Multiple Linear Regression Models Applied to Urban Terrain Assessment - Municipality of Itaqui - RS Case

This work had as objective to identify explanatory variables for modeling and prediction of actual value (ITBI) and venal value of urban plot in the city of Itaqui-RS using the comparative method of Direct market data, developed by multiple linear regression. In a sample of 67 lots, quantitative variables were evaluated: area, front of urban plot, value of m2, market value and the value ITBI and qualitative variables: topography, location and flooring, and for each variable were determined statistical parameters including the data distribution for asymmetry, kurtosis, normality test of Kolmogorov-Smirnov, as well as the Pearson linear correlation and path analysis, which shows the relationships of cause and effect of the explanatory variables on the market and ITBI values, respectively. The importance of each explanatory variable on the model of market and ITBI values was determined by the stepwise method. The market value showed a positive, significant correlation (p≤0.05) and increased with flooring, front of urban plot, area and value of m2, in that order. Already ITBI value was positively correlated, significant (p≤0.05) and with increasing value of m2and flooring, in that order. Path analysis indicated that the area and the value of m2 have cause and effect on the market value and, the value of square meters and flooring shows cause and effect on the value of ITBI. The models: Y = 16,90 x área + 594,69 x value of m 2 - 9526,54 x topography (r2=0,91) and, Y = 868,36 x value of m 2 - 23642,19 x topography + 33633,46 x flooring (r2=0.91) and, (r2=0.70), are suitable for the estimation of market and ITBI values, respectively.

本研究旨在借助多元线性回归构建的直接市场数据比较法，识别伊塔基-里约格兰德州（Itaqui-RS）城市地块的实际价值（ITBI）与交易价值的建模及预测解释变量。研究选取67宗地块作为样本，对两类变量展开评估：其一为定量变量，包括地块面积、城市地块临街面、每平方米价值、市场价值及ITBI价值；其二为定性变量，涵盖地形、区位与地面铺装情况。针对各变量，本研究测定了多项统计参数，包括数据分布的偏度、峰度、柯尔莫哥洛夫-斯米尔诺夫（Kolmogorov-Smirnov）正态性检验结果，同时计算了皮尔逊线性相关系数与通径分析结果——后者可分别揭示解释变量对市场价值与ITBI价值的因果关联。通过逐步回归法确定各解释变量在市场价值与ITBI价值模型中的重要性。研究结果显示，市场价值与地面铺装、地块临街面、面积及每平方米价值呈显著正相关（p≤0.05），且相关性依次增强；ITBI价值则与每平方米价值及地面铺装呈显著正相关（p≤0.05），且相关性依次提升。通径分析结果表明，地块面积与每平方米价值对市场价值存在因果影响，而每平方米价值与地面铺装则对ITBI价值存在因果影响。适用于市场价值与ITBI价值估算的模型分别为：Y = 16.90×面积 + 594.69×每平方米价值 - 9526.54×地形（决定系数r²=0.91），以及Y = 868.36×每平方米价值 - 23642.19×地形 + 33633.46×地面铺装（r²=0.91），另有决定系数r²=0.70的模型，分别适配两类价值的估算需求。