Timeline for Why do physicists use sigma while biologists use p values/posterior probabilities?
Current License: CC BY-SA 4.0
23 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 19 at 14:18 | history | edited | Wrzlprmft | CC BY-SA 4.0 |
added 2 characters in body
|
| May 19 at 14:16 | answer | added | Wrzlprmft | timeline score: 2 | |
| May 16 at 5:44 | comment | added | Simon Crase | Founder effect? Perhaps early particle physicists used σ, their students followed them, then the grandstudents followed the students,... If nobody can come up with a decisive argument for p versus σ, it could come down to contingency and tradition. And each tradition looks weird and outlandish to those who follow the other tradition. | |
| May 15 at 19:59 | history | became hot network question | |||
| May 15 at 16:04 | answer | added | Sextus Empiricus | timeline score: 7 | |
| May 15 at 15:01 | history | edited | User65535 | CC BY-SA 4.0 |
Spelling
|
| May 15 at 14:45 | history | edited | User65535 | CC BY-SA 4.0 |
Added example
|
| May 15 at 14:35 | history | edited | Nick Cox | CC BY-SA 4.0 |
deleted 4 characters in body
|
| May 15 at 14:22 | comment | added | Alexis | @User65535 Wouldn't the CLT apply to estimates of the means in very large data sets, which are, after all, sums of very large numbers of observations (normalized by sample size)? | |
| May 15 at 13:48 | history | edited | User65535 | CC BY-SA 4.0 |
Added link to why so strict question
|
| May 15 at 13:47 | comment | added | Sextus Empiricus | I can imagine that $\sigma$ is popular wherever one primarily describes noise levels or standard errors, and makes comparisons with that. Signal to noise ratio is another popular expression. This statistics and dealing with error has already been done before Pearson and Fisher did their work in biology and eugenics and came with more 'sophisticated' statistics, like p-values and hypothesis testing. | |
| May 15 at 13:41 | comment | added | Ian Sudbery | @StephanKolassa Interseting to consider whether the equivalent to measuring more particles is similar to say sequencing more indeviduals, or sequencing each indevidual to a higher depth in biology. Sequencing to a higher depth increases your precision for that indevidual, and is cheap, but doesn't help you quantify the variation between indeivdauls. | |
| May 15 at 13:24 | comment | added | Stephan Kolassa | My suspicion is that while physics experiments are costly to set up, measuring more data does not cost a lot, because there are a lot of particles floating around. In biology, medicine etc., test subjects are more expensive, or we may have ethical concerns about killing lots of mice. So achieving high precision and high power is comparatively cheaper in physics than in the life sciences. But that is all POOMA from someone whose physics were a long time ago. | |
| May 15 at 13:03 | answer | added | Roger V. | timeline score: 5 | |
| May 15 at 12:57 | comment | added | Ian Sudbery | The difference in thresholds is due to multiple testing. The physicist says you need 5 sigma due to the "look elsewhere effect". The biologist says the corrected p-value has to be less than 0.05 - the physicist changes the threshold, the biologist adjusts the p-value, but they amount to the same thing. | |
| May 15 at 12:55 | comment | added | User65535 | When one makes decisions about how the aggregation is done one can frequently choose how the test statistic will be distributed, perhaps between a log normally or normally. | |
| May 15 at 12:54 | comment | added | Ian Sudbery | Also worth buying that in biology, particularly the parts of biology that deal with small p-values and high levels of multiple testing, distributions are often very non-normal. | |
| May 15 at 12:49 | comment | added | Stephan Kolassa | @User65535: but test statistics are usually computed not on the level of individual observations, but based on large aggregates of observations, and that is where we hope the CLT to kick in. | |
| May 15 at 12:46 | comment | added | User65535 | @Anyon "particle physics tend to have very large data sets, so the ... central limit theorems are often assumed to apply" it is not clear why a large number of individual observations of particles could be assumed to follow the CLT. For that to apply each particle's value would have to be the sum of a large number of individual variables, with is not my understanding of physics. | |
| May 15 at 12:44 | history | edited | Wrzlprmft | CC BY-SA 4.0 |
added 57 characters in body; edited tags
|
| May 15 at 12:42 | comment | added | Anyon | Regarding skewed distributions, note that experimental particle physics (the field in which you're probably most likely to hear about 5-sigma in popular treatments) tend to have very large data sets, so the law of large numbers and central limit theorems are often assumed to apply. Systematic errors is typically a bigger concern (of physicists, anyway). | |
| May 15 at 12:38 | history | migrated | from academia.stackexchange.com (revisions) | ||
| May 15 at 11:59 | history | asked | User65535 | CC BY-SA 4.0 |