In our analysis, we have identified various assumptions and limitations associated with our project. Prior to our multi-criteria evaluation, the rescaling and transformation of data introduces a degree of subjectivity. The choice of transformation functions varied based on our opinion of applicability for each data set leaving some ambiguity in determining whether to favour high or low values or with the selection of a particular function due to the differences in how transformation functions conduct their calculations using the specific input parameters that each consider (i.e. MSSmall vs Small or MSLarge vs Large). By conducting a weighted multi-criteria evaluation, the decisions on the specific weight allocation does in fact play a significant role as they may be considered subjective. While there is a significant 31% overlap of areas identified between our equal-weight model and AHP weighted model, adjustments to particular weights result in a variable preference for differing areas. 

Aside from the subjectivity in weighting, not all available socioeconomic or environmental criteria are considered. Rather, it utilises a range of influential criteria to provide a relatively representative picture for suitability in regards to the LA county population and region. Our comparison analysis with Metro’s expansion was also specific to their morning peak hours while most of their planned expansion spanned a range of times.This is the result of a temporal limitation for our project timeline, as well as, a limitation in general data modelling as perfect or near perfect representations of real world environments are near impossible. The applicability of the identified socioeconomic, and environmental criterion may also be limited to the LA county region. With differing populations, regional development and local environmental conditions, the indicator criteria identified in our analysis may be specific to the analysed region and population and differ from that of other populations or locales. With the analysis of specific population characteristics, overlaps may occur as populations with correlated characteristics would skew and negatively affect the representation of suitability as we did not test for correlations between identified criteria. Assumptions were also made for the existing rail systems and bus stops. With rail buffers, 2 kilometer distances from stations were designated as optimal regions for development and regions below 1 kilometer were not preferentially favoured due to the rail’s service region. With bus stops, a 800 meter or approximately half mile buffer was designated as a sensible walking distance and was used to identify bus service regions for neighbourhoods.