Here's a question from Gary Shannon (City Traffic Engineer, Rochester, MN) and my response regarding typical use of the Institute of Transportation Engineers' Trip Generation report.
Gary's Question about the use of Trip Generation's Average Rate vs. Regression Curve
We have a situation in Rochester concerning a trip generation estimate for a development that is being opposed by a neighborhood. At issue is whether the average trip generation rate or the equation be used to determine the daily trips. (The value being determined happens to fall near one end of the plotted data points).
Would you, as friends and peers in this discipline, please provide me with an explanation concerning the practice within the reviewing government agencies that you are familiar with that is generally followed concerning the rate that is used? This item is going before the City Council next Monday, so if I could hear back from you in the next couple days it would be much appreciated.
The majority (almost all) of the traffic studies I’ve produced and have reviewed in my career in Minnesota have used the average trip generation rate instead of the supplied regression equations to calculate trip generation for developments. The hurdle for using the regression equation is quite high based on the Institute of Transportation Engineers’ Trip Generation Handbook, 2nd Edition, which recommends using the regression equation only when the data sample has at least 20 data points AND an R2 value of 0.75 or higher.
Caveat to allow for the engineer’s judgment – we’re supposed to look at the data plot, find the cluster of developments that are sized similarly to our study development, and consider using the fitted line (average rate or regression) that best matches the cluster of trip generation values. A further caveat to all of this – if the proposed development size lies on the extreme high or low end of the data plot, performing an independent trip generation study of similarly sized/local developments to determine local trip generation characteristics should be considered.