Methods: Map and Table Design

Map and Table Design

To convey information about each county or town in a region, we employed gradient colored geographic maps (more precisely, bipolar color choropleth maps). Every county or town in the region being analyzed was colored according to the indicator being analyzed in that section of the report. With just two exceptions noted below, the measures were normalized as follows. Where the indicator value was “better” (e.g., lower poverty) than the mean of the Park Towns, the indicator was positive. Where the indicator was worse, it was negative. The positive and negative indicator values were linearly transformed to the intervals 0 to +1 and 0 to -1 so that the “best” county or town in the region was +1 and the “worst” was -1. If the indicator was equal to the weighted mean value of the Park Towns, it was transformed to 0. After this transformation, the numerical measures were transformed to color values. The color values were computed in the RGB color model. A numerical value of 0 was transformed to white (maximum values for red, green, and blue). As numerical values increased linearly from there, red and blue were decreased linearly. A numerical value of +1 was transformed to pure green. Similarly, as numerical values decreased from 0, the blue and green were linearly decreased. A numerical value of -1 was transformed to pure red. Counties or towns outside the region of analysis were colored neutral gray. Counties or towns overlapping the Park were colored a darker gray. The two exceptions to this method were as follows. While the zero point for the numerical transformations was generally the weighted mean Park Towns value, that was not the appropriate zero point for the two New York State analyses that included analyses of areas inside the Park. For the map of New York State school district enrollment trends, the analysis included and shaded districts inside the Park. Therefore, we chose the zero point to be exactly zero percent enrollment change. For the map of New York State real property assessed value trends, towns inside the Park were again included and shaded. Therefore, we chose the zero point to be the median increase in assessed value among all New York towns.

The several tables of regional population changes among various age groups employed the same coloring used in the maps. The zero point (white) was always zero percent recruitment of population. The comparison was always with the immediately preceding decennial census, except for the final total change column, where the reference point was the first column. The linear transformation to +1 and to -1 was performed so that the colors would match across the several pages of age group tables. An increase of +19% was mapped to +1, a decrease of -41% was mapped to -1, and no change was mapped to 0. The +19% and -41% changes were chosen to accommodate the largest changes observed across the several tables. Thus, there was no clipping or truncation of color in any of the several tables and each particular color represented exactly the same percent change population across every table in the population trends analysis. ❑