Most cars today don't come with enabled cellular radios (or cellular radios at all for that matter). The luxury ones (like Tesla) do. The others, not so much. The subscriptions are expensive.
RDS? For transmitting what song is playing on FM stations? Hooking that up to do firmware updates on a car's computer sounds like a great idea!
Chat apps aren't exactly hard to install. Most people probably already have several. History has also shown that user retention on chat apps isn't particularly robust. Facebook is discovering this with their own Messenger, in fact, possibly a motivation behind acquiring WhatsApp.
Several hundred million users, some of whom have already pledged to quit since Facebook bought it, and many of whom will quit when the first annual renewal comes around and/or Facebook decides to introduce ads.
Besides, FB already has most of their address books. It's begged for mine often enough I'm surprised I haven't accidentally hit yes yet.
How so? Glass is mostly useless without a smartphone. So it gives you the option to replace your smartphone with a smartphone and a low resolution display you wear on your face.
I suspect many WhatsApp users have it free. I do. Anyone who used it before they "monetized" doesn't pay. If they change that, or if Facebook starts mucking with it, I'll use something else.
There are a LOT of free texting programs, and it takes about a weekend to write another one. Extracting sixteen billion dollars from WhatsApp is going to be an exercise in futility. Hopefully the WhatsApp people are laughing their way to the bank (and selling their FB stock as fast as they can).
Bayesian statistics still needs multiple comparison correction, and it's usually more complicated. Bayesian stats lets you quantitatively combine the results from multiple tests of the same thing, usually in different datasets, i.e. different experiments testing the same hypothesis. If you're doing that with frequentist stats you don't need multiple comparison correction either.
If you're testing multiple different things, usually in the same dataset, you need to do multiple comparison correction, Bayesian or otherwise.
"If you get a bad P value, then it means that the data is unlikey to have come from that hypothesis."
Insignificant p-values don't meet anything beyond "my standard of evidence was not met." It's a common mistake. Suppose you get p=0.1. That's pretty universally considered non-significant. But what it actually means is that there's a 90% chance (assuming no prior information and no screwups) that the alternative hypothesis is true. That's a long way from "the data is unlikely to have come from that hypothesis."
Even (especially) when the p-values get very large, you can't draw any meaningful conclusions. A p-value of 0.99 could mean that the null hypothesis is very likely, OR that you simply have too much unexplained variance and too small a sample. You have to draw out the confidence intervals and find out. Only in the former case, provided the maximum likely effect is smaller than what you consider relevant, you have evidence against the hypothesis. In the latter case you have evidence only that you need to improve your model, measurements and/or collect more data.
You seem to have the rest backwards. You often collect some pilot data (or use someone else's) and then propose a model, but that's not testing your model. Evidence for or against a model comes from data collected AFTER you've proposed it. You generate hypotheses and design experiments to try to show the model is incorrect, collect the data, and test it. If the data doesn't fit, you discard the model. If it does fit, it counts as evidence in its favour. That's not the minimum standard, it is the standard.
Then you're not a statistician. There's a reason data mining is a dirty word in science.
Before you start working you need to have a hypothesis like "woman are shorter on average than men." You then find or collect a dataset of height measurements from a sample of women and men and do a test on the means. That gives you a p-value, which is what you report. If you do it in two separate datasets, you get two p-values and you report both. You don't correct either one for multiple comparisons, and whoever is reading your paper sees that you did an experiment and then replicated it. If both showed a significant difference your evidence is stronger. If they conflicted, it is weaker. If you're actually good at stats you can combine the two with Bayes's theorem and find out quantitatively how much stronger or weaker.
What you're describing is, yes, how a lot of poor research happens. Your hypothesis is something like "there is a difference between these two groups (I hope)", which isn't a proper hypothesis at all. Then you go fishing.
