ABSTRACT
The obesity and alcohol intake categories appear to have been chosen to equate roughly the percentages of patients in these categories. The strongly negative interaction in the cell indicates that patients without obesity or hypertension are negatively associated with high alcohol intake. The statistical significance of the association between obesity and hypertension varies at different levels of alcohol intake. In 1966, cervical cancer was associated with the presence of abnormal cells in smear tests, but not all abnormal cells were related to cervical cancer. The student confined attention to data on patients with one of these grades of abnormality. Kelly McKerrow considered data for two periods, 1981-1986, 1988-1993, split by age and grade of abnormality, but without recording outcome; only a couple of hundred patients from each sample actually developed cervical cancer.
