Author(s)

Shuyi Chen, Jiaoyang Li, Zetian Zhang

Corresponding Author

Shuyi Chen

ABSTRACT

Cervical Cancer is the leading medical cause of death in women globally. It cannot be cured but is preventable. When use of cancer prevention strategies remains low in developing countries like Indonesia where the poverty rate is high. Our study that focuses on exploring the association between social risk behavioral factors with cervical cancer may help women to identify their cancer risk when medical access is limited. There were 72 respondents with 22 of them having cervical cancer responded to the questionnaire with nine questions for different variables. All assumptions of Health Belief Model, Protection Motivation Theory, and Theory of Planned behavior were considered to measure the nine categorical factors. To avoid collinearity, the data was processed by Principal Component Analysis (PCA) to identify factors that are associated with cervical cancer, the first three components are selected. A confusion matrix is built upon the PCA model to examine accuracy. Variables from PC1 with the highest contributions and qualities of representation are considered most significant. The three categorical risk factors are social support, motivation, and empowerment. The logistic equation of cancer risk prediction based on the social determinants has shown an accuracy of 94% from confusion matrix. By identifying most associated social risk factors, it can be used for various purposes including education and further research studies. Also, the time and cost-efficient prediction model allows more women to detect their state of risk early before the disease begins to deteriorate. CCS CONCEPTS• Mathematics of computing ~ Mathematical software ~ Statistical software • Mathematics of Computing ~ Probability and statistics ~ Probability inference problems ~ Maximum Likelihood estimation.

KEYWORDS

Biostatistics, R studio, Principal Component Analysis, Logistic Regression, Confusion Matrix, Cervical Cancer, Social Theory