And we're yet to see free-market capitalism except for brief periods of time - eg the Wild West - it seems to look remarkably like the worst forms of anarchism. If Trump doesn't get impeached we may well see it again,,, briefly.
The Green Party was consistent in being pro EU. They are generally excluded from the media debates as -
(1) the First Past the Post favours the largest parties and the only other party that small is the UKIP - who make far better headlines for the media.
(2) The print media are 80% owned by 5 anti-EU individuals who stand to gain more media power in a post-Brexit UK.
(3) The UK Civil Service (completely anti-EU) are also very antagonistic towards their communitarian ethos.
All very true - something pseudo-random is needed, perhaps, coupled with changing the way we test for driving ability which, in the UK at least, has been getting easier recently.
The requirement for parallel parking was removed in the past decade because so many people found it difficult.
The least likely change is to enforcement I suspect.
My wife gave up driving after spending some time riding motorcycles and noting how 'safe' car drivers seem to feel.
Good points, fortunately in the UK we have very few multiple lane roads and the removals are generally done where accidents show that people tend to speed because they feel too safe and should be reminded of both their own and others' vulnerability.
It's recently happened near where I live, on a long slow bend where people were speeding up just before the crest of a hill and hugging the centre due to not seeing over that crest. The results have been immediate, people now tend towards the sides of the road and drive at a speed appropriate to the dangers.
The 'theorists' predicting the end-of-the-world still claim that the evidence is 'wrong' and probably always will.
Mebbe we should reward the understanding of stats rather than maths?
Great comments:-) - notice how the navy's own page says "The craft shattered all previous endurance records [having previously noted the navy's 40-minute flight in 1924] performed by similar, propeller driven, fossil fuel and battery-powered UAVs by completing an uninterrupted 26-hour flight carrying a five-pound payload."?
The Internationally agreed Radiocarbon calibration curve (IntCal) - co-ordinated from Belfast University - takes info from ice-cores, lake sediment cores, tree-rings, corals, etc from the Southern and Northern hemispheres (there's an offset between them) puts them together (this work is done by statisticians using specially developed methods rather than other scientists using off-the-shelf techniques) and although some scientists would rather that only their work was used (as they can then claim whatever 'accuracy' they wish to claim) independent verification of lab practices is extremely useful in the work. The most recent published work dates back to 50,000 years BP ('before present' where 'present' is 1950) and the next set of curves (IntCal 12) - being worked on at the moment will take it back further. Abstract for IntCal 09 - http://eprints.whiterose.ac.uk/10694/
I'm mainly with you on that one, but SPSS is also a reckless black box as far as I'm concerned. Too many of the options have no statistical legitimacy, merely having been found useful in the past as quick fixes.
Given the complexity of interrelated issues in medical stats, I'd prefer people to get serious about their science and use Bayesian methods so that competing theories can be compared for a wide range of parameters. Harnessing stats to the classical reasoning that has no validity in applications was only responsible when computing power was the limiting factor.
R and C/C++ would be my preferred current vectors of persuasion. Best wishes for your mission...
You have my sympathy:-) It does sound like your boss has a tight grip - hopefully s/he does know that outliers can be random artifacts..!
One of my favourite examples of this sort of thing is a radiocarbon calibration package (OxCal), put together by a physicist at Oxford University which has a popular radio-carbon lab.
This software provides access to an option that was only originally there to do quick and dirty tests of the software itself, it has proved to be the most commonly used function (as cited in journals) despite being a 'black box' with no methodological basis because it 'gives the smoothest curve'! What a life, huh?
in a word, yes, check out almost any medical stats methodology - it looks sort of right if you have only degree level maths but, eg, statisticians have pretty much given up on pointing out that treating binned averages of a population as raw data typically invalidates the method under consideration, rendering the results speculative at best.
researchers will tend to insist that what they have handed over is raw data because they have (or a research associate, or Excel! has) only performed a few simple transformations on it and, that being many months ago, probably have forgotten the fact. one can either keep performing extra (unpaid and unasked for) analyses showing that this distribution verges on the impossible (and risk not be asked for help in future) or shut up and get cited and allow your reputation to grow
having said that, the same is true for many scientific practitioners and, indeed, the majority of published journal papers - the peer review generally doesn't extend to a competent mathematical practitioner (still less frequently a statistician) and most academics do not appear to consider that anything beyond their (often high school- or graduate-level) understanding of mathematics is required, after all (like the paper concerned here) building on previously published and highly cited work of little worth is all that's required for a career
I call Trollshit! The classical frequentist assembly of statistical theorums has only disparate theoretical foundations (which indeed often are mutually antagonistic) - that nice distribution curve and 2-sigma point, what were they based on?
