| Contributing authors | p. xi |
| Foreword: on a manuscript of 1984 | p. xiii |
| Research support, and more | p. xvii |
| National and Regional Models | |
| Multiple Levels, Damages, Forms, Moments and Variables in Road Accident Models | p. 1 |
| Introduction: the [double left angle bracket]Modelling Quartet[double right angle bracket] in this book | p. 1 |
| Problem Formulation | p. 2 |
| A multilevel structure; a multidamage application | p. 2 |
| Perspectives on problem structure | p. 6 |
| The quantification of effects | p. 9 |
| From fixed to flexible mathematical form | p. 9 |
| From monotonic to multitonic forms: the case of alcohol | p. 14 |
| Variables: multimoment, multivariate | p. 19 |
| The dependent variable: from observations to moments | p. 19 |
| The explanatory variables: not a triad, but a quatrain | p. 24 |
| Is pregnancy a risk factor? | p. 27 |
| Other modelling dimensions of interest | p. 29 |
| Multidata | p. 29 |
| Multiple documentation of reference results | p. 30 |
| Conclusion: matching tools to questions | p. 31 |
| Post Scriptum: model acronyms | p. 31 |
| References | p. 32 |
| The Drag-2 Model for Quebec | p. 37 |
| Introduction | p. 37 |
| The structure of the DRAG-2 model | p. 38 |
| A diagram of the model structure | p. 38 |
| Dependent variable graphs | p. 40 |
| The matrix of direct effects of independent variables over dependent variables | p. 45 |
| REsults on mathematical form and particular variables | p. 49 |
| Econometric results | p. 50 |
| Results on elasticity | p. 51 |
| Other results: forecasts for the period of 1997-2004 | p. 63 |
| Other developments | p. 66 |
| The SNUS-2.5 Model for Germany | p. 67 |
| Context | p. 67 |
| Structure of model | p. 68 |
| The dependent variables | p. 68 |
| Visual analysis of the dependent variables | p. 69 |
| Matrix of direct effects | p. 70 |
| Results and their interpretation | p. 73 |
| Statistical results | p. 73 |
| Economic results: overall specific results | p. 73 |
| Decomposition of the impact by variable: results common to other models | p. 75 |
| Results for other variables | p. 79 |
| Deriving other interesting results | p. 82 |
| The analysis of victims: direct, indirect and total elasticities | p. 82 |
| Multiple moments and their marginal rates of substitution | p. 84 |
| Marginal rates of substitution with comparable accident data | p. 88 |
| Marginal rates of substitution with disaggregated accident data | p. 92 |
| Policy implications | p. 94 |
| Higher prices save energy and lives | p. 94 |
| Risk substitution in terms of first moments | p. 94 |
| References | p. 94 |
| The TRULS--1 Model for Norway | p. 97 |
| Introduction | p. 97 |
| Structure of the Model Truls-1 | p. 98 |
| Dependent variables: definitions and relations | p. 99 |
| Visual analysis of dependent variables | p. 100 |
| Matrix of direct effects | p. 107 |
| The casualty subset test | p. 108 |
| Results on form and selected explanatory variables | p. 111 |
| References | p. 126 |
| The DRAG-Stockholm-2 Model | p. 127 |
| Introduction | p. 127 |
| The Dennis agreement | p. 127 |
| The MAD-project | p. 128 |
| The concept of zero fatality | p. 128 |
| Structure of the DRAG model for Stockholm county | |
| Introduction | p. 128 |
| Dependent variables: definitions and relations | p. 129 |
| Visual analysis of dependent variables | p. 129 |
| Matrix of direct effects | p. 133 |
| Model form and explanatory variables | p. 134 |
| Summary of econometric results | p. 134 |
| The demand for road use | p. 134 |
| Comparison between estimated and actual demand for road use | p. 139 |
| The contribution of road infrastructure to road traffic growth | p. 139 |
| The Road accident frequency and gravity models | p. 140 |
| Economic activities | p. 141 |
| Quality of vehicle fleet | p. 141 |
| Road network data | p. 142 |
| Weather data | p. 142 |
| Intervention measures | p. 142 |
| Gasoline price | p. 142 |
| The DRAG-Stockholm-2 model | p. 143 |
| The new model specification | p. 143 |
| Comparison of results between the "old" and "new" specification | p. 143 |
| Comparison of actual and estimated accident risks | p. 146 |
| Specific results on the Drag-Stockholm model | p. 148 |
| Points of interest and conclusion | p. 149 |
| Alcohol consumption: the J-shaped relationship | p. 150 |
| Medicine consumption | p. 151 |
| Pregnancy--a new risk factor | p. 151 |
| Conclusions | p. 154 |
| References | p. 154 |
| The TAG-1 Model for France | p. 157 |
| Introduction | p. 157 |
| Structuring the TAG model | p. 159 |
| Econometric form of the TAG model | p. 163 |
