« Making Beliefs Pay Rent (in Anticipated Experiences) | Main | Belief in Belief »

July 29, 2007


I'm a huge fan of Tyler Cowen but it turns out this particular "fact" is very poorly sourced -- just an opinion piece from Doug Kern on the TCS website that cites no studies or statistics in support of its claims.

It is quite interesting that you and Alex both wrote blog posts on this non-fact fact, and that Tyler decided to put it in his book. Confirmation bias?

Matt, I agree that solid statistics would better support the claim, but I don't think we must refrain from tentatively accepting any claims for which there are no solid statistics.

Robin, but the acceptance seemed to be performed in a more than tentative manner. Matthew C., I think loyalty networks may have more to do with the bias, to the extent it existed. Also, I think there can be an element of heirchical performance to a person asserting something poorly confirmed by data as a fact. To the degree that the asserted fact is accepted as such despite the paucity of data-support, it can be a measure of their raw hierarchical power (and in other people's responses, a measure of their social intelligence/ability to nimbly perform power alignment rather than independent fact assessment).

There doesn't seem to be an official Canadian count, contrary to what some of those sources suggest. Here is one French count: http://www.cnes-geipan.fr/geipan/stats_an.html; it shows decline. The main point is that UFOs are less culturally central and hold less credence; this is I think widely recognized. The example cited is also about abductions, not objects blurring by in the sky. See for instance here: http://www.virtuallystrange.net/ufo/updates/2002/jun/m03-013.shtml.

Tyler's source, says that media attention is down, not that sightings are down.

Sources like Robin's graph are very misleading and they are just measuring the size of one particular organization. I think he should take it down. For some of the early years they are reporting a very scant number of sightings (just a few at the very beginning) but of course that just means the group was a very small one. By prominently displaying the graph as he does, he is subject to considerable bias, probably resulting from a desire to rebut my comment. The general problem is to count the sightings rather than the span of the organization reporting the number of sightings.

Before writing the passage I did I read numerous sources on the question. I believe an account such as this one: http://www.thisishampshire.net/news/hampshirenews/display.var.1509916.0.is_there_anybody_out_there.php

is more accurate than Robin's take. Admittedly it is not statistical, but it is confirmed by my own interactions with popular culture, also post X-Files. Robin is biased toward a quantitative count when there is not a constant or nearly-constant source through time, and that is leading him astray.

Tyler's latest link says "public interest has waned considerably" and "a number of UFO magazines and journals have folded"; this is more evidence that public interest is down, not that sightings are down. The other two sources are a stat-less statement that abductions are down, and a French time series of a few hundred events, while the data I plotted had tens of thousands of events. The French series could also have reflected varying interest in its organization.

Yeah, if you look at that chart, other than the annual fluctuations, which seem to peak during the summer (since people are outside more at that time), you might as well be looking at a chart of Internet usage. It starts increasing around 1995 and plateaus in recent years. It probably just reflects the accessibility of the organization doing the data collection.

I'm surprised no one has advanced the obvious alternative hypothesis that aliens are being scared away by the cellphones.

Off the top of my head, the UFO trend looks like it tracks GSM rollout pretty well; but possibly Internet adoption even better.

I would rather like to see global stats on UFO sightings, rather than sightings by countries.

And I would also be interested specifically in the testimonies of the airline pilots (with still cameras :) ). They are serious people, you could mandate them to have always a still camera in their command cabin, and you can have reliable stats on their UFO sightings.

Psstt... The French source is serious (the CNES = the French NASA), and I believe that every French police station has a UFO sighting form that it sends to the GEIPAN for analysis and statistics. Of course, people have to go to the Police, otherwise their testimony is not recorded.

Psstt #2: You should ask historians. They will tell you that each civilization in the past had some kind of UFO sightings... No technology-oriented "UFOs" but some kinds of mythical animals or persons or else that "appeared" and impressed ancient people... I'm not saying that these things are not real. I'm saying the solution could be found in physics if there's a better understanding of some kind of meta mind--matter interactions... maybe via information traveling backwards in time (which is not completely silly in quantum physics).

The UFO phenomenon is quite interesting.

Many of the phenomena are clearly weather balloons, sightings of Venus, floating lanterns, and the like.

There is also something of a conspiracy cult around the phenomena who refuse to accept the more normal explanations even when they are pretty convincingly correct.

Whether there is something else also going on in some cases -- well I haven't done any substantial research into the phenomena, so I won't venture to guess.

There is now a detailed report on the well-publicized late-2006 Chicago O'Hare UFO sighting available here from an organization called NARCAP.

Looks pretty thoughtful, sane and rational to me, not at all like the UFO conspiracy nut material I've occasionally run across.

beginning preliminaries on correlating public opinion to sighting reports over time. WOuld like to start with Mexico City.
Hard to find credible raw data.
This phenomena may be very well suited to serious statistical analysis because of the 'absurd' element in individual cases, but examined as trends over time, significant patterns may very well appear.

The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30