Statistics Final report

Pepperdine University

Final Report

Introduction

The hedonic price approach is among a few modern techniques used by real estate businesses to appraise houses. It is a quantitative method that uses a regression model to estimate real estate prices (Yeh & Tzu-Kuang, 2018). The case business is a real estate company looking for a better way to price its properties. The firm currently uses the comparative approach, which entails assigning similar real estate the same benefits and prices. However, this method has its limitations, with the main one being human bias while appraising real estate (Bailey, 2020). The purpose of this research is to define a regression model that helps make future-oriented predictions with regards to house prices. Answers are sought in this analysis to questions such as: are there any statistically significant relations between DV and IVs, and is it possible to estimate the price of real estate with the regression model?

Hypothesis

The regression analysis uses t-tests to conduct hypothesis tests on the coefficients of each IV. “Near zero p-values imply that there is strong evidence against the null hypothesis," and we can conclude that a given coefficient is statistically significant to the model (Rose, Spinks, & Canhoto, 2015). The hypothesis statements for the two-sided t-test are expressed as:

, the coefficient of the ith IV is equal to zero.

, the coefficient of the ith IV is not equal to zero.

The F-test helps determine the regression's overall significance relative to an intercept only model (i.e., a model without IVs). Similar to a t-test, a near-zero p-value for the test means that we can accept the alternative over the null hypothesis (Rose, Spinks, & Canhoto, 2015). The statements for the hypothesis tests of the F-test are expressed as:

The fit of the regression is equal to that of a model without IVs.

The fit of the regression is more than that of a model without IVs.

Model

The data used to build the model has a total of seven continuous variables. Given the nature of the data, we choose a multivariate linear regression model with more than one IV, Transaction date (X1), House age (X2), Distance of the nearest MRT station (X3), Number of convenience stores (X4), Latitude (X5), Longitude (X6). The sample used has a total of 414 observations and seven attributes. Table 1 below shows the variables include in the analysis. Based on the direction of correlation (see Table 2) for each IV to the DV, we expect the equation of the model to be. In cases where the coefficient is positive, an increase in the IV suggests that the DV's mean tends to increase. For example, an increase in the transaction date by a day will increase the house price. However, if an IV has a negative coefficient, then an increase in it tends to decrease the DV's mean (Uyanık & Güler, 2013). For instance, an increase in the house's age by one year will result in a decrease in house prices.

Conclusion

The purpose of this study was to build a multivariate linear regression model for a real estate business that uses the hedonic price approach to appraise its properties. The primary hypothesis tests test the model's overall significance in predicting house prices and whether or not each of the seven IVs has a statistically significant relationship with the DV. We expect the equation of the model to be. Positive coefficients indicate that an increase in the IV value tends to increase the mean of the DV. A negative coefficient suggests an increase in the IV value causes a decrease in the DV (Uyanık & Güler, 2013). The hedonic price approach's main benefit is that it does not have the human bias that is common with conventional appraisal methods, which makes the house price more reliable (Yeh & Tzu-Kuang, 2018; Bailey, 2020).

References Bailey, J. (2020). Can Machine Learning Predict the Price of Art at Auction? Harvard Data Science Review. Rose, S., Spinks, N., & Canhoto, A. I. (2015). Management Research: Applying the Principles. Routledge. Uyanık, G. K., & Güler, G. (2013). A study on multiple linear regression analysis. Procedia-Social and Behavioral Sciences, 234-240. Yeh, I.-C., & Tzu-Kuang, H. (2018). Building real estate valuation models with comparative approach through case-based reasoning. Applied Soft Computing, 260-271.

Tables

Table 1

Summary of Variables

Table 2

Correlation of IVs to DV