Skip to contentUnited States Department of Transportation - Federal Highway Administration FHWA Home
Research Home
Report
This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-HRT-08-041
Date: March 2008

Safety Evaluation of Increasing Retroreflectivity of STOP Signs

PDF Version (835 KB)

PDF files can be viewed with the Acrobat® Reader®

Table of Contents

FOREWORD

The goal of this research was to evaluate and estimate the safety effectiveness of increasing retroreflectivity of STOP signs as one of the strategies in the Evaluation of Low-Cost Safety Improvements Pooled Fund Study (ELCSI-PFS), Phase I.

The ELCSI-PFS provides Crash Reduction Factor (CRF) and economic analysis for the targeted safety strategies where possible. The estimate of effectiveness for increasing retroreflectivity of STOP signs was determined by conducting scientifically rigorous before-after evaluations at sites where this strategy was implemented in the United States.

This safety improvement and all other targeted strategies in the ELCSI-PFS are identified as low-cost strategies in the NCHRP Report 500 guidebooks. Participating States in the ELCSI-PFS are Arizona, California, Connecticut, Florida, Georgia, Illinois, Indiana, Iowa, Kansas, Kentucky, Maryland, Massachusetts, Minnesota, Mississippi, Montana, New York, North Carolina, North Dakota, Oklahoma, Pennsylvania, South Carolina, South Dakota, Tennessee, Texas, Utah, and Virginia.

Michael F. Trentacoste

Director, Office of Safety

Research and Development

Notice

This document is disseminated under the sponsorship of the U.S. Department of Transportation in the interest of information exchange. The U.S. Government assumes no liability for the use of the information contained in this document.

The U.S. Government does not endorse products or manufacturers. Trademarks or manufacturers' names appear in this report only because they are considered essential to the objective of the document.

Quality Assurance Statement

The Federal Highway Administration (FHWA) provides high-quality information to serve Government, industry, and the public in a manner that promotes public understanding. Standards and policies are used to ensure and maximize the quality, objectivity, utility, and integrity of its information. FHWA periodically reviews quality issues and adjusts its programs and processes to ensure continuous quality improvement.

Technical Report Do

1. Report No.
FHWA-HRT-08-041

2. Government Accession No.

3. Recipient's Catalog No.

4. Title and Subtitle
Safety Evaluation of Increasing Retroreflectivity of STOP Signs

5. Report Date
December 2007

6. Performing Organization Code

7. Author(s)
Persaud, Bhagwant; Craig Lyon; Kimberly Eccles; Nancy Lefler; and Roya Amjadi

8. Performing Organization Report No.

9. Performing Organization Name and Address

Vanasse Hangen Brustlin, Inc (VHB)
8300 Boone Blvd., Ste. 700 
Vienna, VA 22182-2626     
Persaud Lyon, Inc
87 Elmcrest Road
Toronto, Ontario M9C 3R7
Canada

10. Work Unit No.

11. Contract or Grant No.
DTFH61-05-D-00024 (VHB)

12. Sponsoring Agency Name and Address

U.S. Department of Transportation
Federal Highway Administration
Office of Safety - HSST
1200 New Jersey Avenue, SE
Washington, DC 20590

13. Type of Report and Period

Safety Evaluation
Covered 1997- 2004

14. Sponsoring Agency Code

FHWA

15. Supplementary Notes
The FHWA (Office of Safety Research and Development) managed this study. The project team members were Kim Eccles, Nancy Lefler, Dr. Hugh McGee, Dr. Frank Gross, Dr. Forrest Council, Ram Jagannathan, Dr. Bhagwant Persaud, Craig Lyon, Dr. Raghavan Srinivasan, and Daniel Carter. The FHWA Office of Safety Research and Development Contract Task Order Manager was Roya Amjadi.

16. Abstract
The Federal Highway Administration (FHWA) organized a Pooled Fund Study of 26 States to evaluate low-cost safety strategies as part of its strategic highway safety effort. One of the strategies chosen to be evaluated for this study was STOP signs with increased retroreflectivity. This strategy is intended to reduce the frequency of crashes related to driver unawareness of stop control at unsignalized intersections.

Geometric, traffic, and crash data were obtained at unsignalized intersections for 231 sites in Connecticut and 108 sites in South Carolina. In each case, the strategy was implemented as a blanket application of STOP signs with increased retroreflectivity. Empirical Bayes (EB) methods were incorporated in a before-after analysis to determine the safety effectiveness of increasing the sign retroreflectivity. There was a statistically significant reduction in rear-end crashes in South Carolina. Based on the results of the disaggregate analysis, reductions in crashes were found at three-legged intersections and at intersections with low approach volumes. The analysis also indicated a slight reduction in nighttime- and injury-related crashes in Connecticut and South Carolina, but the results were not statistically significant. It was determined that a sample size much larger than that available would be needed to detect a significant effect in these types of crashes. Given the very low cost of installing STOP signs with increased retroreflectivity, even with conservative assumptions, only a very modest reduction in crashes is needed to justify their use. Therefore, this strategy has the potential to reduce crashes cost-effectively, particularly at lower volume intersections.

17. Key Words: Retroreflectivity, STOP signs, Low cost, Safety improvements, Safety evaluations, Empirical Bayesian, Unsignalized intersections

18. Distribution Statement
No restrictions.

19. Security Classif. (of this report)
Unclassified

20. Security Classif. (of this page)
Unclassified

21. No. of Pages
38

22. Price

Documentation Page

Form DOT F 1700.7 (8-72)                                                                                              Reproduction of completed pages authorized


SI* (Modern Metric) Conversion Factors


TABLE OF CONTENTS

Executive Summary

Introduction

Objective

Study Design

Methodology

Data Collection

Development of SPFs

Results

Economic Analysis

Summary

Conclusion

Appendix A: South Carolina Safety Performance Functions (SPFs)

Appendix B: Connecticut SPFS

Acknowledgements

References

LIST OF FIGURES

Figure 1. Relative Visual Comparison of Sheeting Types.

LIST OF TABLES

 

ABBREVIATIONS AND SYMBOLS

Abbreviations

Injury, incapacitating
AADT
Average annual daily traffic
ADT
Average daily traffic
AASHTO
American Association of State Highway Transportation Officials
ANOVA
Analysis of variance
ASTM
American Society for Testing and Materials
B
Injury, nonincapacitating
C
Possible injury
ChiSq
Chi-Squared
ConnDOT
Connecticut Department of Transportation
DF
Degrees of freedom
EB
Empirical Bayes
FHWA
Federal Highway Administration
ft
Feet
HPMS
Highway Performance Monitoring System
KABCO
Scale used to represent injury severity in crash reporting
K
Fatality
mi
Miles
NCHRP
National Cooperative Highway Research Program
NHTSA
National Highway Traffic Safety Administration
O
Property damage only
Pr
Probability
SCDOT
South Carolina Department of Transportation
SPF
Safety performance functions
Stddev
Standard deviation
TRB
Transportation Research Board
UCONN
University of Connecticut
Var
Variance

 

Symbols

alpha Greek letter Alpha
beta Greek letter Beta
delta Greek letter Delta
lambda Greek letter Lamda
pi Greek letter Pi
theta Greek letter Theta

 

Executive Summary

The Federal Highway Administration (FHWA) organized a Pooled Fund Study of 26 States to evaluate low-cost safety strategies as part of its strategic highway safety effort. The purpose of the FHWA Low-Cost Safety Improvements Pooled Fund Study is to evaluate the safety effectiveness of several of the low-cost strategies through scientifically rigorous crash-based studies. One of the strategies chosen to be evaluated for this study was the implementation of STOP signs with higher retroreflectivity. This strategy is intended to reduce the frequency of crashes related to driver unawareness of stop control at unsignalized intersections. The safety effectiveness of this strategy had not previously been thoroughly documented and this study is an attempt to provide an evaluation through scientifically rigorous procedures.

Geometric, traffic, and crash data were obtained at unsignalized intersections for 231 sites in Connecticut and 108 sites in South Carolina. In each case, the strategy was implemented as a blanket application of STOP signs with increased retroreflectivity. Empirical Bayes (EB) methods were incorporated in a before-after analysis to determine the safety effectiveness of increasing the sign retroreflectivity. For rear-end crashes, there was a statistically significant reduction in crashes in South Carolina. Based on the results of the disaggregate analysis, reductions in crashes were found at three-legged intersections and at intersections with low approach volumes. Installations at three-legged intersections (indiscriminate of urban/rural factor) and three-legged urban intersections in South Carolina were found to have a statistically significant reduction in crashes. In Connecticut, a statistically significant reduction in crashes was found for three-legged rural intersections. The disaggregate analysis also showed that the strategy is more effective at lower volumes for the minor approach. A statistically significant reduction in crashes was found at intersections with approaching volumes of less than 1,200 annual average daily traffic (AADT) in South Carolina and less than 1,000 AADT in Connecticut. The analysis indicated a slight reduction in nighttime- and injury-related crashes in Connecticut and South Carolina, but the results were not statistically significant. It was determined that a sample size much larger than that available would be needed to detect a significant effect in these types of crashes. Given the very low cost of this strategy, even with conservative assumptions, only a very modest reduction in crashes is needed to justify its use. Therefore, this strategy has the potential to reduce crashes cost effectively, particularly at lower volume intersections.

Introduction

Background on Strategy

Intersections account for a small portion of the total highway system, yet in 2005, approximately 2.5 million intersection-related crashes occurred, representing 41 percent of all reported crashes. Intersection-related crashes account for more than 50 percent of total crashes in urban areas and over 30 percent of total crashes in rural areas. In addition, 8,655 fatal crashes (22 percent of the total 39,189 fatal crashes) occurred at or within an intersection environment in 2005.(1) The high frequency of crashes is not surprising, however, due to the fact that intersections present more points of conflict than non-intersections.

Unsignalized intersections often present potential hazards not associated with signalized intersections because of the priority of movement on the major roadway. This is often problematic on two-lane highways. Unsignalized intersections are usually found along low- to moderate-volume roads in rural and suburban areas that are generally associated with higher-speed travel than those in more developed suburban and urban areas.(2)

Driver compliance with intersection traffic control devices is vital to intersection safety. At stop-controlled intersections, drivers on the stop-controlled approach must identify and observe the STOP sign. Therefore, the STOP sign must be visible and conspicuous. This is particularly important during nighttime or other reduced visibility conditions such as rainy weather. One method to increase both the visibility and conspicuity of STOP signs is to use higher retroreflectivity sheeting.

Retroreflectivity is the property of a material that reflects a large portion of the light directly back to the source, through a wide range on angles of incidence of illumination. When applied to a sign, retroreflective sheeting will redirect light from the driver's headlights back to the driver's eyes. The amount of light from an object reaching the driver's eyes will have a great impact on the ability of a driver to see that object. Retroreflective materials use micro-sized glass beads, either enclosed or encapsulated, or microprisms (cube corner reflectors) in the sign sheeting material. Variations in the technology result in differing levels of retroreflectivity. A higher retroreflectivity measure will return a greater amount of light to the driver's eye at night, hence making the retroreflective object more visible.(3) While the difference in sign brightness (retroreflectivity) provided by different sheeting types cannot be illustrated adequately by photography, figure 1 does provide a relative visual comparison of STOP signs with six different grades of retroreflective sheeting. The American Society for Testing and Materials ( ASTM) develops technical standards for industry worldwide. This includes retroreflective sheeting which is included in ASTM's D495—Standard Specification for Retroreflective Sheeting and Traffic Control.(4) ASTM Type I and II are commonly known as Engineering Grade and Super-Engineering Grade, respectively. Both are made with glass bead compositions. ASTM Type III, commonly known as High Intensity, is made with an encapsulated glass bead technology, while Types VII, VIII, and IX are manufactured with microprismatic technology.(4)

Figure 1. Image. Relative visual comparison of sheeting types. The image shows 6 variations on retroreflective sheeting on STOP signs. The samples are numbered 1, 2, 3, 7, 8, and 9 in Roman numerals. Number 1 is the least retroreflective (most difficult to see) and number 9 is the most retroreflective (easiest to see).

Figure 1. Image. Relative Visual Comparison of Sheeting Types.

The strategy to change to STOP signs with higher retroreflectivity was implemented in Connecticut from 1998 to 2001 and in South Carolina from 1997 to 2004 in an effort to reduce crashes at unsignalized intersections.

Background on Study

In 1997, the American Association of State Highway and Transportation Officials (AASHTO) Standing Committee for Highway Traffic Safety, with the assistance of the FHWA, the National Highway Traffic Safety Administration (NHTSA), and the Transportation Research Board (TRB) Committee on Transportation Safety Management, met with safety experts in the field of driver, vehicle, and highway issues from various organizations to develop a strategic plan for highway safety. These participants developed 22 key areas that affect highway safety. One of these areas is unsignalized intersection crashes.

The National Cooperative Highway Research Program (NCHRP) published a series of implementation guides to advance the implementation of countermeasures targeted to reduce crashes and injuries. Each guide addresses one of the 22 emphasis areas and includes an introduction to the problem, a list of objectives for improving safety in that emphasis area, and strategies for each objective. Each strategy is designated as proven, tried, or experimental. Many of the strategies discussed in these guides have not been rigorously evaluated; about 80 percent of the strategies are considered tried or experimental.

The FHWA organized a Pooled Fund Study of 26 States to evaluate low-cost safety strategies as part of this strategic highway safety effort. The purpose of the Pooled Fund Study is to evaluate the safety effectiveness of several tried and experimental low-cost safety strategies through scientifically rigorous crash-based or simulation-based studies. Based on inputs from the Pooled Fund Study Technical Advisory Committee and the availability of data, installing higher retroreflective STOP signs was selected as a strategy that should be evaluated as part of this effort.

Literature Review

The literature review did not uncover any studies that specifically evaluated the safety effects, in terms of crash frequency and severity measures, of increasing retroreflectivity levels of STOP signs. There has been research, however, that shows increased driver visibility distance provided by increased retroreflectivity levels. This research includes a study by Carlson and Hawkins, which investigated the legibility effects of increasing the retroreflectivity of freeway guide signs.(5) In this study, ASTM Type III and Type IX retroreflective sheeting were analyzed. A total of 60 subjects, both young and old, participated in this nighttime study. The measure of effectiveness used in this study was legibility distance. The statistical test used was a mixed-factor repeated-measures analysis of variance (ANOVA). The ANOVA test indicated that sheeting type was statistically significant (F1,116 = 34.69, p-value < 0.0001). The improvement associated with increasing the retroreflectivity was nearly a 10-percent (16.2-m (53.0-ft)) increase in visibility distance. The link between visibility and crashes has not been established; therefore, no safety inference can be made from this finding.

