DUNLAPS METHOD FOR THE MEAN VARIATION 



BUFORD JOHNSON 



From the Psychological Laboratory of the Bureau of Educational Experiments 



In the Psychological Review for March, 1913, Knight Dunlap 

 described a simple method developed by him, of obtaining the 

 mean variation of a series of values, together with the necessary 

 operations for carrying out the process on a calculating machine. 



In the derivation of the formulae there is one obvious misprint 

 in writing (SP + PM) for (SP - PM), but this is correctly 

 printed when the same expression is used in the succeeding line. 



The formulae derived are as follows: 



MV = (SP - PM) +1/2N = (RM - 2R) -- 1/2 N, when 



N = total number of measures 



M = average or mean 



M V = mean variation 



P = number of terms greater in value than the average 



R = number of terms less in value than the average 



SP = sum of the terms greater in value than the average 



2R = sum of the terms less in value than the average 



These same formulae are reproduced, with a slightly different 

 symbolization, in Whipple's 1 Manual of Mental and Physical 

 Tests, Part I: Simpler Processes, which was published in 1914. 

 The following substitutes are used: 



N+M forP 



N-M ior R 



2 +M forSP 



S_ M for 2R 



.5 for 1/2 



1 Whipple's Manual of Mental and Physical Tests, Parti: Simpler Processes, 

 p. 22. Dunlap developed the method itself and not merely the application 

 of the calculating machine technique, as Whipple's footnote (p. 21) might 

 imply. We are not acquainted with any previous development of the method. 



325 



PSYCHOBIOLOGY, VOL. I, NO. 4 



