Simple probability math theory, probability of a single guess success is p = 0.5, then combinations and variance probability formula.
web makes it for you
Binomial Probability Calculator
First row fill in 0.5, last row P(X>=x) is the result. I hope you can fill in "number of trials" and "number of successes".
Oh, OK. I don't know anything about a website. I suppose I'm naive enough to take people at face value and assume they have a legitimate motive. Not the case here?
15:26:23 : Test started.
15:30:43 : 01/01
15:33:05 : 02/02
15:33:33 : 02/03
15:34:41 : 03/04
15:35:37 : 04/05
15:36:25 : 04/06
15:37:55 : 05/07
15:38:28 : 05/08
15:39:03 : 05/09
15:40:16 : 06/10
15:40:16 : Test finished.
i pasted this so i could show what i'm looking at, the numbers at the start of each trial are they not timing each trial? cause if it's merely based on choices why run a timer?
i tried using the data in the binomial calculator 6 correct out of 10 trials...no percentage?
15:30:43 : 01/01
15:33:05 : 02/02
15:33:33 : 02/03
15:34:41 : 03/04
15:35:37 : 04/05
15:36:25 : 04/06
15:37:55 : 05/07
15:38:28 : 05/08
15:39:03 : 05/09
15:40:16 : 06/10
15:40:16 : Test finished.
i pasted this so i could show what i'm looking at, the numbers at the start of each trial are they not timing each trial? cause if it's merely based on choices why run a timer?
i tried using the data in the binomial calculator 6 correct out of 10 trials...no percentage?
Last edited:
Time is doing nothing here. Now put the 10 trials and 6 successes to the chart I linked and you'll get the same result as from foobar.
Sure, they are overrated & people try to assign to these forum run ABX tests some level of 'validity' above sighted listening. That's my simple point but there is great reluctance to accept this.
I have stated, in detail, as has Jakob2, the many issues with these tests - I would refer you to the couple of previous long threads in which this was thrashed out & yet many still bring up the same points here as they did on that thread. I'm not sure you were also involved in those discussions, I think maybe you were? Were you?
So you avoid backing up your claim that I misrepresented what you said - yet again. I find it interesting that when I call for this accusers dodge the issue. This demonstrates that this accusation is a typical dodge to avoid a tricky position they find themselves in.
I will bypass your question as there is a chasm to be crossed to answer it & you don;t seem willing to even take the first step in this
I just asked you for the link, didn't accuse you of lying.
I have ofered concrete information many times
Some questions need elucidation - maybe you misunderstand this as 'rarely answer a direct question'
There is good research behind what I post & I link to it - this i snow considered a fault, is it?[/quote]Now maybe you genuinely believe that unless we all achieve the equivalent of an undergraduate degree in psychoacoustics (which you have?) we have nothing useful to say about audio testing, but I suspect that is not the real issue.[/quote]You have lots of misinformation about perceptual testing but seem reluctant to read any links given or corrections offered.
Sure, I've seen how JC is treated on here - is he under constant moderation?
I see the usual attempted attack on the messenger in many posts that surround this one
The thread was closed at least a month ago, I'm not going to waste time looking for it.
I asked for a link to any conclusive equipment tests that follow an approved protocol and you continue to say I ask the same thing over and over and then just refer again to the same articles. I gather there aren't any?
I read the ITU spec there was no appendix containing examples and results. I am assuming that it is necessary and sufficient and your other articles don't discredit it.
Last edited:
This has not been established, the Mr. M associated with that claims to be an EE. The add copy leaves much to be desired.
No, it's not the messenger who is being attacked. It's the lack of any good evidence or support for claims. If you are just blowing smoke, you'll be challenged on it. Rightfully so.
You simply object to online ABX tests. Fine, we understand that. Please stop trying to embellish that objection.
p > 0.05 😉 (I'm being a pedant, so please read that in a lighthearted manner)
In any case, a LOT of academics are moving away from this being a reliable marker unto itself, as it lets a lot of leakage through. (unless it's supplemented with orthogonal tests, positive/negative controls, etc.)
Of course tests like Pavels here (and I apologize for not taking the time to do the test myself) are highly problematic, because so much of the test is dictated by the end-user. That's obviously nothing PMA or anyone else can necessarily control. I would similarly be reluctant to combine trials, but would also say that it's going to err towards false negative rather than false positive. (unless there was some perverse file-specific playback effect)