Objective

This research examined the change in crash frequency due to increasing the retroreflectivity of STOP signs at unsignalized intersections. The desired objective was to identify sites with crashes related to poor visibility due to the retroreflectivity of the STOP sign in the before period and estimate the expected change in crashes due to increasing the reflectivity using the EB method. While this is a worthy objective, it was not possible to determine those sites that had poor retroreflectivity in the before period because this was a blanket strategy. Although the type of sign (Type I, Engineer Grade) used in the before period was known, the exact condition including age of the signs or any degree of deterioration that occurred on each of the signs was not known. In addition, there were very few nighttime crashes, which made it difficult to identify a sufficient sample of sites that had crashes related to low retroreflectivity. Therefore, the objective was modified to estimate, in general, the safety effectiveness of this strategy as measured by crash frequency. Target crash types considered included the following:

  • All intersection-related crash types.
  • Injury crashes (includes K, A, and B on KABCO scale).
  • Right-angle (side impact) crashes.
  • Rear-end crashes.
  • Daytime crashes.
  • Nighttime crashes.

The range of safety effects was expected to vary by crash type; therefore, a second objective was to estimate, if necessary, the overall effect of the strategy by considering the economic costs by crash type and crash severity using crash costs recently developed for FHWA.(6)

A further objective was to address questions of interest such as:

  • Do effects vary by traffic volumes?
  • Do effects vary by approach speeds?
  • Do effects vary by number of lanes?

Meeting these objectives placed some special requirements on the data collection and analysis tasks. These were:

  • The need to select a large enough sample size to detect, with statistical significance, what may be small changes in safety for some crash types.
  • The need to properly account for traffic volumes changes.
  • The need to pool data from more than one jurisdiction to improve reliability of the results and facilitate broader applicability of the products of the research.

Study Design

The study design involved a sample size analysis and the prescription of needed data elements. The sample size analysis assessed the size of sample required to statistically detect an expected change in safety. Assumptions on the expected safety effects, on the average crash frequency at potential strategy sites in the before period, and on the average number of after period years of available data are basic to estimating sample sizes. Following a literature review and the application of methodology in Hauer, a minimum sample size was estimated.(7)

For this analysis, it was assumed at the time that the study was designed that a conventional before-after study with comparison group design would be used, since available sample size estimation methods are based on this assumption. To facilitate the analysis, it was also assumed that the number of comparison sites is equal to the number of strategy sites. The sample size estimates provided would be conservative in that state-of-the-art EB before-after methodology actually proposed for the evaluations would require fewer sites.

Sample sizes were estimated for various assumptions of likely safety effect and crash frequencies before the strategy was installed. Table 1 provides the crash frequency assumptions used. Rate A is based on a Minnesota study.(8) Rate B is based on an Ohio Study.(9) Rate C is based on Minnesota data from FHWA-RD-03-0037.(10) Right-angle and rear-end proportions were adopted from SafetyAnalyst development data.(11) The literature review provided no sound basis for an assumption on the expected safety effect. Thus, the analysis was based on logical values for this parameter.

Table 1. Before Period Crash Rate Assumptions.

Crash Type

Rate A (crashes/ intersection/year)

Rate B (crashes/ intersection/year)

Rate C (crashes/ intersection/year)

All

3.45

7.62

0.44

Right-Angle (39% of total assumed)

1.35

2.97

0.17

Rear-End (23% of total assumed)

0.79

1.75

0.10

Table 2 provides estimates of the required number of before period intersection-years in the sample for both the 90-percent and 95-percent confidence levels. The calculations assume equal number of intersection-years for strategy and comparison sites and equal length of before and after periods. Intersection-years are the number of intersections where the strategy was applied multiplied by the number of years the strategy was in place at each intersection. For example, if a strategy was applied at nine intersections and has been in place for three years at all 9 intersections, this is 27 intersection-years.

A minimum sample size of 1,076 intersection-years and a desirable sample size of 2,036 intersection-years per period were calculated as shown in bold in table 2. It was expected that these sample sizes could be reduced if the assumption for crashes per intersection-year before strategy implementation turns out to be conservatively low for strategy data, or if more after period years than assumed are available. The desirable sample assumes that the reduction in crashes could be as low as a 10-percent reduction in all crashes and that this is the smallest benefit that one would be interested in detecting with 90-percent confidence. The logic behind this approach is that safety managers may not wish to implement a measure that reduces crashes by less than 10 percent, and the required sample size to detect a reduction smaller than 10 percent would likely be prohibitively large. The minimum sample indicates the level for which a study seems worthwhile (i.e., it is feasible to detect with 90-percent confidence the largest effect that may reasonably be expected based on what is known currently about the strategy). In this case, a 20-percent reduction in right-angle crashes was assumed as this upper limit on safety effectiveness.

These sample sizes could be reduced if the assumption for crashes per intersection-year before strategy implementation turned out to be conservatively low for strategy data or if there are more after period years of data available than assumed.

Table 2. Minimum Required Before Period Intersection-Years for Treated Sites.

Expected Percent Reduction in Crashes

95% Confidence

90% Confidence

A

B

C

A

B

C

All

5

1,629

738

12,773

1,141

516

8,943

10

371

168

2,907

260

118

2,036

20

76

34

594

53

24

416

30

27

12

211

19

9

147

40

12

5

92

8

4

64

Right Angle

5

4,163

1,892

33,060

2,915

1,325

23,146

10

948

431

7,525

663

302

5,268

20

194

88

1,537

135

62

1,076

30

69

31

545

48

22

381

40

30

14

237

21

10

166

Rear End

5

7,114

3,212

56,203

4,981

2,249

39,349

10

1,619

731

12,793

1,134

512

8,956

20

331

149

2,612

232

105

1,829

30

117

53

926

82

37

648

40

51

23

403

36

16

282

Note: Bold denotes the calculated minimum and desirable sample size for intersection-years per period.

Methodology

The EB methodology for observational before-after studies was used for the evaluation.(7) This methodology is rigorous in that it accomplishes the following:

  • It properly accounts for regression-to-the-mean.
  • It overcomes the difficulties of using crash rates in normalizing for volume differences between the before and after periods.
  • It reduces the level of uncertainty in the estimates of safety effect.
  • It provides a foundation for developing guidelines for estimating the likely safety consequences of contemplated strategy.
  • It properly accounts for differences in crash experience and reporting practice in amalgamating data and results from diverse jurisdictions.

In the EB approach, the change in safety for a given crash type at a site is given by:       

  Equation 1. Change in safety. Change in safety equals Lambda minus Pi.
(1)

Where:

lambda is the expected number of crashes that would have occurred in the after period without the strategy.
pi is the number of reported crashes in the after period.

In estimating lambda, the effects of regression-to-the-mean and changes in traffic volume were explicitly accounted for using safety performance functions (SPFs) relating crashes of different types to traffic flow and other relevant factors for each jurisdiction based on untreated sites. Annual SPF multipliers were calibrated to account for the temporal effects on safety of variation in weather, demography, crash reporting, and so on.

In the EB procedure, the SPF is used to first estimate the number of crashes that would be expected in each year of the before period at locations with traffic volumes and other characteristics similar to the one being analyzed. The sum of these annual SPF estimates (P) is then combined with the count of crashes (x) in the before period at a strategy site to obtain an estimate of the expected number of crashes (m) before strategy. This estimate of m is:

  Equation 2. M equals the product. m equals the product of w sub 1 and x plus the product of w sub 2 and P.
(2)

Where:

w1 and w2   are estimated from the mean and variance of the SPF estimate as:

 

  Equation 3. W sub 1. w sub 1 equals P divided by the quantity of P plus the quotient of 1 divided by k.
(3)

 

  Equation 4. W sub 2, w sub 2 equals the inverse of the product of k and x, where x equals P plus the inverse of k.
(4)

Where:

k is a constant for a given model and is estimated from the SPF calibration process with the use of a maximum likelihood procedure. (In that process, a negative binomial distributed error structure is assumed with k being the dispersion parameter of this distribution.) 

