To further this research, I would have liked to include some more detailed constraints based on where high schools were located, and perhaps income based by zip codes.  A prime location site, outside of the existing malls, could be revealed by simply adding zip code information to my analysis, however, this proved to be too big a scope for my project.  For example, to include every GVRD zip code, I would have to first decipher which codes belonged with each municipality, then I would have to heads-up digitize all of this information onto a based that was large enough to handle such fine detail (or just create that, too), and finally, I would have to go back and add all of the attribute information that I wanted to analysis.  You can see why I opted to forget it.

 
    As far as my Boolean analysis of the GVRD, I ran into a few minor problems.  For instance, when I began the process of running Edit/Assign on my raster data, Idrisis had a problem with creating the legend because my class values exceeded the integer range (ex. Household Income = $48791.00).  Since, I could not use real numbers to run assign, I initially thought that I would have to run the analysis and fix the colouring problem using different software.  I did find an Idrisi solution, though.  By changing the palette from qual256 to Idrisi256, the new image delivered a range of colours to encompass my values, and present the information as a continuous variable.

    The Burnaby MCE, had a few minor problems as well.  Initially, generating an overlay of the roads within the Burnaby municipality produced all roads from the GVRD.  This caused a problem until I realized that I could run another overlay to highlight only those roads which were within the municipality.  From this new layer, I simply reclassed the unwanted roads out.  A bigger problem came with running the analysis only on Burnaby.  As you already know this problem was solved by using the Window operation to 'cut' out the desired area.  However, as this solution came about after a weekend of my own trouble shouting, I felt it was worth mentioning.  You see, since all of Burnaby malls are situated so close to one another, Idrisi would simply return a black image for this area, regardless of which distance analysis was run.  Thankfully, on the following Monday, Andreas gave me the simple solution.  Finally, I had expectations of including things like bus routes and other vector data into my analysis.  However, the problem arose that the files that I wanted to rasterize did not have an associated raster file from which to take the parameters.  Therefore, I could not include bus routes, skytrain, and DEM of the area into the analysis.  However, just today I learned that you can compute your own base image and set the parameters that would coincide with the vector files.  I think one thing that can be learned from this project is that there is still much to learn.