RESPIRATION 



V. THE EVALUATION OF ANTHROPOMETRIC DATA. 



A large proportion of the problems that the medical man has to 

 solve involves the finding of averages of a large number of observa- 

 tions. This is sure to be true in all anthropometric problems. In 

 the course of the preceding lesson valuable anthropometric data 

 were collected and recorded upon cards. The value of this material 

 is purely potential. Before the data will furnish a basis for drawing 

 conclusions it is necessary to subject them to a process of evaluation. 

 This process consists, first, in grouping; second, in getting the average 

 or the median value for each measurement; and, third, in graphically 

 representing the averages. In the previous lesson the observer noted 

 upon each card whether the subject had lived in a hilly or flat country; 

 further, whether he had lived a physically active or inactive life. 

 This gives one an opportunity for four groups when the cards for 

 the whole class are collected. 



Group I. Active men from a hilly country. 



Group II. Active men from a flat country. 



Group III. Inactive men from a hilly country. 



Group IV. Inactive men from a flat country. 

 Until recently it has been customary simply to write the data for 

 any group in columns and "strike an average" of each column. 

 If there are only 10 to 20 or 30 individuals in each group this method 

 does not entail the unnecessary expenditure of much energy, but it 

 is not reliable, for one "giant" or "dwarf" in any group would 

 vitiate the whole result. If there are 100 or 1000 individuals in a 

 group, then the use of the old method of finding the arithmetical 

 average is exceedingly wasteful of both time and energy. It must 

 be added, however, that when the number of observations is large 

 the chances are that there will be as many dwarfs as giants, thus 

 making the average approximate closely the median value. It is 

 the latter we are seeking, viz., the median value; this may be defined 

 as that value which is so located in the whole series of observations 

 in a single measurement of any group, that there are as many below 

 it as above it i. e., that the number of values which it exceeds equals 

 the number of values which exceed it. 



Let us take a concrete case. In a group of 316 seventeen-year-old 

 boys certain physical measurements were recorded upon individual 

 cards. Let us take, for example, the girth of the head recorded in 

 centimetres and tenths. Instead of writing in a column the 316 

 head-girths, each expressed in three figures, adding and averaging, 

 let us adopt the new method, first suggested by the Belgian astronomer 

 and anthropologist, Quetelet, and later elaborated by Galten, the 

 London anthropologist. Arrange the cards in piles, placing in one 

 pile all of the cards having girth of head 51 cm., in another pile all 