A factor is then applied to m to account for the length of the after period and differences in traffic volumes between the before and after periods. This factor is the sum of the annual SPF predictions for the after period divided by P, the sum of these predictions for the before period. The result, after applying this factor, is an estimate oflambda. The procedure also produces an estimate of the variance of lambda.

The estimate of lambda is then summed over all sites in a strategy group of interest (to obtain lambdasum) and compared with the count of crashes during the after period in that group (pi sum). The variance of λ is also summed over all sites in the strategy group.

The Index of Effectiveness (theta) is estimated as:

  Equation 5. Theta equals x, Theta equals x divided by y, where x equals sum of the Pi divided by the sum of the Lambdas, and y equals 1 plus the quotient of the variance of the sum of the Lambdas divided by the sum of the Lambdas squared.
(5)

The standard deviation of theta is given by:

  Equation 6. The standard deviation of Theta. The standard deviation of Theta equals the square root of the product of Theta squared and a plus b divided by the quantity of 1 plus b, where a equals the quotient of variance of sum of Pi divided by sum of Pi squared, and b equals variance of sum of Lambda divided by sum of Lambda squared.
(6)

The percent change in crashes is calculated as 100(1-theta); thus a value of theta= 0.7 with a standard deviation of 0.12 indicates a 30-percent reduction in crashes with a standard deviation of 12 percent.

Data Collection

A survey was conducted to collect data for several low-cost strategies. Two States, Connecticut and South Carolina, responded that they had installed a large number of STOP signs with increased retroreflectivity as a blanketed effort across the State to improve safety. In addition to the locations and dates of the STOP signs, additional data including roadway geometry, traffic, and crash data were collected in order to conduct the evaluation. This section provides a summary of the data assembled for the analysis.

Connecticut Data Collection

Background

The Connecticut Department of Transportation (ConnDOT) replaced over 7,000 STOP signs (R1-1) on State highways and town roads approaching State highways from December 1998 to May 2001. The signs were replaced as part of a comprehensive replacement program. The overall motivation for the effort was traffic safety. However, individual locations were not selected based on crash experience. Instead, the replacement was a blanketed effort at all stop-controlled intersections.

The existing STOP signs were made up of Type I, Engineer Grade reflective sheeting. The exact condition before replacement of each of the signs, including the age of the signs and the degree of deterioration, was unknown. The sheeting was upgraded to a material that provides relatively high retroreflectivity at large observational angles, which was designated by the ASTM as Type IX sheeting. (At the time the signs were installed there was not a Type IX reflective sheeting ASTM designation.)

ConnDOT provided installation data, roadway, and traffic data for use in this study. The data collected were entered into a database, designed specifically for use in this evaluation, and matched to crash data supplied by the University of Connecticut (UCONN). Details on the data are provided in the following sections.

Installation Data

The installation data provided by ConnDOT contained the town name, route number, intersecting road name, size of the sign, and the date the sign was replaced. Of the 7,000 sign locations provided by ConnDOT, 231 intersections were included in the evaluation. The primary motivation for selecting these 231 intersections was the availability of traffic volume data. This is discussed in the section on traffic data.

Of the intersections used in the evaluations, 762-mm (30-inch) STOP signs were installed at 218 intersections, 1,219.2-mm (48-inch) STOP signs were installed at 11 intersections, and a combination of 762-mm (30-inch) and 1,219.2-mm (48-inch) STOP signs were installed at the remaining 2 intersections.

Roadway Data

ConnDOT provided access to a 2004 electronic photo log of the roadway system. Roadway data were collected for each intersection from the photo log. This included the intersection log mile, number of intersection approaches, number of roadway lanes per approach, presence of a shoulder on each approach, presence of a median on each approach, presence of other warning measures (e.g., STOP AHEAD warning signs), and intersection illumination.

There was a concern that some of the STOP signs used in the evaluation had received subsequent strategies such as a signal or a flashing beacon. Based on the photo logs, 18 intersections were identified that had been signalized or a flashing beacon had been installed since the STOP signs were replaced. A list of energized signals, provided by ConnDOT, helped to identify other intersections that had received a signal, and the signalization date. ConnDOT also provided the dates of the flashing beacons installations.

Traffic Data

The primary reason many intersections were excluded from the evaluation was the lack of traffic volume data. In order to be included in the evaluation, traffic volume on the major road was needed both before and after the sign was replaced. In addition, there had to be a traffic count on the minor roadway in at least the before or after period, although both were preferred.

Volume data are available from three sources in Connecticut: average daily traffic (ADT) maps, electronic count data in a spreadsheet format, and special counts. The ADT maps are available in both hard copy and electronic (.pdf) formats from 1999 to 2004. Traffic counts are conducted every three years to develop these maps. The count locations vary from year to year; not all locations were counted on each map. The electronic count data contained the same information as the ADT maps but for a longer period. Electronic count data are available from 1995 to 2006. For the purposes of extrapolating counts from nearby intersections, spatial count maps (i.e., the ADT maps) are preferable to tabular count data.

The third source of volume data are special counts. ConnDOT provided paper copies of special counts. These are volume counts that are requested for a variety of reasons including signalization studies, citizen complaints, and traffic operations analysis.

The stop-controlled roadway was considered the minor roadway for this study. In most cases, this was also the lower volume roadway. There were a few three-legged intersections where the stop-controlled approach had a higher ADT than the nonstop-controlled approach. Therefore, there were a few intersections in the database where the minor roadway had a higher ADT than the major roadway.

Crash Data

The Connecticut Transportation Institute at the UCONN provided crash data from 1995 to 2004 for this study. These data were originally provided to UCONN by the ConnDOT. UCONN formatted the raw data into a more user-friendly version. These data included all crashes on State-maintained roadways and crashes on non-State-maintained roadways that occurred within 0.02 km (0.01 mi) of an intersection with a State-maintained roadway. Therefore, all intersections in this study included at least one State-maintained roadway.

During the evaluation, it was discovered that there were log-mile changes throughout the study period. That is, the same intersection could have two different log miles in two different years. This was due to changes in the Connecticut roadway system. ConnDOT supplied a file of where log mile changes have occurred. These were used to resolve the log mile changes.   

South Carolina Data Collection

Background

District 1 of the South Carolina Department of Transportation (SCDOT) conducted a comprehensive replacement of over 6,000 STOP signs from 1997 to 2005. District 1 is located in central South Carolina and is comprised of Aiken, Kershaw, Lee, Lexington, Richland, and Sumter Counties. Data from Kershaw and Lexington counties were used for this study.

The existing signs were made of Type I Engineer Grade reflective sheeting. The exact age and condition of the signs prior to replacement was unknown. They were replaced with signs that had Type III high-intensity reflective sheeting. The signs were replaced as part of a comprehensive replacement program.

SCDOT provided installation data, roadway, traffic, and crash data for use in this study. The data collected were entered into a database, designed specifically for use in this evaluation, and matched to crash data. Details on the data are provided in the following sections.

Installation Data

SCDOT provided a database of sign installations. For each sign, the database included the county, route, milepoint, direction, installation date, and sheeting type. Out of the more than 6,000 signs, 108 intersections were used in the evaluation. As with the Connecticut data, this was largely due to the availability of the traffic data. Of the 108 intersections, 93 had 762-mm (30-inch) STOP signs installed, and 15 had 1,219.2-mm (48-inch) STOP signs installed.

Roadway Data

SCDOT provided a copy of the roadway and traffic data that were collected for the Highway Performance Monitoring System (HPMS). The HPMS is a national highway information system that includes data on the extent, condition, performance, use, and operating characteristics of the Nation's highways. This database provided data on the land use (i.e., urban or rural), number of lanes, lane width, presence of a shoulder, shoulder width, presence of a median, and median type for each intersection approach. The speed limit was also available for 35 of the intersections.

