The relationship between the choice of deterrence function and the goodness of fit of a singly constrained spatial interaction model is examined as a basis for improving model performance. The results show that there is no significant improvement in model goodness of fit until a deterrence-function characterisation is used which is based on a family of functions, with the spatial domain of each function being determined in an approximately optimal manner. These findings are consistent with theoretical research on microlevel trip behaviour and can be used to identify descriptive models which possess maximum levels of performance.
Get full access to this article
View all access options for this article.
References
1.
AshfordNCovaultD O, 1969“The mathematical form of travel time factors”Highway Research Record28330–43
2.
BlackJ ESalterR J, 1975“A statistical evaluation of the accuracy of a family of gravity models”Proceedings of the Institute of Civil Engineers Part 2 591–20
FletcherR, 1971“A modified Marquardt subroutine for non-linear least squares” AERE R6799 (HMSO, London)
5.
HathawayP J, 1975“Trip distribution and disaggregation”Environment and Planning A771–97
6.
HymanG M, 1969“The calibration of trip distribution models”Environment and Planning1105–112
7.
HymanG M, 1976“Gravity model representation of heterogenous spatial interaction processes” CES-WN-432 (Centre for Environmental Studies, London)
8.
HymanG MGleaveD, 1976“A reasonable theory of migration” CES-RP-22 (Centre for Environmental Studies, London)
9.
KirbyH R, 1974“Theoretical requirements for calibrating gravity models”Transportation Research897–104
10.
LeeseM N, 1977“A statistical test for trip distribution models”Transport Studies Group, University College London (mimeo)
11.
MarchL, 1971“Urban systems: A generalised distribution function” in London Papers in Regional Science 2. Urban and Regional Planning Ed. WilsonA G, (Pion, London) pp 157–170
12.
OpenshawS, 1976“An empirical study of some spatial interaction models”Environment and Planning A823–41
13.
OpenshawS, 1977“Optimal zoning systems for spatial interaction models”Environment and Planning A9169–184
14.
SouthwellW H, 1976“Fitting data to nonlinear functions with uncertainties in all measurement variables”Computer Journal1969–73
15.
SouthworthF, 1974“The calibration of a trip-distribution–modal-split model with origin-specific decay parameters” WP-40, Institute for Transport Studies, University of Leeds, Leeds, England
16.
SouthworthF, 1975“A highly disaggregated modal-split model” WP-58, Institute for Transport Studies, University of Leeds, Leeds, England
17.
StetzerF, 1976“Parameter estimation for the constrained gravity model: A comparison of six methods”Environment and Planning A8673–684
18.
WilsonA G, 1973“Further developments of entropy maximising transport models”Transportation Planning and Technology1183–193
19.
WoottonH JPickG W, 1967“A model for trips generated by households”Journal of Transport Economics and Policy1137–153
