ESTIMATION AND TESTING OF FINER GRADATIONS 129 



cases before these had taken place. The estimates, 

 then, were not affected by the ranking in the school 

 tests, and the rather high correlation could therefore 

 be regarded as a reliable expression of the degree of 

 correspondence between intelligence and the work 

 done in the examinations. Burt also had estimates 

 of intelligence of the same pupils made by different 

 teachers and by disinterested school-mates of the 

 pupils. The correlations between these estimates 

 are very high, but since all those who estimated set 

 out from the already known rank-order of the pupils, 

 which they had merely to correct, it follows that this 

 high correspondence is nothing remarkable and that 

 it has no scientific value. 



In the course of discussions of this subject in the 

 meetings of the Psychological Seminary at Breslau, 

 during the winter semester 1911-12, the need became 

 evident of clearing up the methodological aspects of 

 the whole subject of estimating intelligence by trials 

 of our own. Fortunately, two of our members, who 

 were engaged in practical school work, were ready to 

 secure new material. 3 The results thus obtained are 

 worth noting because they very clearly demonstrate 

 the methodological difficulties and the way to over- 

 come them and also bring out the necessary differ- 

 ence between the procedure in secondary and in ele- 

 mentary schools. 



Principal Eindfleisch had the teachers in charge 

 of a boys' Volksschule prepare for their classes lists 

 that showed both the ranking of the pupils on the 



"Hearty thanks are due to Principal Rindfleisch (Liegnitz) and 

 Dr. Scheifler (a high-school teacher at Gorlitz) for their great 

 pains and for their courtesy in placing the material at my disposal. 



