U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

 
REPORT
This report is an archived publication and may contain dated technical, contact, and link information
Back to Publication List        
Publication Number:  FHWA-HRT-14-020    Date:  January 2015
Publication Number: FHWA-HRT-14-020
Date: January 2015

 

Factors Influencing Operating Speeds and Safety on Rural and Suburban Roads

Appendix C. Safety Effects of Lane-Width and Shoulder-Width Combinations on Rural, Two-Lane Roads (Additional Modeling Results)

Table 107. Negative binomial regression model estimation results for property damage only crashes (all severities).

Variable

Coefficient

Standard Error

z

P > z

95-Percent Confidence Interval

ln_aadt

0.649

0.061

10.700

< 0.001

0.530

0.768

illinois

1.195

0.108

11.080

< 0.001

0.984

1.407

Lane10

1.047

0.227

4.610

< 0.001

0.602

1.493

Lane11

0.431

0.166

2.610

0.009

0.107

0.756

shoulder

-0.052

0.024

-2.140

0.032

-0.100

-0.004

ln10shld

-0.119

0.060

-1.980

0.048

-0.238

-0.001

ln11shld

-0.037

0.032

-1.160

0.246

-0.099

0.025

barrier

4.346

1.807

2.400

0.016

0.804

7.889

drvwy_den

0.016

0.007

2.340

0.019

0.003

0.029

solid_CL

0.563

0.215

2.620

0.009

0.143

0.984

dash1_CL

0.992

0.334

2.970

0.003

0.338

1.646

curve

0.253

0.076

3.340

0.001

0.105

0.402

_cons

-5.401

0.486

-11.120

< 0.001

-6.352

-4.449

ln(Length)

1

(exposure)

/lnalpha

-0.86727

0.117387

-1.09734

-0.63719

alpha

0.420098

0.049314

0.333757

0.528774

Log likelihood = -1473.751
Number of Observations = 877
LR chi2(12) = 425.44
Prob > chi2 = 0
Pseudo R2 = 0.1261
Likelihood-ratio test of alpha=0: chibar2(01) = 254.30 Prob>=chibar2 = 0.000
— Indicates Not Available

Table 108. Negative binomial regression model estimation results for single-vehicle crashes
(all severities).

sv_all

Coefficient

Standard Error

z

P > z

95-percent Confidence Interval

ln_aadt

0.514

0.059

8.67

< 0.001

0.398

0.631

illinois

1.128

0.103

10.9

< 0.001

0.925

1.331

Lane10

0.896

0.219

4.08

< 0.001

0.466

1.326

Lane11

0.413

0.160

2.58

0.010

0.099

0.727

shoulder

-0.064

0.024

-2.71

0.007

-0.111

-0.018

ln10shld

-0.090

0.058

-1.56

0.120

-0.202

0.023

ln11shld

-0.030

0.031

-0.99

0.323

-0.091

0.030

barrier

3.597

1.760

2.04

0.041

0.147

7.046

drvwy_den

0.017

0.007

2.630

0.009

0.004

0.030

solid_CL

0.478

0.208

2.300

0.022

0.070

0.886

dash1_CL

0.923

0.324

2.850

0.004

0.288

1.559

curve

0.246

0.074

3.330

0.001

0.101

0.391

_cons

-4.174

0.469

-8.890

< 0.001

-5.094

-3.254

ln(Length)

1

(exposure)

/lnalpha

-0.94318

0.120937

-1.180212

-0.7061473

alpha

0.389388

0.047092

0.3072135

0.493542

Log likelihood = -1493.2046
Number of Observations = 877
LR chi2(12) = 408.55
Prob > chi2 = 0
Pseudo R2 = 0.1203
Likelihood-ratio test of alpha=0: chibar2(01) = 229.65 Prob>=chibar2 = 0.000
— Indicates Not Available

Table 109. Negative binomial regression model estimation results for multiple-vehicle crashes (all severities).

