IMPROVING VALIDITY OF SELF REPORTS OF CANCER SCREENING

本建议的具体目的是进行一个复杂的析因实验来评估收集自传体信息的四种方法，我们假设这些方法可能有助于减少过度报告健康促进和疾病预防行为。评估的方法包括：A)从被调查者那里收集数据的方式，b)在询问行为是否发生之前询问获得程序的意图（一种社会期望的行为），c)在询问他们是否收到程序之前询问可能阻止他们获得程序的障碍或问题（不执行行为的理由），d)询问他们作为年度身体检查的一部分可能报告的定期检查的例外情况，作为提高他们估计的一种方式。我们将把研究重点放在三个特定的癌症筛查程序上：巴氏涂片检查、乳房x光检查和乳房检查。数据将从800名55岁及以上的妇女中收集，这些妇女居住在伊利诺斯州的香槟-厄巴纳，并在该地区居住了至少五年。在使用计算机辅助电话访谈（CATI）或音频计算机辅助自我访谈（ACASI）完成访谈后，我们将要求受访者允许提取他们的医疗记录，以验证他们的自我报告或癌症筛查检查。分析将集中于评估每个实验操作的实际效用，以提高自我报告的准确性。

The specific aim of this proposal is to conduct a complex factorial experiment to evaluate four approaches to collecting autobiographical information that we hypothesize may help reduce over reporting of health promoting and disease prevention behavior. The methods to be evaluated include: a) the mode in which the data are collected from the respondent, b) asking about intention to obtain the procedure (a socially desirable behavior) before asking whether the behavior occurred, c) asking about barriers or problems that might keep them from getting the procedure (justification for not performing the behavior) before asking about whether they had received it, and d) asking about exceptions to the regularity with which they might report getting the examination as part of their annual physical examination, as a way of improving their estimates. We will focus this research on three specific cancer screening procedures: Pap tests, mammograms and breast examinations. Data will be collected from 800 women aged 55 and older who live and receive their health care in Champaign-Urbana, Illinois and who have lived in the area for at least five years. After completing interviews using either computer-assisted telephone interviewing (CATI) or audio computer- assisted self interview (ACASI), we will ask respondents for permission to abstract their medical records in order to validate their self reports or cancer screening examinations. Analyses will focus on assessments of the practical utility of each experimental manipulation for improve the accuracy of self-reports.