Right, just like the Jews developed resistance to Xyklon-B, or how the victims of the Spanish Inquisition developed resistance to fire and trauma.
The thing about bacteria is that for many species, a "generation" is on the order of 20 or 30 minutes. They can evolve "slowly" over thousands of generations, but that might only be a few months of our time.
I want a candidate with character, morals, one who represents my beliefs on abortion and on the deficit and whether or not we should reduce spending or increase taxes.
To be honest that's probably the kind of thinking you should be engaing in.
Good scientists make terrible leaders, as they either tend to believe themselves to be experts in everything despite only knowing much about the migration habits of snow geese, or being so balanced and equivocal (ie scientific) about every issue that they will never be able to make a decision.
You can not though compare them to the people who sneak here and pop out babies and look for welfare with forms printed in their native language. These are not the people that strengthen a country.
I don't watch, and don't much care, and my life is far better for it.
I had started this reply to disagree with you about this but I'm struggling to come up with a good reason. I keep watching the news (from a lot of different sources), but mostly I think because I find it interesting.
It's quite a thorough analysis, and a totally different spin than anything I had heard on BBC, RTE (Ireland) or US news sources. I guess all news sources are biased, and you need to take more than one point of view if you want to be able to form your own balanced opinion.
Ironically, I've found many computer science majors are not very versed in the ramification of statistics either. I think it has something to do with the binary world that they envisage or something like that.
The most common example is the "1-in-a-million" mentatlity many computer science majors have when talking about bugs or special-case code paths. You'd think they'd know better as they can often quote all sorts of statistical sort or database traversal, O(log n), big-o little-o, etc, but when you get them with a common sense thing about code performance issue, they appear to get some sort of temporary lobotomy.
...especially as regards the use of mathematics in the interpretation of 'data' where the soft sciences have such a 'hand wavy' approach to cause and effect.
To me, economics is a prime example. Forgive me if I'm off base in in my belief that economics is both sociological and soft(headed), but tyring to measure human behavior in the absence of an accounting for political corruption within this purely human realm and leaving the so-called black market beyond it's consideration leaves the inclusion of economics within the realm of 'science' suspect.
I would haved greatly appreciated any attempt by a professor to explain the difference between soft science and hard science, especially if it included an math based explanation of the nuance between these different domains.
Soft sciences are typically about trying to solve 'wicked' problems, which are those that are generally impossible to completely solve (end poverty or health inequality, understand crime, migration, or human behaviour in general etc). Hard scientists typically try to solve problems that are relatively much easier because they have a simple concrete goal (put a man on the moon, make a bomb, cure some disease)
Soft scientists need a much stronger theoretical framework to interpret their data, because of the absence of any really testable mechanisms for the effects they observe. This can come across as 'hand wavy' but it really isn't. Your economics example isn't entirely fair, some economic models will include corruption and black markets etc and others wont, just as some physics models include relativistic effects and others don't. A good scientist has to choose the right model to approah any problem, regardless of discipline.
I've been working in an inter-disciplinary group and have had the opportunity to see medics and economists try to work together. The two cultures are very different in their scientific approach, both consider the other to be unnecessarily picky about some aspects of the work while not being rigorous enough in others. Eg economists spend a huge amount of their time trying to prove causation in observational data, while medics will typically wave this away if they think the causal effect is likely enough. On the other hand economists tend not to contextualise their results well enough, while medics will see the bigger picture in terms of building on existing science.
He might end up losing some of his elitist attitude before the course is over. It would be better if lost the attitude ahead of time, and approached the experience like he was at least teaching the same species.
Indeed. I teach statistics to mathematicians, biologists, psychologists and social scientists and I would say the social scientists 'get' the principles of statistics better than the 'hard' scientists do. The main reason is that soft scientists (which is a horrible term) can think about uncertainty and its consequences, whereas hard scientists (mathematicians included) are unhappy if they don't have a yes/no answer to a question. Obviously this is a generalisation but it may inform your approach to teaching.
Also, statistics is not 'just math'. I know this because I can do statistics but I can't do math(s) any more.:-)
Likewise, using genetech to produce insuline should be stopped immediately. The only way to cut down on diabetes is to improve your diet. Hell, I throw healthy veggies away daily because everybody wants to eat hamburgers and drink coke all the time. Diabetes prone people will die off naturally over time and than we have solved the problem.
