“…Because there were 3 sequential levels of the adherence variables (needs to obtain initial screening, overdue for rescreening, up-to-date with rescreening), we used continuation ratios to model 1) whether a woman has or has not been screened for cancer (ever vs never screened), and 2) whether a woman is up-to-date or not with repeat screening (up-to-date vs overdue repeat screening). 41,42 On the basis of results from past studies conducted with Latina samples, 18,22,23,26,30,33,37,[43][44][45] the following predictors were considered: demographic variables, exposure to cancer education, and interactions among these variables. Demographic variables were age, formal education (less than 6, 6, 7-11, 12, more than 12 years completed), annual household income ($0-1,500; 1,501-5,000; 5,001-7,000; 7,001-10,000; 10,001-14,000; 14,001-25,000; more than 25,000), recruitment site (TX, CT, NJ, DC, MW), country of origin (US, Mexico, Puerto Rico, South America, Central America and the Caribbean, other), proportion of life in US (.10, .25 and .43, which correspond to the 25th, 50th, and 75th percentiles, respectively), and having had a child.…”