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-12-048    Date:  November 2013
Publication Number: FHWA-HRT-12-048
Date: November 2013

 

Pavement Marking Demonstration Projects: State of Alaska and State of Tennessee

CHAPTER 3. OPERATIONAL EFFECTS OF WIDE EDGE LINES

INTRODUCTION

This chapter describes the activities conducted to understand the operational effects of wide edge lines. Researchers summarized the literature and then conducted a before-after study on horizontal curves in Tennessee to determine the operational impacts of wider edge lines.

LITERATURE REVIEW

Measures such as speed and lateral position in the travel lane are surrogate measures for safety that are commonly used in the absence of crash data. The following subsections describe research that relates to the operational effects of pavement markings.

Vehicle Speed

While there have been several studies that used vehicle speed as a measure of pavement marking performance, most show no significant effect in absolute speed difference or, perhaps more importantly, speed variance (which is correlated with crash rates).(35,36) For instance, in 2004, Van Driel et al. performed a meta-analysis of vehicle operating speeds based on edge line presence.(37) The range of reported before-after results was -3 mi/h (reduction in mean speed) to +8.1 mi/h. An overall increase in mean speed after installing edge lines on roadways that previously only had a centerline was less than 0.5 mi/h. The authors came to the conclusion that the net speed effect was essentially zero.

In 2005, researchers from Louisiana reported on a before-after study of adding edge lines to narrow two-lane highways (with pavement widths of 20 to 22 ft).(38) Conclusively, the researchers found that the addition of an edge line on narrow two-lane highways did not impact vehicle speeds, day or night.

A recent study performed by Donnell et al. focused on the effectiveness of pavement marking delineation on curves to induce consistency in vehicle speed and lateral position based on a nighttime driving experiment.(39) Based on the results of the experiment, the use of brighter or wider pavement markings did not improve speed consistency between an approach tangent and the midpoint of a horizontal curve.

Tsyganov et al. conducted a before-after study on rural two-lane highways where edge line markings were added.(40) The highways had lane widths of 9, 10, and 11 ft. The researchers discovered that there were no significant differences in vehicle speeds before and after adding edge lines to the narrow highways. They also learned that there were no statistical differences in vehicle speeds when considering daytime versus nighttime conditions. The researchers’ findings consistently showed that speeds increased slightly in all conditions after edge lines were applied, but the differences were not statistically significant. They also showed that absolute speed standard deviations were all less than 1 mi/h.

Many experts believe that drivers reduce speeds based solely on their perceived risk. For instance, if drivers perceive sharp curves, narrow lanes or shoulders, steep roadside drop-offs, low side friction, etc., they will lower their speeds accordingly.

Lateral Vehicle Position

While research shows that the variance of vehicle lateral placement is strongly correlated with crash rates, findings related to the effect of pavement markings have been inconsistent.(41,42) A meta-analysis of lateral vehicle position was performed by Van Driel et al.(37) Based on research conducted in the United States, the change in mean lateral position after installing edge lines on roadways that previously only had a centerline was approximately 0.5 inches toward the centerline. The range of reported before-after results was a -10.5-inch shift (toward the centerline) to a +14-inch shift away from the centerline. The authors came to the conclusion that the net lane position effect was essentially zero.

The work by Donnell et al. found little evidence to show that enhanced pavement markings change the way in which motorists transition from a tangent into a curve.(39) As such, the authors concluded that the use of enhanced pavement markings does not improve driver lane position differential between an approach tangent and the midpoint of a horizontal curve.

Conversely, Cottrell compared the lateral vehicle position of vehicles using 4- and 8-inch-wide edge lines.(43) The results indicated that lateral vehicle position variance was unchanged at locations with a 4-inch edge line but was lowered both during the day and at night for the 8-inch edge line condition.

