Stack Exchange
log in

cbeleites

 Discussion on answer by Maarten Buis:…

Imported from a comment discussion on academia.stackexchange.c...
cbeleites
Mar 10 19:17
@Graham: while the vast majority of these jobs do not require physical strength, physical strength requirements are discussed as one reason (among others) for surgery and ortthopeadics still having almost 5 male for every female practitioner - while onverall the medical profession has reached a 1 : 1 ratio, and 1 : 2 male vs female students.
cbeleites
Mar 7 15:27
hard sciences: biology -> chemistry -> physics are perceived with increasing hard-scienceness, and also have decreasing female fractions (subtantially more females in biology, chemistry: somewhere in the 40 - 50 %, depending a bit on who is counted in [e.g. where I am: chemical economics -> chemist, food chemist -> separate]. Physics: much more males. But when we add maths, which is (at least by us chemists) considered "harder" than physics, even though not a natural science: that has a male : female ratio of about 1 : 1. (hard here refering to black/white true/wrong, as opposed to hard work)
 

 Discussion on question by Foxy: Why i…

Imported from a comment discussion on academia.stackexchange.c...
cbeleites
Mar 7 10:37
I "stole" your source and added an answer.
cbeleites
Mar 7 07:10
@RuthMcT: similar for Germany: aerzteblatt.de/archiv/… I think this is a highly relevant point for the thoughts here because MD has higher prestige than science or engineering professions (typically among the top 3, after firemen and nurses), and in addition has the highest earnings these top 3. The hypothesis of high status -> male thus may play some role, but there must be important other factors.
 

 Discussion on answer by Ian Sudbery: …

Imported from a comment discussion on academia.stackexchange.c...
cbeleites
Jan 28 14:57
IMHO all these issues (blinding, not being a good control, conflicts of interest, ...) can nicely be resolved by volunteering for a clinical trial at a neighbouring institute where one is selected according to the proper statistical criteria of an IRB approved study without any problems. And, if one feels more safe with self-experimentation than being a subject like anyone else to someone else's trial: that's then how your study participants feel as well.
cbeleites
Jan 28 14:57
@JoshuaZ: well I guess the lab head in question would have been the director of the whole research institute. At that level, you have a decent chance of finding people who think the IRB should report to them, and in general that normal social-type rules don't apply to them. Which curiously means, that the IRB blanket ban at least statistically speaking may have had good reasons...
cbeleites
Jan 28 14:57
On questioning further, this applied even to the rather trivial self experimentation of taking one's own capillary blood the way every diabetic patient does several times per day, and considering that oneself is the source of blood that comes with the lowest infection risk. That being said, volunteering healthy students are a suitable control group for some experiments, but for all the research we did, we were decidedly no good as controls from a statistical point of view (it's just not a relevant clinical diagnostic question to distinguish healthy 25-yos from terminally ill 80+ year olds).
cbeleites
Jan 28 14:57
@JoshuaZ: there is also a concern that the volounteer may not be all that volountary. At least that point was brought up very prominently when I was in clinical diagnostic method development: The vast majority of researchers are PhD students and postdocs who are rather low down in the command chain. Our ethics training told us that the IRB would refuse any "volunteers from the lab" or self experimentation proposals since they considered the risk of (undetected) pressure by supervisors or peer group pressure to high.
cbeleites
Jan 28 14:57
@JoshuaZ: If I understand Azor Ahai correctly, the point is that regardless of who the subject is, the same ethical standards should be applied. I.e., IRB approval. And since medical science has learned that both treating doctor and patient should better be blinded in experiments, conflating those roles (plus potentially skipping any blinding) doesn't sound too promising in terms of conflicts of interest.
 

 Discussion on question by Hydrazin: H…

Imported from a comment discussion on academia.stackexchange.c...
cbeleites unhappy with SX
Apr 4, 2024 19:19
@user4052054: I don't think it would necessarily violate employee rights here in Germany: the relevant right is: when working more than 6 hours, there must be a breat of not less than 30 min, but there is no right to have this (lunch) break at a specific time. Most (though not all*) people will be fine with occasionally shifting their lunch break by an hour.
* this includes people who for health reasons need to stick to a regular schedule - which would trigger particular protection from a labour law perspective.
 

 Discussion on answer by Nobody: Why a…

Imported from a comment discussion on politics.stackexchange.c...
cbeleites unhappy with SX
Dec 19, 2023 09:10
In addition to point 2, the article question is about change in inflation rate, i.e. 2nd derivative of price/cost. It is notoriously easy to get people confused about this. In addition to the "general" confusion caused by change-of-change, it is even further removed the price/cost which is the relevant measure for most people. And, the 2nd derivative is of course far more wiggly/noisy and thus more suitable for generating click-bait.
 

