Weighted Least Square Logistic Regression Modelling For Aquatic Benthos

摘要

This paper illustrated an alternative weighted logistic regression as a technique for modeling of aquatic benthos through SAS algorithm. Weighted logistic regression analysis extends the technique of regression analysis to research situation in which the outcome variables is a categorical. Through the analysis, the aim of this paper is to evaluate the marine aquaculture factors that influencing the population of benthos in Bidong Island, Kuala Terengganu, Terengganu, Malaysia. The results shows that benthos size (OR= 0.099; 95% CI= -3.6703, -0.9479), flattened body factor (OR = 4.93; 95% CI=0.2576, 2.9317), body form factor (OR=0.012;95% CI= -5.9579, - -2.9152), weight (OR = 0.0744;95% CI= -4.0338, -1.1613), and distribution factor (OR = 3.947;95% CI= -0.1942, - 2.9405). All the listed factors were significant at α = 0.05.