Taken at face value, that is certainly one conclusion that might be drawn, at least as to four Southern States, but the figures merit a closer look. What Klineberg did, as Garrett has shown, was to take the four Southern States where the white medians were _lowest_ and compare them with the four Northern States where the Negro medians were _highest_. Beyond demonstrating that Negroes in some Northern States scored higher than whites in some Southern States, this widely reproduced table tells us little. Moreover, Klineberg--and Montagu, and Benedict, and others who are so fond of this data--do not present the figures from the four Northern States that might truly have significance in terms of local problems of public education. Garrett, whose computations of medians differ slightly from Klineberg’s, puts the data together in this fashion: --------------------------------------------------------- _Number Tested_ _White_ _Negro_ _State_ _White_ _Negro_ _Median_ _Median_ --------------------------------------------------------- Pennsylvania 3,089 498 64.6 41.5 New York 2,843 850 64.0 44.5 Illinois 2,056 578 63.0 46.9 Ohio 2,318 152 66.7 48.8 --------------------------------------------------------- Garrett then makes the self-evident point that Negroes in these four States scored as far below white soldiers _from the same States_ as they scored below whites in the country as a whole. The student who wants to dig more deeply into these World War I findings will find them fully reported in professional literature. Audrey Shuey’s _The Testing of Negro Intelligence_ summarizes the data and provides an extensive bibliography of work done on the figures. It is curious that so much labor has been spent on the World War I figures, and relatively so little on the more up-to-date data from World War II and Korea. Yet from one point of view this is not so curious either: In the thirty-six years between 1917 and 1943, the American Negro experienced prodigious gains in educational, cultural, economic, and social opportunities. Surely, it might be thought, these gains would have been reflected in some significant improvement in his military test scores. No such improvement can be detected. Nathaniel Weyl has summed up the facts: