logoalt Hacker News

kitdtoday at 11:39 AM1 replyview on HN

The survey excluded veterans with no commercial ties. Maybe those publications not mentioned used them instead.


Replies

JdeBPtoday at 12:28 PM

Yes, and I was slightly disappointed, from a statistical point of view, that they didn't tell us how many people they excluded from the original sample for this reason. That could have told us things.

If (say) they threw out 967 to leave those 33, then one possible explanation that that leaves the door open for is that journalists are so used to there being no conflict of interest, it being the case the majority of the time, that they don't check in the minority of cases where there is.

I suspect that they didn't throw out anywhere near as many as that, though. But, still, I would have liked to have been told the figure.