[ExI] few bits per second

[Rafal Smigrodzki]

On Mon, Jun 28, 2010 at 12:14 PM, Damien Broderick <thespike at satx.rr.com> wrote:
> On 6/28/2010 1:37 PM, Rafal Smigrodzki wrote:
>
>> wouldn't you agree that the
>> PEAR record is actually quite disappointing as evidence of psi goes?
>
> The PEAR work wasn't just interested in looking for evidence of psi for the
> sake of it, over time it was more concerned with the process (and hence were
> prepared to accept low effect size). So even low effect size can be
> harnessed to investigate aspects such as whether number of bits per REG
> output unit influenced success rate,** gender correlates, etc.
>
>> I mean, 2 events out of 100,000 is pretty low.
>
> Some of their results were better than that (the remote viewing efforts, for
> example--but those have been seriously criticized from within the parapsych
> community); I think their insistence on getting the mechanisms bulletproof
> against critics meant that the tasks were unbearably drab and repetitive.
> Other protocols that are more natural and engaging, such as Ganzfeld and
> Remote Viewing, have shown much higher deviations from chance expectation
> without compromising their robustness.
>
>> I am about 99% sure I
>> have not seen credible evidence of any psi phenomena. I am about 85 -
>> 95% percent sure that physics as we know it does not provide a basis
>> for their existence...  What
>> are your percentages here?
>
> 99% convinced. If it turns out that they're all lying or making some
> frightfully subtle procedural error, I'll change my mind, of course. And
> I've seen arguments by physicists who claim that psi looks to be consonant
> with the sorts of extensions that are needed to bridge the current gap
> between GR and QT.
>
> **see e.g. http://www.scientificexploration.org/journal/jse_12_3_ibison.pdf
>
### With an effect size this small you don't need all people lying -
all you need is one guy skewing the data a few times. The inverse
correlation between the resistance of protocol to being "bulletproof
against critics" and the effect size is quite telling. Really, low
effect size in the absence of predictive (this is very important,
predictive, not post-dictive) theory is a very serious obstacle to
obtaining reliable knowledge.

One thing - my prior regarding existence of anything (and I mean
really anything, from my belly to the Milky Way) is "It almost
certainly didn't happen". Only as I stub my toe on it am I grudgingly
admitting to its existence. When I say I am 99% sure I haven't seen
convincing evidence of a phenomenon it doesn't mean that something
swayed me into disbelief but rather that my initial 99.999 disbelief
wasn't yet crushed by a force my mind (or toes) cannot resist. With
psi there are faint inkllings, enough to push the needle on my
existence-meter from -99.999% all the way to -99% but not any farther.
To go from "99% didn't happen" to "99% happened" I need a lot more
stubbing. Did the dry PEAR numbers sway you towards "99% happened"
from a position of disbelief, or did you set out from a certain
position of longing and eagerness towards psi?

Rafal