Hacker News Reader

Comment by idm | Hacker News Reader

This is an open access article and the statistics are published. See p.41 for a per-country analysis of the effect on literacy.

I would be ashamed to publish criticism about an open access scientific article without having any understanding of statistics.

View on HN

It looks like variance is increasing as distance increases. This is bad statistics.

Variance can increase with time without invalidating the trend. Signals can attenuate with time and nonetheless still be a signal. This is very frequently the case with longitudinal studies that track outcomes from initial conditions.

And how you validate Bartlett’s test in this case?

Personally I prefer levene's test (it's more robust in situations where the distribution may not be normal). Either way though, typical practice would be to run the tests both assuming equal variance and again not assuming equal variance. If both are statistically significant then it's a moot point, the results are statistically significant regardless of whether the variance in populations is equivalent.

> It looks like variance is increasing as distance increases.

Given the topic, how would it not?

And how would you validate Bartlett’s test?

This is like saying a stone doesn’t really cause ripples because they fade away before the hit the far shore.

You don't think confounding variables would accumulate as the area being considered greatly increases? Why not?

What statistics does one need to understand to know that the correlation on that scatter graph is terrible? Hell you don't even need statistics - just use your eyes.

The fact that the x-axis variable only explains a small amount of the variance in the y-axis does not mean that the effect is not real or not statistically significant. In this case, it's exactly what you would expect, even if the effect is real; surely you don't think it's reasonable that there are no other factors affecting literacy.

Saying "just use your eyes!" instead of actually doing analysis can lead you to all kinds of incorrect conclusions -- both false positives and false negatives.

The first couple of weeks of my STATS 101 course was my professor showing examples where “your eyes” were wrong.

"Don't need statistics just use your eyes"

Yet the statistic supports the conclusion. Just because the signal is weak doesn't mean there's no signal.

Reply on news.ycombinator.com