January 17, 1919] 



SCIENCE 



55 



Bricklayer 



Cobbler 



Caterer 



65 to 69 

 Laundrymau 

 Plumber 

 Auto repairman 

 General pipe-fitter 

 Auto engine mechanic 

 Auto assembler 

 General mechanic 

 Tool and gauge maker 

 Stock checker 

 Detective and policeman 

 Tool-room expert 

 Ship carpenter 

 Gunsmith 



Marine engine man 

 Hand riveter 

 Telephone operator 



85 to 89 

 Photographer 



95 to 99 

 General clerk 

 FUing clerk 



105 to 109 

 Mechanical engineer 



115 to 119 

 Stenographer 

 Typist 

 Accountant 

 Civil engineer 



125 and over 

 Army chaplains 

 Engineer officers 



This table shows conclusively that in the 

 sort of ability measured by the test (1) skilled 

 mecfianics and tradesmen, men who work with 

 tools, are in frcneral very closely alike and very 

 low — near the level of the imskilled laborer; 

 (2) clerical workers are in general very high — 

 near the level of professional men. Either 

 the clerical worker is a man of much greater 

 general intelligence than the blacksmith, car- 

 penter, locomotive engineer, machinist, tool 

 maker, gunsmith or assembler; or the ability 

 measured by the test is very much specialized; 

 or both of these statements are true in a more 

 moderate form. The matter is one of great 

 importance. In proportion as it is true that 

 the more intelligent men seek clerical work 

 rather than work in skilled trades, an es- 

 sentially invidious class distinction will tend 



Bridge carpenter 

 Railroad conductor 

 Bailroad shop mechanic 

 Locomotive engineer 



70 to 74 

 Truckm aster 

 Farrier and veterinarian 



75 to 79 

 Receiving clerk 

 Shipping clerk 

 Stock keeper 



80 to 84 

 General electrician 

 Telegrapher 

 Band musician 

 Concrete construction 

 foreman 



90 to 94 

 RaUroad clerk 



100 to 104 

 Bookkeeper 



110 to 119 

 Mechanical draughts- 



120 to 125 

 Y. M. C. A. secretaries 

 Medical officers 



to have a real basis in fact; and the manage- 

 ment of business concerns will tend to fall 

 into the hands of men trained in the office 

 and salesroom rather than in the shop. In 

 proportion as this representative of our stand- 

 ard tests of intelligence is specialized, over- 

 weighting ability to think with words and 

 symbols in comparison with ability to think 

 with materials and mechanisms, our whole 

 procedure in measuring intelligence requires 

 a critical review; and probably the common 

 view of intelligence requires reconstruction. 



No less significant is the variability within 

 each occupational group. Taking the meas- 

 urements as they stand, the 75 percentile im- 

 skilled laborer is up to the level of the median 

 general mechanic, tool-room expert, or auto- 

 mobile mechanic, and up to the level of the 

 25 percentile mechanical engineer. The 75 

 percentile railroad clerk is at the level of the 

 average accountant or civil engineer. The 75 

 percentile receiving or shipping cleark is at the 

 level of the 25 percentile physician. This 

 variability would be reduced by longer and 

 repeated tests, but, unless the test as given 

 has a very large probable error, it would still 

 be enormous. It would still imply that there 

 were in the occupations supposed to demand 

 a high minimum standard of intelligence, a 

 very large number of dull men; and in the 

 occupations supposed to give little opportunity 

 for the use of intellect, a very large number 

 of gifted men and consequently a large un- 

 used surplus of intellect. Further informa- 

 tion concerning the exact nature of the abili- 

 ties of which the test is symptomatic is evi- 

 dently important here. 



As one considers the use of intelligence 

 tests in the army, the question at once arises, 

 "If for the sake of war we can measure 

 roughly the intelligence of a third of a million 

 soldiers a month, and find it profitable to do 

 so, can we not each year measure the intelli- 

 gence of every child coming ten years of age, 

 and will not that be still more profitable?" 

 A more extended test such as will place an in- 

 dividual on tlie scale for intellect for his age 

 with an average error .of not over 0.2 the 

 mean square deviation for his age, would 