The difference is that in the first case you're testing the same thing, that you planned to look at in advance, in one or more datasets. In the second you're testing a bunch of different things, in one or more datasets, and if you find one that works you'll then claim a discovery. In the first case, in random data, if your threshold is a=0.05 you won't get more than 1 in 20 positive results, and everyone will see that. In the second, you expect to find a difference in 1 in 20 experiments; the multiple positive results don't strengthen each other because they're testing different things.
In either case, if you lie and say you only did one test, you're committing fraud.
Human trials are incredibly expensive. You don't just do 100 and take the best one. Also, various people, including regulatory agencies, are wise to that. If you have a bunch of registered trials without results it's going to be looked at with suspicion.
There have been cases of abuse, but pharma generally doesn't want to outright fake results. Developing drugs is expensive, but getting sued into oblivion is too.
Not usually. Most people's shoes are quite good insulators. The car discharges when you touch it with the metal filling hose. The problem is that YOU don't always discharge before the gas starts flowing. It takes a few things to go wrong at the same time but it has cussed explosions. There has never been a reliable report of a cell phone causing such a thing.
Seriously, the thing that causes gas pumps to explode is static electricity, usually from entering and exiting your car. While everyone is worried about cell phones, how many people make sure to touch their car with bare hands before pumping?
You don't need to correct for experiments done on other datasets. If it's multiple experiments on the same dataset, whoever is in charge of that dataset should be keeping track. Even then, you only need to correct for multiple tests of the same, or similar, hypothesis. That's mostly a problem when your hypothesis is "something happened," which it should never be, but it is all too often.
It's also impossible to tell if the other guy made the whole thing up. Fraud is detected via replication. Generally though, it's pretty easy to detect people doing it inadvertently - they publish all their "p-values." I suppose if people actually get better at doing stats some of the inadvertent stuff will turn into harder to detect fraud.
Clinical trials are actually required to be pre-registered with one of a few tracking agencies if they are to be accepted by the FDA and other similar agencies. There are a few problems, but it's much better than it used to be.
Any statistical test requires that you apply it appropriately. If you don't do so the result is called a "mistake," not a "coincidence" OR a "correlation".
Please tell me they don't really call them 'q-values'?
A p-value IS corrected for multiple comparisons. If you did multiple comparisons and you didn't correct it, it ain't a p-value. A good term for those would be "the result of my fishing expedition."
A confidence interval is completely equivalent to a statement of p-value, mean and type of distribution. In fact, CIs are almost always calculated from that trio. It's just another way of showing the same information.
No. If someone writes a paper and claims that "This is true because p 0.05" then they are wrong regardless. The correct conclusion is "this result supports our hypothesis because p the threshold we set for minimal evidence." The point the article makes is correct, but it's not a problem with p-values, it's a problem with the conclusions people draw from them. His professor is absolutely correct, supposing that the paper's he's talking about aren't meaningless bullshit to start with.
Um, no. Your criticisms on't make sense. You're falling into a misunderstanding that is perpetuated because theoretical statisticians are so careful about how they define things, particularly when Bayesians might be looking over their shoulders.
A p-value is the probability that accepting your statistical hypothesis (rejecting the null hypothesis) would be an error. This is equivalent to saying that the p-value is the probability that, picking a random run out of many runs of your experiment, you'd expect to get your result or something more extreme, purely by chance.
It shouldn't be, and I hope it never will be. If you use a non-informative prior for your Bayesian analysis in most cases you're just doing extra work to get the same result. If you use an informative prior you're colouring your results with your preconceptions. The reader, or the author of a meta-analysis, is the one who should be doing the Bayesian analysis, using their own priors. By all means, report a Bayes factor to make the meta-analysis easier, but also report a p-value, which is a good metric for a single experiment.
There's no such thing. A correlation test always comes with a p-value that gives you an idea of how likely your observations are to be a coincidence rather than a correlation.
The biggest problem with the way most people do statistics is that they don't have adequate statistical reasoning skills. The problem is in the design of experiments and analyses, before you ever get to punching the buttons in your stats package of choice. The differences you get from punching the wrong button are really very minor compared to things that happen all the time, like drawing conclusions based on tests you didn't do (the difference of differences error is an excellent example: half of all high impact neuroscience papers that can make this mistake do so).
