Dear stata users, I would like to examine the influence of different determinants on the price of real estate. For example, attributes such as traffic volume, lake view, size of the property, etc. Now I wonder if I could do the analysis for "high-priced" properties by creating a subgroup where only those properties with the highest 20% price are included. Given that the price of these "luxury" properties could probably be driven more by the location and less by the view. Can I form subgroups based on the dependent variable, or does this lead to an econometric disaster?