 Discussion on answer by Trish: Would …

Imported from a comment discussion on law.stackexchange.com/qu...
cbeleites unhappy with SX
Nov 24, 2023 02:30
We don't know whether the TV suffers from planned obsolescence. It is still possible that it suffers from "bad luck" or (more relevant) from bad design without the intention to make it fail early. (Actually, we don't even know whether it is obsolescent, OP did not mention that they unsuccessfully tried to repair.)
 

 Discussion between gerrit and user253751

Imported from a comment discussion on money.stackexchange.com/...
cbeleites unhappy with SX
Dec 15, 2022 19:03
@gerrit: (a notary is useful when you need/want to have the money in an escrow/trust account. But your practical problem is staying the same, you need to wire the money to the trust account) Banks can do larger transactions, it's just the online limits that are lower. If you go to the bank in person and have them wire the money, they can send much larger amounts. (They don't use different IT for business accounts, just different conditions). They may also be able to manually raise your online transaction limit to a higher amount than you can do yourself online - but that I don't know for sure.
cbeleites unhappy with SX
Dec 15, 2022 18:03
Also, as I understand, the already existing anti-money-laundering law means quite a bit more work for e.g. car dealers (who are in a special "at risk" category) than just know who you deal with.
cbeleites unhappy with SX
Dec 15, 2022 18:00
@gerrit: I asked my bank 2 a ago (actually buying a light truck) about real-time transaction, and it turned out they do it even without fee - BUT they have a hard limit on it at 10 k€. So you can use that (if the dealer agrees and their bank offers to receive in real-time) only to replace cash transactions that would also be OK under the proposed cash limit.
 

 Discussion on answer by Nobody: My em…

Imported from a comment discussion on workplace.stackexchange....
cbeleites unhappy with SX
Oct 9, 2022 18:32
@nvoigt: however, OP is a researcher on a fixed term contract in a foreign country. Unemployment benefits may be less of a problem than a visum ending with the fixed term contract.
cbeleites unhappy with SX
Oct 9, 2022 18:27
Just for comparison, I'd expect a university mensa to have lunch prices for employees around 4 - 6 €
 

 Discussion on question by Alice: How …

Imported from a comment discussion on academia.stackexchange.c...
cbeleites unhappy with SX
Oct 3, 2022 11:03
In addition, any German income they have is taxable in Germany, regardless where the tax residency is.
cbeleites unhappy with SX
Oct 3, 2022 11:03
@Alice: You cannot conclude much from the accidental knowledge that the professor keeps tax residency in Germany. E.g. I was working for a couple of years in Italy (with a contract subject to Italian law) but the German tax office nevertheless decided that my tax residency stayed in Germany. From what they wrote, they consider it typcial for scientists to move around the world for a while :-) and the 180-days-rule is only one indicator of tax residency.
cbeleites unhappy with SX
Oct 3, 2022 10:57
@gnasher729: for employees of German employers there are limits to working abroad by the social security/social insurance system, leading e.g. to some harsh surprises when people would like to do some mobile/remote work physically being in some other part of the EU. BUT: AFAIK, they could be resolved by getting the appropriate insurance also for their remote location in addition - which is of course expensive - but maybe worth while for a professor who likes to work from Brazil.
 

 Discussion on answer by PMF: Is there…

Imported from a comment discussion on law.stackexchange.com/qu...
cbeleites unhappy with SX
Sep 27, 2022 18:58
@MichaelHall and Greendrake: the (German) employee also has the right to get the employer's consent to a side job unless the employer can show that the side job harms their interests (including e.g. leads to violation of maximum working hours.) Wrt holiday side jobs, it seems to be a case-to-case decision. Sounds like a inland bookkeeper working a holiday job as barkeeper at a famous seaside place is OK vs. a PhD student taking holidays to work long hours as barkeeper in their home town is not.
cbeleites unhappy with SX
Sep 27, 2022 18:58
@MichaelHall: indeed why not. In Germany, the answer would still be that both jobs combined need to be within the maximum allowed working time (per day and week). Iow, they'd need to be part-time jobs.
 

 Discussion on question by JonathanRee…

Imported from a comment discussion on academia.stackexchange.c...
cbeleites unhappy with SX
Sep 23, 2022 13:16
@JonathanReez: my point (and that of a number of other people here) is: no subterfuge is needed to remove the unfair advantage of the high profile name/institution/whatever. Anonymization would be sufficient and in contrast to the subterfuge also ethical. In cases where anonymization/double blind review doesn't work, also the subterfuge wouldn't work. So there's IMHO nothing to be gained, but there is the sure loss due to being unethical in itself.

