.05 stat sig
Championed by Fischer but also agreed upon via mutual consent. In the social sciences, we were taught that this was a decision that was made via social process (and that most, if not all "decisions" in science are social processes-see Kuhn, "The Structure of Scientific Revolutions") and that that the key issue with the .05 level was to strike a balance between type one and type two errors when analyzing statistical data.
To provide just a little perspective here, I recently had a conversation with an engineering friend who works for a machine tool co. The tools are now manufacturing products to a 0.0001 mm tolerance. Every once in a while, an outlier product measures at 0.001 mm and is rejected, even though the measure error is statistically irrelevant as the measuring tool has an error rate of 0.001.
Clearly, the thing being measured has everything to do with the tolerance in the measurement error.
Dew's quiz was clever and required some "outside of the box" thinking, but really doesn't reflect what would have occurred if humans were hexapedal, (at least imo.) We have 10 fingers, insects have 2 antennae, one could argue that they would be more comfortable using base 2 rather than base 6 (or base 10.)
JM2C
aj
