First let me deal with my sighting report (see middle of this post http://www.internationalskeptics.com/forums/showthread.php?postid=7444230#post7444230):
There are currently two proposed explanatory hypotheses: The “birds” hypothesis and the “multiple satellite” hypothesis.
Considering birds, one has to explain how birds can be illuminated so that they become practically indistinguishable from stars/satellites.
Considering the multiple satellite hypothesis, one has to explain the period (about 1 minute) and also the (regular and predictable) oscillation of the leading pair about a central point between them.
Unless those things can be explained, then the “birds” and the “multiple satellite” hypotheses remain implausible.
Regarding the “null hypothesis” testing where:
and
The current argument is about whether this is a legitimate test or whether there are factors that would potentially skew the results.
For example the proposal has been put that the hoax/psychological category of reports might skew the results.
First the generally accepted rate of hoax/psychological report is actually very small – in the order of 1-2% (see The Hynek UFO Report - http://www.scribd.com/doc/43531895/J-Allen-Hynek-The-Hynek-UFO-Report-1977 p.259 for example). Meaning of course that in any statistical analysis the effect of this category of reports on the outcome will be very small. To put that in perspective we are talking just 1 or 2 in every 100 cases - and that is simply not enough to skew the distribution significantly.
Further, as we are talking about Chi-Square analysis (number observed v. number expected) you could control for that percentage by using 1 or 2 cases less on the unknown category side for every 100 cases in that category examined.
However, cutting to the chase, as AstroP points out, the research has already been conducted (here http://www.ufocasebook.com/pdf/specialreport14.pdf - pp 60-76 ).
In that study the characteristics of colour, number (of objects), shape, duration (of sighting), speed and brightness were examined. In the first pass, all categories except brightness showed a significant difference. In the second pass (where all astronomical cases were removed from the known category of reports) ALL results were significant.
This means of course that the, according to the study hypothesis, the unkowns are not drawn from the same population as the knowns.
Interestingly, despite the positive (significant) result, the authors of the report state:
Here of course the authors are simply ignoring the statistically significant result of their own analysis! That is, either the result is significant (in which case it indicates the population from which each caegory is drawn is not the same) or it is not (and the population is the same). There can be no “halfway” result here. There is no “inconclusive” – the results are conclusive – in that a statistically significant result was obtained. The authors of the report simply could not bring themselves to acknowledge it! LOL.
This then was a test of the hypothesis that all UFO reports are of mundane origin – and the result falsified that hypothesis.
Now of course that does not explain the unkowns – it is just that, according to this analysis, they don’t seem explicable in terms of the mundane. In other words, as I have been contending all along, they defy (plausible) mundane explanation.
There are currently two proposed explanatory hypotheses: The “birds” hypothesis and the “multiple satellite” hypothesis.
Considering birds, one has to explain how birds can be illuminated so that they become practically indistinguishable from stars/satellites.
Considering the multiple satellite hypothesis, one has to explain the period (about 1 minute) and also the (regular and predictable) oscillation of the leading pair about a central point between them.
Unless those things can be explained, then the “birds” and the “multiple satellite” hypotheses remain implausible.
Regarding the “null hypothesis” testing where:
H0: All UFO sightings are of mundane origin
and
If the H0 is true, then we would expect no difference on defined characteristics between known category reports and unknown category reports.
The current argument is about whether this is a legitimate test or whether there are factors that would potentially skew the results.
For example the proposal has been put that the hoax/psychological category of reports might skew the results.
First the generally accepted rate of hoax/psychological report is actually very small – in the order of 1-2% (see The Hynek UFO Report - http://www.scribd.com/doc/43531895/J-Allen-Hynek-The-Hynek-UFO-Report-1977 p.259 for example). Meaning of course that in any statistical analysis the effect of this category of reports on the outcome will be very small. To put that in perspective we are talking just 1 or 2 in every 100 cases - and that is simply not enough to skew the distribution significantly.
Further, as we are talking about Chi-Square analysis (number observed v. number expected) you could control for that percentage by using 1 or 2 cases less on the unknown category side for every 100 cases in that category examined.
However, cutting to the chase, as AstroP points out, the research has already been conducted (here http://www.ufocasebook.com/pdf/specialreport14.pdf - pp 60-76 ).
In that study the characteristics of colour, number (of objects), shape, duration (of sighting), speed and brightness were examined. In the first pass, all categories except brightness showed a significant difference. In the second pass (where all astronomical cases were removed from the known category of reports) ALL results were significant.
This means of course that the, according to the study hypothesis, the unkowns are not drawn from the same population as the knowns.
Interestingly, despite the positive (significant) result, the authors of the report state:
”The results of these tests are inconclusive since they neither confirm nor deny that the UNKNOWNS are primarily unidentified KNOWNS, although they do indicate that relatively few of the UNKNOWNS are actually astronomical phenomena.” (p. 76)
Here of course the authors are simply ignoring the statistically significant result of their own analysis! That is, either the result is significant (in which case it indicates the population from which each caegory is drawn is not the same) or it is not (and the population is the same). There can be no “halfway” result here. There is no “inconclusive” – the results are conclusive – in that a statistically significant result was obtained. The authors of the report simply could not bring themselves to acknowledge it! LOL.
This then was a test of the hypothesis that all UFO reports are of mundane origin – and the result falsified that hypothesis.
Now of course that does not explain the unkowns – it is just that, according to this analysis, they don’t seem explicable in terms of the mundane. In other words, as I have been contending all along, they defy (plausible) mundane explanation.
