics is the Mind Killer.
Thanks! I like this idea a lot, and have changed the relevant example accordingly.