We cannot hope at present to get nearly so accurate a judgment about the frequency of deficiency in groups by means of any school test as by the psychological tests. Nevertheless, I believe that it may furnish us some supplementary evidence. The main difficulty in formulating any general rule for interpretation of the school level is that very different plans of promotion prevail in different school systems. It is not uncommon, for example, to find that a child will be promoted to a higher grade regardless of his ability provided that he has spent two years with the same teacher. This practise, of course, makes it impossible to judge a particular individual's ability by the school grade he has attained without knowing how he reached it. Nevertheless, spending two years in each grade will begin to show in a general distribution of pupils by the time we deal with 12-year-olds. I have gone over the tables of school retardation of pupils provided by Strayer for several hundred cities in the United States and I find that the percentage method of approach gives us at least a rough cue as to what might be expected by any general principle of interpretation (_189_). Using age 7 as satisfactory in the first grade, 8 in the second, and so on, we find that among 319 cities of all sizes, half of them had 2% or more retarded four or more years in school position. This condition was about the same for cities less than 25,000 as with the larger cities. On the basis of school position for groups of children of all the school ages it would, therefore, be safer to make a low estimate of the frequency of mental deficiency on the basis of five or more years of scholastic retardation in the groups and regard 4 years or more of school retardation as a maximum estimate. Since most children leave school at 14 it is generally best to regard all older as only 14 years of age when estimating deficiency. I have not been able to check this by school and test records on a group of children through all the grades. Goddard's published records do not give the mental ages for those four or more years retarded scholastically. Moreover, he only included those in the sixth grade and below. For a group of young children this estimate would undoubtedly be too low. The delinquent groups, however, are all older. Most of them, if they lived in this country have gone to school until they were at least 14 years of age. Wallin (_211_) and Strong (_190_) also give records of school position to check the Binet rating. TABLE XV. PERCENTAGES OF PUPILS 12 AND 13 YEARS OF AGE MOST SERIOUSLY RETARDED IN SCHOOL