Subject:
|
Re: Personality test vs. Religion
|
Newsgroups:
|
lugnet.off-topic.debate
|
Date:
|
Fri, 29 Oct 2004 19:26:23 GMT
|
Viewed:
|
2187 times
|
| |
|
|
In lugnet.off-topic.debate, David Eaton wrote:
> > Forgive me, but so far I can't think of anything that *would* convince you to
> > abandon this baseless test. Because the test is inherently non-falsifiable,
> > it is worthless as an empirical instrument.
>
> To make me disbelieve it, you'd have to find some people that I knew pretty
> well, whom I could classify into multiple M/B types. Then, after I had made
> predictions, you'd have to have them take the test and show that I was only 60%
> or less correct (or some percentage reasonably close to 50% basically).
> Essentially the same proposition I came up with later (the psychiatrists
> evaluating people, and being wrong or arbitrarily right)
This is part of the problem. You're implicitly assuming that the test is a
valid instrument, and that therefore the only way to disprove the validity of
the test is to take the test (which is designed not to yield falsifiable
results) and make it yield false results. This is a logical impossibility.
I have repeatedly pointed out several of the many ways that this test is
methodologically unsound, and for any of these reasons the test is invalid.
However, you choose to embrace the test regardless of any argument to the
contrary, so I assert that you're staking your claim at the wrong point.
The issue, as far as I'm concerned, is not whether the subjective test
accurately describes people who in your (and their) subjective description match
the test's description. Instead, the issue is that the test does not allow the
possibility for a definitive true/false assessment of the results. You yourself
have stated this clearly, when you dismissed the final summary. The final
summary is, in fact, the essence of the test's results, so you're saying in
effect "I know that the test is invalid, but I'm still choosing to believe that
it's valid."
> I'm not sure I follow-- if the 4 dimensions aren't accurate, wouldn't we find
> people who match both traits (or neither), hence invalidating that dimension
> since it will most likely be arbitrarily picked by the person and/or
> psychiatrists? Basically, if someone has a tendancy to BOTH
> "jump-to-conclusions" AND "never make any conclusions", won't they be
> arbitrarily picked as one or the other more-or-less arbitrarily by the
> psychiatrists and by themselves on the test, hence creating a higher degree of
> error, and invalidating that dimension, or at the very least the test's ability
> to *measure* that dimension?
Yes, but every single question on the test yields a similar objection, even the
seemingly opposite "open" and "closed" are problematic: does "open" mean
vulnerable? Or receptive? Or uncritical? Or lascivious? Or awake? Or
generous? How do you know? And how do you contrast this multi-dimensional and
multiply-connotative word with "closed," when "closed" is similarly faceted?
Rather than the true binary choice that M/B pretends, you're facing a field of
say 36 choices, no one of which can be called "true" or "false" with any
certainty.
In addition, every time that the test asks for a pseudopoetic, interpretive
decision between "Choice A" and "Choice B" (since B is not a clear opposite of
A), then the test forfeits its objective validity, which is what the test's
owners really use as the main selling point, and it's also what a critical user
should bear in mind while taking the test.
> As for the dimensions being right, what would you accept as "proof"?
Actually, this is similar to the problems of biblical "prophecy." The so-called
"prophecies" in the bible are so shamelessly open-ended and non-specific that
any of a zillion results can be claimed to "prove" the "prophecy."
So here's what I propose: The test as it exists is hopelessly untestable and
should be scrapped. In its place, I would suggest developing a test that makes
specific descriptions that can be shown to be false or true. Further, as a
scientific instrument, the tester should be able to make testable predictions
about future behaviors as described by the test. Since the test makes bold
claims about a range of behaviors, then I suggest that at least several examples
of each of the many possible results must be tested in this way.
> Let's say some biologists claim to have found the "decisive" gene in human
> genetics. They remove it from 1,000 people, and add it to 1,000 people. How
> would you test whether someone in your 2,000 people was "decisive" or not to
> see if it worked?
I'm afraid that this is too hypothetical to be useful. How did the geneticists
identify the target gene as the "decisive gene?" Presumably the testing would
have identified it before that point, right? So why not examine that same test
as it pertains to the post-gene-removal subjects?
> > > And assume that the subjects rated themselves on M/B, and compared the
> > > results to their psychiatrists' ratings? Assuming a good match (say, 90+%),
> > > would you say it's any more useful of a test, now that it's not necessarily
> > > up to personal self-bias?
> >
> > I think that this would be an interesting starting point, if we could
> > demonstrate the validity of the 4 M/B dimensions. However, you're describing
> > a single-blind experiment, since in your description the psychiatrists know
> > what they're testing. To increase the validity of the experiment you'd need
> > to ensure that the psychiatrists' results were not themselves tainted.
>
> Single blind-- how?
Because in your description the psychiatrists are testing with awareness of what
the test's intent is: namely, to achieve positive correlation between the
subject's answers and the testers' observations of the subjects.
If you propose that the testers should make only general observations of the
subjects, then you face the problem of correlating these observations with the
M/B results, which is another subjective, interpretive process.
> But if you're asking for them to test the M/B characteristics
> whithout knowing they're testing them, well... I'm not sure that's possible.
Ideally, the subjects shouldn't know that they're being tested, and the
psychiatrists shouldn't know *what* they're testing. However, any reasonable
psychiatrist (which, parenthetically, neither Myers nor Briggs was) would
recognize that the testing parameters of Myers-Briggs are inherently
non-scientific and non-falsifiable, and the psychiatrist would reject the test
as folly.
Dave!
|
|
Message has 1 Reply:
 | | Re: Personality test vs. Religion
|
| (...) Not really-- because as I've said I've seen what I believe to be evidence of it yielding *correct* results. And, as I've said, it IS (for my part) falsifiable, because if I had measured someone (say) as indecisive, and they took the test and (...) (20 years ago, 29-Oct-04, to lugnet.off-topic.debate)
|
Message is in Reply To:
| | Re: Personality test vs. Religion
|
| (...) Well, here's the juicy bit. It's sort of like a Christian believing in God. You can't disprove it, they can't prove it, but they have what they feel is evidence, based on "a feeling" or "an intuition": With me, I'm pretty gosh-darn indecisive. (...) (20 years ago, 29-Oct-04, to lugnet.off-topic.debate)
|
53 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
