A Spatial Panel Data Modeling for COVID-19 Areal Data in Sarawak

Saira, Aslam and Phang, Piau (2025) A Spatial Panel Data Modeling for COVID-19 Areal Data in Sarawak. Masters thesis, UNIMAS.

	PDF dsva_Saira Aslam.pdf Restricted to Repository staff only Download (332kB)
	PDF MSc Thesis_Saira Aslam.pdf Download (2MB)

Abstract

The world has been facing its biggest infectious disease outbreak, the COVID-19 pandemic, since 2020. As of December 23, 2022, over 650 million confirmed cases and over 6 million deaths have been recorded worldwide. In Malaysia, with a population of 32.78 million, there have been over 5 million cases and 36,000 deaths. One of the essential properties of pandemics is their spatial spread, which depends on the disease transmission mechanism, human mobility, and control strategies. Although the disease was well controlled in Sarawak in 2020, by the end of 2021, Sarawak recorded over 250,000 infections and 1,619 fatalities, making it the second state with the highest cumulative cases among Malaysia's 13 states and three federal territories. This research quantifies the spatial autocorrelation of Sarawak COVID-19 district data across different pandemic waves to understand the spatial spread of COVID-19 in Sarawak. Regression models such as ordinary least square(OLS),spatial lag model (SLM), and spatial error model (SEM) were used to determine the spatial relationship between district-wise COVID-19 cases and various demographic, social, and economic factors. The study found that the percentage of households with garbage collection facilities, population density, and proportion of males are key factors driving COVID-19 incidence rates in Sarawak. Furthermore, this research examined spatial autocorrelation at two levels: within the state of Malaysia and the district of Sarawak. The impact of flight routes on COVID-19 transmission from state to district was analyzed using supra-adjacency matrices to compute local and global Moran's I values. Centrality measures identified pivotal nodes in districts, states, and state-to-district connections, revealing spatial dependence across these levels. Moran's I statistics indicated positive spatial autocorrelation for all levels except for states in wave 1. Significant p-values (< 0.05) for districts, two-level, and state-to-district comparisons highlighted the non-random nature of the spatial patterns, suggesting that neighboring areas share similarities in incidence rates. Kuala Lumpur and Kuching exhibited the highest significance in state-to-district connections, Pahang at the state level, and Sibu and Sri Aman at the district level. The main contribution of this research lies in its detailed analysis of spatial patterns and factors influencing COVID-19 spread in Sarawak. The findings can assist the Government of Malaysia in devising targeted interventions and policies to control future outbreaks by understanding the spatial dynamics of disease transmission. Keywords: COVID-19 pandemic, Sarawak, spatial autocorrelation, spatial regression model, spatial lag, spatial error model

Actions (For repository members only: login required)

View Item