Well, as I’ve said, my prior for p(site issue) is fairly low. Especially given that several other people have suggested (in other browsers) that they’re seeing things fine. And I don’t mean to imply certainty. You’re correct—the only evidence I have is what has been posted in the comments here. I’m just trying to propose an alternative hypothesis, that you hadn’t seemed to consider. If you had, then I apologize for the redundancy, but you made no mention of this possibility in your OP.
Other people have said they’re seeing things fine in other browsers. Sorry. Edited for clarity.
No need. The point is, the fact that other people see things fine in other browsers (which, btw, I already knew) does not help to establish why IE7 previously worked but recently stopped, nor suggest that nothing changed on LW’s end.
The two are not mutually exclusive. And both are relevant.
No, likelihood ratios are the only concern for a Bayesian, unless I have a particular reason to put credence in your priors.
the fact that other people see things fine in other browsers (which, btw, I already knew) does not help to establish why IE7 previously worked but recently stopped, nor suggest that nothing changed on LW’s end.
In fact, it does. Unless you’re suggesting that compatibility with IE7 and other browsers is statistically independent, p(site problem | no problem in other browsers) < p(site problem | problem in other browsers). It may not be strong evidence, but it is evidence. Although, at this point, it’s moot, since your bug has been duplicated.
No, likelihood ratios are the only concern for a Bayesian, unless I have a particular reason to put credence in your priors.
I didn’t say they were relevant to the question asked, but they were relevant given the context in which they were mentioned.
Well, as I’ve said, my prior for p(site issue) is fairly low. Especially given that several other people have suggested (in other browsers) that they’re seeing things fine. And I don’t mean to imply certainty. You’re correct—the only evidence I have is what has been posted in the comments here. I’m just trying to propose an alternative hypothesis, that you hadn’t seemed to consider. If you had, then I apologize for the redundancy, but you made no mention of this possibility in your OP.
Really? Who mentioned seeing the comments just fine in IE7 over the past few days, rather than yesterday but not today?
More accurately, you’re telling me your priors. I’m interested in likelihood ratios though, and you had nothing to offer on that front.
Other people have said they’re seeing things fine in other browsers. Sorry. Edited for clarity.
The two are not mutually exclusive. And both are relevant.
No need. The point is, the fact that other people see things fine in other browsers (which, btw, I already knew) does not help to establish why IE7 previously worked but recently stopped, nor suggest that nothing changed on LW’s end.
No, likelihood ratios are the only concern for a Bayesian, unless I have a particular reason to put credence in your priors.
In fact, it does. Unless you’re suggesting that compatibility with IE7 and other browsers is statistically independent, p(site problem | no problem in other browsers) < p(site problem | problem in other browsers). It may not be strong evidence, but it is evidence. Although, at this point, it’s moot, since your bug has been duplicated.
I didn’t say they were relevant to the question asked, but they were relevant given the context in which they were mentioned.