The wild world of nutrition recommendations is mostly the way it is because it's all made up. The scientific evidence amounts to "get the basic amounts of macro and micro-nutrients to avoid disease", "eat a variety of foods", "eat vegetables" and a bunch of very basic, and specific things. Those basic and specific results are then wildly extrapolated, mostly by talk show hosts, celebrities, and people looking to cash in on gullible dieters.
Other way around, and not quite true for a properly formulated hypothesis.
Frequentist statistics involves making a statistical hypothesis, choosing a level of evidence that you find acceptable (usually alpha=0.05) and using that to accept or reject it. The statistical hypothesis is tied to your scientific hypothesis (or it should be). If the standard of evidence is met or exceeded, the results support the hypothesis. If not, they don't mean anything.
HOWEVER, if you specify your hypothesis well, you include a minimum difference that you consider meaningful. You then calculate error bars for your result and, if they show that your measured value is less than the minimum you hypothesized, that's evidence supporting the negative (not the null) hypothesis: any difference is so small as to be meaningless.
I am not a fan of everyone using Bayesian techniques. A Bayesian analysis of a single experiment that gives a normally distributed measurement (which is most of them) with a non-informative prior is generally equivalent to a frequentist analysis. Since scientists already have trouble doing simple frequentist tests correctly, they do not need to be doing needless Bayesian analyses.
As for informative priors, I don't think they should ever be used in the report of a single experiment. Report the p-value and the Bayes factor, or the equivalent information needed to calculate them. Since an informative prior is inherently subjective, the reader should be left to make up his own mind what it is. Reporting the Bayes factor makes meta-analyses, where Bayesian stats SHOULD be mandatory, easier.
Slashdot Mirror
There are a large number of apps to learn French.
Most cars today don't come with enabled cellular radios (or cellular radios at all for that matter). The luxury ones (like Tesla) do. The others, not so much. The subscriptions are expensive.
RDS? For transmitting what song is playing on FM stations? Hooking that up to do firmware updates on a car's computer sounds like a great idea!
Chat apps aren't exactly hard to install. Most people probably already have several. History has also shown that user retention on chat apps isn't particularly robust. Facebook is discovering this with their own Messenger, in fact, possibly a motivation behind acquiring WhatsApp.
Several hundred million users, some of whom have already pledged to quit since Facebook bought it, and many of whom will quit when the first annual renewal comes around and/or Facebook decides to introduce ads.
Besides, FB already has most of their address books. It's begged for mine often enough I'm surprised I haven't accidentally hit yes yet.
It's like GMail. Except you have to pay. It will be invite only for five years.
How so? Glass is mostly useless without a smartphone. So it gives you the option to replace your smartphone with a smartphone and a low resolution display you wear on your face.
I suspect many WhatsApp users have it free. I do. Anyone who used it before they "monetized" doesn't pay. If they change that, or if Facebook starts mucking with it, I'll use something else.
There are a LOT of free texting programs, and it takes about a weekend to write another one. Extracting sixteen billion dollars from WhatsApp is going to be an exercise in futility. Hopefully the WhatsApp people are laughing their way to the bank (and selling their FB stock as fast as they can).
Sigh. Know why they call the threshold for significance alpha?
http://en.wikipedia.org/wiki/T...
Bayesian statistics still needs multiple comparison correction, and it's usually more complicated. Bayesian stats lets you quantitatively combine the results from multiple tests of the same thing, usually in different datasets, i.e. different experiments testing the same hypothesis. If you're doing that with frequentist stats you don't need multiple comparison correction either.
If you're testing multiple different things, usually in the same dataset, you need to do multiple comparison correction, Bayesian or otherwise.
"If you get a bad P value, then it means that the data is unlikey to have come from that hypothesis."
Insignificant p-values don't meet anything beyond "my standard of evidence was not met." It's a common mistake. Suppose you get p=0.1. That's pretty universally considered non-significant. But what it actually means is that there's a 90% chance (assuming no prior information and no screwups) that the alternative hypothesis is true. That's a long way from "the data is unlikely to have come from that hypothesis."
Even (especially) when the p-values get very large, you can't draw any meaningful conclusions. A p-value of 0.99 could mean that the null hypothesis is very likely, OR that you simply have too much unexplained variance and too small a sample. You have to draw out the confidence intervals and find out. Only in the former case, provided the maximum likely effect is smaller than what you consider relevant, you have evidence against the hypothesis. In the latter case you have evidence only that you need to improve your model, measurements and/or collect more data.
You seem to have the rest backwards. You often collect some pilot data (or use someone else's) and then propose a model, but that's not testing your model. Evidence for or against a model comes from data collected AFTER you've proposed it. You generate hypotheses and design experiments to try to show the model is incorrect, collect the data, and test it. If the data doesn't fit, you discard the model. If it does fit, it counts as evidence in its favour. That's not the minimum standard, it is the standard.
Then you're not a statistician. There's a reason data mining is a dirty word in science.
Before you start working you need to have a hypothesis like "woman are shorter on average than men." You then find or collect a dataset of height measurements from a sample of women and men and do a test on the means. That gives you a p-value, which is what you report. If you do it in two separate datasets, you get two p-values and you report both. You don't correct either one for multiple comparisons, and whoever is reading your paper sees that you did an experiment and then replicated it. If both showed a significant difference your evidence is stronger. If they conflicted, it is weaker. If you're actually good at stats you can combine the two with Bayes's theorem and find out quantitatively how much stronger or weaker.
What you're describing is, yes, how a lot of poor research happens. Your hypothesis is something like "there is a difference between these two groups (I hope)", which isn't a proper hypothesis at all. Then you go fishing.
The difference is that in the first case you're testing the same thing, that you planned to look at in advance, in one or more datasets. In the second you're testing a bunch of different things, in one or more datasets, and if you find one that works you'll then claim a discovery. In the first case, in random data, if your threshold is a=0.05 you won't get more than 1 in 20 positive results, and everyone will see that. In the second, you expect to find a difference in 1 in 20 experiments; the multiple positive results don't strengthen each other because they're testing different things.
In either case, if you lie and say you only did one test, you're committing fraud.
Human trials are incredibly expensive. You don't just do 100 and take the best one. Also, various people, including regulatory agencies, are wise to that. If you have a bunch of registered trials without results it's going to be looked at with suspicion.
There have been cases of abuse, but pharma generally doesn't want to outright fake results. Developing drugs is expensive, but getting sued into oblivion is too.
Not usually. Most people's shoes are quite good insulators. The car discharges when you touch it with the metal filling hose. The problem is that YOU don't always discharge before the gas starts flowing. It takes a few things to go wrong at the same time but it has cussed explosions. There has never been a reliable report of a cell phone causing such a thing.
Intelligence?
Seriously, the thing that causes gas pumps to explode is static electricity, usually from entering and exiting your car. While everyone is worried about cell phones, how many people make sure to touch their car with bare hands before pumping?
You don't need to correct for experiments done on other datasets. If it's multiple experiments on the same dataset, whoever is in charge of that dataset should be keeping track. Even then, you only need to correct for multiple tests of the same, or similar, hypothesis. That's mostly a problem when your hypothesis is "something happened," which it should never be, but it is all too often.
It's also impossible to tell if the other guy made the whole thing up. Fraud is detected via replication. Generally though, it's pretty easy to detect people doing it inadvertently - they publish all their "p-values." I suppose if people actually get better at doing stats some of the inadvertent stuff will turn into harder to detect fraud.
Clinical trials are actually required to be pre-registered with one of a few tracking agencies if they are to be accepted by the FDA and other similar agencies. There are a few problems, but it's much better than it used to be.