Traffic Data

The majority of the traffic volumes used for this study were extracted from the HPMS files. These values came from a segment of roadway ranging from 0.16 to 8.05 km (0.1 to 5.0 mi) or more in length. The range was less in urban areas and greater in rural areas. Therefore, in rural areas, the volume count used to describe the volume entering the intersection may be collected from a point up to 8.05 km (5 mi) from the intersection. There were no records of where in the segment the count was actually collected.

For those locations where AADTs were not available through the HPMS, the AADT numbers were calculated from turning movement counts. Factors supplied by SCDOT were used to calculate the AADTs from the raw turning movement count data.

Crash Data

SCDOT supplied crash data in two databases. One database contained crashes occurring from 1994 to 2000. The second database contained crashes occurring from 2001 to 2005. The second database system was necessary because the crash data report and some associated variables were modified in 2001. In addition, prior to 1997, there was no threshold on reporting property damage only crashes. Starting in 1997, only crashes involving an injury or property damage greater than $1,000 were reported in the system.

Based on guidance from the SCDOT, the crash milepost was not used to locate crashes. Instead, the variable "base-offset distance" was used to identify crashes occurring at intersections.

Summary of Data

The analysis included a total of 3,323.8 intersection-years of data (2,038.6 intersection-years from CT and 1,285.2 intersection-years from SC). This sample was greater than the 1,076 intersection-years estimated in the study design required to detect a 20-percent reduction in right-angle crashes and the 2,036 intersection-years required to detect a 10-percent reduction in all crashes.

Table 3 provides crash definitions used in the two States. This information is crucial in applying the safety effect estimates in other jurisdictions.

Table 3. Definitions of Crash Types.

State

Intersection-Related

Injury

Right-Angle

Rear-End

Daytime

Nighttime

SC

Within 264 ft of intersection

K, A or B on KABCO scale

Defined as angle

Defined as rear-end

Daylight, dawn, dusk

Dark

CT

Within 264 ft of intersection, within
0.01 mi on minor

K, A or B on KABCO scale

Defined as angle or turning-intersecting paths

Defined as rear-end

Daylight, dawn, dusk

Dark

1 ft = 0.305 m
1 mi = 1.61 km

Table 4 and table 5 provide summary information for the data collected. This information should not be used to make simple before-after comparisons of crashes per site-year since such an analysis would not account for factors other than the strategy that may cause safety to change between the two periods. Such comparisons are properly done with the EB analysis as presented in subsequent sections.

Table 4. Data Summary for South Carolina Sites (n = 108).

Variable

Mean

Minimum

Maximum

Months before

100.7

45.0

144.0

Months after

42.1

2.0

99.0

Crashes/site-year before

2.1

0.0

16.1

Crashes/site-year after

2.0

0.0

18.9

Injury crashes/site-year before

0.7

0.0

3.8

Injury crashes/site-year after

0.6

0.0

6.0

Right-angle crashes/site-year before

0.8

0.0

7.1

Right-angle crashes/site-year after

0.7

0.0

6.4

Rear-end crashes/site-year before

0.7

0.0

7.1

Rear-end crashes/site-year after

0.7

0.0

12.9

Daytime crashes/site-year before

1.7

0.0

13.9

Daytime crashes/site-year after

1.6

0.0

15.4

Nighttime crashes/site-year before

0.4

0.0

2.5

Nighttime crashes/site-year after

0.4

0.0

3.7

Major road AADT before

9,847

413

53,587

Minor road AADT before

2,017

218

7,970

Major road AADT after

10,414

344

57,353

Minor road AADT after

2,139

206

9,178

 

Table 5. Data Summary for Connecticut Sites (n = 231).

Variable

Mean

Minimum

Maximum

Months before

59.7

48.0

84.0

Months after

46.2

3.0

60.0

Crashes/site-year before

1.9

0.0

18.9

Crashes/site-year after

2.4

0.0

32.0

Injury crashes/site-year before

0.7

0.0

5.9

Injury crashes/site-year after

0.8

0.0

6.7

Right-angle crashes/site-year before

0.5

0.0

3.6

Right-angle crashes/site-year after

0.6

0.0

4.0

Rear-end crashes/site-year before

0.6

0.0

10.6

Rear-end crashes/site-year after

0.9

0.0

11.8

Daytime crashes/site-year before

1.4

0.0

14.1

Daytime crashes/site-year after

1.8

0.0

22.0

Nighttime crashes/site-year before

0.5

0.0

5.1

Nighttime crashes/site-year after

0.6

0.0

10.0

Major road AADT before

7,690

929

29,816

Minor road AADT before

2,033

68

18,074

Major road AADT after

8,021

969

31,267

Minor road AADT after

2,122

71

18,879

Development of SPFs

This section presents the SPFs that were developed for use in the EB methodology. Generalized linear modeling was used to estimate model coefficients using the software package SAS® and assuming a negative binomial error distribution, which is consistent with the state of research in developing these models.

SPFs were calibrated separately for South Carolina and Connecticut. The reference groups used to develop SPFs were the same as the strategy groups since the installations were blanketed across the jurisdictions. The approach taken was as follows:

  1. Combine the before and after period data to develop SPFs.
  2. Recalibrate each SPF separately for the before and after periods to develop yearly multipliers.

Since the installations were over a multiyear period, it was possible to represent yearly trends in crash counts in an unbiased way that would not be possible if all installations occurred in the same year.

The primary form of the SPFs is:

  Equation 7. Crashes per year. Crashes per year equals Alpha times x times y, where x equals the major AADT to the power of Beta one, and y equals minor AADT raised to the power of Beta two.    
(7)

Where:

maj is major road entering AADT.
min is minor road entering AADT.
alpha, beta1 and beta2 are parameters estimated from data in the SPF calibration process.

In some cases, the separate exponents could not be estimated with significance and the following Safety Function (SF) form was used:

  Equation 8. Crashes per year, second version. Crashes per year equals Alpha times x, where x equals total AADT raised to the power of Beta zero.    
(8)

Where:

AADT is the total entering AADT.
Beta 0 is a parameter estimated from data in the SPF calibration process.

Using additional variables did not significantly improve the models. In specifying a negative binomial error structure, the "dispersion" parameter, k, which relates the mean and variance of the SPF estimate and is used in equations 3 and 4 of the EB procedure, is iteratively estimated from the model and the data. The value of k is such that the smaller its value, the better a model is for a given set of data.

The SPFs developed are presented in appendix A. Note the following in interpreting the output:

  • The value of alphais obtained as the e superscript ln(alpha),where ln(alpha) is the model output.
  • The value of the parameter k is used in the EB approach.
  • The value for Pr > ChiSq gives the level at which the estimate is significant. For example, Pr > ChiSq = 0.05 indicates that the parameter estimate is statistically significant at the
    5-percent level (or, alternatively, that the 95-percent confidence interval does not include a value of 0).

SPFs were estimated for the following crash classifications:

  • Total (all severities and types combined).
  • Injury (all crash types combined).
  • Right angle (all severities combined).
  • Rear end (all severities combined).
  • Day (all severities and types combined).
  • Night (all severities and types combined).

Results

Based on the data, two sets of results were calculated and are presented in the following sections. One set contains aggregate results for each jurisdiction and for the two combined; the other set is based on a disaggregate analysis that attempted to discern factors that may be most favorable to increasing STOP sign retroreflectivity.

Aggregate Analysis

The aggregate results are shown in tables 6 through 8. Results significant at the 95-percent confidence level are bolded. Note that a negative sign indicates an increase in crashes.

The results indicate that there may be a slight effect in South Carolina, but this effect is too small to detect with statistical significance, as evidenced by the relatively large standard errors (i.e., substantially greater than one-half of the estimated effect). The exception is for rear-end crashes, for which the reduction in crashes is significant at the 95-percent confidence level as shown in table 6.

