Josh Miller points to this:
“We offered mTurk workers $0.50–$0.75 to complete the survey.”
Why would someone who spent $20k+ on their wedding be filling out a survey on mTurk? Maybe things didn’t turn out so well?
Josh continues:
I didn’t read the paper or the empirical section, just the abstract and I quickly looked at their data source and stopped.
I don’t think mTurk is always bad, just in this case the interaction could be a source of selection bias, and produce an effect mechanically.
I guess you gotta make up that $20K somehow. . . .
I think I need to remind myself that there is always a risk of being quoted!
Let me re-phrase that: It is an intriguing (published) result that’s a nice fit for your blog and would be fun to discuss. I had a knee-jerk concern, but I stopped short of reading the details cos I didn’t have time. My concerns could be completely misplaced.
one tid-bit: apparently turks spend far less on their weddings than the general population, not a really a surprise.
What’s the average that mTurkers spend? I can’t find that number in the paper under discussion.
Yeah, I was going to say $20k is not a lot of money for a wedding. I mean, there are often on the order of 100 people that come to a typical wedding, so that’s $200 per person. It’s possible to have a cheaper or smaller wedding, but this isn’t outlandish or anything.
turks have fewer friends?
The real headline is “It’s 2017 and economists still haven’t heard of graphs”
+1e6
FYI, a point of jargon: MTurk is populated by Turkers, only some of whom are Turks. (“Turks” was actually confusing to me, but I’ve also been using MTurk for data eval for a while, so I’ve kinda forgotten what the Turk in MTurk originally was.)
Yes, Turks are people from Turkey, Turkers are people who sell their services on Mechanical Turk. Using “Turks” really does change the meaning.
I think Amazon’s mechanical turk is named for this mechanical turk:
https://en.wikipedia.org/wiki/The_Turk
One can recruit a large number of genuinely voluntary participants without any compensation so paerhaps perhaps the reason Turkers participated in this study is not monetary.