124 PSYCHOLOGICAL METHODS OF TESTING INTELLIGENCE 



vail decided differences with regard to the class- 

 marks. The plan, which was formerly quite gener- 

 ally followed, whereby each pupil had a "class- 

 place," which determined his rank among his class- 

 mates for a quarter of a year, is now becoming less 

 and less common. Sometimes the statement of 

 standing is accompanied by a designation of place in 

 the class, e. g., "promoted as 15th in a class of 27," 

 without laying any further special stress on the 

 ranking ; sometimes even this designation is lacking, 

 so that there exists no positive school rank-order at 

 all. Of course, even where there is a school rank- 

 order, it is to be desired that the teacher work out 

 his estimation of intelligence as far as possible inde- 

 pendently of this school ranking. On this account 

 it is in every way objectionable to proceed, as is often 

 done because it is the easiest way, by taking the 

 school ranking as a starting point and simply shift- 

 ing the position of those children whose rank in this 

 list has been displaced by some special circumstance, 

 like illness, transfer of school, evident laziness, etc. 

 Burt, for example, worked in that way. To be sure, 

 such a "corrected school ranking " is doubtless bet- 

 ter than an uncorrected one for psychological pur- 

 poses, but it by no means presents a correct ranking 

 of intelligence. 



To secure as impartial a ranking of intelligence as 

 possible, the following procedure may be recom- 

 mended. Write the names of each of the pupils to 

 be ranked on a separate card, and arrange these first 

 in alphabetical order. Then and only then sort out 

 the cards into different groups for their intelligence, 

 and finally try to settle upon rankings within each 