Same argument can be made about most avenues of progress. You sir, are anti-progressive. Look into a mirror, you do not deserve a/. id (so says the AC).
People generally reproduce before they die of type-2 diabetes related complications, so I don't think natural selection will do it for us here.
It will happen with improved education and rights for women. Birth rate in Europe (1.59 births per female) / US (2.0 births per female) / Japan (1.37 births per female) is far below replacement rate (which is 2.1), evidence from South America is that women given the choice and access to edcuation and contraception have fewer children and birth rates are falling there (Brazil at 1.86 from 2.81 in 1990). The same will be true for the rest of the world (eg India 2.63 from 3.92 in 1990).
I know there are all kinds of privacy concerns with this, but it would be great if the technology could be used to identify people that are depressed and get them treatment.
I'm not advocating that for the general public, but maybe for special situations, like people that have a history of depression and want monitoring. Or people taking medicine for depression but their doctors are still calibrating the dosage.
Guys its simple. If I'm reading Slashdot at 1.30am then something must be wrong.
Most of the people I know who take sleeping pills are not necessarily the most stable people in the world to begin with. Sorry to all you Ambien fans.
Theoretically, yes. In practice I don't think so.
This 'confounding by indication' is one of the biggest problems in pharmacoepidemiology. We know that people take meds because there is something wrong with them. We also suspect that taking certain meds over a long period of time is bad for you, particularly if you are already at high risk. So how can you separate those effects? A lot of statisticians spend a lot of time thinking about this, and 'adjusting for everything you can think of', propensity scoring and very tight matching of cases and controls seem to be the most often used solutions. None of these is satisfactory as they obviously don't adjust for things you can't measure. Use of instrumental variables is another possibility but there is rarely a good instrument to use.
Ideally you would run a randomised trial of a med to check whether death rates or adverse drug reactions are higher in the group taking them, but this is impractical because often the required trial would be enormous (massively expensive and time consuming), would have to recruit many of the 'high risk' people that are the groups most at risk of excess mortality but are usually not recruited into trials, and could only really examine one compound at a time. Also trials exclude people taking many other medications, or with comorbid medical conditions, because these may be unsafe and would again dilute the true effects - however it is likely that unknown drug-drug interactions are the cause of a lot of the problems we think we are seeing.
It's easy to snipe at this kind of research since its 'correlation not causation' but this really is the best that is possible at the moment when trying to answer these extremely important questions regarding drug safety. If anybody has any better ideas we'd be glad to hear them.
These granted software patents are ridiculous. Patenting detection of trending topics and search queries? Jeez. Companies like Google and Apple are collectively abusing the system and patenting every single thing they can think of, most of which are outright obvious.
It's not the corporations' fault in this case.. the fact that software patents exist mean that you have patent the obvious before somebody else does.
R is good for analysis (although it has a steep learning curve) - but it seems to be that the poster has more of a data management problem than an analysis one. 'Administrators' are unlikely to be wanting a inference or projections, they will just be wanting informative series of data on usage (nice graphs and tables etc). So a good database solution is probably the most important step, then exporting tables into something that will make nice reports (Excel might be okay) is the next.
I'm a statistician so I know the strengths and limitations of most stats packages - and I don't think a stats package is the correct approach here. But I would agree with GP, we need to know exactly what you are trying to do.
Indeed, give fair notice and make the move if you think the new company is a good match for you.
Loyalty is a good thing, but sometimes it also holds back the people you are being loyal to.
Agreed - you may not be as indespensible as you think you are, and your manager should have a contingeny for if you (say) get hit by a bus. Another compromise may be to give a longer notice period and offer to consult for the old place for a while?
I got my own grant money so it should be between me and my funders how I spend it (and how I assign my own copyright).
If your grant money is funded, in whole or in part, by the taxpayer then the taxpayer has a right to the fruits of the money he has spent on your research.
If it's a privately funded grant, then by all means do as you wish.