The aggregate effects are negligible and statistically insignificant for Connecticut, and for the two jurisdictions combined. There are no detectable effects for nighttime crashes, the primary targets of this measure, which is likely a result of the reality that there are relatively few of these crashes at the strategy sites.

These inconclusive results and the fact that they are based on non-selective implementations emphasize the need for a disaggregate analysis to see if significant effects can be detected for specific conditions. This analysis is presented in the next section.

Table 6. Results for 108 South Carolina Strategy Sites.

 

Right-Angle

Rear-End

Night

Day

Injury

Total

EB estimate of crashes expected in the after period without strategy

266.5

257.4

134.5

559.6

220.1

692.9

Count of crashes observed in the after period

247

213

141

515

200

656

Estimate of percent reduction

7.6%

17.5%

-4.4%

9.1%

9.4%

5.4%

Standard error

7.6

7.3

10.8

5.3

8.1

4.9

Notes: Bold denotes results significant at the 95% confidence level. The negative sign indicates an increase in crashes.

Table 7. Results for 231 Connecticut Strategy Sites.

 

Right-Angle

Rear-End

Night

Day

Injury

Total

EB estimate of crashes expected in the after period without strategy

483.3

663.6

510.8

1494.8

700.1

2,019.2

Count of crashes observed in the after period

512

729

478

1543

659

2025

Estimate of percent reduction

-5.8%

-9.7%

6.6%

-3.2%

6.0%

-0.2%

Standard error

6.2

5.7

5.5

3.6

4.8

3.1

Note: The negative sign indicates an increase in crashes.

 

Table 8. Combined Results for 339 South Carolina and Connecticut Strategy Sites.

 

Right-Angle

Rear-End

Night

Day

Injury

Total

EB estimate of crashes expected in the after period without strategy

749.8

921.0

645.3

2054.4

920.2

2712.1

Count of crashes observed in the after period

759

942

619

2058

859

2681

Estimate of percent reduction

-1.2%

-2.2%

4.4%

-0.1%

6.7%

1.2%

Standard error

5.3

4.8

6.0

2.7

4.5

2.7

Note: The negative sign indicates an increase in crashes.

Disaggregate Analysis

Table 9 presents the results of the disaggregate analysis. Nighttime crashes are the primary targets of this measure and should be the basis for this analysis; however, there are too few of these crashes to facilitate a disaggregate analysis. The results of the disaggregate analysis are based on all crashes combined. Significant results at the 95-percent confidence level are shown in bold.

The three factors that provided indications of an association with crash effects are environment (urban versus rural), number of approach legs, and minor road entering AADT.

Table 9. Results of the Disaggregate Analysis.

Intersection Type

Sites

EB estimate of crashes expected in the after period without strategy

Count of crashes observed in the after period

Estimate of percent reduction (standard error)

SC urban

47

333.9

288

13.7% (6.7)

SC rural

61

360.0

368

-2.0% (7.0)

SC three-legged

48

354.7

299

15.9% (6.3)

SC four-legged

60

338.2

357

-5.3% (7.4)

SC three-legged, urban

20

172.9

128

26.3% (8.3)

SC four-legged, urban

27

160.0

160

0.05% (10.6)

SC three-legged, rural

28

181.8

171

6.3% (9.4)

SC four-legged rural

33

178.2

197

10.2% (10.2)

CT urban

190

1,789.5

1,830

-2.2% (3.3)

CT rural

41

229.7

195

15.4% (8.1)

CT three-legged

172

1,458.0

1,399

4.1% (3.5)

CT four-legged

59

559.2

625

-11.6% (6.3)

CT three-legged, rural

29

152.6

118

23.1% (9.2)

CT four-legged, rural

12

75.2

76

-0.2% (15.8)

SC < 1200 minor AADT

42

219.0

165

24.9% (7.2)

SC > 1200 minor AADT

66

473.9

491

-3.4% (6.3)

CT < 1000 minor AADT

90

509.0

437

14.3% (5.6)

CT >1000 minor AADT

141

1,510.7

1,588

-5.1% (3.7)

Notes: Bold denotes results significant at the 95% confidence level. The negative sign indicates an increase in crashes.

For the urban versus rural factor, there are opposing indications from the two States, with the more favorable effects for rural installations in Connecticut and urban installations in South Carolina.

For number of approaches (i.e., legs), a more consistent pattern emerges. For both States, in particular for the favored environment, installations at three-legged intersections appear to be more effective than at four-legged intersections.

For minor entering AADT, there is a consistent pattern that this strategy is more effective at lower volumes. The boundaries of 1,200 AADT in South Carolina and 1,000 in Connecticut were chosen to provide the most discrimination between upper and lower AADT levels in order to indicate the effect of this factor. Therefore, these numbers should not be used in decisions on whether or not to install a sign. The analysis does suggest, however, that lower minor road locations should be given higher priority if there is a need to prioritize locations (as should three-legged intersections).

Speculation on reasons for the differential effects found is undertaken in a discussion section later in the report. However, it should be pointed out that further investigation was undertaken to ensure that the effects found were not due to biases in the analysis. This further investigation involved an examination of the results of a naïve before-after study that simply compared crash frequencies pre- and post-strategy and did not use safety performance functions. The naïve before-after study yielded similar conclusions to the EB study regarding the influence of the three factors, but different magnitudes for the crash effects for the various groupings in table 9. The project team also investigated whether the findings regarding the differential effects for one factor may have been confounded by co-linearity of this factor with another for which similar effects were found. For example, this investigation revealed that the conclusion regarding minor road AADT was equally relevant for three-legged and four-legged intersections and for urban and rural intersections. As shown in table 9, the finding regarding three-legged versus four-legged intersections is equally valid for urban and rural environments.

Data were available for an analysis of other possible factors that might influence crash effects. However, no such effects could be ascertained. The other factors examined were sign size (762 mm (30 inches) versus 1,219.2 mm (48 inches)), the presence of lighting (for Connecticut), the presence of other measures such as STOP AHEAD signs, the major road entering volume, and the expected crash frequency prior to strategy. For sign size, there were very few that were of the 1,219.2-mm (48-inch) variety and so, statistically, it was difficult to detect different crash effects for the two sign sizes, even if such differences exist.

Economic Analysis

The purpose of this analysis was to determine the economic feasibility of applying this strategy. The life-cycle costs of the strategy were estimated and expressed as an annual cost. The crash benefits required to offset these costs were estimated using the most recent FHWA unit crash cost data for unsignalized intersections. The results of the aggregate and disaggregate analysis of crash effects were used to make a judgment on the circumstances that would be favorable to ensuring economic feasibility (i.e., circumstances that may yield a benefit cost ratio of at least 2:1).

Cost data provided by the two States suggest a conservatively high initial cost of about $200 per intersection, considering the mix of three-legged and four-legged intersections and sign sizes. State sources also suggest an expected sign life of 8 years, again conservatively estimated. Costs would be even lower if the marginal costs of replacing the signs were used. As of 2007, the approximate costs of sheeting are as follows:

  • Type I sheeting is $0.75 per square foot.
  • Type II sheeting is $1.25 per square foot.
  • Types VII, VIII, and IX are $3.50 per square foot.

These would reflect the costs if a jurisdiction used higher retroreflective materials as part of its routine maintenance program, as opposed to replacing all of the existing signs across a jurisdiction at one time regardless of the condition of the existing signs, as was done in Connecticut and South Carolina.

