Haggerty: Studies in Arithmetic 



63 



Table I groups the results from the different systems so that 

 they are comparable. Figures 1 and 2^ show the standing of the 

 county systems compared with the average of the twenty Indiana 

 cities tested in 1914, which average is represented by the lines 

 marked ''Indiana Standard". I wish to call attention to but three 

 points here: the improvement shown by the second testing of 

 Warren county, the uniformity of the results from the different 

 systems, and what it means for a system to be up to standard. 



Superintendent Harry Evans of Warren county, in discussing 

 the improvement shown by the second testing, states that after 

 giving the first tests he gave the results to the teachers and 

 pointed out a few weaknesses, but did not introduce any con- 

 structive work. He thinks, however, that the pointing out of 

 the weaknesses probably served as a stimulus to both teacher and 

 pupil. He also mentions the fright of the children at the first 

 testing. 



As for the second point, an examination of Table I shows but 

 a slight degree of variation in results from different systems as 

 compared with the twenty Indiana cities^ tested in 1914. There 

 are of course many factors entering into this, some of which are 

 discussed later in this Study, but supervision probably has as 

 much to do with the variations in the case of the city scores as 

 any other one factor. Might, then, the uniformity in the systems 

 here given be due to their comparative lack of supervision? 



The point as to the meaning of the standard remains. It 

 will be noticed that in a great many places the scores of the county 

 systems exceed the Indiana Standard. If we are to take this 

 standard in any case as the minimum score which a child should 

 have, and the maximum necessary for him to have, we might con- 

 clude that no further work is needed in these classes in funda- 

 mentals of arithmetic. This conclusion would be erroneous for 

 when the median score of a group is equal to the standard only 

 50 per cent of the group have a score equal to or exceeding the 

 standard. There is still work to be done with the low 50 per cent. 

 These individual cases should be examined carefully to discover 

 whether the low scores are the results of wrong grading or of 

 some obstacle. Attention can be given to these while the standard 



3ln figures 1 and 2 the greater the height of the left end of the graph, the greater 

 the "speed". The greater the slope upward toward the right the greater the per 

 cent of accuracy or "dependability". Equal heights represent equivalent speeds 

 thruout. Equal slopes represent equivalent dependabilities. 



■•See Second Annual Conference on Educational Aleasurements, {Indiana University 

 Bulletin, XIII, No. 11), p. 26. 



