Hello, I am a real estate appraiser and use the Multiple Variable Regression Analysis and Firecasting template. I have a question regarding subjective independent variables. For instance, # of bathrooms and # of bedrooms is cut and dry for the analysis. However, quality and condition is not. How would you recommend I employ a numeric value to the quality and condition of each independent variable used in the analysis? I've noticed when just 1 or 2 is used as an independent variable indicator, the results are often quite high. But ranges of 100-200 seem to adjust more proportionately. I hope I'm explaining and asking this in an understandable manner..?? Thank you to all who can help me!
By adding more granularity to the factor level for qualitative variables we can get more accurate results. This depends, of course, on the number of levels available in the data but more is preferable.
If we have logistic data e.g. en-suite bathroom yes or no, then we can only have 2 levels: 1 and 0 but if we can gauge and create multiple levels of quality then we have a better way to predict the dependent variable when these values are specified for prediction.