home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!usc!aludra.usc.edu!not-for-mail
- From: augustho@aludra.usc.edu (August Horvath)
- Newsgroups: rec.aviation.military
- Subject: Aero beauty survey RESULTS
- Date: 3 Jan 1993 15:48:03 -0800
- Organization: University of Southern California, Los Angeles, CA
- Lines: 158
- Message-ID: <1i7tvkINNkvi@aludra.usc.edu>
- NNTP-Posting-Host: aludra.usc.edu
-
- Thanks to those who contributed responses to my survey of
- aeronautical beauty a few weeks ago. 43 of you submitted
- responses, a bit less than I hoped for but adequate for some
- purposes. This note contains the results I promised to post.
-
- According to the survey, the top 20 of the 55 aircraft, for the
- whole sample, were:
-
- 1. North American XB-70 Valkyrie
- 2. Supermarine Spitfire - Griffon (Mks. XII-24)
- 3. Supermarine Spitfire - Merlin (Mks. I-XI, XVI)
- 4. Lockheed SR-71 Blackbird
- 5. BAC/Aerospatiale Concorde
- 6. de Havilland Mosquito
- 7. North American P-51 Mustang
- 8. Northrop YF-23
- 9. Sukhoi Su-27
- 10. Lockheed Constellation
- 11. Northrop T-38 Talon
- 12. Convair B-58 Hustler
- 13. Lockheed F-104 Starfighter
- 14. General Dynamics F-16 Fighting Falcon
- 15. Lockheed P-38 Lightning
- 16. North American F-86 Sabre
- 17. Avro Vulcan
- 18. Hawker Hunter
- 19. Dassault Mirage III
- 20. Grumman F-14 Tomcat
-
- Noting that the XB-70 beat the Spitfire by a narrow margin (mean
- score 16.17 to 16.03), I suspected nationalist bias from the 28
- out of 43 judges who responded from nodes in the United States.
- Hoping to unmask such bias I repeated the analysis for U.S. vs.
- non-U.S. respondents. What happened was unexpected:
-
- U.S. non-U.S.
- ---- --------
- 1. Spitfire (Merlin) Spitfire (Griffon)
- 2. XB-70 XB-70
- 3. SR-71 SR-71
- 4. Spitfire (Griffon) Spitfire (Merlin)
- 5. P-51 Concorde
- 6. Concorde Mosquito
- 7. Mosquito F-16
- 8. T-38 CF-105 Arrow
- 9. F-104 YF-23
- 10. Constellation Su-27
- 11. F-86 Hunter
- 12. B-58 Vulcan
- 13. P-38 P-51
- 14. Su-27 B-58, Constellation [tie]
- 15. YF-23
- 16. Albatros Comet, Draken [tie]
- 17. F-16
- 18. Mirage F7F Tigercat
- 19. Nieuport 28 T-38
- 20. F-14 Mirage
-
- U.S. respondents thought the short-nosed Spitfires are prettier
- than the XB-70, while non-U.S. respondents thought the long-nosed
- ones are nicest; but both also rated the XB-70 very high, and as
- a result it "won" the overall competition. It should be pointed
- out that the top 4 aircraft are all well within each other's
- standard measurement error for this sample size, so these results
- probably are not very stable.
-
- Although the XB-70 is not the subject of a nationalist bias,
- such a bias is present in the list more generally. The XB-70,
- Spitfire, SR-71, Concorde, and Mosquito were universally well
- regarded, but every other type in the top 20 did appreciably
- better among one group than the other. Notably, the U.S. top 20
- is studded with American types like the P-51, T-38, F-104, F-86,
- and P-38 that are evidently not so highly admired (at least
- aesthetically) overseas, while the foreign list has more
- nationalities represented (the Canadian Arrow, British Hunter,
- Vulcan, and Comet, and Swedish Draken).
-
- My survey was designed to control for respondents' familiarity
- with the various aircraft by allowing them to leave the spaces
- blank for planes they did not recognize. This produced some
- interesting results in itself. Of the 43 judges, only 5 entered
- a score for every type. The most-recognized aircraft were the
- P-51 Mustang and F-4 Phantom, which got 42 judgements; the least
- familiar was the Travel Air 4000, which got only 9 responses.