With Bayes' theorum, models are created such that rival theories can be compared, with a common framework, to all *available* data *plus* theoretical biases, more data can be added and different theories applied. In almost no circumstances is this 'safe' mathematically (although it's often done) in the frequentist world. As for confidence intervals... as these are based on what the 'experts' of the time will wear, these are simply loose adoptions of a part of the Bayesian paradigm without the rigour.
I must express an interest, here, my partner is a Professor (in the British sense, rather than 'lecturer' as I believe it to be used in the US) of Bayesian statistics having lectured in archaeology (a discipline in its' own right here in Europe) and seen how dreadful was the state of maths in the physical sciences that handled data that lead to 'results' for archaeologists.
It's still as bad in most subjects where maths is the prerequisite discipline: maths is most often a discipline for people with a belief system resembling the quote above. Statisticians have early on to learn that the specialist they are taking data from has often, for example, binned and averaged the data already (by the use of specific lab protocols or, sometimes, software) and doesn't begin imagine that this will have an impact on the stats methodology and so will not mention the fact. Often this invalidates the stats method used but no-one will know and so the project is wasted effort but published nevertheless and enters the canon of 'wisdom'.
Seeing the state of climate research just makes the numerate weep - the same old frequentist methods, modified slightly, trotted out without examination by people who know that you won't get sacked for using a 'tried and trusted' method even where it's patently wrong to statistical experts who don't stick their neck above the parapet. It's no wonder that the results of very few science papers in 'learned journals' can be repeated, often the method claimed hasn't been used (their 'data' wasn't raw) even though the authors sincerely believe that it has.
When I was in the British military (in the early 1980's) we took it for granted that the main difference, as it affected world security, between the USA and USSR was that only a tiny proportion of the Soviet people believed their state propaganda. We certainly believed little that came out of the BBC.
Slashdot Mirror
Southern Reach trilogy, psychologically thrilling...
And we're yet to see free-market capitalism except for brief periods of time - eg the Wild West - it seems to look remarkably like the worst forms of anarchism. If Trump doesn't get impeached we may well see it again,,, briefly.
The Green Party was consistent in being pro EU. They are generally excluded from the media debates as - (1) the First Past the Post favours the largest parties and the only other party that small is the UKIP - who make far better headlines for the media. (2) The print media are 80% owned by 5 anti-EU individuals who stand to gain more media power in a post-Brexit UK. (3) The UK Civil Service (completely anti-EU) are also very antagonistic towards their communitarian ethos.
All very true - something pseudo-random is needed, perhaps, coupled with changing the way we test for driving ability which, in the UK at least, has been getting easier recently. The requirement for parallel parking was removed in the past decade because so many people found it difficult. The least likely change is to enforcement I suspect. My wife gave up driving after spending some time riding motorcycles and noting how 'safe' car drivers seem to feel.
Good points, fortunately in the UK we have very few multiple lane roads and the removals are generally done where accidents show that people tend to speed because they feel too safe and should be reminded of both their own and others' vulnerability. It's recently happened near where I live, on a long slow bend where people were speeding up just before the crest of a hill and hugging the centre due to not seeing over that crest. The results have been immediate, people now tend towards the sides of the road and drive at a speed appropriate to the dangers. The 'theorists' predicting the end-of-the-world still claim that the evidence is 'wrong' and probably always will. Mebbe we should reward the understanding of stats rather than maths?
No wonder you're anonymous.
Both posts by the same AC?
Great comments :-) - notice how the navy's own page says "The craft shattered all previous endurance records [having previously noted the navy's 40-minute flight in 1924] performed by similar, propeller driven, fossil fuel and battery-powered UAVs by completing an uninterrupted 26-hour flight carrying a five-pound payload."?
