Within this week, the overall impact geography can have on data was touched upon. Geography is crucial because the geography of a problem must be considered to produce an effective analysis and many geographical elements of a study such as scale, boundaries and spatial dependence can impact results produced. In regards to landscape ecology, there isn’t a “natural” scale at which ecological studies should be undertaken. While most ecological phenomena have a spatial temporal domain, the definition most appropriate scale must be considered carefully. The Modifiable Areal Unit problem (MAUP) was again reviewed, which describes how scale or boundaries impact the data being collected.
MAUP must be considered in all GIS analyses to evaluate potential sources of error and uncertainty. One concept which highlights the impact geography can have on an analysis or social data collection is Gerrymandering, which refers to the practice of manipulating electoral district boundaries in order to create zones where a particular party would have a political advantage. The Simpson’s paradox was also explained as a confounding variable that is not taken into consideration in the analysis as some variables vary in correlation with another (e.g. high unemployment and other socio-economic characteristics) making it difficult to obtain a reliable estimate of true correlation between two variables.