In this lecture we tackled the question “Why is Geography important?” and went over some of the major issues in geospatial analysis. As I’ve come to know throughout my years in the geography department the role of geography in any analysis is vital and that many research questions, whether the researchers know it or not, have a strong geographical component. This is why in any study it is important to first eliminate the geographical explanation.
In doing so there are some important concepts to understand to make sure your explanation recognizes biases and limitations. In this lecture we were re-introduced to the Modifiable Areal Unit Problem (MAUP), Simpson’s Paradox, and the importance of defining the scale, grain, and extent of a study.
The MAUP concens what constitutes the object of spatial study and how different areal arrangements of the same data can produce different results. This is easily demonstrated through the example of a demographic study involving census tracts and dissemination areas as was done in a lab in GEOB 370. The lab demonstrated that the results of a study are dependent on the spatial unit used. The results of the lab showed that using different spatial units revealed different patterns across the landscape.
Simpson’s Paradox states that if the value of one variable varies in correlation with another it may not be possible to obtain a true estimate of the real correlation between the two variables. This can be seen in action when looking at crime and income levels of census tracts.