mv_all

Coefficient

Standard Error

z

P > z

95-Percent Confidence Interval

ln_aadt

1.482

0.110

13.430

< 0.001

1.266

1.698

illinois

0.628

0.171

3.680

< 0.001

0.294

0.962

Lane10

0.769

0.470

1.640

0.102

-0.153

1.690

Lane11

0.199

0.306

0.650

0.515

-0.400

0.798

shoulder

0.007

0.040

0.160

0.872

-0.073

0.086

ln10shld

-0.223

0.132

-1.690

0.092

-0.481

0.036

ln11shld

-0.006

0.053

-0.110

0.910

-0.110

0.098

barrier

5.690

2.653

2.150

0.032

0.491

10.889

drvwy_den

-0.002

0.012

-0.160

0.875

-0.024

0.021

solid_CL

0.422

0.343

1.230

0.218

-0.250

1.094

dash1_CL

0.641

0.542

1.180

0.237

-0.421

1.703

curve

-0.013

0.126

-0.100

0.921

-0.259

0.234

_cons

-13.401

0.909

-14.750

< 0.001

-15.182

-11.620

ln(Length)

1

(exposure)

/lnalpha

-1.1353

0.372991

-1.866348

-0.4042488

alpha

0.321326

0.119852

0.1546876

0.667478

Log likelihood = -649.6117
Number of Observations = 877
LR chi2(12) = 213.52
Prob > chi2 = 0
Pseudo R2 = 0.1411
Likelihood-ratio test of alpha=0: chibar2(01) = 12.89 Prob>=chibar2 = 0.000
— Indicates Not Available

Table 110. Negative binomial regression model estimation results for key lane width/shoulder width crashes (all severities).

key_kabco

Coefficient

Standard Error

z

P > z

95-Percent Confidence Interval

ln_aadt

0.606

0.059

10.27

< 0.001

0.490

0.721

illinois

1.083

0.099

10.98

< 0.001

0.890

1.277

Lane10

0.903

0.223

4.04

< 0.001

0.465

1.341

Lane11

0.519

0.161

3.23

0.001

0.205

0.834

shoulder

-0.052

0.023

-2.26

0.024

-0.097

-0.007

ln10shld

-0.076

0.056

-1.37

0.172

-0.185

0.033

ln11shld

-0.060

0.030

-1.97

0.049

-0.119

0.000

barrier

4.305

1.888

2.28

0.023

0.604

8.006

drvwy_den

0.012

0.007

1.88

0.060

-0.001

0.025

solid_CL

0.471

0.209

2.26

0.024

0.062

0.880

dash1_CL

0.800

0.323

2.47

0.013

0.166

1.433

curve

0.231

0.074

3.14

0.002

0.087

0.376

_cons

-4.613

0.466

-9.9

<0.001

-5.527

-3.700

ln(Length)

1

(exposure)

/lnalpha

-0.72081

0.102919

-0.92253

-0.5191

alpha

0.486356

0.050055

0.397512

0.595057

Log likelihood = -1684.8577
Number of Observations = 877
LR chi2(12) = 395.990
Prob > chi2 = 0.000
Pseudo R2 = 0.105
Likelihood-ratio test of alpha=0: chibar2(01) = 395.21 Prob>=chibar2 = 0.000
— Indicates Not Available

Table 111. Negative binomial regression model estimation results for single-vehicle crashes (fatal-plus-injury).

sv_fi

Coefficient

Standard. Error

z

P > z

95-Percent Confidence Interval

ln_aadt

0.562

0.112

5.01

< 0.001

0.342

0.783

illinois

0.427

0.189

2.26

0.024

0.057

0.798

Lane10

-0.035

0.428

-0.08

0.935

-0.874

0.804

Lane11

0.193

0.305

0.63

0.528

-0.406

0.791

shoulder

-0.116

0.045

-2.58

0.010

-0.203

-0.028

ln10shld

0.063

0.106

0.6

0.551

-0.144

0.271

ln11shld

-0.023

0.061

-0.39

0.699

-0.142

0.095

barrier

0.576

3.244

0.18

0.859

-5.783

6.934

drvwy_den

0.007

0.013

0.590

0.553

-0.017

0.032

solid_CL

0.530

0.369

1.440

0.151

-0.194

1.253

dash1_CL

0.345

0.615

0.560

0.575

-0.860

1.550

curve

0.081

0.141

0.580

0.563

-0.194

0.357

_cons

-5.385

0.882

-6.100

<0.001

-7.114

-3.656

ln(Length)

1

(exposure)