BTW, I come from a region of the world where reputation of a university [or quality of education one can get there] plays *much* less of a role…
cbeleites unhappy with SX
Sep 23, 2022 11:32
And it beats me how Lomonosov university should be an example for a low prominence university... Are you aware that it is much higher up e.g. in the times ranking than Chapman where the Nobel laureate of the paper is? I'd expect a severe political bias, particularly right now, but that's IMHO yet a different question. Maybe Kazan University would be a better example. Or TU Madrid if you want to stay away from the political bias and discuss general author prominence...
cbeleites unhappy with SX
Sep 23, 2022 11:31
As for the answers you get here, IMHO it does play a role that you suggest a clearly unethical approach to gain an advantage over the anonymous baseline.
cbeleites unhappy with SX
Sep 23, 2022 11:11
(=> ask for clarification in revision) rather than assuming the whole thing is BS (=> reject). The high acceptance for the prominent author being revealed may have such components. Which would be systematic, but not necessarily erroneous.
cbeleites unhappy with SX
Sep 23, 2022 11:09
... and for whom I'd certainly trust that a manuscript submitted with them as supervisor has undergone an honest and reliable "internal review" before it was submitted - and others, where I think I'd better check carefully these and those details. And, there are prominent names in my field where if I get a manuscript where I don't understand the worthiness I'd certainly think it may very well be me who didn't understand ...
cbeleites unhappy with SX
Sep 23, 2022 11:05
How much of the latter is in general systematic error and how much is justified (across papers), we don't know. There's also the question what the reviewers should estimate: the quality of the manuscript given the manuscript only, or the probability of the manuscript being good for publication in that journal. The latter may legitimately include different priors. I'd certainly say that there are some prominent researchers in my field, who I know e.g. from conferences, ...
cbeleites unhappy with SX
Sep 23, 2022 11:00
@JonathanReez: I've had a look at the paper. Notice: we have a strong bias observed in favor of a prominent author for n = 1 paper. The paper states that for the low prominence author, (at least) two effects happen that counteract each other compared to anonymization, resulting in a small bias towards rejection compared to the anonymized baseline (52 % LL vs. 48 % for anonymized paper AA). The high prominence author did get systematicaly more favorable reviews (20 % reject, HH and AH).
cbeleites unhappy with SX
Sep 19, 2022 07:25
Iow, the IMHO legitimate goals of avoiding unfair advantages or disadvantages, can be achieved by pseudonyms or double (triple) blinding. I don't see taking someone else's name/affiliaton giving any legitimate advantage over that baseline.
cbeleites unhappy with SX
Sep 19, 2022 07:23
@JonathanReez: the bias gives some papers an unfair advantge: Papers that are not so good that they get the heads up regardless of who wrote them, and are not so bad that they are rejected also with MIT affiliation. So on average, too many mediocre MIT papers are accepted. Personally, I don't see any reason to change that situation in a way that "corrects" the unfair MIT advantage by publishing more mediocre papers from other institutions. I'd rather see the mediocre MIT paper rejected as well.
cbeleites unhappy with SX
Sep 18, 2022 12:49
@NateEldredge: Very good examples. I'd like to generalize them: the expected gain for the author is also the higher, the worse their own work is. I.e., OP's "proposal" suffers from a serious bias as well. That bias alone would (and should) make MIT (or you or me: anyone who thinks they have a reputation to lose) adverse to their name being "borrowed".
 

 Discussion on answer by wimi: Is it e…

Imported from a comment discussion on academia.stackexchange.c...
cbeleites unhappy with SX
Sep 19, 2022 12:42
Moreover, if you want to compensate for a possible bias, it is sufficient to remove it. Which can be achieved by a pseudonym. A false identity that is presumably seen favorably would overcompensate. So in addition to the ethical problems of misrepresentation and using someone else's name [without even asking them], it would be trying to get a free ride on the (hopefully honestly earned) good reputation of the impersonated.
 

 Discussion on question by alexhroom: …

Imported from a comment discussion on travel.stackexchange.com...
cbeleites unhappy with SX
Aug 15, 2022 18:52
A large number of police making groups of people go certain ways (but usually not lining them up for frisking) on a weekend may happen when police escort soccer fans... Also not so nice for a traveler to end up in between those lines.
 

