148 BUREAU OF AMERICAN ETHNOLOGY [Bull. 197 



Comparing the scores obtained from the enrolled Navaho popula- 

 tions with those obtained in the decennial censuses, we find that the lat- 

 ter appear to produce somewhat better age data. It must be recalled, 

 however, that these differences may be fortuitous in view of the smaller 

 frequencies obtained from the former source. 



The lowest and highest scores obtained among the Navaho age distri- 

 butions are sufficiently variant to merit special consideration. The 

 lowest scores, obtained from the Kamah data for the period 1920 to 

 1948, are not surprising in view of the exceptional care with which the 

 ages of the numbers of this population were obtained and verified. 

 These procedures would tend to minimize distortions due to "age 

 heaping," "digit preference" or other manifestations of inaccurate 

 procedures in recording age data. Furthermore, the Ramah popula- 

 tion itself appears to have been sufficiently free of important migratory 

 influences or other disturbances to maintain a relatively smooth age 

 distribution in this period.^^ 



In contrast, the highest scores are found for the Fort Defiance Agen- 

 cy population in 1915, and for the Indian populations enumerated in 

 1960 in the six counties of Arizona, Utah, and New Mexico which in- 

 clude the Navajo Eeservation. The former cases appear to reflect 

 the crude procedures whereby ages were recorded (or assigned) by the 

 responsible officials. In the case of the 1960 census returns, the large 

 excess of persons reported in the age group 55 to 59 yields a consider- 

 able increase in the resultant scores. 



In regard to the relative merits of the age data provided by the 

 Bureau of Indian Affairs as against those of the Bureau of the Census, 

 tentative conclusions can be offered. 



In the first place, neither source appears to provide data on the 

 Navaho population that are better than "fair" in quality. This con- 

 clusion, of course, rests on the assumption that the actual age distribu- 

 tion of both the enrolled and the enumerated Navaho populations, like 

 those of the total U.S. Indian populations, are not marked by major 

 irregularities at any point. This assumption is supported by our 

 general knowledge that none of these populations has experienced 

 severe depletion because of wars or epidemics during the past two 



same excess among Indians residing on reservations. It is also conceivable that large 

 numbers of older Indians reported their age as "about GO," or were recorded as about 60 

 by the enumerators. If, in such cases, the enumerator filled the circles on the census 

 schedule for the 1900 decade and for year 0, the computer would have allocated the quarter 

 of birth to such persons at random, with the result that three-fourths of these persons 

 would have been tabulated as 59 years old, and one-fourth as 60. However, such a 

 procedure could not possibly account for such a large excess of persons in the 55 to 59 

 group without producing a smaller but perceptible excess in the 60 to 64 group as well. 

 The puzzle remains unsolved as of this writing. 



99 The smoothness of the Ramah age distribution is apparently enhanced by the fact that 

 each age-specific total is the sum of the figures reported annually for a 16-year period. 

 These synthetic totals would thus minimize the effect of year-to-year fluctuations in the 

 numbers given. The relatively high sex ratio score obtained from this population, on the 

 other hand, may reflect actual peculiarities in the sex ratios of this small population. 