Any statistical test requires that you apply it appropriately. If you don't do so the result is called a "mistake," not a "coincidence" OR a "correlation".
Please tell me they don't really call them 'q-values'?
A p-value IS corrected for multiple comparisons. If you did multiple comparisons and you didn't correct it, it ain't a p-value. A good term for those would be "the result of my fishing expedition."
A confidence interval is completely equivalent to a statement of p-value, mean and type of distribution. In fact, CIs are almost always calculated from that trio. It's just another way of showing the same information.
No. If someone writes a paper and claims that "This is true because p 0.05" then they are wrong regardless. The correct conclusion is "this result supports our hypothesis because p the threshold we set for minimal evidence." The point the article makes is correct, but it's not a problem with p-values, it's a problem with the conclusions people draw from them. His professor is absolutely correct, supposing that the paper's he's talking about aren't meaningless bullshit to start with.
Um, no. Your criticisms on't make sense. You're falling into a misunderstanding that is perpetuated because theoretical statisticians are so careful about how they define things, particularly when Bayesians might be looking over their shoulders.
A p-value is the probability that accepting your statistical hypothesis (rejecting the null hypothesis) would be an error. This is equivalent to saying that the p-value is the probability that, picking a random run out of many runs of your experiment, you'd expect to get your result or something more extreme, purely by chance.
It shouldn't be, and I hope it never will be. If you use a non-informative prior for your Bayesian analysis in most cases you're just doing extra work to get the same result. If you use an informative prior you're colouring your results with your preconceptions. The reader, or the author of a meta-analysis, is the one who should be doing the Bayesian analysis, using their own priors. By all means, report a Bayes factor to make the meta-analysis easier, but also report a p-value, which is a good metric for a single experiment.
There's no such thing. A correlation test always comes with a p-value that gives you an idea of how likely your observations are to be a coincidence rather than a correlation.
The biggest problem with the way most people do statistics is that they don't have adequate statistical reasoning skills. The problem is in the design of experiments and analyses, before you ever get to punching the buttons in your stats package of choice. The differences you get from punching the wrong button are really very minor compared to things that happen all the time, like drawing conclusions based on tests you didn't do (the difference of differences error is an excellent example: half of all high impact neuroscience papers that can make this mistake do so).
The wild world of nutrition recommendations is mostly the way it is because it's all made up. The scientific evidence amounts to "get the basic amounts of macro and micro-nutrients to avoid disease", "eat a variety of foods", "eat vegetables" and a bunch of very basic, and specific things. Those basic and specific results are then wildly extrapolated, mostly by talk show hosts, celebrities, and people looking to cash in on gullible dieters.
Other way around, and not quite true for a properly formulated hypothesis.
Frequentist statistics involves making a statistical hypothesis, choosing a level of evidence that you find acceptable (usually alpha=0.05) and using that to accept or reject it. The statistical hypothesis is tied to your scientific hypothesis (or it should be). If the standard of evidence is met or exceeded, the results support the hypothesis. If not, they don't mean anything.
HOWEVER, if you specify your hypothesis well, you include a minimum difference that you consider meaningful. You then calculate error bars for your result and, if they show that your measured value is less than the minimum you hypothesized, that's evidence supporting the negative (not the null) hypothesis: any difference is so small as to be meaningless.
I am not a fan of everyone using Bayesian techniques. A Bayesian analysis of a single experiment that gives a normally distributed measurement (which is most of them) with a non-informative prior is generally equivalent to a frequentist analysis. Since scientists already have trouble doing simple frequentist tests correctly, they do not need to be doing needless Bayesian analyses.
As for informative priors, I don't think they should ever be used in the report of a single experiment. Report the p-value and the Bayes factor, or the equivalent information needed to calculate them. Since an informative prior is inherently subjective, the reader should be left to make up his own mind what it is. Reporting the Bayes factor makes meta-analyses, where Bayesian stats SHOULD be mandatory, easier.