search for books and compare prices
Tables of Contents for Observational Before-After Studies in Road Safety
Chapter/Section Title
Page #
Page Count
Preface
xi
1
Glossary
xii
1 Introduction
1
8
Suggested Readings
5
1
Endnotes
6
3
PART I: ESSENTIALS
9
50
2 What is the question?
11
6
Suggested Readings
15
1
Endnote
15
2
3 Defining safety
17
14
3.1 Underbrush
17
3
3.2 Safety as a property of an entity
20
6
3.3 Frequency or rate?
26
2
3.4 Chapter summary
28
3
Endnote
29
2
4 Counting accidents
31
20
4.1 What is being counted
31
9
a. What is an accident?
31
1
b. The question of reportability
32
4
c. Incomplete reporting
36
2
d. Errors
38
1
e. Summary
39
1
4.2 Target accidents
40
7
4.3 Chapter summary
47
4
5 Prediction and estimation
51
8
5.1 Prediction of what safety would have been
51
4
5.2 Estimation of what safety was after the treatment
55
1
5.3 Chapter summary
56
3
PART II: ADAPTATIONS OF CONVENTIONAL APPROACHES
59
112
6 Basic building blocks
61
12
6.1 The Four-Step
61
8
6.2 Statistical differentials
69
1
6.3 Chapter summary
70
3
Suggested readings
71
1
Endnote
71
2
7 The Native Before-After study
73
22
7.1 Statistical analysis of the Naive Before-After study
75
5
Derivations
79
4
7.2 Separating the wheat from the chaff
80
2
7.3 Study design considerations
82
7
Derivations
88
1
7.4 Signal heads and intergreen times - on reading and learning
89
4
7.5 Chapter summary
93
2
8 Improving prediction I: Factors measured and understood
95
20
8.1 Accounting for change in traffic
96
4
8.2 The traffic flow correction in the Four-Step
100
4
Derivations
104
1
8.3 The estimation of r(tf)
104
4
Derivations
106
2
8.4 Coefficients of variation for AADT estimates
108
2
8.5 Illustrations and discussion
110
3
8.6 Chapter summary
113
2
9 Improving prediction II: Using a comparison group
115
38
9.1 Statistical analysis
119
8
Derivations
125
2
9.2 Study design considerations for the `C-G method'
127
6
Derivations
132
1
9.3 Estimation of VAR {w}
133
5
Derivations
137
1
9.4 A case study: Replacing STOP signs by YIELD signs
138
5
9.5 When different entities have different comparison ratios
143
4
9.6 The modified comparison ratio
147
2
9.7 Chapter summary
149
4
Endnote
151
2
10 The variability of treatment effect
153
18
10.1 The expanded `Four-Step'
155
4
Derivations
156
3
10.2 An illustration: Raised pavement markers
159
3
10.3 Application to Meta Analysis
162
6
10.4 Chapter summary
168
3
PART III: ELEMENTS OF A NEW APPROACH
171
104
11 Back to the starting point: The Empirical Bayes approach
175
48
11.1 The shaky foundation and how to shore it up
175
3
11.2 The regression-to-the-mean phenomenon
178
7
11.3 Two clues to safety
185
2
11.4 The mathematics of mixing the two clues
187
9
Derivations
193
3
11.5 How to estimate E{k} and VAR{k}
196
9
a. The Method of Sample Moments
197
3
b. The Multivariate Regression Method
200
4
Derivations
204
1
11.6 The proof of the pudding
205
3
11.7 Two case studies
208
4
a. Conversion from `two-way' to `four-way' stop control in San Francisco
208
2
b. The safety effect of warning devices at highway-rail grade crossings
210
2
11.8 Naive and C-G studies revisited
212
3
a. The EB Naive Study
212
2
b. The EB Comparison-Group study
214
1
11.9 Additional applications
215
3
11.10 Chapter summary
217
6
Endnote
218
5
12 A more coherent approach?
223
48
12.1 Uses of multivariate models of accident counts
225
2
12.2 The model equation: Meaning, form and assumptions
227
7
12.3 Likelihood function for parameter estimation
234
7
Derivations
239
2
12.4 An illustration
241
5
12.5 How to estimate the k(1), k(2), k(3), ...k(y) for some entity?
246
11
a. Stage 1: Maximum likelihood estimation
247
3
b. Empirical Bayes estimation
250
4
Derivations
254
3
12.6 How to predict the K(i), Y+1,..., K(i), Y+Z
257
4
12.7 The safety effect of road resurfacing in New York State
261
8
12.8 Chapter summary
269
2
13 Closure
271
4
References
275
8
Index
283