* AED13.do March 2015 for Stata version 12 capture log close // capture means program continues even if no log file open log using AED13.txt, text replace ********** OVERVIEW OF AED13.do ********** * STATA Program * copyright C 2015 by A. Colin Cameron * Used for "Analyis of Economics Data: An Introduction to Econometrics" * by A. Colin Cameron (2015) W.W. Norton * To run you need file * AED_HOUSE.DTA Same as chapter 8 * in your directory ********** SETUP ********** set more off version 12 clear all set mem 10m * set linesize 82 set scheme s1manual /* Graphics scheme */ ************ * This STATA does analysis for Chapter 13 * 13.1 EXAMPLE: HOUSE PRICE AND CHARACTERISTICS * 13.2 TWO-WAY SCATTERPLOTS * 13.3 CORRELATION * 13.4 REGRESSION LINE * 13.5 ESTIMATED PARTIAL EFFECTS * 13.6 MODEL FIT * 13.7 COMPUTER OUTPUT ********** DATA DESCRIPTION * House sale price for 29 houses in Central Davis in 1999 * 29 observations on 9 variables **** 13.1 EXAMPLE: HOUSE PRICE AND CHARACTERISTICS use AED_HOUSE.DTA describe * Table 13.1 summarize price size bedrooms bathroom lotsize age monthsold * Table 13.2 list price size bedrooms bathroom lotsize age monthsold, clean * Regression with and without control regress price bedrooms regress price bedrooms size **** 13.2 TWO-WAY SCATTERPLOTS * Figure 13.1 graph matrix price size bedrooms age, scale(1.2) graph export AED13FIG1.wmf, replace **** 13.3 CORRELATION * Table 13.3 correlate price size bedrooms bathroom lotsize age monthsold pwcorr price size bedrooms bathroom lotsize age monthsold, sig star(.05) set textsize 150 **** 13.4 REGRESSION LINE use AED_HOUSE.DTA * Multivariate regression regress price size bedrooms bathroom lotsize age monthsold * Demonstrate that can get from bivariate regression on a residual * First regress size on all other regressors to get the residual regress size bedrooms bathroom lotsize age monthsold predict res_size, resid * regress price on the residual regress price res_size **** 13.5 ESTIMATED PARTIAL EFFECTS * Partial Effects regress size bedrooms scalar dsdb = _b[bedrooms] regress price bedrooms size scalar totaleffect = _b[bedrooms] + _b[size]*dsdb di "total effect " totaleffect regress price bedrooms * Same adding a regressor regress size bedrooms scalar dsdb = _b[bedrooms] regress age bedrooms scalar dadb = _b[bedrooms] regress price bedrooms size age scalar totaleffect = _b[bedrooms] + _b[size]*dsdb + _b[age]*dadb di "total effect " totaleffect regress price bedrooms ***** 13.6 MODEL FIT regress price size bedrooms bathroom lotsize age monthsold * R-squared is squared correlation between yhat and y regress price size bedrooms bathroom lotsize age monthsold di e(r2) predict pprice correlate price pprice di r(rho)^2 * Compute adjusted R-squared di "Adjusted R-squared = " e(r2) - (e(df_m)/e(df_r))*(1-e(r2)) * Information criteria estat ic * Information criteria manually di "AIC = " e(N)*ln(e(rss)/e(N)) + e(N)*(1+ln(2*_pi)) + 2*e(rank) di "BIC = " e(N)*ln(e(rss)/e(N)) + e(N)*(1+ln(2*_pi)) + e(rank)*ln(e(N)) ***** 13.7 COMPUTER OUTPUT * Table 13.4 regress price size bedrooms bathroom lotsize age monthsold ********** CLOSE OUTPUT log close