It was required that we create two map layouts, the second of which I will provide below, one which showed the percent of the population in each tract that was over the age of 65, and the other which conveyed the quantity of individuals in each tract over the age of 65. With the latter, we were required to normalize the data, meaning that the amount of individuals had to be divided by the area of the tract to communicate the density of applicable individuals in the tract.
In short, the four methods of classification we examined were natural breaks, equal interval, quantile, and standard deviation. The natural breaks method of classification is where our ArcGis Pro software uses an algorithm to lump like values together based on witnessed discrepancies in our clusters of data. The equal interval method takes the data value range (numerically, the lowest to the highest data value) and divides them by the number of classes we wish to represent. This results in an even stratification which visually represents all values that fall within the calculated data range. Quantile classification methods take the entirety of our observations (each observed data value, for example 521 recorded data observations as utilized in this set) and divides them by the desired number of classes. This provides even segments of observations and our values are then compiled by their grouping within these classes. Lastly, standard deviation takes the average value of our data and subsequently illustrates the variation in data value per each standard deviation away from the average both negatively and positively.
Each classification method has its own pros and cons contingent on the intent the cartographer is attempting to communicate. As we were tasked with illustrating the population distribution of senior citizens, we had to chose between utilizing the raw population data (number of individuals in the age bracket per each tract) or the percent of the population over 65 in each tract.
If I were in a setting where I had to communicate to an audience the population density of senior citizens in a given area, I would unequivocally select the usage of normalized total population. When categorized under any of these methodologies, it is clear that is easier to discern how many seniors make up an area when you are utilizing their true numbers per area comparatively with other regions' density. This shows value relativity to other regions. If I had to communicate which area has the highest number of seniors relative to its overall population, the natural choice would be utilizing percent population. Though, I must admit, I am not entirely keen with the communicative efficacy of either of these cartographical strategies. The reason for this is a glaring lack of specificity.
In the map above, in the areas that are shaded dark red, meaning that there is the largest numbers of senior citizens, the values extend from roughly 2000 to 13,000. In attempting to collect precise data, this is far too large of a swath. Translated to percent value, as seen below, our swaths in dark red polygons range from roughly 20 percent to 70 percent. If I were targeting seniors for communicative or marketing reasons, for instance, allocation a budget towards a region with a 70 percent per population value would be a cogent justification of funding. If it were only 20 percent, a firm may want to redirect their fiscal efforts. Therefore, map design is contingent on usage and this population swath does not define a clear percentile. This is where refinement through these different methods could help.
In the aforementioned marketing scenario, and for most purposes, in the percent population map layout, the equal interval map communicates the greatest discernable accuracy. If I had to communicate which areas simply had the most seniors numerical compared to one another, the best choice for a classification method would be the natural breaks methodology.
Comments
Post a Comment