16 MASS. EXPERIMENT STATION BULLETIN 310 



should be given to the qualitative evaluation and how much to bod\' weight and 

 the amount of subcutaneous tissue in such cases is a very nice question to decide. 

 Ratings made in this manner, of course, are subjective and severe criticism of 

 their use in analysis of data would be justifiable had they been made by poorlj' 

 trained or inexperienced observers. Nevertheless, if made by experienced workers 

 who know children as the laboratory man knows experimental animals, who 

 are able to recognize signs of health and well-being on the one hand and of ill- 

 ness and malnutrition on the other, and who have had sufificient experience to 

 enable them to develop standards and power of discrimination, such ratings 

 would seem to be a more accurate index of nutritional status than any obtained 

 by the more objective measures available at the present time. Certainly, incorrect 

 or even absurd evaluations were not apparent, as was true in some cases when 

 any of the conventional methods were used. This should not be interpreted 

 as a contention that the ratings of the general nutritional condition of the children 

 are absolutely accurate. Some variations are inherent because of the personal 

 equation, and the ratings would be expected to differ somewhat if made by another 

 group of workers. They do, however, represent the best combined judgment of 

 two observers on the basis of the data at hand and, the writers feel, give a reason- 

 ably accurate basis for interpreting the results of the study. The choice lay 

 between accepting the results from arbitrary methods which have been shown 

 to be inadequate in many instances or from a common-sense evaluation of the 

 best in all methods. The latter seemed the more reasonable procedure. Further- 

 more, as all the evaluations were made by the same individuals, those for the 

 same child at different seasons and for the various groups should be entirely 

 comparable. 



Correlation Between Various Ratings of Nutritional Status 



Assuming the more comprehensive general evaluation of the status of our 

 subjects to be the most accurate, it seemed pertinent to make a statistical analysis 

 of the data to determine whether the deductions as to the comparative value of 

 the different methods arrived at by clinical observation were confirmed. For 

 779 records in which hip breadth was included, the nutritional status of the 

 children was estimated by the use of the Baldwin-Wood and the Lucas-Pryor 

 tables as well as from the medical examinations. These ratings and the amount 

 of subcutaneous tissue, which is the clinical evaluation of the thinness of the 

 child, were all correlated with the ratings of the "general" nutritional condition 

 at the same examination. As had been anticipated the highest correlation was 

 found between the "general" nutritional condition and that obtained from an 

 evaluation of the medical records (qualitative nutritional condition), the co- 

 efficient being -f.93 (Table 3). This indicates that on the whole these two ratings 

 of nutritional condition do not differ greatly. 



The other three correlations did not show so close a relationship and confirm 

 our first observation that none of the quantitative measures of nutrition are 

 dependable guides. A coefficient of even -J-. 72, as was found between the rat- 

 ing for general nutritional condition and the percentage deviation from the 

 Baldwin-Wood standard, does not indicate a close agreement between the 

 results of the two methods for evaluating the same conditions. Somewhat less 

 agreement was shown between general nutritional condition and the amount 

 of subcutaneous tissue (r = -|-.64), and still less with the deviations from the 

 Lucas-Pryor standards (r = -|-.49). It is somewhat surprising that the relation- 

 ship between this last standard and the general nutritional condition should 