Another study using lateral vehicle position was conducted on a closed-course study in the early 1980s and showed improvements in vehicle positioning measures for an 8-inch edge line versus a 4-inch edge line on curved roadways using alcohol-impaired versus non-impaired drivers.(44) In this study, 16 males in their early 20s drove on an isolated section of a two-lane roadway in New Jersey between 12 and 3 a.m. Each person drove the course twice, the first time after consuming a placebo drink (0.0 percent blood alcohol content) and the second time after consuming either a placebo drink or controlled alcohol dosage (0.05 or 0.08 percent blood alcohol content). Fewer centerline encroachments, more central positioning within the lane, and less variability in positioning among drivers were observed for the wider edge lines (6 and 8 inches) versus the 4-inch edge lines. The authors concluded that the improved driving performance of the test subjects in the presence of wide edge lines indicates that strengthening the visual signal at the road edge may help partially compensate for visual impairments, although benefits are provided to all drivers.

Research conducted in Louisiana also investigated lateral placement as a function of adding edge lines to rural two-lane highways.(38) The before-after measurements showed that edge lines helped drivers confine their traveling path, particularly at night. They found that with edge lines, centralization of a vehicle’s position was more apparent at nighttime and that drivers generally positioned their vehicles away from the edge line, irrespective of the roadway alignment.

Tsyganov et al. evaluated lateral placement after adding edge lines to narrow two-lane highways.(40) They discovered a reduction in vehicle lateral placement variability, meaning vehicles were more consistently following a specific path. The exact location of that path depended on the overall lane width. For the 9-ft lane width, the vehicle path shifted closer to the newly installed edge line, especially in the curve sections. For 10-ft lane widths, there were no consistent changes noted. However, for the 11-ft lane width highways, the majority of the drivers moved closer to the centerline, especially on the curve sections. However, all changes were subtle.

TENNESSEE HORIZONTAL CURVE STUDY

A crash surrogate study was designed to detect possible operational impacts of 4- versus 6-inch pavement marking edge lines on horizontal curves on rural two-lane two-way (RTLTW) highways. Through the literature review and team discussions, three curve site selection criteria (curve radius, posted speed limit, and presence of paved shoulder) were identified as having the greatest potential impact on the effectiveness of wider edge lines. The crash surrogate study employed a before-after technique to reduce site-to-site variability using operational measures of effectiveness as surrogates for crashes. It was assumed that driver-to-driver (or vehicle-to-vehicle) variability would be less than variability caused by installation of wider lines. The literature review, combined with the expert opinion of the research team, lead to the decision to study the impacts of wider pavement markings exclusively on horizontal curves. The operational measures of effectiveness that were studied included the following:

Even with a before-after technique, it is possible that some uncontrolled extraneous factor may impact the data. As a result, the research team chose to have comparison sites. Comparison sites are curves that have similar geometric and traffic flow characteristics to the treatment site curves and where the pavement marking width is left unchanged between the before and after periods. Use of comparison sites helped ensure internal validity of the study by reducing confounding between the effect of treatment and the effects of uncontrollable extraneous variables. Examples of uncontrollable extraneous variables in this measure of effectiveness study might have included changes in drivers, driver behavior, and observers between the before and after period.

Study Site Selection

Based on a review of the literature regarding safety problem areas, all horizontal curve test sites were established on RTLTW highways. Approximately 60 potential sites within Tennessee were visited to assess the geometric and operational characteristics of the candidate curves (see table 20).

Table 20. Safety-related controls for curve study.

Geometric Operational
  • Lane width (10-12 ft).
  • Grade (≤ 4 percent).
  • Approach tangent length (≥ 0.25 mi).
  • Curve length (vehicle time in curve, t ≥ 3 s).
  • Ambient lighting (none).
  • Vehicle headway (≥ 5 s).
  • On-coming vehicles (none).
  • Approach speeds (≥ posted speed limit minus 10 mi/h).
  • Curve speeds (≥ posted advisory speed minus 10 mi/h).

 

As a result of these site visits, the researchers recommended that a total of 19 horizontal curves be studied in Tennessee, with 10 treatment sites and 9 comparison sites. The black dots in figure 6 represent the location of the 19 horizontal curve study sites. The researchers verified that no roadway improvements were planned for the 19 study sites for the duration of the study. While efforts were made to select only isolated horizontal curves, two of the horizontal curves were located within winding roadway segments.