/lnalpha

-1.10368

0.491638

-2.06727

-0.1400852

alpha

0.331649

0.163051

0.1265307

0.8692842

Log likelihood = -579.36383
Number of Observations = 877
LR chi2(12) = 64.28
Prob > chi2 = 0
Pseudo R2 = 0.0526
Likelihood-ratio test of alpha=0: chibar2(01) = 6.35 Prob>=chibar2 = 0.006
— Indicates Not Available

Table 112. Negative binomial regression model estimation results for multiple-vehicle crashes (fatal-plus-injury).

mv_fi

Coefficient

Standard Error

z

P > z

95-percent Confidence Interval

ln_aadt

1.142

0.162

7.060

< 0.001

0.825

1.459

illinois

0.557

0.248

2.240

0.025

0.071

1.043

Lane10

0.427

0.992

0.430

0.667

-1.517

2.371

Lane11

0.762

0.481

1.580

0.114

-0.182

1.705

shoulder

0.146

0.059

2.470

0.013

0.030

0.262

ln10shld

-0.185

0.219

-0.840

0.398

-0.613

0.244

ln11shld

-0.100

0.078

-1.290

0.196

-0.252

0.052

barrier

5.476

4.069

1.350

0.178

-2.498

13.450

drvwy_den

-0.011

0.019

-0.570

0.570

-0.048

0.026

solid_CL

0.129

0.559

0.230

0.817

-0.966

1.225

dash1_CL

0.533

0.836

0.640

0.524

-1.106

2.173

curve

0.126

0.186

0.680

0.498

-0.238

0.490

_cons

-12.279

1.343

-9.140

<0.001

-14.911

-9.646

ln(Length)

1

(exposure)

/lnalpha

-1.18079

0.886305

-2.917915

0.5563366

alpha

0.307036

0.272128

0.0540463

1.744271

Log likelihood = -374.64352
Number of Observations = 877
LR chi2(12) = 79.46
Prob > chi2 = 0
Pseudo R2 = 0.0959
Likelihood-ratio test of alpha=0: chibar2(01) = 1.73 Prob>=chibar2 = 0.094
— Indicates Not Available

Table 113. Negative binomial regression model estimation results for key lane-width/shoulder-width crashes (fatal-plus-injury).

key_kabc

Coefficient

Standard Error

z

P > z

95-Percent Confidence Interval

ln_aadt

0.598

0.114

5.27

< 0.001

0.376

0.821

illinois

0.523

0.180

2.9

0.004

0.170

0.875

Lane10

0.080

0.458

0.17

0.862

-0.819

0.978

Lane11

0.552

0.319

1.73

0.084

-0.073

1.178

shoulder

-0.070

0.044

-1.6

0.109

-0.155

0.015

ln10shld

0.068

0.105

0.65

0.518

-0.138

0.274

ln11shld

-0.093

0.060

-1.55

0.121

-0.211

0.025

barrier

0.314

3.738

0.08

0.933

-7.014

7.641

drvwy_den

0.009

0.013

0.67

0.501

-0.017

0.034

solid_CL

0.288

0.397

0.72

0.469

-0.490

1.066

dash1_CL

0.416

0.629

0.66

0.508

-0.815

1.648

curve

0.034

0.147

0.23

0.818

-0.255

0.322

_cons

-5.392

0.888

-6.07

<0.001

-7.132

-3.651

ln(Length)

1

(exposure)

/lnalpha

0.481668

0.153744

 

0.180335

0.783001

alpha

1.618773

0.248877

1.197619

2.18803

Log likelihood = -817.20472
Number of Observations = 877
LR chi2(12) = 66
Prob > chi2 = 0.0000
Pseudo R2 = 0.0388
Likelihood-ratio test of alpha=0: chibar2(01) = 144.74 Prob>=chibar2 = 0.000
— Indicates Not Available

Table 114. Negative binomial regression model estimation results for single-vehicle crashes (PDO).

