Modeling of Some Common Cancers in Rivers State of Nigeria, by Applying Markov Switching Intercept Vector Autoregressive (MSI-VAR) Model
Nnoka Love Cherukei *
Department of Statistics, Captain Elechi Amadi Polytechnic, Rumuola, P.M.B. 5936, Port Harcourt, Nigeria.
Ettuk, Ette Harrison
Department of Mathematics, Rivers State University, Nkpolu Oroworukwo, P.M.B. 5080, Port Harcourt, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
The study applied a non linear modeling technique; Markov Switching Vector Autoregressive (MS-VAR) model in modeling common cancers data in Rivers State of Nigeria. The data for the study was obtained from the Cancer Registry Unit, University of Port Harcourt Teaching Hospital. The study spanned from January, 2009 to December, 2020 consisting of 144 monthly observations. The objectives of the study may include ; to model and estimate the linear interdependence between the identified common cancers, to determine the direction of causality and the significance of causality among the study variables and lastly to determine the probability of transitioning from one state to the other and the duration of stay in a particular regime. The study ascertained the stationary conditions of the three variables. Breast cancer and Cervical cancers were stationary at levels while Prostate cancer was stationary after first difference.
The Augmented Dickey Fuller Unit Root Test was used to test for stationary behavior of the variables while Philip Peron Confirmatory test was applied. The study also tested for long-run relationship among the study variables using Johansen co-integration test which however indicated no long-run relationship according to Maximum- Eigen value and Trace test results. The stability of the model was tested by evaluating the inverse roots of the characteristic polynomial, since all the roots lied inside the unit circle, it was concluded that stability condition was satisfied.
Keywords: Common cancers, Markov switching intercept vector, metastasis