28 



MASS. EXPERIMENT STATION BULLETIN 310 



As would be Inferred from such coefficients of correlation, the scatter in every 

 case was wide, indicating that many factors are involved in the picture of an 

 individual's nutritional condition even from a qualitative standpoint, and also 

 that the observations made by the physician are affected by other influences as 

 well as nutrition. Particular attention may be called to the relationships in- 

 dicated for the condition of the lungs and tonsils and the amount of subcutaneous 

 tissue with the nutrition ratings, which are shown graphically in Charts 2, 3, 

 and 4. 



The scatter plot (Chart 2) indicates that children might be rated from excellent 

 to very poor nutritionally and at the same time have lungs in which no defects 

 were found. But children whose lungs were in very poor or poor condition, for 

 the most part also had other symptoms which marked them as poor from a 

 nutritional standpoint. 



Subcuteneou: 



Chart 3. Correlation of Amount of Subcutaneous Tissue with "Qualitative" Nutritional Condi- 

 tion: showing that while the amount of subcutaneous tissue varied from very much deficient to 

 excessive when the children were in good condition as indicated by a medical examination, it was 

 almost invariably more or less deficient when they were in poor condition. 



The scatter for amount of subcutaneous tissue (Chart 3) indicates that, while 

 a child who had no clinical symptoms of malnutrition and was rated as good 

 might have subcutaneous fat padding varying in amount from excessive to 

 much deficient, those who were in poor condition almost without exception had 

 little subcutaneous tissue. 



The story of tonsils (Chart 4) was much the same. The child might show 

 all other signs of being in good and even excellent condition and still have poor 

 tonsils, but children in poor condition of nutrition did not have good tonsils and 

 in most cases not even fair ones. 



No attempt has been made to work out an objective scale for evaluating records 

 of medical examinations, weighting the points observed on the basis of the amount 

 of correlation found. Such an attempt did not seem justified at this time for 

 several reasons. For one thing, because of inadequate descriptions, correlations 