Based on the Office of Management and Budget suggested discount rate of 7 percent, and on the expected service life (8 years), the initial costs per intersection were converted to annual costs using the standard economics formula for a capital recovery factor. The more conservative $200 initial cost translates into an annual cost of around $33 over the 8-year cycle, requiring an annual crash saving of more than $66 per intersection for a benefit cost ratio of at least 2:1.

The most recent FHWA mean comprehensive costs per crash for unsignalized intersections are $13,238 for rear-end and $61,114 for right-angle crashes.(6) Comprehensive crash costs represent the present value, computed at a discount rate, of all costs over the victim's expected life span that result from a crash. The major categories of costs used in the calculation of comprehensive crash costs include medical-related costs, emergency services, property damage, lost productivity, and monetized quality-adjusted life years.(6) By applying the more conservative figure, $13,238, a $66 saving would require a reduction of approximately 0.005 crashes per intersection per year. This is a reduction of approximately 0.5 percent for rural Connecticut intersections, which have an annual crash frequency of 1.11, the lowest of the four State/environment groups. This reflects the more conservative costs of replacing all existing signs across a jurisdiction at one time with signs with retroreflective material regardless of condition of the existing signs.

Even with the conservative assumptions made, just a very modest reduction in crashes is required to justify this strategy economically. The evidence suggests that this reduction is easily achievable, in particular, under the circumstances identified from the disaggregate analysis.

Summary

The objective of this study was to evaluate the safety effectiveness as measured by crash frequency of higher retroreflective sheeting on STOP signs at unsignalized intersections. The study was designed to detect a 10-percent reduction in all crashes with 90-percent confidence. The study also examined the effects of higher retroreflectivity on specific crash types. While it is desirable to evaluate the effectiveness of this strategy on related crashes (i.e., nighttime, low-visibility), there was not a sufficient number of related crashes to determine an effect with confidence.

The aggregate analysis indicates that higher retroreflective STOP signs may affect the likelihood of crashes at unsignalized intersections, but the effect is not detectable with the study design and available sample size. The exception is for rear-end crashes in South Carolina, where there was a significant reduction.

The disaggregate analysis provided further insight into the circumstances where crash reductions were identified. Installations at three-legged intersections (indiscriminate or urban/rural factor) and three-legged urban intersections in South Carolina were found to have a statistically significant reduction in crashes. In Connecticut, a statistically significant reduction in crashes was found for three-legged rural intersections. The disaggregate analysis also showed that the strategy is more effective at lower volumes for motorists approaching the intersection along the minor road. Statistically significant reduction in crashes were found at intersections with approaching volumes of less than 1,200 in South Carolina and less than 1,000 in Connecticut. This volume related finding is expected. At higher volume intersections, there are more visual cues for the approaching minor road motorist that the intersection is stop-controlled. Most notably, other traffic stopped in front of the driver on the approach is a visual cue.

For the urban versus rural factor, there are opposing indications from the two States, with the more favorable effects for rural installations in Connecticut and urban installations in South Carolina. There was no explanation available for these inconsistent results between the two States.

There are no detectable effects for nighttime crashes. As discussed previously, this might be because there are relatively few of these crashes at the strategy sites. It is also likely that this is because these are blanket installations and the significant benefits at relatively few nighttime crash problem locations become diluted by the negligible effects at other locations. To establish the benefits for nighttime crashes with statistical significance would require a database with a substantial number of sites at which this strategy was implemented because of a high frequency of nighttime crashes perceived to be "correctable" by this strategy. The sample size required for such a special database would be of a similar order of magnitude to that required for the database for the blanketed installations.

It should be noted, however, that the study results do not support the degradation of signs below any desired retroreflectivity requirements. The results are based on a blanket improvement with no knowledge of the previous sign conditions. This being the case, it is difficult to determine the safety effectiveness of more highly retroreflective sheeting on STOP signs for specific conditions. There was not a large enough sample size to detect any significant effects. The sample size required to detect a significant effect would be outside the scope of this project. As indicated in the FHWA Supplemental Notice of Proposed Amendments, improving sign retroreflectivity will be a benefit to all drivers, including older drivers.(12) All drivers need legible signs in order to make important decisions at key locations, such as intersections and exit ramps on high speed facilities. This is particularly true for regulatory and warning signs.

Conclusion

A minimal reduction in crashes can be expected with the installation of higher retroreflective STOP signs. However, given the very low cost of this strategy, even with conservative assumptions, only a very modest reduction in crashes is needed to justify their use. Therefore, this strategy has the potential to reduce crashes cost effectively, particularly at lower volume intersections.

Appendix A: Safety Performance Functions (SPFs)

Table 10. Total—All Severities.

Parameter

Rural

Urban

Estimate

Standard Error

Pr > ChiSq

Estimate

Standard Error

Pr > ChiSq

ln(α)

-7.9320

1.7537

<.0001

-10.5902

2.6984

<.0001

β0

 

 

 

 

 

 

β1

0.5990

0.1309

<.0001

0.6639

0.2299

0.0039

β2

0.4331

0.1652

0.0087

0.6460

0.2170

0.0029

k

0.6494

0.1253

 

1.1429

0.2385

 

Note: The negative sign indicates an increase in crashes.

Table 11. Injury—All Types.

Parameter

Rural

Urban

Estimate

Standard Error

Pr > ChiSq

Estimate

Standard Error

Pr > ChiSq

ln(αalpha)

-7.2994

2.0709

0.0004

-11.5587

2.7736

<.0001

β0

 

 

 

 

 

 

β1

0.4866

0.1491

0.0011

0.6890

0.2386

0.0039

β2

0.3401

0.1912

0.0753

0.5893

0.2091

0.0048

k

0.7322

0.1652

 

0.8016

0.2096

 

Note: The negative sign indicates an increase in crashes.

Table 12. Right—Angle—All Severities.

Parameter

Rural

Urban

Estimate

Standard Error

Pr > ChiSq

Estimate

Standard Error

Pr > ChiSq

ln(α)

-8.9674

2.1553

<.0001

-10.0403

3.0644

0.0011

β0

 

 

 

 

 

 

β1

0.6588

0.1516

<.0001

0.5194

0.2731

0.0572

β2

0.3689

0.2044

0.0711

0.6448

0.2642

0.0147

k

0.8566

0.1899

 

1.5344

0.3539

 

Note: The negative sign indicates an increase in crashes.

Table 13. Rear—End—All Severities.

Parameter

Rural

Urban

Estimate

Standard Error

Pr > ChiSq

Estimate

Standard Error

Pr > ChiSq

ln(α)

-15.3916

3.1187

<.0001

-16.4839

4.3121

0.0001

β0

 

 

 

 

 

 

β1

1.0693

0.2289

<.0001

1.0228

0.3342

0.0022

β2

0.6968

0.2495

0.0052

0.8386

0.2947

0.0044

k

1.2372

0.2744

 

1.7207

0.4234

 

Note: The negative sign indicates an increase in crashes.

Table 14. Day—All Severities and Types.

Parameter

Rural

Urban

Estimate

Standard Error

Pr > ChiSq

Estimate

Standard Error

Pr > ChiSq

ln(α)

-9.6967

1.9103

<.0001

-9.0236

2.9690

0.0024

β0

 

 

 

 

 

 

β1

0.7080

0.1413

<.0001

0.5947

0.2581

0.0212

β2

0.5015

0.1761

0.0044

0.5143

0.2401

0.0322

k

0.7423

0.1498

 

1.4078

0.2902

 

Note: The negative sign indicates an increase in crashes.

Table 15. Night—All Severities and Types.

