[math-fun] Ex. What Not To Do.
Screen Cap: https://0x0.st/i1zI.png of Fig. 4.4 from: https://scholarworks.uark.edu/etd/2456/ Look how horrible these _published_ fits are! This is a U of A certified Physics PhD, in fact, certified by the two people whose sole decision it was to make me a victim of the March purge, against the best wishes of my committee. The histograms themselves are asymmetric, but the fit function is a symmetric Gaussian. Instead of rushing ahead to incorrectly repeat standard blah-blah-blah, if that's all they can do after so many years, looks like the author needs to stop and ask "WTF?" or "What's The Function?" If the author is still ignorant to this day, maybe it would be a good idea to study the recently released "Summer Insect Pandemics": https://demonstrations.wolfram.com/SummerInsectPandemicsInTheUnitedStates/ Sorry, but this is much better data analysis, which accounts for asymmetry. It's not even for a PhD, it's just for fun and to cope with stress of being treated like an insect by these "big shot" faculty members! --Brad PS. Where are the error bars on extracted parameters? What about covariance? etc. etc...
See also: https://0x0.st/i164.png . The author did not make this easy for me, but I was able to extract data points and the "fit" function. In "Insect Pandemics", I mentioned that an analogy to Quantum mechanics allows that dot product can be used as an accuracy statistic. I also gave examples with accuracy better than 99%, some better than 99.7%. Roughly these are good fits, with acceptable fits better than 98%, and anything less than, say 95% completely failing. The accuracy statistic for the "PhD Dissertation" fit, as computed by the extraction above, was only 90%, much, much, much, lower than anything that could ever be reported, even as "satisfactory". This is obvious when looking at the graph, so the student, and the professors who signed their approval (William Oliver, Julia Kennefick, Etc.) should have known better. In case this analysis is not standard enough, I also calculated the chi-squared statistic, it was 984 on 29 degrees of freedom, in other words, the so-called "fit" is a __complete and total lie__. Let me tell you, in practice, what happens at the University of Arkansas. The chair and the vice chair sign their names on dissertations containing blatant lies, and they also sign their names on dismissal letters for students who are telling the truth, and give no factual explanation as to why they would do so. The meritocracy is dead, but politics of the day will continue to live, from the March purge, to the November purge, and to all the purges thereafter. --Brad On Sat, Apr 18, 2020 at 9:46 AM Brad Klee <bradklee@gmail.com> wrote:
Screen Cap: https://0x0.st/i1zI.png of Fig. 4.4 from: https://scholarworks.uark.edu/etd/2456/
Look how horrible these _published_ fits are!
participants (1)
-
Brad Klee