I don't quite agree - there's a subtle but important distinction between being employed by the taxpayer and being grant funded by the taxpayer. An employee is being asked by the taxpayer to do something specific that they are interested in, and the government retains control of the direction of the work and the ownership of the results. A grant funded scientist is being enabled by the taxpayer to do work that they (both parties) think is important (it's a charitable donation in a sense) - much like grant funded artists or museums there is no expectation that all of the output is then owned by the taxpayer.
The public do have a right to ensure that the results are disseminated in a way that is useful to them, this does not necessarily mean open access for primary research findings - as in some cases this can do more harm than good (although as I've said earlier I think in most cases open access is a good thing).
That's why what Princeton is doing is such a great thing. It allows you to still submit and be published in a prestigious journal, but hide behind the university's legal team when it comes to posting your publication where everyone can access it. Google scholar does an amazing job of finding publicly available copies of scientific publications on a researcher's personal website, etc, so this is a big step towards open-access scientific publication without having to sacrifice your career.
Depends if the journals will accept this. It would be no great loss to any journal in particular to not accept work from Princeton - it may only (in the short term) harm their own researchers if other universities don't follow (though I see from the summary there is a waiver - will be interesting to see how that works out).
Although I'm in favour of open access I get a bit pissed off with Universities dictating publication policy like this. I got my own grant money so it should be between me and my funders how I spend it (and how I assign my own copyright).
I'm currently applying for grad schools - and nothing is more frustrating than finding all of a professor's research "hidden" behind pay-journals... what a step in the right direction.
Trinity College Dublin (my institution) also does this in a way. We are obliged to send anything we publish to the college's open access server. A lot of institutions and funding bodies have similar policies or are putting them in place.
Slashdot Mirror
Right, just like the Jews developed resistance to Xyklon-B, or how the victims of the Spanish Inquisition developed resistance to fire and trauma.
The thing about bacteria is that for many species, a "generation" is on the order of 20 or 30 minutes. They can evolve "slowly" over thousands of generations, but that might only be a few months of our time.
Bacteria also have Horizontal Gene Transfer which speeds things up a lot.
What's with all that sciency guff?
I want a candidate with character, morals, one who represents my beliefs on abortion and on the deficit and whether or not we should reduce spending or increase taxes.
To be honest that's probably the kind of thinking you should be engaing in.
Good scientists make terrible leaders, as they either tend to believe themselves to be experts in everything despite only knowing much about the migration habits of snow geese, or being so balanced and equivocal (ie scientific) about every issue that they will never be able to make a decision.
>
You can not though compare them to the people who sneak here and pop out babies and look for welfare with forms printed in their native language.
These are not the people that strengthen a country.
No, but the babies probably are.
I don't watch, and don't much care, and my life is far better for it.
I had started this reply to disagree with you about this but I'm struggling to come up with a good reason. I keep watching the news (from a lot of different sources), but mostly I think because I find it interesting.
or any other news channel
RT's coverage of Syria shooting down Turkish plane is enlightening on that point:
Turkey's downed jet: NATO action in disguise?
It's quite a thorough analysis, and a totally different spin than anything I had heard on BBC, RTE (Ireland) or US news sources. I guess all news sources are biased, and you need to take more than one point of view if you want to be able to form your own balanced opinion.
Good point. I've never actually taught social science students, only social science professionals, which is always going to be easier.
Ironically, I've found many computer science majors are not very versed in the ramification of statistics either. I think it has something to do with the binary world that they envisage or something like that.
The most common example is the "1-in-a-million" mentatlity many computer science majors have when talking about bugs or special-case code paths. You'd think they'd know better as they can often quote all sorts of statistical sort or database traversal, O(log n), big-o little-o, etc, but when you get them with a common sense thing about code performance issue, they appear to get some sort of temporary lobotomy.
Hence 'data-mining'.
...especially as regards the use of mathematics in the interpretation of 'data' where the soft sciences have such a 'hand wavy' approach to cause and effect.
To me, economics is a prime example. Forgive me if I'm off base in in my belief that economics is both sociological and soft(headed), but tyring to measure human behavior in the absence of an accounting for political corruption within this purely human realm and leaving the so-called black market beyond it's consideration leaves the inclusion of economics within the realm of 'science' suspect.
I would haved greatly appreciated any attempt by a professor to explain the difference between soft science and hard science, especially if it included an math based explanation of the nuance between these different domains.
