[EDIT TO CORRECT FATAL OMISSION OF A NEGATION - Oops!]
Note that Zander's aim isn't [THAT'S THE EDIT] to establish in the eyes of the world that astrology works (or doesn't, or that the jury is still quite rightly out) *in general*. As I read it, this is a focussed test to see, basically, does astrology (in the form picked for the test, 'cos of course there are many) work *for Zander*.
That said, you're right that to be a valid test, a single trial is insufficient, but in this case the way to multiply the data points available so we can filter real results from the noise isn't to test 100 different people, but to test Zander 100 times (or at least, as many times as we feasibly can).
Suggestion: We do three tests this year. If people are stil interested, we do three more this time next year, and so on - make it an annual event. Then maybe by about 2030 we'd be able not only to say "is Zander significantly Astrologically Correlated?", but even make a first pass at "is Zander becoming more or less Astrologically Correlated over time?" =:o}
N.B. Test conditions would have to be as near identical as possible each time. Could be tricky if the unacquainted astrologer and Zander happen to be introduced to each other somewhere down the line, at one of those big all-night parties he loves to frequent... =;o}
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
branchandroot
no subject
Date: 2011-01-18 09:32 am (UTC)Note that Zander's aim isn't [THAT'S THE EDIT] to establish in the eyes of the world that astrology works (or doesn't, or that the jury is still quite rightly out) *in general*. As I read it, this is a focussed test to see, basically, does astrology (in the form picked for the test, 'cos of course there are many) work *for Zander*.
That said, you're right that to be a valid test, a single trial is insufficient, but in this case the way to multiply the data points available so we can filter real results from the noise isn't to test 100 different people, but to test Zander 100 times (or at least, as many times as we feasibly can).
Suggestion: We do three tests this year. If people are stil interested, we do three more this time next year, and so on - make it an annual event. Then maybe by about 2030 we'd be able not only to say "is Zander significantly Astrologically Correlated?", but even make a first pass at "is Zander becoming more or less Astrologically Correlated over time?" =:o}
N.B. Test conditions would have to be as near identical as possible each time. Could be tricky if the unacquainted astrologer and Zander happen to be introduced to each other somewhere down the line, at one of those big all-night parties he loves to frequent... =;o}