 Discussion on question by Janos: Deal…

Imported from a comment discussion on workplace.stackexchange....
cbeleites unhappy with SX
Aug 8, 2022 21:31
If the upper management is concerned about putting lots of money into training people that leave soon after (a fair concern IMHO), a pretty direct solution to this are contracts that specify the employee has to reimburse the company for the costs of the training (incl. their salary and travel) iff they leave on their own decision, less 1/n of that sum for each month they stayed after finishing the training. (Here in Germany, court decisions limit n depending on the duration of the training, ranging from 6 months for a 1 month training to 5 years after a 2 year training)
 

 Discussion on answer by J D: Is the u…

Imported from a comment discussion on philosophy.stackexchange...
cbeleites unhappy with SX
May 27, 2022 17:39
@DdogBoss: "From a metaphysical perspective, I guess science can be wrong." However, a more practical point to keep in mind is: there is also bad science out there (within the belief system of science)...
 

 Discussion on answer by David A. Crav…

Imported from a comment discussion on workplace.stackexchange....
cbeleites unhappy with SX
May 3, 2022 07:18
Here's the corresponding EU directive for railway transport: eur-lex.europa.eu/legal-content/EN/TXT/?uri=LEGISSUM:l24003
cbeleites unhappy with SX
May 3, 2022 07:18
@GregoryCurrie: here in Germany, if you are "washed ashore" on some station on your way and there is no connection until the next day, the railway pays for that hotel room (and in turn tells you which hotel to use).
cbeleites unhappy with SX
May 3, 2022 07:18
@lalala: in terms of OP's income tax, I don't think so, because the nature of a compensation for damage is to compensate a loss. VAT is a different story, but that's for the railway company to wrap their head around, and also their profit suffers which has tax consequences ;-)
 

 Discussion on answer by Kilisi: Is it…

Imported from a comment discussion on workplace.stackexchange....
cbeleites unhappy with SX
Apr 30, 2022 10:13
It's seen as analogous to cold or hot beverages the railway must provide for free if the AC doesn't work in summer or the heating doesn't work in winter that offset the inconvenience of sweating or freezing. In OP's case, a fixed fraction of the ticket price is used (per legislation) as practical shortcut to put a sensible amount to the compensatiion.
cbeleites unhappy with SX
Apr 30, 2022 10:08
@PoloHoleSet: I'd expect that it may vary between legislations what this money exactly is. In the EU, it is compensation for the inconveniences caused to OP, and that inconvenience is considered as bound to the person (the passenger rather than the buyer of the ticket). It is therefore also not subject to income tax since it doesn't lead to a gain but compensates a damage. The same as compensation for pain and suffering after an injury is not subject to income tax.
cbeleites unhappy with SX
Apr 30, 2022 10:03
In particular in a forum that routinely tells people that HR is not their friend I'm rather astonished if disclosure to the employer is said to be mandatory when it may actually not be.
cbeleites unhappy with SX
Apr 30, 2022 10:00
My take on the question is that this is (one of the lines of inquiry) how OP tries to find out whether it is their money or not. For which there is no ethical or legal obligation to disclose to their employer. The requirement is only for OP to find out, and, once they know, to act accordingly.
cbeleites unhappy with SX
Apr 30, 2022 09:58
@PhilipTinney: sure it's a qualitfication. I just wanted to point out/add to your comment that there exist very good and completely legal and ethical reasons for not disclosing. Namely, that Killisi's statement that the money is not OP's is wrong. Which would be the case where I am and moreover I have reasons for an educated guess that OP is in a sufficiently similar if not the same legislation.
cbeleites unhappy with SX
Apr 28, 2022 12:08
(@Voo incidentally, in case of an accident for which a 3rd party is responsible and that rendered OP unable to work, here in Germany the employer needs to be told because they'd have to pay sick leave for OP and in turn would be compensated for that from the one who caused the accident.) But that's off topic here...
cbeleites unhappy with SX
Apr 28, 2022 11:59
@PhilipTinney IMHO iff it is clear that the compensation is OP's and not the employer's, that would be a very good reason for not disclosing.
 

 Discussion between cbeleites unhappy …

Imported from a comment discussion on stats.stackexchange.com/...
cbeleites unhappy with SX
Apr 22, 2022 18:31
BTW thanks for the discussion - helps me to better formulate my thoughts in future as well :-)
cbeleites unhappy with SX
Apr 22, 2022 18:31
And the latter would require huge amounts of samples of the fault/out-of-spec class in my scenario, since it covers most of the feature space.
cbeleites unhappy with SX
Apr 22, 2022 18:29
When I said ill-defined above: such a low sample density means for a binary model that the boundary is uncertain because of the uncertainty in that class. So to me, here, it's not a black-and-white question of having samples of the 2nd class or not: it's from no samples (=> one-class) over insufficient samples (still one-class) to having sufficient so that sampling density is good everywhere, and no surprises expected (=> binary).
cbeleites unhappy with SX
Apr 22, 2022 18:21
I don't think one can say, one-class is better than discriminative. The question is what the application scenario calls for.