The Internationally agreed Radiocarbon calibration curve (IntCal) - co-ordinated from Belfast University - takes info from ice-cores, lake sediment cores, tree-rings, corals, etc from the Southern and Northern hemispheres (there's an offset between them) puts them together (this work is done by statisticians using specially developed methods rather than other scientists using off-the-shelf techniques) and although some scientists would rather that only their work was used (as they can then claim whatever 'accuracy' they wish to claim) independent verification of lab practices is extremely useful in the work. The most recent published work dates back to 50,000 years BP ('before present' where 'present' is 1950) and the next set of curves (IntCal 12) - being worked on at the moment will take it back further. Abstract for IntCal 09 - http://eprints.whiterose.ac.uk/10694/
meson@f2s(dot)com - many thanks...
difficult to reply to this.... did you mean something when you typed it? keep taking the tablets ;-)
No, I can't believe it either.
the robotic overlords invite us to the news of the millenia!
Given the complexity of interrelated issues in medical stats, I'd prefer people to get serious about their science and use Bayesian methods so that competing theories can be compared for a wide range of parameters. Harnessing stats to the classical reasoning that has no validity in applications was only responsible when computing power was the limiting factor. R and C/C++ would be my preferred current vectors of persuasion. Best wishes for your mission...
Hardly any peer reviewers know enough about how to rate the maths part of the methodology, even where the Principle Investigator mentions one.
Mainly you get a bit of waffle that might mention a software package or previous piece of uninspiring work. Given no alternatives they get funded.
One of my favourite examples of this sort of thing is a radiocarbon calibration package (OxCal), put together by a physicist at Oxford University which has a popular radio-carbon lab.
This software provides access to an option that was only originally there to do quick and dirty tests of the software itself, it has proved to be the most commonly used function (as cited in journals) despite being a 'black box' with no methodological basis because it 'gives the smoothest curve'! What a life, huh?
researchers will tend to insist that what they have handed over is raw data because they have (or a research associate, or Excel! has) only performed a few simple transformations on it and, that being many months ago, probably have forgotten the fact. one can either keep performing extra (unpaid and unasked for) analyses showing that this distribution verges on the impossible (and risk not be asked for help in future) or shut up and get cited and allow your reputation to grow
having said that, the same is true for many scientific practitioners and, indeed, the majority of published journal papers - the peer review generally doesn't extend to a competent mathematical practitioner (still less frequently a statistician) and most academics do not appear to consider that anything beyond their (often high school- or graduate-level) understanding of mathematics is required, after all (like the paper concerned here) building on previously published and highly cited work of little worth is all that's required for a career
how many lamp-posts there are in London? He's a well-meaning right-wing buffoon.
Finally, someone who has done their research.
With Bayes' theorum, models are created such that rival theories can be compared, with a common framework, to all *available* data *plus* theoretical biases, more data can be added and different theories applied. In almost no circumstances is this 'safe' mathematically (although it's often done) in the frequentist world. As for confidence intervals... as these are based on what the 'experts' of the time will wear, these are simply loose adoptions of a part of the Bayesian paradigm without the rigour.
I must express an interest, here, my partner is a Professor (in the British sense, rather than 'lecturer' as I believe it to be used in the US) of Bayesian statistics having lectured in archaeology (a discipline in its' own right here in Europe) and seen how dreadful was the state of maths in the physical sciences that handled data that lead to 'results' for archaeologists.
It's still as bad in most subjects where maths is the prerequisite discipline: maths is most often a discipline for people with a belief system resembling the quote above. Statisticians have early on to learn that the specialist they are taking data from has often, for example, binned and averaged the data already (by the use of specific lab protocols or, sometimes, software) and doesn't begin imagine that this will have an impact on the stats methodology and so will not mention the fact. Often this invalidates the stats method used but no-one will know and so the project is wasted effort but published nevertheless and enters the canon of 'wisdom'.
Seeing the state of climate research just makes the numerate weep - the same old frequentist methods, modified slightly, trotted out without examination by people who know that you won't get sacked for using a 'tried and trusted' method even where it's patently wrong to statistical experts who don't stick their neck above the parapet. It's no wonder that the results of very few science papers in 'learned journals' can be repeated, often the method claimed hasn't been used (their 'data' wasn't raw) even though the authors sincerely believe that it has.
Or the US... Fox!
Someone, please, mod this up!
With your attitude why wait?
Lots of posts here with '4' or '5' - few as well considered.
When I was in the British military (in the early 1980's) we took it for granted that the main difference, as it affected world security, between the USA and USSR was that only a tiny proportion of the Soviet people believed their state propaganda. We certainly believed little that came out of the BBC.