WORDS AND THINGS

1906 BMJ (Clinical Research Edition)  
an error in the last-series of ttibles, though the po'int affected is ot no real importance. Nothing, indeed, in the way of mathematical criticism to, which the tables can be subjected can disturb their great outstanding feature, namely, the extreme and quite abnormal difference shown between certain large classes of, candidates in the number of men who fail to reach the qualifying standard. This is the point of real interest in the tables, for though, obviously, all candidates cannot win
more » ... es cannot win commissions, at any rate, all might pass the qualifying standard, and in the winning of commissions candidates from the three divisions of the country-England, Scotland, and Irelandmight be expected to show a proportionately equal degree of success. So far, however, from this being the case,. there are material differences in. the latter respect, while in the former the percentages of total failures or of mn who fail to pass the qualifying, standard range from about 3 to 43. As an explanation of the very great difference in the degree of success obtained by different classes of candidates, Professor Dixon suggests that, in the earlier part of tlWe period covered by the returns, the scope of the examinations was not equally widely known in all parts of the kingdom; if this were the case the differences would, of course, tend to disappear, and the author by certain calculations of his own endeavours to show that they are doing so, thus proving the trath of his explanation. It is, doubtful, however, whether many persons will be convinced by his reasoning on this point, for, though it is true that in the last year orsothenumberof Irish candidates has augmented and that they have gained proportionately more commissions, Professor Dixon's own tables show that the limits of total failure have widehed, not diminished.
doi:10.1136/bmj.2.2391.1140 fatcat:77tp56iibzdzra3vu7wihytfnq