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-17-086    Date:  January 2018
Publication Number: FHWA-HRT-17-086
Date: January 2018

 

Safety Evaluation of Multiple Strategies at Stop-Controlled Intersections

Chapter 2. Literature Review

Literature on the stop-controlled intersection strategies of interest is limited. The following provides a summary of the salient research related to specific strategies. There were very few studies identified that investigated the effects of multiple strategies.

Doubling and Oversizing Advance Warning Signs, Stop Signs, and Yield Signs

The Institute of Transportation Engineers published the results of two evaluations concerning doubling stop signs (i.e., placement of a stop sign on the left side of the stop approach road to complement the existing stop sign on right side).(7) The first evaluation was solely for the installation of double stop signs. This countermeasure had an estimated reduction of 11 percent on total crashes. The study neglected to estimate a standard error for this crash reduction. The other evaluation was for a combination of treatments. The combination was adding a centerline, moving the stop line to the extended curb lines, and doubling stop signs. The estimated effect on total crashes was a nine percent decrease, and there was no observed change in right-angle crashes.

Polanis conducted an evaluation of another combination of intersection treatments.(8) The combination was adding a centerline, adding a stop line, and replacing existing 24-inch stop signs with 30-inch stop signs. The results indicate a 55-percent reduction in right-angle crashes with a standard error of 52 percent. Therefore, the reductions were not statistically significant. Not only were the results of Polanis’s study statistically insignificant, the design of the study was not rigorous. The study employed a simple before–after methodology, which fails to account for traffic volume changes, RTM, and a host of other potentially confounding factors. Furthermore, only 10 sites were used in the study, and these sites were all from the same municipality in North Carolina: Winston-Salem.(8)

A case study by the Federal Highway Administration (FHWA) further examined the effect of doubling and oversizing stop signs at intersections in Winston-Salem, NC.(9) At four stop-controlled intersections with a history of high crash frequency, stop signs were doubled using 30-inch signs as opposed to the conventional 24-inch variety. Based on a simple before–after comparison, the results estimated a 48-percent reduction in total crashes. As with the Polanis study, this case study employs a weak study methodology (i.e., simple before–after) and uses a small sample size, which lacks geographic diversity. Furthermore, this case study exhibits a clear selection bias, making the regression to the mean (RTM) phenomenon a serious concern.

Fluorescent Yellow Sheeting

There have been multiple studies that examined the use of fluorescent yellow sheeting on warning signs. Eccles and Hummer focused on the use of fluorescent yellow sheeting as an inexpensive method of increasing the conspicuity of signs without violating the provisions contained in the Manual on Uniform Traffic Control Devices (MUTCD).(10) Multiple studies confirm the superiority of fluorescent signs in terms of conspicuity. Jenssen et al. conducted a comparative evaluation of fluorescent and nonfluorescent signs on a closed track in Norway.(11) Subjects seated in moving railcars were asked to indicate when they could detect and recognize the size, shape, and content of fluorescent and nonfluorescent signs. The performance of the fluorescent signs proved to be superior because the subjects were able to detect and recognize the fluorescent signs well before they could detect and recognize the nonfluorescent signs. Burns and Johnson studied fluorescent and nonfluorescent materials and found that the photometric properties of fluorescent materials explained their superior visibility and conspicuity.(12)

RetroReflective Sign Posts

A recent study by the Virginia Department of Transportation directly relates to the retroreflective sign post strategy.(13) This study examined the effectiveness of retroreflective material on stop sign posts with respect to visibility and driver compliance. The authors measured performance with respect to visibility using a video survey in which participants were asked to pinpoint when a stop sign with retroreflective material on the post and another without retroreflective material on the post could be detected. The results indicated that during daytime conditions, the vast majority of participants could detect the stop sign without retroreflective material on the post sooner than the stop sign with retroreflective material on the post. In contrast, during nighttime conditions, the vast majority of participants could detect the stop sign when retroreflective material is on the post sooner than the stop sign without retroreflective material on the post. In terms of compliance, the behavior of drivers approaching a stop sign with retroreflective material on the post was not observed to be different from that of a driver approaching a stop sign without retroreflective material on the post.

Refreshing Existing Pavement Markings

The research team did not identify any existing studies on the safety effects of refreshing existing pavement markings at stop-controlled intersections.

Stop Lines

The installation of a stop line has also been studied in recent years. The installation of a stop line on the minor approach of an unsignalized intersection is intended to address angle crashes in which drivers are unaware of the presence of an intersection or fail to stop at the stop sign. Golembiewski and Chandler estimated that this countermeasure reduced total crashes by 19 percent.(14) This study did not provide an estimate of the standard error of the crash reduction.

The Institute of Transportation Engineers estimated that the installation of stop lines on minor road approaches with short segments of centerline reduced total crashes by 19 percent and reduced right-angle crashes by 47 percent. Once again, the standard errors of these crash reductions were not estimated.(7)

Polanis evaluated the combination of adding a centerline, adding a stop line, and replacing a 24-inch stop sign with a 30-inch stop sign.(8) This combination of treatments was estimated to reduce right-angle crashes by 67 percent with a standard error of 11 percent. Therefore, the reduction is statistically significant.

A FHWA study analyzed the effects of installing stop lines and a short interval of double yellow centerlines at intersections in Winston-Salem, NC.(9) The treatment group consisted of four stop-controlled intersection with a history of high crash frequency. The study used a simple before–after method to estimate the crash effects on total crash frequency. The results indicated that total crashes decreased by 53 percent. It should be noted that this study employed a weak study design (i.e., simple before–after) and a small sample size.

This strategy is a potentially effective countermeasure at locations with a history of crashes in which drivers are unaware of the presence of the intersection. However, the authors noted that the stop line should be installed in such a way that it can be seen from a significant distance by approaching drivers.(14) Aside from its significant crash reduction potential, this countermeasure is appealing because of its relatively low cost.

Limitations of Previous Research

Literature on the stop-controlled intersection strategies of interest was scarce in some cases and nonexistent in others. In the cases of fluorescent yellowing sheeting, retroreflective sign posts, and refreshing existing pavement markings, the research team did not identify studies that quantify the effect on crashes of implementing the strategies. With regard to the strategy of doubling and oversizing advance warning signs, stop signs, and yields signs, the research team identified several studies that estimate crash effects. However, the studies employed study designs that lacked statistical rigor and frequently neglected to estimate standard errors for the crash reductions. The standard error of a crash reduction enables one to judge the statistical significance of the crash effect. Therefore, the omission of standard errors in these studies poses a major obstacle to a meaningful interpretation of the results. With respect to the strategy of installing stop lines, some of the studies also lacked standard error estimates. Furthermore, none of the previous studies conducted a comprehensive evaluation with regard to crash type and severity. The previous studies generally reported the effect on total crashes or angle crashes, and virtually none estimated the effect on injury crashes. Thus, there is a need for additional research of the stop-controlled strategies of interest that employs rigorous study designs and analyzes a full range of crash types and severities.

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