97 



false impression of their food consumption; in some cases b}- purchas- 

 ing larger quantities than usual or by reporting larger amounts than 

 were actualh' purchased, and in other cases by omitting to mention some 

 of the purchases made. For instance, it appears that families who 

 had formerly been in more comfortable circumstances would some- 

 times be ashamed to let an outsider know how economically the}' were 

 now living, and perhaps how insufficienth' the}' were nourished. In 

 such cases there might be a tendency to procure more food during the 

 time of the study than ordinarily. Other families, suspecting an 

 opportunity for pecuniary assistance, might be tempted to purchase 

 less food than usual, or to conceal food already on hand. Another 

 possible source of error is in the failure to make proper record of 

 the number of meals taken by each memljer of the family or by any 

 visitors. 



It is noticeable that quite generally the results of dietary studies 

 among' poor families, where the statistics are recorded by the families 

 themselves, indicate a larger food consumption than is found in the 

 more reliable studies in which the food materials were actually weighed 

 by those conducting the investigations. This was very forcibly illus- 

 trated by the results of dietary studies in Chicago in 1895 and 1896, 

 described in a former publication of this Office.'^' In 25 studies in 

 which the data were collected entirely b}' the investigators the 

 average cost of the diet per man per day was 17.9 cents, and the 

 average quantity of protein 116 grams, and of energy 3,160 calories. 

 The persons in charge of these studies also conducted at the same 

 time 28 others, in which the statistics were kept by the families them- 

 selves. These were made with families in the same localities and 

 living under the same circumstances as the others, but the average 

 cost of the diet per man per day as recorded was 22.1 cents, the 

 average quantit}' of protein 11-7 grams, and of energy 3,550 calories, 

 A comparison of the individual studies shows that where the statistics 

 were furnished by the families the differences in results with different 

 families were very much wider than in the studies made entireh' by 

 the investigators. 



In the dietarv studies in Philadelphia reported in these pages the 

 families were not very intelligent and were in destitute circumstances; 

 the chances for errors were therefore comparatively large. The fami- 

 lies in the Chicago studies here reported were more intelligent and 

 were in more comfortable circumstances, so that the possibilities of 

 error in this case seem smaller. Bearing these facts in mind, it is 

 evident that too sweeping conclusions should not be drawn from the 

 results of the studies themselves or from the averages as compared 

 with those of later studies of families in similar circumstances. 



«U. S. Dept. Agr., Office ot Experiment Stations Bui. 55. 

 25580— No, 129—03 7 



