68 x 065 x 227r 245r versus fits response is year

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: NGE OF X IS FROM 2256 to 7018 we cannot predict for universities charging outside that range nalysis of Variance ource egression esidual Error otal DF 1 31 32 SS 143190926 39866422 183057349 MS 143190926 1286014 F 111.34 P 0.000 nusual Observations bs 1 2 7 10 Year 2000 7018 7002 4994 4715 Year 2008 13706 13642 12106 6385 Fit 13010 12983 9584 9112 SE Fit 494 491 235 214 Residual 696 659 2522 -2727 St Resid 0.68 X 0.65 X 2.27R -2.45R Versus Fits (response is Year 2008) 3000 2000 Residual 1000 0 -1000 -2000 -3000 4000 5000 6000 7000 8000 9000 10000 Fitted Value 11000 12000 13000 RESIDUALS: no curves, no wedges MEANING: errors are about the same for any values of X we decide to use R‐SQUARED: redictor onstant ear 2000 Coef 1132.8 1.6924 = 1134.03 SE Coef 701.4 0.1604 R-Sq = 78.2% T 1.61 10.55 P 0.116 0.000 R-Sq(adj) = 77.5% MAXIMUM ERROR : 2*1134.03, about $2,268 THAT IS THE MODEL’S MAX ERROR PREDICT A TUITION VALUE: if in 2000 University XWY, charged $5,000, how much are they likely to charge (average of all those who charged $5,000) in 2008? A REGRESSION MODEL PREDICTS BETTER WHEN THE “X” VALUE IS NEAR THE AVERAGE OF ALL X’s because of that WE WISH WE COULD HAVE A VERY SMALL STANDARD ERROR (there are ways to DECREASE THE ERROR, AN EASY ONE IS ELIMINATE AN OUTLIER). DEC 5TH: ANOVA (CHAPTER 12) ECEMBER 5TH, LAST CLASS: ANOVA. CENARIO: OU WANT TO FIND DIFFERENCES IN DOSAGES (means). THE...
View Full Document

This document was uploaded on 01/28/2014.

Ask a homework question - tutors are online