Re:Research Papers That Caught the Public's Attention in 2019
by js8 on Monday December 30, 2019 @01:14PM
(#21915411)
> Out of 330 Game of Thrones characters, 186 die by the end of season 7, with survival ranging from 11 seconds to 57 hours and 15 minutes. This study of their characteristics reveals great potential for preventing violent deaths in the Game of Thrones world.
Looking forward to a follow-up study which will hopefully include season 8..
Re:Research Papers That Caught the Public's Attention in 2019
by jimbob45 on Monday December 30, 2019 @11:42AM
(#21914582)
>A Randomized Trial of E-Cigarettes versus Nicotine-Replacement Therapy
Seems like this is going to be the most cited and controversial paper here.
>Association Between Push-up Exercise Capacity and Future Cardiovascular Events Among Active Adult Men
They had the push-ups done to a metronome but I don't see anywhere that it states that the participants were made to use one strict form of push-up. Not all push-ups are created equal (get your butt down!).
There is nothing wrong with statistical significance as long as no one is misled by it. Unfortunately, there is plenty of confusion around it and lots of people being misled. Here are two of the concerns mentioned in the article:
1) P > 0.05 is mistakenly taken as evidence of no effect. Tradition is that P > 0.05 should be interpreted as no evidence of anything. For example, run your study with two people and you will find that p > 0.05 for a relationship between exercise and obesity or between teeth brushing and cavities.
2) 0.05 is mistakenly perceived as an indicator of whether things are 100% likely to exist or 0% likely to exist. The challenge here is that the user is not comfortable thinking about probabilities. So a study finding that the difference in cholesterol between Seinfeld fans and Chris Rock fans is statistically significant at 0.0499 is reported as "Seinfeld raises cholesterol."
I think that the central issue here is an objection to having to participate in a research community with people whose training in statistics did not allow them to understand the statistical tools they use in their work.
Re:Research Papers That Caught the Public's Attention in 2019
by yellow_lead on Monday December 30, 2019 @01:21PM
(#21915460)
Then these articles saying scientists are "Against statistical significance" are misleading. It sounds like scientists are more pro statistical significance than ever, so much so that they want it to be defined and interpreted correctly.
I guess many people would benefit from taking a high school statistics class.
Re:Research Papers That Caught the Public's Attention in 2019
by scarejunba on Monday December 30, 2019 @01:57PM
(#21915750)
Well, in isolation p is problematic. For instance, without hypothesis pre-registration you have the obvious problem with negative results and publication bias.
Looking forward to a follow-up study which will hopefully include season 8..
Seems like this is going to be the most cited and controversial paper here.
>Association Between Push-up Exercise Capacity and Future Cardiovascular Events Among Active Adult Men
They had the push-ups done to a metronome but I don't see anywhere that it states that the participants were made to use one strict form of push-up. Not all push-ups are created equal (get your butt down!).
There is nothing wrong with statistical significance as long as no one is misled by it. Unfortunately, there is plenty of confusion around it and lots of people being misled. Here are two of the concerns mentioned in the article:
1) P > 0.05 is mistakenly taken as evidence of no effect. Tradition is that P > 0.05 should be interpreted as no evidence of anything. For example, run your study with two people and you will find that p > 0.05 for a relationship between exercise and obesity or between teeth brushing and cavities.
2) 0.05 is mistakenly perceived as an indicator of whether things are 100% likely to exist or 0% likely to exist. The challenge here is that the user is not comfortable thinking about probabilities. So a study finding that the difference in cholesterol between Seinfeld fans and Chris Rock fans is statistically significant at 0.0499 is reported as "Seinfeld raises cholesterol."
I think that the central issue here is an objection to having to participate in a research community with people whose training in statistics did not allow them to understand the statistical tools they use in their work.
I guess many people would benefit from taking a high school statistics class.
https://www.nature.com/articles/d41586-019-00874-8
https://www.fharrell.com/post/pval-litany/
https://www.scientificamerican.com/article/the-significant-p...
1) Work out 3+ hour/week with as lease 2 weight training sections.
2) No processed food, no added sugar. More nuts, fruit, vegetables.
3) The world is fxxked because of climate change.
4) Sleep more.