This figure shows a map of 19 curve study sites in central Tennessee. Nashville is near the center of the map. Black dots indicate the location of the study sites, which are all outside the greater Nashville area.
Figure 6. Illustration. Map of 19 curve study sites in Tennessee

The researchers categorized the horizontal curves based on three factors that were identified through the literature review and team discussions as having the greatest potential impact on the effectiveness of wider edge lines. The sites were selected based on the radius of the curve (two levels), the posted speed limit (two levels), and the presence of a paved shoulder (two levels). Table 21 shows the study matrix that includes 2-by-2-by-2 levels of those factors. The curves were split into treatment and comparison sites in such a way as to have comparisons for each combination of selection criteria. Note that sites for one of the eight combinations could not be identified.

Table 21. Study site matrix.


Speed Limit

Curve Design Safety Rating (Radius)*

Radius ≤ 700 ft
(Degree of Curvature ≥ ~8.0)

Radius ≥ 800 ft
(Degree of Curvature ≤ ~7.0)

Presence of Paved Shoulder**

Presence of Paved Shoulder**

Yes

No

Yes

No

≥ 55 mi/h

1/1

2/2

2/1

1/1

≤ 50 mi/h

0

2/2

1/1

1/1

*2/1 indicates that there will be at least two treatment sites and one comparison site for each category.
**For this project, a paved shoulder is present when there is at least 36 inches of usable pavement beyond the inside edge of the edge line. A paved shoulder is absent when there is less than or equal to 24 inches of usable pavement beyond the inside edge of the edge line.

Data Collection

Data were collected along the 19 rural horizontal curves using traffic classifiers. The before data collection took place over a 5-week period from August to September 2007, and the after data collection took place over a 5-week period from July to August 2008. Traffic classifiers were installed on a Monday and retrieved on a Thursday in the same week by a team of two to four researchers. Approximately 96 h of data were collected at each study site for the before and after periods.

During the before data collection period, the curves had 4-inch-wide pavement markings. During the after period, the edge lines were restriped with 6-inch-wide pavement markings along the edge lines but not the centerlines. Centerlines were restriped with 4-inch-wide markings. Driver eye scanning studies show that drivers use the adjacent pavement marking edge line to negotiate curves regardless of whether they are in the inside or outside lane.(45)

Every effort was made to minimize differences between the periods of data collection and pavement marking installations. The average retroreflectivity of the edge lines in the before period was 200 mcd/lux/m2, with none of the sites below 100 mcd/lux/m2, while the average edge line retroreflectivity for the after period was 288 mcd/lux/m2. The pavement markings were installed in late May 2008 for the after period. After the pavement markings were installed, at least 1 month was provided to allow drivers to acclimate to the new markings.

Equipment Setup

The traffic classifiers recorded when a vehicle passed through a particular curve, the classification of the vehicle (i.e., passenger car or tractor trailer), the lateral position of the vehicle, and the speed of the vehicle. Piezoelectric road sensors were used in conjunction with traffic classifiers. The traffic classifiers enabled researchers to collect raw data with time stamp precision of 0.001 s.

Four traffic classifiers were used at each study site to track the movements of the vehicles traveling through the outside of each horizontal curve. These locations are defined as follows and are shown in figure 7:

This diagram contains the layout of four traffic counters along a curving roadway. It identifies the location of the upstream counter (U), about 500 ft upstream of the start of the treatment (ST), the start of treatment, about 500 ft prior to the advance warning sign, the counter located at the advanced curve warning sign (W), the counter located at the point of the curve (PC), and the counter located at the midpoint of the curve (MC).
Figure 7. Illustration. Horizontal curve traffic classifier layout

