I'm extremely skeptical of the study you linked about magnetic alignment and sleep latency. Just looking at table 6 (the only table related to the only associa

I'm extremely skeptical of the study you linked about magnetic alignment and sleep latency. Just looking at table 6 (the only table related to the only association stated significant), the stats for South-North sleepers look noticeably better than the stats for North-South sleepers. I'm no statistician, but just look at it. The South-North sleepers sample, versus the North-South (supposedly best) sleepers, has proportionately three times more people experiencing zero days weekly of difficulty getting to sleep, two-thirds as many people having even one day of difficulty, two-thirds as many people having two days of difficulty... and 50% more people having three days of difficulty, but even then, we're talking only four people each among a group of 47 vs a group of 35. Again, I'm no statistician, but does that last and smallest column overshadow the three much bigger columns with the opposite trend, to the point of being a 1 in 1000 result assuming the null hypothesis? Furthermore, the sleep measures were self-reported, through the Pittsburgh Sleep Quality Inventory (PSQI), which I quote: >In Iran, Tehran psychiatry institute assessed the validity and reliability of the Farsi version of this questionnaire with 89.6% for sensitivity and 86.5% for specificity. Did the math behind this eye-catching <0.001 p-value take into account that the questionnaire used was measured to have a 10.4% false negative rate and a 13.5% false positive rate? I find it hard to believe. But I'd love to see a statistician who actually knows how to interpret these numbers tell me that I'm wrong.