Soft sciences are typically about trying to solve 'wicked' problems, which are those that are generally impossible to completely solve (end poverty or health inequality, understand crime, migration, or human behaviour in general etc). Hard scientists typically try to solve problems that are relatively much easier because they have a simple concrete goal (put a man on the moon, make a bomb, cure some disease)
Soft scientists need a much stronger theoretical framework to interpret their data, because of the absence of any really testable mechanisms for the effects they observe. This can come across as 'hand wavy' but it really isn't. Your economics example isn't entirely fair, some economic models will include corruption and black markets etc and others wont, just as some physics models include relativistic effects and others don't. A good scientist has to choose the right model to approah any problem, regardless of discipline.
I've been working in an inter-disciplinary group and have had the opportunity to see medics and economists try to work together. The two cultures are very different in their scientific approach, both consider the other to be unnecessarily picky about some aspects of the work while not being rigorous enough in others. Eg economists spend a huge amount of their time trying to prove causation in observational data, while medics will typically wave this away if they think the causal effect is likely enough. On the other hand economists tend not to contextualise their results well enough, while medics will see the bigger picture in terms of building on existing science.
My thoughts exactly.
He might end up losing some of his elitist attitude before the course is over. It would be better if lost the attitude ahead of time, and approached the experience like he was at least teaching the same species.
Indeed. I teach statistics to mathematicians, biologists, psychologists and social scientists and I would say the social scientists 'get' the principles of statistics better than the 'hard' scientists do. The main reason is that soft scientists (which is a horrible term) can think about uncertainty and its consequences, whereas hard scientists (mathematicians included) are unhappy if they don't have a yes/no answer to a question. Obviously this is a generalisation but it may inform your approach to teaching.
Also, statistics is not 'just math'. I know this because I can do statistics but I can't do math(s) any more. :-)
Likewise, using genetech to produce insuline should be stopped immediately. The only way to cut down on diabetes is to improve your diet. Hell, I throw healthy veggies away daily because everybody wants to eat hamburgers and drink coke all the time. Diabetes prone people will die off naturally over time and than we have solved the problem.
Same argument can be made about most avenues of progress. You sir, are anti-progressive. Look into a mirror, you do not deserve a /. id (so says the AC).
People generally reproduce before they die of type-2 diabetes related complications, so I don't think natural selection will do it for us here.
hah, sane population control? not fucking likely.
It will happen with improved education and rights for women. Birth rate in Europe (1.59 births per female) / US (2.0 births per female) / Japan (1.37 births per female) is far below replacement rate (which is 2.1), evidence from South America is that women given the choice and access to edcuation and contraception have fewer children and birth rates are falling there (Brazil at 1.86 from 2.81 in 1990). The same will be true for the rest of the world (eg India 2.63 from 3.92 in 1990).
I know there are all kinds of privacy concerns with this, but it would be great if the technology could be used to identify people that are depressed and get them treatment.
I'm not advocating that for the general public, but maybe for special situations, like people that have a history of depression and want monitoring. Or people taking medicine for depression but their doctors are still calibrating the dosage.
Guys its simple. If I'm reading Slashdot at 1.30am then something must be wrong.
you just went full retard.
At least he can 'make his own cables and such'.
Most of the people I know who take sleeping pills are not necessarily the most stable people in the world to begin with. Sorry to all you Ambien fans.
Theoretically, yes. In practice I don't think so.
This 'confounding by indication' is one of the biggest problems in pharmacoepidemiology. We know that people take meds because there is something wrong with them. We also suspect that taking certain meds over a long period of time is bad for you, particularly if you are already at high risk. So how can you separate those effects? A lot of statisticians spend a lot of time thinking about this, and 'adjusting for everything you can think of', propensity scoring and very tight matching of cases and controls seem to be the most often used solutions. None of these is satisfactory as they obviously don't adjust for things you can't measure. Use of instrumental variables is another possibility but there is rarely a good instrument to use.
Ideally you would run a randomised trial of a med to check whether death rates or adverse drug reactions are higher in the group taking them, but this is impractical because often the required trial would be enormous (massively expensive and time consuming), would have to recruit many of the 'high risk' people that are the groups most at risk of excess mortality but are usually not recruited into trials, and could only really examine one compound at a time. Also trials exclude people taking many other medications, or with comorbid medical conditions, because these may be unsafe and would again dilute the true effects - however it is likely that unknown drug-drug interactions are the cause of a lot of the problems we think we are seeing.
