Multiple Linear Regression Analysis Applied for Modeling in Torrens Community Disaster Resilience Score Card for Flood-Prone Schools in the Schools Division of Misamis Oriental, Northern Mindanao – the Philippines

Abstract

Accurate assessment of the level and extent of disaster resilience of flood-prone schools and their communities is a complicated task. This is given the various means to measure disaster resilience. This is further complicated by the fact that disaster resilience is context-specific. Therefore, applying one of these measures for disaster resilience modeling proves to be a practical approach, as this model has the advantage of having context-specific viability in predicting disaster resilience. One of the statistical methods that can be used for testing a model in disaster resilience is Multiple Linear Regression (MLR). Therefore, this work aims to test a model for disaster resilience. For this, 26 schools in the Schools Division of Misamis Oriental were identified as flood-prone areas based on the geo-hazard maps of the Department of Environment and Natural Resources. With the use of the Torrens Community Disaster Resilience Score Card, the schools and their communities were rated by their respective School Administrator/Principal; School Disaster Risk Reduction Management Coordinator; Municipal Disaster Risk Reduction Management Officer; President of the General Parents-Teachers Association; President of the Student School Government; and the Barangay Chairperson. The ratings served as the data. Multiple Linear Regression was applied to generate a new model with four explanatory variables, i.e. community-connectedness, risk and vulnerability, planning and procedures, and available resources; and the response variable is Disaster Resilience. The MLR model obtained an Adjusted R Square of 0.965, which means that the four explanatory variables explained 96.5% of the variability of Disaster Resilience. The Goodness of Fit of the model had shown a right-tailed, F=175.9, p-value=7.77156e-16, thus, rejecting the null hypothesis that the model is not a good fit. All four variables are significant predictors of Disaster Resilience, hence, attesting to the adequacy of the model. Thus indicative that in the context of these flood-prone school and their communities, the model justifies that the application of the said scorecard is of great importance to these schools and their communities as well as those people responsible for disaster risk and reduction management regular monitoring and assessment for disaster resilience.