Sample Size

A power analysis was used to determine the sample size (the number of vehicles) needed to detect a practically important minimum difference in effects of increasing the pavement marking width and among the interaction effects between the pavement marking width and the day/night factor at each site. The procedures given in Wheeler, Nelson, and Bratcher et al. were used for the sample size calculation.(46-48) Because the necessary sample size (n) varies with the desired significance level (α), the desired power, the standard deviation (σ) of the response variable, and the minimum difference of practical importance (Δ), those values were predetermined before the sample size calculation. By convention, the desired significance level and the desired power were set to 0.05 and 0.90, respectively. Previous research indicates that the approximate standard deviations in speed and lateral placement in curves similar to those used in this study are 8 mi/h and 20 inches, respectively.(49) The minimum difference of interest before and after installation of wider lines was determined to be 3 mi/h for the mean speeds and 6 inches for the mean lateral placements based on engineering judgment and previous research.(49,50) It is believed that 6 inches is the minimum change in mean lateral position that would be a practically significant change for at least two reasons: (1) field experience has shown that striping installations vary in width as much as ±0.5 inches and restriping can be misaligned by more than 1 inch, which may result in wide variability between pavement marking installations, and (2) previous research supported 6 inches.(49) The 3-mi/h minimum difference of interest was selected as a value between the values chosen by other researchers because it is believed that a change of 3 mi/h would be the minimum change that would influence changing posted speed limits or advisory speeds.(49,50)

The minimum sample size (nspeed) necessary for detecting a mean speed difference (Δ) of 3 mi/h with a in speed of 8 mi/h before and after installation of wider lines at each site is shown in figure 8, where r is the number of levels of a factor.

n subscript speed equals open parenthesis 3 times r times sigma divided by delta close parenthesis squared equals open parenthesis 3 times 2 times 8 divided by 3 close parenthesis squared equals 256.
Figure 8. Equation. Power analysis for sample size to detect a speed difference of 3 mi/h

The minimum sample size (nip) necessary for detecting a mean lateral placement difference (Δ) of 6 inches with σ of 20 inches before and after installation of wider lines at each site is shown in figure 9.

n subscript lp equals open parenthesis 3 times r times sigma divided by delta close parenthesis squared equals open parenthesis 3 times 2 times 20 divided by 6 close parenthesis squared equals 400.
Figure 9. Equation. Power analysis for sample size to detect a lateral placement difference of 6 inches

The minimum sample size necessary for detecting a mean speed difference of at least 3 mi/h in any two interactions means between pavement marking width and day/night at each site is shown in figure 10, where v is the number of interaction degrees of freedom, c is the number of factor-level combinations for the factors that are involved in the interaction, k is the number of factors involved in the interaction, and δ is the minimum difference of interest among the interaction effects.

n subscript speed equals 9 times sigma squared times open parenthesis nu plus 1 close parenthesis times c divided by delta squared times 2 raised to the power of k minus 2 times open parenthesis 1 divided by 2 close parenthesis equals 9 times 8 squared times open parenthesis 1 plus 1 close parenthesis times 4 divided by 3 squared times 2 raised to the power of 2 minus 2 times open parenthesis 1 divided by 2 close parenthesis equals 256.
Figure 10. Equation. Power analysis for sample size to detect a speed difference of 3 mi/h with two interactions

The minimum sample size necessary for detecting a mean lateral placement difference of at least 6 inches in any two interactions means between pavement marking width and day/night at each site, is shown in figure 11.

n subscript lp equals 9 times sigma squared times open parenthesis nu plus 1 close parenthesis times c divided by delta squared times 2 raised to the power of k minus 2 times open parenthesis 1 divided by 2 close parenthesis equals 9 times 20 squared times open parenthesis 1 plus 1 close parenthesis times 4 divided by 6 squared times 2 raised to the power of 2 minus 2 times open parenthesis 1 divided by 2 close parenthesis equals 400.
Figure 11. Equation. Power analysis for sample size to detect a lateral placement difference of 6 inches with two interactions

A sample size of 400 vehicles was selected to assure the power of the tests to be at least 0.90 for both mean speed difference and mean lateral placement difference. Thus, the desired number of vehicles to be observed for each daytime and nighttime condition and for each before and after installation of wider lines at each site was at least 100 vehicles.

