I've had a chance to sit down and correct my earlier tally of the 3105 kHz signal receptions on the Signal Catalog spreadsheet. I did a manual count (includes approximates) and came up with the following: 85 total, 44 (52%) credible, 22 (26%) uncertain, and 19 (22%) not credible.
Joe, your total of 85 includes 19 messages that have “Itasca” as the Agency/Person. These lie between message 10 and message 73, inclusive. Your total also includes message 75, which relates to 3105 kHz only in that it reports nothing has been heard on that frequency for two hours. Remove those 20 messages and your reduced total of 65 should then have a breakdown something like this: 40 (~61%) credible, 9 (~14%) not credible, and 16 (~25%) uncertain.
The reason I originally undertook the exercise was to investigate a poster's claim that a signal received on 3105 kHz stood an inordinately high chance of receiving a rating of credible.
Something like a 40 to 9 ratio of credible to not credible assessments may have motivated that poster’s claim.
Chuck
Chuck,
Thanks for your persistence. The signals you mention between signals 10 and 73, inclusive, are receptions
by the Itasca, not transmissions from the Itasca. I assigned to any documented transmissions from any party a credibility rating of "n/a". One should not exclude these signals from the list of 85 alleged receptions on 3105 kHz. For message 75, I will admit that I did not look at this signal as closely as you have. The only signal heard in this reception was a weak carrier wave on 6210 kHz. Therefore, I will exclude this signal, rated as "not certain," to obtain the following:
3105 kHz credibility = uncertain 21 25%
3105 kHz credibility = credible 44 52%
3105 kHz credibility = not credible 19 23%
total: 84 100%
I would still maintain that with a bare majority of signals registering as credible, this hardly reveals a bias of the magnitude you cite. You excluded the uncertain signals from your ratio formula. Your ratio was calculated as (Credibles minus all credible Itasca receptions) / (not credibles minus all not credible Itasca receptions). My ratio was calculated as (Credibles including credible Itasca receptions) / (uncertains including Itasca uncertains + Not credibles including Itasca not credibles). I concede, however, how it might be considered acceptable to exclude uncertains. I cannot see, however, how Itasca receptions can logically be excluded. If one takes a compromise formula, such as (Credibles) / (not credibles), one arrives at a ratio of 44 to 19. That's still perhaps, in your view, a high number of credible 3105 kHz signals, but what's the alternative? Discount all 3105 kHz signals as too close to Earhart's known nighttime transmitting frequency and thus the province of likely hoaxers? If Earhart were calling for help in that region, what was she supposed to have done, broadcast on random frequencies to prove she was really her?
In any case, to my understanding, Bob Brandenburg did not use a single criteria for the acceptance of any one signal. He used many, including SNR, message content, response timed with request for information, response timed with Earhart's pre-takeoff schedule of broadcasts, and many more.
Joe Cerniglia
TIGHAR #3078 ECR