Tool Scoring Methodology
A primary objective of the Transportation and Health Tool (THT) is to help users understand the connection between transportation and health. Many of the indicators used in the tool are technical. Users who are not transportation and health specialists might wonder if the indicator values for their state or urban area are “good” or “bad.” For some indicators, higher values are better. For other indicators, lower values are better. The range of results varies widely between indicators.
To make results easier to interpret, the website provides scores for states, metropolitan statistical areas, and urbanized areas on a scale of 0 to 100, where higher values are better. The score for a given state or area represents its percentile value. This is the percent of states or regions that score below it.
These scores were created by first assigning standardized scores (“z-scores”) to each state or region. Z-scores measure the difference between the value for a given state or region and the national average in terms of the standard deviation. Standard deviation shows the amount of variation in a dataset. Z-scores work best when data are normally distributed, so the THT team made adjustments to rein in extreme values (“outliers”).
For example, the public transit commute mode share in the New York–Newark–Jersey City metropolitan area is twice that of the metropolitan area with next-highest value, so it is an outlier.
We calculated z-scores for each state and region by dividing the difference between the state or region value and the national average by the standard deviation.
We adjusted the results so that all final z-scores show values as “higher is better.”
We then translated the z-scores into a percentile value to get the final scores for each state or region.