Abstract:
Multiple linear regression were used in environmental studies to investigate the relationship among factors. In this study, flood cases in Kuamut, Sandakan and three factors were taking into account which are rainfall (X1), temperature (X2) and humidity(X3) in the area. The possibilities of interaction between factors that may contribute to the flood cases were also being emphasised. There were 12 possible models have been identified, however only 2 models were selected as significant models using the global test. The Eight selection criteria for best model selection and normal test for residuals confirmed that the best model to explain the flood cases is
Y = 183.016 – 11.910(X2) + 0.003(X1X2) + ε
Where:
Y = Flood
X2 = temperature
X1X2 = interaction between rainfall and temperature
This showed that the flood cases in Kuamut were not just affected by the temperature but also the interaction between the amount of rainfall and temperature were more important rather than the rainfall or humidity itself