- The immortal Stearman received only 17 scores, making it the 3rd-
- least familiar plane. Obviously this newsgroup crowd does not
- belong to the age cohort that follows the dictum, "Real Airplanes
- have Two Wings and Round Engines."
-
- The 2nd-least recognized plane was the CF-105 Arrow with 16
- responses. You XB-70 fans should really get to know this
- airplane. Like the Valkyrie, it was a big white delta airplane
- that flew great but cost too much. It was cancelled in 1959,
- after which there was a massive migration of Canadian aerospace
- talent to the U.S., and it occurs to me that since the XB-70
- first flew in 1964, there may be some engineers running around
- with both the CF-105 and XB-70 on their resumes. If I find one,
- I'll buy him a beer to listen to his stories.
-
- I had originally intended to do a factor analysis of responses to
- the aircraft beauty survey, seeking underlying cognitive
- "dimensions" of aeronautical aesthetics. But 43 responses is way
- too few for even an informal analysis. (I tried, but couldn't
- get stable factors.) So instead I resorted to an informal
- network analysis, in which the correlations between judgements of
- various aircraft are made into a map, showing which aircraft are
- conceptually linked to which others for the purpose of drawing
- aesthetic conclusions. The map is very interesting, but I can't
- figure out how to post it so I'll summarize it briefly.
-
- I was hoping that the clusters of correlated aircraft would be
- interpretable based on visual characteristics of the aircraft
- (e.g. delta wings, elliptical wings, etc.) that might contribute
- to a hypothesis about underlying aesthetic properties. This
- happened, but only secondarily; the dominant clustering pattern
- was determined by functional and time-era categories (e.g. there
- were clusters of fast jets, of civil transports, of WW2 birds,
- of SAC bombers, etc.). Some of the secondary clustering did seem
- to reflect aesthetic characteristics; transports broke out into
- straight vs. curvaceous, WW2 types sub-clustered into
- (apparently) radial vs. inline-engined types, and a couple of the
- delta fighters broke out together.
-
- Besides cluster membership, the identity of the aircraft that
- form bridges between clusters or that fall with the "wrong"
- cluster of other aircraft can be enlightening. For example, the
- P-38 loaded with the jets instead of with the other warbirds,
- suggesting that it is liked by jet buffs and hence seen as more
- advanced than many of its contemporaries; on the other hand, the
- F-86 clustered with the WW2 types. The bridges (aircraft
- connected to both clusters) between the WW2 and jet clusters were
- the B-26 and P-51 on the WW2 side, and the F-100 on the jet side.
- There was a cluster of WW1 types (Albatros, Nieuport) and one of
- between-the-wars lightplanes (Vega, Cub, Tiger Moth) sharing a
- single member in common--appropriately, the Jenny. A few of the
- "wrong" loadings were not surprising: the B-29 clustered with the
- B-47 and B-52 rather than with its WW2 contemporaries, and the
- Convair B-58 hung in with the jet fighters rather than with its
- fellow bombers. And, there was a very strong correlation between
- the Boeing B-52 and the Junkers Ju-87. I'll let you figure that
- last one out for yourself, but here's a clue: they ranked 49th
- and 55th, respectively, out of the 55 aircraft.
-
- The most widely connected aircraft in the entire network was the
- Lockheed 12 Electra, which was linked to civil transports (DC-3,
- DC-8, Gulfstream), to the B-29 in the bomber cluster, to the P-51
- in the warbird cluster, and to the A-4 and F-100 in the jet
- cluster. This performance means either that (1) the old Lockheed
- is seen by respondents as a quite advanced type, which it surely
- was, or that (2) some respondents thought that I was referring to
- the postwar L.188 Electra. If anyone did respond thinking it was
- the L.188, I'd appreciate hearing from you.
-
- Hope you've found this interesting; thanks again to all who
- participated.
-
- August Horvath <augustho@usc.edu>
-
-