“…Many approaches have been used and developed to identify and examine the effects of driving factors on urban expansion, including bivariate regression (BR) (Haregeweyn, Fikadu, Tsunekawa, Tsubo, & Meshesha, 2012;Wu & Zhang, 2012), multiple linear regression (MLR) (Dewan & Yamaguchi, 2009;Müller, Steinmeier, & Küchler, 2010;Seto et al, 2011), analytic hierarchy process (AHP) (Thapa & Murayama, 2010), adaptive Monte Carlo (aMC) (Chen et al, 2002), redundancy analysis (RDA) (Hietel, Waldhardt, & Otte, 2007), canonical correspondence analysis (CCA) (Fu et al, 2006), and logistic regression (Dendoncker, Rounsevell, & Bogaert, 2007;Dubovyk et al, 2011;Long et al, 2012;Reilly, O'Mara, & Seto, 2009). Of these methods, the most widely used is logistic regression, which has the following advantages: 1) it is an effective method to handle binary dependent variables, which is the case in LULC change (change or no change); 2) there is no assumption of normality or a linear relationship between the dependent and independent variables (Cheng & Masser, 2003); 3) the results of logistic regression can be directly used to predict the locations of future urban expansions (Dubovyk et al, 2011;Hu & Lo, 2007).…”