Cover Image

Not for Sale

View/Hide Left Panel
Click for next page ( 74

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 73
62 I N N O VAT I O N S I N T R AV E L D E M A N D M O D E L I N G , V O L U M E 2 TABLE 2 Aggregation Levels at Which Variables in Synthetic Population Have Reasonably High Accuracy and Precision Base Year Usage Forecast Year Usage Variable (Version 316) (Version 52) Household income (major control categories) Tract Tract Household income (subcategories) PUMA PUMA Person poverty status PUMA ? Household size (major control categories) Tract PUMA Household size (subcategories) Tract PUMA Household workers (major control categories) Tract PUMA Household workers (subcategories) PUMA PUMA Person employment status PUMA PUMA Person weekly work hours (35+ category) Tract PUMA Person weekly work hours (other categories) PUMA PUMA Household with holder age 65+ Tract PUMA Household presence/absence of own children age 017 Tract PUMA Person age category PUMA PUMA Household family status Tract PUMA Household housing type (major categories) PUMA PUMA Household housing ownership (major categories) PUMA PUMA Person gender Tract Tract Person race and Hispanic status PUMA ? Person school enrollment category PUMA ? variables are synthesized much less accurately, even in accurate, because (a) the quantity of forecast controls is the base year. Second, the accuracy drops when results smaller and (b) they are more aggregate than the base- are examined at a more detailed level of aggregation. year controls. The conclusion to be drawn is that it is Third, even for controlled variables, the accuracy is not indeed important to implement validation procedures perfect; in the ARC base-year case, the rounding proce- that provide the user of a PopSyn with the information dures used after IPF, before the households are drawn, needed to use it appropriately. It is also valuable to introduce a substantial amount of noise; in the forecast implement a flexible PopSyn that can be adjusted and case, the use of averages (variables that do not match the improved in response to the validation information. IPF categories) and regional values all degrade accuracy, With this version of the PopSyn in hand, ARC is in a precision, or both. The accuracy and precision of the good position to continue validating and improving it, forecast population are less than those of the base-year even as ARC incorporates it into the demand models and population, even assuming that the forecast inputs are uses it for analysis.