5 ms·
The IPK being 2nd lightest out of 12 on the given graph is not statistically significant: even before accounting for post-hoc analysis. The even distribution s
by ronaldx 13y ago
The IPK being 2nd lightest out of 12 on the given graph is not statistically significant: even before accounting for post-hoc analysis.
The even distribution suggests (to my eye) that this is as likely random as by some systematic effect.
- blahedo 13y agoIt's third lightest---there's an outlier that's way lower. But, again just from eyeballing, it looks like there's two clusters, one that has tracked with the IPK and one that is now heavier than the IPK. Based only on that, it strikes me as more likely that some of the national copies have gained mass(/weight), due to storage conditions or whatever, while others have kept their same weight. Of course this "doesn't matter" in that whatever the mechanism, it's still a problem for the science!
- ronaldx 13y ago> It's third lightest---there's an outlier that's way lower. wow, good spot, thanks :) Although it could be a problem for science, somehow it's nice that we can get a sense of margin of error by this method.