|
Julia M. Rohrer@dingdingpeng.the100.ci |
Followed by the insight "oh this number will usually be high because ppl describe ppl similarly", so then you're trying to tweak the measure so that it no longer counts that as similarity, and you end up in a messy place.>
1 replies 0 reposts 0 likes
|
|
Julia M. Rohrer
@dingdingpeng.the100.ci
[ View ] |
I think the proper solution would be to realize "why does it bother me when the number is always high", and then conclude something like "actually, I want to know whether A being agreeable has an effect on B rating A as agreeable.">
1 replies 0 reposts 0 likes