sv_o

Coefficient

Standard Error

z

P > z

95-Percent Confidence Interval

ln_aadt

0.515

0.065

7.9

< 0.001

0.387

0.643

illinois

1.307

0.117

11.17

< 0.001

1.077

1.536

Lane10

1.029

0.240

4.3

< 0.001

0.560

1.499

Lane11

0.427

0.175

2.44

0.015

0.083

0.770

shoulder

-0.054

0.026

-2.04

0.041

-0.105

-0.002

ln10shld

-0.113

0.064

-1.77

0.077

-0.238

0.012

ln11shld

-0.031

0.034

-0.9

0.370

-0.097

0.036

barrier

4.386

1.928

2.27

0.023

0.606

8.166

drvwy_den

0.018

0.007

2.490

0.013

0.004

0.032

solid_CL

0.499

0.230

2.170

0.030

0.048

0.951

dash1_CL

1.031

0.356

2.900

0.004

0.333

1.728

curve

0.288

0.081

3.570

<0.001

0.130

0.447

_cons

-4.589

0.519

-8.840

<0.001

-5.606

-3.572

ln(Length)

1

(exposure)

/lnalpha

-0.76859

0.121186

-1.006113

-0.5310743

alpha

0.463665

0.05619

0.3656375

0.587973

Log likelihood = -1390.4041
Number of Observations = 877
LR chi2(12) = 397.58
Prob > chi2 = 0
Pseudo R2 = 0.1251
Likelihood-ratio test of alpha=0: chibar2(01) = 239.28 Prob>=chibar2 = 0.000
— Indicates Not Available

Table 115. Negative binomial regression model estimation results for multiple-vehicle crashes (PDO).

mv_o

Coefficient

Standard Error

z

P > z

95-Percent Confidence Interval

ln_aadt

1.727

0.144

12.010

< 0.001

1.445

2.009

illinois

0.671

0.222

3.020

0.003

0.236

1.105

Lane10

0.686

0.542

1.270

0.206

-0.376

1.749

Lane11

-0.198

0.383

-0.520

0.605

-0.948

0.552

shoulder

-0.095

0.052

-1.810

0.071

-0.197

0.008

ln10shld

-0.186

0.167

-1.120

0.264

-0.514

0.141

ln11shld

0.068

0.070

0.980

0.329

-0.069

0.205

barrier

5.742

3.298

1.740

0.082

-0.722

12.207

drvwy_den

0.004

0.014

0.290

0.775

-0.024

0.032

solid_CL

0.568

0.420

1.350

0.176

-0.254

1.391

dash1_CL

0.653

0.682

0.960

0.338

-0.684

1.991

curve

-0.096

0.161

-0.590

0.552

-0.412

0.220

_cons

-15.436

1.179

-13.090

<0.001

-17.747

-13.124

ln(Length)

1

(exposure)

/lnalpha

-0.79447

0.433001

-1.643133

0.0542011

alpha

0.451823

0.19564

0.1933733

1.055697

Log likelihood = -469.42118
Number of Observations = 877
LR chi2(12) = 178.04
Prob > chi2 = 0
Pseudo R2 = 0.1594
Likelihood-ratio test of alpha=0: chibar2(01) = 9.54 Prob>=chibar2 = 0.001
— Indicates Not Available

Table 116. Negative binomial regression model estimation results for key lane-width/shoulder-width crashes (PDO).

key_o

Coefficient

Standard Error

z

P > z

95-Percent Confidence Interval

ln_aadt

0.615

0.066

9.37

< 0.001

0.486

0.743

illinois

1.276

0.113

11.25

< 0.001

1.054

1.498

Lane10

1.059

0.245

4.33

< 0.001

0.579

1.539

Lane11

0.494

0.177

2.8

0.005

0.148

0.841

shoulder

-0.046

0.026

-1.8

0.072

-0.097

0.004

ln10shld

-0.104

0.063

-1.67

0.096

-0.227

0.018

ln11shld

-0.049

0.034

-1.45

0.147

-0.115

0.017

barrier

5.253

2.057

2.55

0.011

1.221

9.286

drvwy_den

0.013

0.007

1.85

0.065

-0.001

0.028

solid_CL

0.544

0.232

2.35

0.019

0.090

0.999

dash1_CL

0.913

0.358

2.55

0.011

0.211

1.614

curve

0.292

0.081

3.6

< 0.001

0.133

0.451

_cons

-5.167

0.523

-9.88

< 0.001

-6.192

-4.141

ln(Length)

1

(exposure)

/lnalpha

-0.5778

0.107196

-0.7879

-0.3677

alpha

0.561129

0.060151

0.454797

0.692322

Log likelihood = -1524.6018
Number of Observations = 877
LR chi2(12) = 391.03
Prob > chi2 = 0
Pseudo R2 = 0.1137
Likelihood-ratio test of alpha=0: chibar2(01) = 363.82 Prob>=chibar2 = 0.000
— Indicates Not Available

Table 117. Multinomial logit model estimation results for total crashes (all types and severities)—base outcome: PDO.

