It should be P(H0|significant) != 1 - P(significant|H0)
If you have a sample without real differences you have P(H0|significant = 0) which is the statistic used in the article.
This would result in:
"Yes, I agree. If a p-value of 0.05 actually "means" 1 when evaluated, then any sane frequentist will tell you that things are fucked, since the limiting probability does not match the nominal probability (this is the definition of frequentism)."
Johnson found that a P value of 0.05 or less — commonly considered evidence in support of a hypothesis in many fields including social science — still meant that as many as 17–25% of such findings are probably false (PDF).
. Found? Was he unaware that using a threshold of 0.05 means a 20% probability that a finding is a chance result - by definition ?
More interesting, IMO, is that statistical doesn't tell you what the scale of an effect is. There can be a trivial difference between A and B even if the difference is statistically significant. People publish it anyway.
Ofcourse it was found. The 20% are not by definition but a function of the percentage of studies done based on correct/incorrect H1. You could have 0% if you only did studies on correct H1s.
P(H0|significant) != P(significant|H0)
A Unified development path? Oh please, if MS wanted one, they COULD have made ALL their games available for both their console AND windows AND their phones ALREADY! They don't because MS isn't a company it is a number of departments involved in century old feuds. I know MS isn't centuries old but there departments are sure feuding like their great-great-great grand-daddy's have done.
MS can't do a unified approach because it is not a unified company. Just examine its countless position changes on whether Windows is or is not a gaming platform. In a way MS is even killing itself with it. The only thing I would need windows for is gaming. No windows games? Then I could just as well run Linux or a Mac. In fact, I do run Linux because more and more games are available on Linux or at least Linux friendly.
And no, I don't own a x-box. But smart move MS, make your own platform less relevant. Oh wait, then there is Games for Windows. Oh then it is not. Why do you think Valve is going ahead with Steam OS? Because they like building a OS more then building games? No because they are fucking tired of being depended on a company that is schizophrenic about its own OS.
The only reason valve is building Steam OS is because they are now in direct competition with MS app store.
A modern library implementing dataflow programming is: https://rx.codeplex.com/
The relationship between the actor model and flow based programming is discussed here: http://c2.com/cgi/wiki?ActorsAndFlowBasedProgrammingDiscussion
Personally i like using f# with rx using this wrapper: https://github.com/fsharp/FSharp.Reactive
I see FBP as a generalized form of functional programming where you can, and have to, specify how the expressions are put together temporally.
It gives you the advantage of automatic (in rx you have to specify the scheduler) concurrency but is cumbersome for stuff that is easily expressed in a synchron way (List processing...).
I guess you would want the most coarse grained granularity.
A nice book on FBP is: http://www.jpaulmorrison.com/fbp/ (the first edition is freely available)
As for the idea that FBP is coupled to visual programming and will enable non programmers to write complex programs, well:
Visual programming makes writing programs for easy problems even easier and for complex problems intractable.
And guess what: The easy problems were already easy to begin with!!!
Do we really want corporations decide what laws they should follow?
Do corporations really have, and should they have, a moral compass and responsibility?
If so, which one and how should they operate in different cultures?
If the laws suck then we, the people, have a responsibility to fix it through democratic measures.
I do not think it legitim to condemn an organisation and the people working there for following the law.
Oh, many researchers do know better and there are theoretical models that are quite stringent and have metric properties:
The problem is that you can then test if your measurement axioms hold and most likely they won't so why do it
replying to undo moderation.
but if you are able to detect if photons are superimposed then you are able to transmit information faster than light violating causality. Or do you mean detecting by trying to communicate with differing keys?
I agree about the need for folks to do a head-check before reacting to something said off-the-cuff, but there's one sticking point...
Because people fuck up sometimes?
Seems to me, Humphrey actually put the good of his Country ahead of personal and party gain. This is a far cry from what we've become as a Nation.
Afaik Humphrey didn't expose Nixon because polls told him he would win anyway and that there was no need to steep that low.
And what would the use have been after having lost.
Better to wait for the rematch and use it then.
I can be 95% correct predicting if a somebody is gay: