(2019-01-19, 09:00 AM)Chris Wrote: [ -> ]I wonder if malf could indicate whether he can see any aspect of the statistical analysis which he feels wasn't completely specified in advance - according to what Roger Nelson says - which the rest of us have missed.
That was never my point. The flexibility I’m querying is around the selection of events that the analyses are applied to (according to Bancel, that’s where the ‘magic’ is happening). Looking through this thread, this was the same issue that vexed you (in discussion with Arouet) on page 3.
(2019-01-19, 05:16 AM)Laird Wrote: [ -> ]Linda, the BS is all from you. Utter lie: "you weren't able to address the points I/we raised". I've spent many, many words addressing in excruciating detail the points you've raised, only to have you come back and either ignore what I've written or spin it away.[ It's not "name-calling" to point this out; it's a necessary part of withdrawal from the conversation, to make it clear that and why meaningful dialogue has become impossible.
Of course the comment "you weren't able to address the points I/we raised" would have applied to whether you responded to the last post(s) made on the subject, not to a post you made over 2 months ago. Your response to my last post was "
semantic obfuscation, manipulation, and (wilful?) obtuseness", which obviously makes no attempt whatsoever to address what I said.
I also have spent many, many words addressing in excruciating detail, specific points which are subsequently ignored. Skeptiko mods, like David Bailey even made a point of deleting some of these posts so they couldn't be read. It's just something that happens on a discussion forum (well, maybe not the deletion part - that was special treatment), and one learns to accept it. If you find a discussion has become meaningless, of course it is reasonable for you to withdraw from that conversation. I frequently find myself doing this. What is unreasonable is for you to not only lay the blame for this on me, but to attribute all sorts of negative characterizations to my participation in this discussion. I am trying, in great detail and in good faith, to explain my perspective using my background knowledge and experience. Just because you don't agree with me and don't understand all that I am saying does not make what I say "spin" or a "lie" or "manipulation" or "obtuseness".
The discussion did not stall because I am any of those things. I am not.
It is not a necessary part of your withdrawal to engage in defamation. The only necessary part of withdrawal is to simply stop responding. Common decency might require a statement that you are withdrawing, although given that you had withdrawn a couple of months ago without saying anything, and nobody was upset about it, suggests you didn't need to.
I avoid discussion with you because you commonly resort to defamation. I don't think that is fair or decent of you, but more importantly, it poisons the possibility of reasonable discussion on a subject which greatly interests me.
Linda
(2019-01-19, 10:11 AM)malf Wrote: [ -> ]That was never my point. The flexibility I’m querying is around the selection of events that the analyses are applied to (according to Bancel, that’s where the ‘magic’ is happening). Looking through this thread, this was the same issue that vexed you (in discussion with Arouet) on page 3.
malf, your point was this:
"In terms of competing hypotheses, one must be that, once presented with colossal amounts of noisy data, mathematics can always produce some statistical significance."
Do you agree that if the statistical tests were fixed in advance, such a hypothesis would in no way explain the findings?
Regarding the discussion on page 3, as I'm sure you understand very well, Peter Bancel's conclusion was that experimenter psi played a role in the selection of the events. Not that the results could be explained by "mathematics" producing significant results from "colossal amounts of noisy data".
Those are completely different explanations, and I don't believe you are genuinely unable to distinguish between them. So I don't understand why you are giving the appearance of not being able to distinguish between them.
(2019-01-19, 03:12 PM)Chris Wrote: [ -> ]malf, your point was this:
"In terms of competing hypotheses, one must be that, once presented with colossal amounts of noisy data, mathematics can always produce some statistical significance."
Do you agree that if the statistical tests were fixed in advance, such a hypothesis would in no way explain the findings?
Regarding the discussion on page 3, as I'm sure you understand very well, Peter Bancel's conclusion was that experimenter psi played a role in the selection of the events. Not that the results could be explained by "mathematics" producing significant results from "colossal amounts of noisy data".
Those are completely different explanations, and I don't believe you are genuinely unable to distinguish between them. So I don't understand why you are giving the appearance of not being able to distinguish between them.
I can’t understand how you don’t see the connection. Without the huge amounts of data you don’t get as many chances for (fortuitous/psi induced) statistical hits. Mathematically those choices turn out to be the most important factor.
What do you think is going on? Bancel’s ‘experimenter psi’, Radin’s ‘love’ affecting distant arbitrary electronic devices, or some sort of ‘p-hacking’?
Does anyone know if the protocols and hypothesis analysis have been through any sort of peer review?
malf
I am genuinely trying to understand what you are getting at here. That's why I asked the question in my last post. It's a question the answer to which would help me to know what's in your mind. Please could you have a go at answering it? Thanks.
(2019-01-19, 06:04 PM)Chris Wrote: [ -> ]malf
I am genuinely trying to understand what you are getting at here. That's why I asked the question in my last post. It's a question the answer to which would help me to know what's in your mind. Please could you have a go at answering it? Thanks.
I don’t think I can be any clearer.
You want to keep the ‘event selection’ seperate from the statistical analysis. I don’t think you can do that given the key importance of the ‘event selection’. Does that help?
(2019-01-19, 06:20 PM)malf Wrote: [ -> ]I don’t think I can be any clearer.
You want to keep the ‘event selection’ seperate from the statistical analysis. I don’t think you can do that given the key importance of the ‘event selection’. Does that help?
I don't know whether you're even reading what I'm writing.
Let me repeat:
It doesn't mean that for each individual event the statistic wasn't fixed in advance of looking at the data. What they say is that the definition of the statistic, and of course the times covered by the event, were always fixed before they looked at the data.
(2019-01-19, 06:33 PM)malf Wrote: [ -> ]Who chooses the event?
No, malf, the question is whether you understand what I'm telling you - that the statistical tests, including the times of the events, are claimed to have been fixed before the data were examined?
And if so, do you understand that if that is true, that rules out your hypothesis "
that, once presented with colossal amounts of noisy data, mathematics can always produce some statistical significance."
That is what I'm asking you to clarify, please.