THE DATA FOR THIS SURVEY WERE OBTAINED FROM MARRIAGE LICENSES ON FILE in the Bureaus of Vital Statistics in Los Angeles, Kern, and Stanislaus counties. A two-stage sampling procedure was followed. As a first step; random samples of 200 to 250 marriage licenses were drawn for each county and year to be studied. These provided estimates of the representation of Southwesterners, natives, and other social groups in the marrying population. The number of Southwesterners in these initial samples was, however, too small for reliable analysis. Consequently a second random sample was taken, consisting only of licenses where one or both of the parties was a white Southwesterner. Since the pattern of male and female choices appeared to be quite comparable in the initial survey, I did not sample them separately. The size of each combined sample of Southwesterners and the margin of error associated with estimates of their marriage choices are reported below.
|County||Year||Sample size||Margin of error*|
|* At the 95 percent confidence level.|