1. If the example which Thernstrom set by depositing his data at the Historical Data Archives of the Inter-University Consortium for Political and Social Research at the University of Michigan could be as widely followed as his path-breaking studies of historical mobility were, the profession would benefit greatly. As we hope to demonstrate, secondary data analyses, too seldom performed by historians, may uncover new facets of the data. We obtained Data Set ICPSR # 7550 from the consortium. Naturally, neither Thernstrom nor the ICPSR bears any responsibility for the analyses we performed
2. The occupational classifications are Thernstrom's. Age, of course, could be treated as an interval level variable. We cut it into categories only in order to illustrate this particular form of log-linear analysis. The number of cases in the sample was cut from 3,362 to 1,724 by our decision to exclude the thirty-five Negroes, white men for whom any data was missing, and, most importantly, all males under fourteen years of age in 1880. Of our exclusions, 84 percent (1,362 of the 1,628) were because of age. One indication that eliminating cases for which there was missing data did not seriously distort our findings was that the proportion “found” in the smaller sample differed from that in the larger sample by less than 1 percent. The age and status variables were collapsed into four and two categories, repsectively, to simplify the presentation. Loglinear runs on a 100-celi table with status broken into three categories and age into five produced results very similar to those presented below
3. Jensen , Richard J. “Found: Fifty Million Missing Americans,” . paper delivered at the Social Science History Association Convention . 8 November . On this point, see