Since the analysis of this project involve using census data, it is always preferable to use the most update version of census data available.  The most recent version of the census data is the DA (Dissemination area) census 2001 data downloaded from the University of Toronto census data website.  Census 2001 was the first time for Statistics Canada to use the Dissemination area to measure the census.  This posts some limitations to my analysis.  Since the previous census data are based on EA (enumeration area) but not Dissemination Area, I cannot determine the population growth rate for each DA polygon.  Population growth rates, if available, are very useful in predicting communities that need a better transit coverage in the future. 

With the exception of skytrain stations, B-line bus, peak hour bus and bus exchange, all the other bus routes are assigned the same value in the MCE analysis.  This cannot fully reflect the real life situation because different bus routes have different frequency services.  A higher frequency routes are obviously more attractive than a low frequency routes.  Thus it is useful if we can gather data based on the bus routes frequencies.

Time lag of the data – because Translink is constantly changing its service area, the bus route data may not be a hundred percent correct. 

Operational Errors

One of the operational problems encountered in this project is that when I want eliminate parks area when doing the analysis, the outlines of some of the parks still show up ( Fig 1.).  Although these outlines did not quite affect the result of the analysis, they are potential sources of errors if I want to perform more analysis.  The possible reason is that the landuse polygon I imported from ArcGIS did not match the municipal (Vancouver in this case) polygon.    This proves  that  quality of data from differenent sources are sometimes variable.