Severity

Variable

Coefficient

Standard Error

z

P > z

95-Percent Confidence Interval

0

(base outcome)

1

ln_aadt

0.103

0.190

0.540

0.588

-0.270

0.476

illinois

-2.615

0.301

-8.700

0.000

-3.205

-2.026

Lane10

-0.993

0.914

-1.090

0.277

-2.784

0.797

Lane11

-0.125

0.546

-0.230

0.819

-1.195

0.945

shoulder

-0.079

0.071

-1.110

0.267

-0.218

0.060

ln10shld

0.216

0.173

1.250

0.211

-0.123

0.555

ln11shld

0.023

0.088

0.260

0.797

-0.150

0.195

barrier

-5.154

7.644

-0.670

0.500

-20.137

9.829

drvway_den

-0.008

0.025

-0.330

0.739

-0.058

0.041

solid_CL

-0.905

0.767

-1.180

0.238

-2.409

0.599

dash1_CL

0.857

1.031

0.830

0.406

-1.163

2.877

curve

-0.377

0.237

-1.590

0.111

-0.842

0.087

_cons

-1.450

1.448

-1.000

0.316

-4.288

1.387

2

ln_aadt

0.038

0.134

0.280

0.780

-0.226

0.301

illinois

-0.007

0.257

-0.030

0.980

-0.511

0.498

Lane10

-1.107

0.630

-1.760

0.079

-2.341

0.127

Lane11

-0.172

0.368

-0.470

0.641

-0.893

0.550

shoulder

0.082

0.053

1.540

0.124

-0.023

0.186

ln10shld

0.133

0.152

0.880

0.380

-0.164

0.431

ln11shld

-0.032

0.070

-0.450

0.650

-0.170

0.106

barrier

1.097

4.403

0.250

0.803

-7.532

9.726

drvway_den

-0.022

0.017

-1.300

0.192

-0.054

0.011

solid_CL

-0.331

0.495

-0.670

0.504

-1.301

0.639

dash1_CL

-0.595

0.771

-0.770

0.441

-2.107

0.917

curve

-0.218

0.168

-1.300

0.194

-0.547

0.111

_cons

-2.529

1.099

-2.300

0.021

-4.683

-0.375

3

ln_aadt

0.257

0.168

1.530

0.126

-0.073

0.587

illinois

0.613

0.393

1.560

0.119

-0.157

1.383

Lane10

-0.057

0.693

-0.080

0.934

-1.416

1.302

Lane11

0.443

0.481

0.920

0.357

-0.500

1.386

shoulder

0.085

0.073

1.160

0.248

-0.059

0.229

ln10shld

-0.069

0.215

-0.320

0.747

-0.491

0.352

ln11shld

-0.097

0.096

-1.010

0.312

-0.284

0.091

barrier

-1.734

5.888

-0.290

0.768

-13.273

9.806

drvway_den

-0.022

0.019

-1.130

0.258

-0.060

0.016

solid_CL

0.884

0.485

1.820

0.068

-0.067

1.836

dash1_CL

-1.950

1.012

-1.930

0.054

-3.934

0.034

curve

0.179

0.205

0.870

0.382

-0.222

0.580

_cons

-5.679

1.441

-3.940

0.000

-8.503

-2.855

4

ln_aadt

0.228

0.351

0.650

0.516

-0.460

0.917

illinois

-0.130

0.700

-0.190

0.853

-1.501

1.242

Lane10

1.177

1.687

0.700

0.485

-2.129

4.484

Lane11

1.441

1.082

1.330

0.183

-0.681

3.562

shoulder

0.207

0.156

1.320

0.186

-0.099

0.513

ln10shld

-0.671

0.962

-0.700

0.486

-2.557

1.216

ln11shld

-0.211

0.189

-1.110

0.265

-0.581

0.160

barrier

-3.203

12.442

-0.260

0.797

-27.589

21.183

drvway_den

-0.054

0.050

-1.090

0.276

-0.152

0.043

solid_CL

0.672

1.006

0.670

0.504

-1.300

2.644

dash1_CL

0.528

1.968

0.270

0.788

