This section provides a brief review of my spatial analysis. In addition, I will go through each of my sections of this project to identify any shortcomings and assess whether or not changes could have been made to alleviate any of these problems. With my spatial analysis I was able to accurately answer most of the questions I had initialy set out to answer, as well as infer a few extra points about parking near Skytrain Stations in particular neighbourhoods.

How has the occurance of petty auto-crimes changed in the six years from 2006-2011? There has been a steady decline in the occurance of petty auto-crimes in the city of Vancouver in this time frame. In regards to the location of these crimes, Skytrain Stations beyond the downtown core of the city have shown the highest decline. Downtown, which has the highest concentration of Skytrain Stations, shows little redution in petty auto-crime rates.

What is the relationship, if any, to auto-crimes within a set distance of Skytrain Stations? Because of the saturation of petty auto-crime in Vancouver it is difficult to assess whether or not Skytrain Stations actually effect these crimes. By looking at my data, particularily the SHP files in ArcMAP, petty auto-crimes appear to be concentrated both Downtown and down major corridors such as Kingsway Boulevard, Main Street, Fraser Street, and Hastings Street, as well as the neighbourhoods of Kitsilano, Kerrisdale, and South Marpole. None of which are serviced by Skytrain. Although South Marpole now has Marine Drive Station within a few Kilometers, which was outside my selected factor range of 500m. From this I have concluded that Skytrain Stations alone do not account for an increased concentration of petty auto-crimes, thus, there must be other factors effecting the crimes in these areas.

Has population change effected auto-crime in any way? As stated in my report, because population growth is happening across the city, it was difficult to assess whether increase or decrease in population had any effect on petty auto-crimes. These crimes are happening across the city, regardless of a change in population.

The final point of my spatial analysis is that it shows the relationship of petty auto-crimes near Skytrain Stations in relation to other Stations. This makes for an accurate assessment of what stations are the safest to park near as well as what season you would be more likely to have your car broken in to at the station you have parked.

---

Analyzing Errors

I feel that my data collection method was accurate, although slightly incomplete. Having more data on particular neighbourhoods, such as income rates, may have helped me make better assessments on vulnerable neighbourhoods as it appears that both the wealthy and poor neighbourhoods are effected more than those of middle wealth standing.

I spent most of my time on the methodology for this project, as it took me several weeks and several failed attempts to get my data to a point where I could manipulate it with IDRISI. First and foremost, starting my project sooner would have alleviated much of the stress that came with my data manipulation. In addition, it may have allowed me to complete my spatial analysis sooner, giving me more time to collect additional data from which I may have been able to run a more accurate spatial analysis. The biggest problem I found was converting my data from SHP to RST, continued errors while importing my dataset to IDRISI left me frustrated and helpless. Persistence and internet research eventually led me to the point I reached with my methodology, I learnt more about GIS software in the weeks I spent working on this project, than the entire semester of lab work.

The report I came up with from my spatial analysis was an accurate assessment on comparing Skytrain Stations to each other. From my data collection and manipulation, the report basically spoke for itself. Once I had a graph from which to think about how petty auto-crime had changed in the years from 2006 to 2011, I felt confident with the maps I created through IDRISI's macro modeler, with my constrains and factors in place. Of course, with the collection of more data and the subsequent creation of more constraints and factors, I feel as though I could have narrowed down my report to a higher level of both confidence and accuracy.