Arithmetic: A Cooperative Study in Educa- 

 tional Measurements 



By M. E. Haggerty, Associate Professor of Psychology, Director of the 

 Psychological Laboratoj-y,, and Director of the Bureau of 

 Cooperative Research, Indiana University. 



PART I 



A COMPARATIVE STUDY OF TWENTY INDIANA CITIES 

 Tests and Methods 



All of the cities ^ave the Courtis Standard Test, Arithmetic, 

 Series B. Some gave Test 7 in Series A, but for several reasons 

 the results of the latter are not included in this study. In the 

 case of Series B, the tests were given as directed, the individual 

 scores were computed, the results were transferred to score sheet 

 No. 1, the median and per cent of variability were computed on 

 this sheet, and these two items were then transferred to score 

 sheet No. 3. All of the above work was done by the teaching- 

 corps of the several cities. All of the material — individual score 

 sheets, record sheets No. 1 and record sheets No. 3 — were then 

 sent to the University. These records were then rechecked to see 

 whether the work had been properly done. The general method 

 was as follows : Sample individual score sheets were taken at 

 random from each city. If these proved to be correctly done it 

 was assumed that the class distributions on record sheet No. 1 

 were correct. If these individual score sheets were in error they 

 were corrected and new distributions were made out. After this, 

 the computations of medians and variabilities were verified and 

 record sheets No. 3 were corrected. Owing to a rather general 

 misunderstanding of some of the methods used the amount of 

 corrective work was considerable; the compensation is the fund of 

 highly dependable data here presented. 



(389) 



