OKLAHOMA ACADEMY OF SCIENCE 145 



but also the checking, compiling, and tabulating were so executed 

 as to remove any suspicion of unreliability. .All who took part in 

 the experiment were interested in the accuracy of the results. 



The exercise in automatic spelling was Contest 15 in Patterson's 

 Thirty Contests in Spelling (published by Educator Supply Co., 

 Mitchell, S. D.), and the exercise in rational spelling was the list of 

 the same fifty words used as contest words in Contest 15, each word 

 being pronounced twice. 



The detailed results of the experiment are found in the fol- 

 lowing table. 



Table I. A Comparison of Median Scores for Automatic and 

 Rational Spelling; (Ten Classes With Total of 286 Pupils). 



-3 PL, k '-^ Oh 



TOTALS 286 832 831 



An analysis of the above table shows that the medians for 

 autom.atic and rational spelling are almost the same, being in fact 

 identical for the three classes in grade 6. It also shows that there 

 is a tendency for the medians to differ less as the grade increases, 

 since the average difference for the fourth grade is 4.5%, for the 

 fifth, 2.3% ; and for the sixth, 0%. 



There is one generalization which can not be neglected. A 

 careful study of the above table shows that the higher median is 

 ' always that of the exercises which came second. The only excep- 

 tion to this is in the three classes in grade 6, where the medians are 



