Since the three data layers that I started with were  good quality, I did not encounter any problems with them or any of the many layers I derived from them.  For the most part, I had few problems with the spatial analyses portion of the project, because I had a fairly clear conceptual idea of what I wanted to do from the beginning of the project (after the many hours spent searching data sites for ideas to start with...).  Good organization was certainly critical to managing the more than 180 files I created in IDRISI - especially the crosstab results.  Most of my difficulties were associated with creating and manipulating all the files needed to produce the website.

Throughout this project, there have been countless instances where I have learned how to perform various tasks much more efficiently than the way in which I was originally doing them.  Inevitably this type of learning involves performing a task the difficult way for a while and then realizing that there was either a much easier way to accomplish the same end or, more disappointingly, that whatever it is that you were doing actually serves no purpose.  Such an example can be found by comparing the results tables for scenarios 3 and 4.  In Scenario 3's table, all three category description factors are written out in a single cell.  This is a result of the fact that I had painstakingly entered all 87 categories into the legend in IDRISI before running the crosstab calculations.  After doing this I realized that it had been a pointless endeavour because there was no way that I would ever display a legend with 87 categories.  In Scenario 4's table, by dividing the crosstab results into two columns and entering soil, landuse and slope descriptions in three separate columns, I was able to achieve the same end (assigning the correct description to each crosstab class so they could be matched with runoff coefficients) by a much, much easier method - sorting the table by crosstab results and pasting in the names.