There was no curvature in this graph. .
Evaluation: Asking Price and average room size .
Outlier .
.
In this graph we could not see a curvature either, but we could detect an outlier, which was house 16 again. Nevertheless, we decided to keep house 16 in our sample and consider it in the development process of forecasting a model for the Asking Price since appearing as an outlier only two times did not justify omitting this sample.
We observed all two-dimensional scatter plots of the variables and could not find curvature in the relationships; we neither had to consider transforming the variables nor did we have to eliminate outliers. .
Amy still had not really identified what I was trying to do. I explained to her that we still were at the beginning of the process and the next thing to do was the regression of the data. .
Regression of Data.
Ozan: "Amy, to form a model that will help price our house, a linear multiple-regression of the data needs to be completed. The variables you chose GAR (garage size), SQFT (square footage), TONMKT (time on market), RMS (number of rooms), FP (number of fireplaces), BATH (number of baths), HEAT (heating type), CITY (which city) and SQFEET (average room size), will be used in the regression. But do not worry Excel will do the computing for us."".
.
The Summary Output for our data looked as follows: .
.
Colors were added by the author for further explanation. This is not a result of the procedure explained above.
It goes without saying that Amy asked what all those figures meant. For her understanding I pinpointed a few very important figures displayed on this summary output. .
R Square (R2): .
"Amy, can you imagine that the smaller the variability of the residual values around the regression line relative to the overall variability, the better is our prediction? Sorry, I know this is a sentence I barely comprehend myself. I will try it a little differently.
For example, if there is no relationship between the x above (Asking Price) and y (e.