Abstract:
The decision to undertake a screening test is conditional upon awareness of screening. From an econometric
perspective there is a potential selection problem, if no distinction is made between aware and unaware nonscreeners.
This paper explores this problem through analysis of the determinants of cervical screening in Australia.
Cervical cancer is one of the most preventable and curable forms of cancer. Since 1991 there has been a concerted
effort in Australia to recommend and encourage women to have Pap smears every two years. The success of this
program can be partly gauged by exploring the determinants of screening for cervical cancer. Using unit record data
from the 1995 National Health Survey, an econometric model is developed for whether women have ever screened or
not. A proportion of women in the sample contend that they have never heard of a Pap test. The analysis
characterizes this group of women and accounts for their presence in the modelling. The paper demonstrates failing
to model awareness can result in inconsistent parameter estimates even when the degree of censoring in the sample is
relatively small.