It's easy to snipe at this kind of research since its 'correlation not causation' but this really is the best that is possible at the moment when trying to answer these extremely important questions regarding drug safety. If anybody has any better ideas we'd be glad to hear them.
I think the larger question is, does.much of anyone in the first world even really give it a second thought?
Isn't this the.price they pay for.taking these.jobs.from us where it used to be done for a.living wage and fair.working conditions?
We do care when it comes to clothes or food - there's a huge 'fair trade' market for these commodities and so there should be for electronics also.
posting to undo accidental bad moderation
The Lancet (an Elsevier journal) does actually pay some reviewers, this could be the issue.
These granted software patents are ridiculous. Patenting detection of trending topics and search queries? Jeez. Companies like Google and Apple are collectively abusing the system and patenting every single thing they can think of, most of which are outright obvious.
It's not the corporations' fault in this case.. the fact that software patents exist mean that you have patent the obvious before somebody else does.
R is good for analysis (although it has a steep learning curve) - but it seems to be that the poster has more of a data management problem than an analysis one. 'Administrators' are unlikely to be wanting a inference or projections, they will just be wanting informative series of data on usage (nice graphs and tables etc). So a good database solution is probably the most important step, then exporting tables into something that will make nice reports (Excel might be okay) is the next.
I'm a statistician so I know the strengths and limitations of most stats packages - and I don't think a stats package is the correct approach here. But I would agree with GP, we need to know exactly what you are trying to do.
That's an interesting point of view.
Wish i had mod points.
What a redundant comment. Wish I had mod points
Umm? People get guns pulled on them/mugged for pocket change every day...
I back this claim up with real world evidence of it happening to multiple of my friends.
I suggest you move.
Indeed, give fair notice and make the move if you think the new company is a good match for you.
Loyalty is a good thing, but sometimes it also holds back the people you are being loyal to.
Agreed - you may not be as indespensible as you think you are, and your manager should have a contingeny for if you (say) get hit by a bus. Another compromise may be to give a longer notice period and offer to consult for the old place for a while?
I got my own grant money so it should be between me and my funders how I spend it (and how I assign my own copyright).
If your grant money is funded, in whole or in part, by the taxpayer then the taxpayer has a right to the fruits of the money he has spent on your research.
If it's a privately funded grant, then by all means do as you wish.
I don't quite agree - there's a subtle but important distinction between being employed by the taxpayer and being grant funded by the taxpayer. An employee is being asked by the taxpayer to do something specific that they are interested in, and the government retains control of the direction of the work and the ownership of the results. A grant funded scientist is being enabled by the taxpayer to do work that they (both parties) think is important (it's a charitable donation in a sense) - much like grant funded artists or museums there is no expectation that all of the output is then owned by the taxpayer.
The public do have a right to ensure that the results are disseminated in a way that is useful to them, this does not necessarily mean open access for primary research findings - as in some cases this can do more harm than good (although as I've said earlier I think in most cases open access is a good thing).
That's why what Princeton is doing is such a great thing. It allows you to still submit and be published in a prestigious journal, but hide behind the university's legal team when it comes to posting your publication where everyone can access it. Google scholar does an amazing job of finding publicly available copies of scientific publications on a researcher's personal website, etc, so this is a big step towards open-access scientific publication without having to sacrifice your career.
Depends if the journals will accept this. It would be no great loss to any journal in particular to not accept work from Princeton - it may only (in the short term) harm their own researchers if other universities don't follow (though I see from the summary there is a waiver - will be interesting to see how that works out).
Although I'm in favour of open access I get a bit pissed off with Universities dictating publication policy like this. I got my own grant money so it should be between me and my funders how I spend it (and how I assign my own copyright).
I'm currently applying for grad schools - and nothing is more frustrating than finding all of a professor's research "hidden" behind pay-journals... what a step in the right direction.
Trinity College Dublin (my institution) also does this in a way. We are obliged to send anything we publish to the college's open access server. A lot of institutions and funding bodies have similar policies or are putting them in place.