| The estimates produced by the TAG model | p. 167 |
| Model of road transport demand | p. 168 |
| Constructing a model of average speed | p. 170 |
| Analysis of the results by risk indicator | p. 173 |
| Analysis of results by explanatory factor | p. 175 |
| Conclusion | p. 181 |
| References | p. 182 |
| The TRACS-CA Model for California | p. 185 |
| Introduction | p. 185 |
| TRACS-CA model structure | p. 186 |
| Exposure and crash losses | p. 186 |
| Historical trends | p. 188 |
| Determining variables included in the TRACS-CA structure | p. 191 |
| Estimation results | p. 194 |
| Statistical summary | p. 194 |
| Common variable results | p. 195 |
| Specific variable results | p. 198 |
| Further results | p. 200 |
| Discussion and future directions | p. 200 |
| References | p. 203 |
| Comparing Six Drag-Type Models | p. 205 |
| Risk exposure | p. 206 |
| Driver behaviour | p. 209 |
| Speed | p. 209 |
| Seatbelt wearing | p. 212 |
| Consumption of alcohol | p. 214 |
| Consumption of medicines | p. 216 |
| Economic variables | p. 217 |
| Households' economic and financial situation | p. 217 |
| Fuel prices | p. 220 |
| Competing supply from public transport | p. 222 |
| Conclusion | p. 222 |
| References | p. 224 |
| Other Models and Issues | |
| The Road, Risk, Uncertainty and Speed | p. 225 |
| Risk, uncertainty and observed road accident outcomes | p. 225 |
| Model structure: simultaneity and perceived risk | p. 226 |
| Selected results: accident frequency and severity | p. 230 |
| Selected results: speed | p. 233 |
| Conclusion | p. 235 |
| References | p. 235 |
| The Res Model by Road Type in France | p. 237 |
| Introduction | p. 237 |
| Structure of the model | p. 238 |
| General outline | p. 238 |
| The data base | p. 239 |
| Economic formulation | p. 242 |
| Econometric specification | p. 243 |
| Algorithm | p. 244 |
| The Results | p. 245 |
| Tests of functional form | p. 245 |
| Measuring elasticities | p. 246 |
| Short and medium term simulations | p. 248 |
| Conclusion | p. 248 |
| References | p. 249 |
| Postface and Perspectives | p. 251 |
| Relevance of models for understanding the influence of risk factors | p. 252 |
| Outlook for research in constructing risk models | p. 258 |
| Data extraction | p. 258 |
| Adding levels to the structure | p. 259 |
| Breakdown of indicators by user and road types | p. 260 |
| Disaggregation by location or vehicle x driver | p. 260 |
| Relevance of the models for managing road safety | p. 260 |
| References | p. 262 |
| Algorithms and Detailed Model Outputs | |
| The Trio Level-1.5 Algorithm for Bc-Gauheseq Regression | |
| Introduction and statistical model | p. 263 |
| Introduction | p. 263 |
| Log-likelihood function | p. 265 |
| Computational aspects | p. 270 |
| Model types | p. 275 |
| Model estimation | p. 276 |
| Estimation results | p. 278 |
| Definitions of moments of the dependent variable | p. 278 |
| Derivatives and elasticities of the sample and expected values of the dependent variable | p. 283 |
| Derivatives and elasticities of the standard error of the dependent variable | p. 288 |
| Derivativies and elasticities of the skewness of the dependent variable | p. 293 |
| Ratios of derivatives of the moments of the dependent variable | p. 296 |
| Evaluation of moments, their derivatives, rates of substitution and elasticities | p. 301 |
| Student's t-statistics | p. 305 |
| Goodness-of-fit measures | p. 306 |
| Special options | p. 309 |
| Correlation matrix and table of variance-decomposition proportions | p. 309 |
| Analysis of heteroskedasticity of the residuals | p. 310 |
| Analysis of autocorrelation of the residuals | p. 311 |
| Forecasting: maximum likelihood and simulation forecasts | p. 314 |
| References | p. 321 |
| The Irposkml Procedure of Estimation | p. 325 |
| Accident frequency model specification | p. 325 |
| Severity model specification | p. 331 |
| References | p. 334 |
| Turning Box-Cox Including Quadratic forms in Regression | p. 335 |
| Model with two Box-Cox transformations on a same independent variable | p. 335 |
| Solution | p. 335 |
| First-order conditions | p. 336 |
| Second-order conditions | p. 337 |
| Special case: quadratic form | p. 339 |
| Model with powers [lamda subscript 1] and [lamda subscript 2] only on a same independent variable | p. 339 |
| First-order conditions | p. 340 |
| Second-order conditions | p. 341 |
| Special case: quadratic form | p. 343 |
| Two-step transformations on a same independent variable | p. 343 |
| References | p. 346 |
| Appendix 1. Detailed Model Outputs | p. 347 |
| Table of Contents provided by Syndetics. All Rights Reserved. |