Jo Heui-Sug, Jung Su-Mi
Department of Health Policy and Management, School of Medicine, Kangwon National University, Kangwon, South Korea.
Asian Pac J Cancer Prev. 2011;12(4):869-74.
This study was designed to segment the audience group of '10 lifestyle for cancer prevention' based on demographic characteristics and the level of knowledge about each guideline for cancer prevention among the community in South Korea.
Participants were chosen through stratified random sampling according to the age and gender distribution of Gangwon province in South Korea. A telephone survey was conducted from 6 to 15 calls among 2,025 persons on October 2008. A total of 1,687 persons completed the survey (response rate: 83.3%). Survey items were composed of socio-demographic characteristics such as age, gender, income, education, and residence area and the knowledge level of '10 guidelines for cancer prevention', developed by 'Korean Ministry of Health and Welfare' and covering smoking cessation, appropriate drinking, condom use, and regular physical activity and so on. We selected the priority needed to promote awareness and segmented the audience group based on the demographic characteristics, homogeneous with respect to the knowledge level using Answer Tree 3.0 with CHAID as a data mining algorithm.
The results of analysis showed that each guideline of ' 10 lifestyle for cancer prevention' had its own segmented subgroup characterized by each demographic. Especially, residence area, city or county, and ages were the first split on the perceived level of knowledge and these findings suggested that segmentation of audiences for targeting is needed to deliver more effective education of patients and community people. In developing the strategy for effective education, the method of social marketing using the decision tree analysis could be a useful and appropriate tool.
The study findings demonstrate the potential value of using more sophisticated strategies of designing and providing health information based on audience segmentation.
本研究旨在根据韩国社区人群的人口统计学特征以及对各项癌症预防指南的知晓程度，对“癌症预防的10种生活方式”的受众群体进行细分。
根据韩国江原道的年龄和性别分布，通过分层随机抽样选取参与者。2008年10月，对2025人进行了6至15次电话调查。共有1687人完成了调查（回复率：83.3%）。调查项目包括年龄、性别、收入、教育程度和居住地区等社会人口统计学特征，以及由“韩国卫生福利部”制定的“癌症预防10项指南”的知晓程度，这些指南涵盖戒烟、适量饮酒、使用避孕套和定期体育锻炼等内容。我们选择了提高知晓度所需的优先事项，并使用以CHAID作为数据挖掘算法的Answer Tree 3.0，根据人口统计学特征对受众群体进行细分，使知识水平保持一致。
分析结果表明，“癌症预防的10种生活方式”中的每项指南都有其各自以不同人口统计学特征为特点的细分亚组。特别是，居住地区、市或县以及年龄是在知识认知水平上的首要划分因素，这些结果表明需要对受众进行细分，以便更有效地为患者和社区人群提供教育。在制定有效教育策略时，使用决策树分析的社会营销方法可能是一种有用且合适的工具。
研究结果证明了基于受众细分设计和提供健康信息的更复杂策略的潜在价值。