Parameter

Rural

Urban

Estimate

Standard Error

Pr > ChiSq

Estimate

Standard Error

Pr > ChiSq

ln(α)

-3.8185

1.6073

0.0175

-7.2487

3.1635

0.0219

β0

0.3381

0.1728

0.0504

0.6526

0.3382

0.0536

β1

 

 

 

 

 

 

β2

 

 

 

 

 

 

k

0.6292

0.1496

 

1.0298

0.2840

 

Note: The negative sign indicates an increase in crashes.

APPENDIX B: CONNECTICUT SPFS

Table 16. Total—All Severities.

Parameter

Rural

Urban

Estimate

Standard Error

Pr > ChiSq

Estimate

Standard Error

Pr > ChiSq

ln(α)

-8.9117

2.7596

0.0012

-7.2564

1.1264

<.0001

β0

1.0156

0.3127

0.0012

 

 

 

β1

 

 

 

0.6607

0.1069

<.0001

β2

 

 

 

0.2883

0.0649

<.0001

k

1.1312

0.2959

 

1.1736

0.1312

 

Note: The negative sign indicates an increase in crashes.

Table 17. Injury—All Types.

Parameter

Rural

Urban

Estimate

Standard Error

Pr > ChiSq

Estimate

Standard Error

Pr > ChiSq

ln(α)

-7.8518

3.1177

0.0118

-8.6687

1.1143

<.0001

β0

0.7967

0.3529

0.0240

 

 

 

β1

 

 

 

0.6715

0.1096

<.0001

β2

 

 

 

0.3149

0.0628

<.0001

k

1.2010

0.3581

 

0.9519

0.1266

 

Table 18. Right—Angle—All Severities.

Parameter

Urban

Estimate

Standard Error

Pr > ChiSq

ln(α)

-7.1373

1.3875

<.0001

β0

 

 

 

β1

0.4303

0.1297

0.0009

β2

0.3670

0.0817

<.0001

k

1.2848

0.1734

 

Note: The negative sign indicates an increase in crashes.

A model for rural night, all severities and types, crashes could not be estimated. A proportion of 26.7 percent of total crashes was used; that is, a factor of 0.267 to the rural total crash SPF was applied.

Table 19. Rear—End—All Severities.

Parameter

Rural

Urban

Estimate

Standard Error

Pr > ChiSq

Estimate

Standard Error

Pr > ChiSq

ln(α)

-18.4262

3.7098

<.0001

-12.4637

1.3882

<.0001

β0

 

 

 

 

 

 

β1

1.1696

0.3556

0.0010

1.0315

0.1300

<.0001

β2

0.9914

0.2073

<.0001

0.3829

0.0746

<.0001

k

0.9103

0.3719

 

1.4777

0.1906

 

Note: The negative sign indicates an increase in crashes.

Table 20. Day—All Severities and Types.

Parameter

Rural

Urban

Estimate

Standard Error

Pr > ChiSq

Estimate

Standard Error

Pr > ChiSq

ln(α)

-10.0048

2.3582

<.0001

-7.8070

1.1413

<.0001

β0

 

 

 

 

 

 

β1

0.3965

0.2239

0.0766

0.6752

0.1085

<.0001

β2

0.8869

0.1495

<.0001

0.3056

0.0653

<.0001

k

0.6789

0.2349

 

1.1850

0.1359

 

Note: The negative sign indicates an increase in crashes.

Table 21. Night—All Severities and Types.

Parameter

Urban

Estimate

Standard Error

Pr > ChiSq

ln(α)

-7.9631

1.1996

<.0001

β0

 

 

 

β1

0.6405

0.1155

<.0001

β2

0.2202

0.0671

0.0010

k

1.0040

0.1377

 

Note: The negative sign indicates an increase in crashes.

A model for rural night, all severities and types, crashes could not be estimated. A proportion of 24.0 percent of total crashes was used; that is, a factor of 0.267 to the rural total crash SPF was applied.

ACKNOWLEDGEMENTS

This report was prepared by Vanasse Hangen Brustlin, Inc. (VHB) for the FHWA, Office of Safety under Contract DTFH61-05-D-00024. The current FHWA COTM for this project is Roya Amjadi. Kerry Perrillo Childress served as FHWA COTM from September 2005 until December 2006. Kimberly Eccles, P.E., of VHB was the study principal investigator. Dr. Bhagwant Persaud and Craig Lyon, subcontractors to VHB, conducted the analysis of the strategy and are the primary authors of the report. Nancy X. Lefler of VHB led the data collection for the study and is a supporting author. Other significant contributions to the study were made by Dr. Hugh McGee, Dr. Forrest Council, Michelle Scism, Nitesh Gupta, and Vicki Glenn, all of VHB.

The project team acknowledges the participation of the following organizations and individuals for their assistance in this study:

  • The Connecticut Department of Transportation, particularly Mr. John F. Carey, Mr. Al Iallonardo, Mr. Brad Overturf, Mr. Sebastian Puglisi, Ms. Kerry Ross, Mr. Gene Interlandi, and Ms. Julie Annino.
  • Connecticut Transportation Institute at the University of Connecticut (UCONN), particularly Mr. Thomas Jonsson.
  • The South Carolina Department of Transportation, particularly Ms. Terecia Wilson, Ms. Amelia Glisson, and Mr. Bryan Jones.

References

  1. Traffic Safety Facts. DOT HS 810 631, National Center for Statistics and Analysis of the National Highway Traffic Safety Administration, Washington, DC, 2005.
  2. Neuman, T.R., R. Pfefer, K. L. Slack, K. K. Hardy, D. W. Harwood, I.B. Potts, D.J. Torbic, E.R.K. Rabbini. Guidance for Implementation of the AASHTO Strategic Highway Safety Plan: A Guide for Addressing Unsignalized Intersection Collisions. NCHRP Report 500, Volume 5, Transportation Research Board, Washington DC, 2003.
  3. McGee, H. W. Chapter 7: Sign Materials in Traffic Signing Handbook. IR-092, Institute of Transportation Engineers, Washington, DC, 1997.
  4. ASTM International Standard Specification for Retroreflective Sheeting for Traffic Control. ASTM D4956-05, ASTM International, West Conshohocken, PA, 2007.
  5. Carlson, P.J. and Gene Hawkins. Legibility of Overhead Guide Signs with Encapsulated Versus Microprismatic Retroreflective Sheeting. No. 1844, Transportation Research Record, Washington, DC, 2003.
  6. Council, Forrest , E. Zaloshnja, T. Miller, B. Persaud. Crash Cost Estimates by Maximum Police-Reported Injury Severity Within Selected Crash Geometries. FHWA-HRT-05-051, Federal Highway Administration, McLean, VA, 2005. 
  7. Hauer, E. Observational Before-After Studies in Road Safety: Estimating the Effect of Highway and Traffic Engineering Measures on Road Safety. Pergamon Press, Elseviser Science Ltd., Oxford, U.K., 1997.
  8. Stackhouse, S; Cassidy, P. Warning Flashers At Rural Intersections. Report MN/RC-1998/01; Final Report, Minnesota DOT, Minneapolis, MN, 1996.
  9. Pant, P; Park, Y; Neti, S; Hossain, A. Comparative Study of Rural Stop-Controlled and Beacon-Controlled Intersections. No. 1692, Transportation Research Record, Washington, DC, 1999.
  10. Washington, S., B. Persaud, C. Lyon, and J. Oh. Validation of Crash Models for Intersections. FHWA-RD-03-037, Federal Highway Administration, Washington, DC, 2005.
  11. Federal Highway Administration. "SafetyAnalyst" McLean, VA. Accessed online: November 2006. (http://www.safetyanalyst.org/).
  12. Supplemental Notice of Proposed Amendments. FHWA Docket No. FHWA-2003-15149, Federal Highway Administration, Washington, DC, 2006.
ResearchFHWA
FHWA
United States Department of Transportation - Federal Highway Administration