12.07.2015 Views

1 Studies in the History of Statistics and Probability ... - Sheynin, Oscar

1 Studies in the History of Statistics and Probability ... - Sheynin, Oscar

1 Studies in the History of Statistics and Probability ... - Sheynin, Oscar

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Ermolaeva’s results <strong>and</strong> concluded that, <strong>in</strong>stead <strong>of</strong> refut<strong>in</strong>g <strong>the</strong>Mendelian law, she completely confirmed it. There also, withoutm<strong>in</strong>utely analys<strong>in</strong>g En<strong>in</strong>’s paper, Kolmogorov implied that his resultsare doubtful because <strong>the</strong>y confirmed that law too f<strong>in</strong>ely.In a popular scientific booklet, I [iii] thought it expedient to rem<strong>in</strong>dreaders about Kolmogorov’s paper <strong>and</strong> supplemented it by treat<strong>in</strong>gEn<strong>in</strong>’s results. Alimov treated <strong>the</strong> same data o<strong>the</strong>rwise <strong>and</strong> formulateda number <strong>of</strong> objections. He directed <strong>the</strong>m to me alone although a part<strong>of</strong> <strong>the</strong>m to <strong>the</strong> same extent concerned Kolmogorov’s calculations. Ibeg<strong>in</strong> with <strong>the</strong> objection which I underst<strong>and</strong> <strong>and</strong> consider essential.He notes that <strong>in</strong> many cases <strong>the</strong> families considered by Ermolaevawere small (not more than 10 observations). Then <strong>the</strong> normalapproximation <strong>of</strong> <strong>the</strong> frequencies <strong>of</strong> a certa<strong>in</strong> phenotype <strong>in</strong>troduced byKolmogorov ought to be very rough. In particular, <strong>the</strong> presence <strong>of</strong>normed frequencies smaller than − 3 which I [iii] considered assignificant deviations from <strong>the</strong> Mendelian law can be expla<strong>in</strong>ed. asAlimov believes, by <strong>the</strong> asymmetry <strong>of</strong> <strong>the</strong> b<strong>in</strong>omial law. Alimovdeclared that my conclusion was wrong (that was somewhat hastily, heshould have said unjustified). Any student <strong>of</strong> a technical <strong>in</strong>stitute, as hestates, would have avoided such a mistake caused by <strong>the</strong> generalcorruption <strong>of</strong> concepts due to <strong>the</strong> application <strong>of</strong> <strong>the</strong> non-Miseslanguage <strong>and</strong> <strong>the</strong> rituals <strong>of</strong> ma<strong>the</strong>matical statistics.Actually, everyth<strong>in</strong>g is much simpler. Before prepar<strong>in</strong>g my booklet,I did not acqua<strong>in</strong>t myself with Ermolaeva’s paper which was notreadily available. Now, however, s<strong>in</strong>ce her data became an object <strong>of</strong>discussion, I had a look at that source. The data on <strong>the</strong> assortment <strong>in</strong>separate families are provided <strong>the</strong>re <strong>in</strong> Tables 4 <strong>and</strong> 6. In Table 4 <strong>the</strong>families are numbered from 1 to 100, but for some unknown reasonnumbers 50 <strong>and</strong> 87 are omitted. In Table 6, <strong>the</strong> number<strong>in</strong>g beg<strong>in</strong>s with22 <strong>and</strong> cont<strong>in</strong>ues until 148, but numbers 92, 95, 115, 127, 144 areabsent. At <strong>the</strong> same time, <strong>the</strong> table show<strong>in</strong>g <strong>the</strong> total, states 100 <strong>and</strong>127 families respectively.Kolmogorov <strong>in</strong>serted a venomous pert<strong>in</strong>ent remark; he counted 98families <strong>in</strong> <strong>the</strong> first, <strong>and</strong> 123 (actually, 122) <strong>in</strong> <strong>the</strong> second table. Thegeneral style <strong>of</strong> her contribution, let me say it frankly, is abom<strong>in</strong>able.The author obviously does not underst<strong>and</strong> <strong>the</strong> mean<strong>in</strong>g <strong>of</strong> <strong>the</strong> errorscalculated by biometric methods for <strong>the</strong> number <strong>of</strong> assortments. It isquite clear that her data do not really deserve to be seriouslyconsidered.However, if only discuss<strong>in</strong>g Ermolaeva’s tables such that <strong>the</strong>y are,Alimov is still unjustly reproach<strong>in</strong>g me for discover<strong>in</strong>g non-existentdeviations from <strong>the</strong> Mendelian law. Indeed, Table 4 <strong>in</strong>cludes a result<strong>of</strong> assortment 0:17, <strong>and</strong> 0:10 <strong>in</strong> Table 6 <strong>in</strong>stead <strong>of</strong> <strong>the</strong> expected ratio3:1. Their probabilities are 4 −17 <strong>and</strong> 4 −10 respectively so that, hav<strong>in</strong>g200 plus trials, such events could not have occurred.Concern<strong>in</strong>g both Kolmogorov’s <strong>and</strong> my own treatment, I wouldlike to <strong>in</strong>dicate that, <strong>in</strong> spite <strong>of</strong> Alimov’ op<strong>in</strong>ion, correct scientificresults are possibly <strong>of</strong>ten obta<strong>in</strong>ed not because we do everyth<strong>in</strong>gproperly, but ow<strong>in</strong>g to some special luck.I did not underst<strong>and</strong> <strong>the</strong> mean<strong>in</strong>g <strong>of</strong> Alimov’s objection to <strong>the</strong>calculation <strong>of</strong> <strong>the</strong> confidence level. From <strong>the</strong> times <strong>of</strong> Laplace, after136

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!