Your Cart

Your cart is empty

Home  >  Volume 38

30. Multivariate Time Series Modelling of Malaria Incidence in Gombe, Nigeria by W.B. Yahya, M.B. Mohammed2 and M.K. Garba Volume 38, pp 207 – 218
Sale price: $5.00


In this study, multivariate time series techniques were employed to model the reported cases of malaria among the residents in Gombe Local Government area of Gombe State, Nigeria. These data were extracts from the records of the Specialist Hospital in Gombe State on the reported cases of malaria from 2009 to 2013. For convenience, the entire residents in this locality were grouped into four comprising the male (adult), the female (adult), the pregnant women and the paediatrics groups. Multivariate time series model selection was performed with special consideration given to all the possible models’ parameters. The results showed evidence of cointegration among the observed series on the four groups of residents in the study area. As a result, a Vector Error Correction (VEC) model of lag two with one co-integration vector was found to best fits the data. As part of checks for models’ adequacy, the Portmanteau test for serial correlation showed no serial correlation among the residuals (p = 0.9814) while the Jaque-Bera test showed that the residuals are multivariate normally distributed (p = 0.2232). Further results showed that reported cases of malaria in Gombe State is generally erratic (non-stationary) over the period with male (adult) having the highest number of malaria cases in 2009 while the paediatrics (children) had the highest number of malaria cases in 2010, 2011, 2012 and 2013.  Consequently, the proportion of deaths due to malaria disease was the highest among the children throughout the study periods as compared to other groups. Results from this work would provide useful guides to the stakeholders at providing necessary measures at arresting the high mortality rate due to malaria among the most vulnerable group (paediatrics) in Gombe State of Nigeria.

Keywords: Multivariate time series,Malaria cases, Cointegration, Vector error correction(VEC) model