-3.330

4.386

curve

0.403

0.425

0.950

0.343

-0.430

1.236

_cons

-7.442

3.027

-2.460

0.014

-13.376

-1.508

Log likelihood = -1543.1747
Number of observation = 2397
LR chi2(48) = 198.440
Prob > chi2 = 0.000
Pseudo R2 = 0.060
— Indicates Not Available

Table 118. Multinomial logit model estimation results for fatal-plus-injury crashes (all types)—base outcome: possible injury.

severity_rev

Variable

Coefficient

Standard Error

z

P > z

95-Percent Confidence Interval

1

(base outcome)

2

ln_aadt

-0.273

0.231

-1.180

0.237

-0.726

0.180

illinois

2.531

0.388

6.530

<0.001

1.771

3.290

Lane10

0.011

1.163

0.010

0.992

-2.267

2.290

Lane11

0.044

0.689

0.060

0.949

-1.305

1.394

shoulder

0.163

0.091

1.780

0.075

-0.016

0.342

ln10shld

-0.191

0.250

-0.770

0.444

-0.682

0.299

ln11shld

-0.063

0.118

-0.530

0.594

-0.294

0.168

barrier

10.549

9.177

1.150

0.250

-7.438

28.537

drvway_den

-0.006

0.031

-0.200

0.840

-0.067

0.054

solid_CL

0.618

0.866

0.710

0.475

-1.079

2.316

dash1_CL

-1.659

1.326

-1.250

0.211

-4.258

0.940

curve

0.147

0.304

0.490

0.627

-0.448

0.742

_cons

0.604

1.785

0.340

0.735

-2.895

4.103

3

ln_aadt

-0.057

0.263

-0.220

0.829

-0.572

0.459

illinois

3.230

0.505

6.400

<0.001

2.240

4.220

Lane10

1.030

1.226

0.840

0.401

-1.372

3.432

Lane11

0.758

0.781

0.970

0.332

-0.773

2.289

shoulder

0.184

0.111

1.660

0.096

-0.033

0.401

ln10shld

-0.389

0.299

-1.300

0.192

-0.974

0.196

ln11shld

-0.155

0.141

-1.100

0.273

-0.432

0.122

barrier

7.135

10.398

0.690

0.493

-13.245

27.514

drvway_den

-0.005

0.033

-0.150

0.884

-0.069

0.059

solid_CL

1.739

0.882

1.970

0.049

0.010

3.468

dash1_CL

-3.028

1.529

-1.980

0.048

-6.025

-0.030

curve

0.542

0.336

1.610

0.107

-0.116

1.199

_cons

-2.654

2.127

-1.250

0.212

-6.823

1.515

4

ln_aadt

-0.047

0.406

-0.120

0.907

-0.843

0.748

illinois

2.492

0.779

3.200

0.001

0.966

4.017

Lane10

2.603

2.183

1.190

0.233

-1.676

6.882

Lane11

1.840

1.247

1.480

0.140

-0.604

4.284

shoulder

0.310

0.178

1.740

0.082

-0.039

0.660

ln10shld

-1.145

1.307

-0.880

0.381

-3.707

1.417

ln11shld

-0.289

0.217

-1.330

0.183

-0.715

0.137

barrier

4.281

15.898

0.270

0.788

-26.879

35.442

drvway_den

-0.041

0.057

-0.720

0.472

-0.152

0.070

solid_CL

1.370

1.227

1.120

0.264

-1.035

3.775

dash1_CL

-0.305

2.244

-0.140

0.892

-4.702

4.092

curve

0.757

0.496

1.530

0.127

-0.216

1.730

_cons

-4.744

3.408

-1.390

0.164

-11.423

1.936

Log likelihood = -464.69741
Number of Observations = 426
LR chi2(36) = 112.18
Prob > chi2 = 0
Pseudo R2 = 0.1077
— Indicates Not Available

 

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101