Statistical Analysis Methodology

The horizontal curve study to compare 4- and 6-inch pavement marking edge lines along an isolated RTLTW highway was a field experimental before-after study. Researchers collected continuous quantitative data from traffic classifiers. Two primary treatments (that correspond to the levels of the main study factor, edge line width) were studied: (1) curves marked with 4-inch-wide edge lines and (2) curves marked with 6-inch-wide edge lines. Other factors were the posted speed limit, the curve radius, the shoulder width, and the period of the day. The dependent variables were vehicle speed and vehicle lateral placement. The changes in mean speed, speed variability, mean lateral position, and lateral position variability before and after installation of wider edge lines were the main interests of the study. In addition, the mean differences in the speed and lateral position between the different traffic classifier locations were investigated, such as between the data collected at the PC and the MC. Evaluation criteria included the following:

The statistical analyses included descriptive statistics, graphical analysis, and hypothesis testing. The descriptive statistics calculations included minimums, maximums, ranges, means, medians, quartiles, and 85th percentile values. Box plots, histograms, scatter plots, and cumulative distributions were used to investigate the distribution of the data and to identify any trends or outliers in the data that would impact the testing methods used to conduct the hypothesis testing. The analysis of variance (ANOVA), specifically a split-plot design analysis, was used to test equality of mean speed and equality of mean lateral position of vehicles before and after the installation of wider edge lines.

Descriptive Statistical Analysis

The descriptive statistics are separated into several tables. Table 22 contains summary statistics with respect to the sample size. While each study site had ample volume to provide 100 vehicles for each condition, some of the sample sizes for the nighttime data were less than desired once the researchers removed all of the unusable data. Unusable data were defined based on the following criteria:

Table 22. Sample size summary.


Statistic

Before

After

Day

Night

Day

Night

Minimum

279

43

613

56

Mean

1,012

113

901

130

Median

890

84

828

100

Maximum

2,770

354

1,403

274

Table 23 shows summary statistics for the general trends. The values were calculated from the difference in the before and after period mean and standard deviation values. A positive value for a change in mean lateral placement indicates that drivers in the after period were driving closer to the centerline, while a negative value for the change in standard deviation in lateral placement indicates that the drivers were more centrally located within their respective lane of travel.

Table 23. Change in speed and lateral position statistics for the treatment sites.


Speed Limit

Change in Statistical Measure

Curve Design Safety Rating (Radius)

Radius ≤ 700 ft (Degree of Curvature ≥ ~8.0)

Radius ≥ 800 ft (Degree of Curvature ≤ ~7.0)

Presence of Paved Shoulder

Presence of Paved Shoulder

Yes

No

Yes

No

Speed (mi/h)

Lateral Position (inches)

Speed (mi/h)

Lateral Position (inches)

Speed (mi/h)

Lateral Position (inches)

Speed (mi/h)

Lateral Position (inches)

≥ 55 mi/h

Mean

1.6

3.8

-0.1

4.0

0.0

-0.4

0.2

-0.4

Std. dev.

0.1

0.0

0.4

1.9

0.2

0.4

0.1

0.4

≤ 50 mi/h

Mean

N/A

N/A

-0.7

2.1

0.7

-1.5

-1.1

0.0

Std. dev.

N/A

N/A

0.1

1.6

0.7

1.1

0.2

1.1

N/A = Not applicable.

Table 24 through table 28 contain the detailed mean and standard deviation values for the speed and lateral position data collected between the before and after periods for all 19 study sites. Table 24 provides the sample size for the crash surrogate study, while table 25 through table 28 provide the speed data by location, change in speed data by location, lateral position data by location, and change in lateral position data by location, respectively.

Other descriptive statistics, such as range and variance, were investigated, but they are not reported herein because they did not enhance the information already provided through the mean and standard deviation. There are no apparent trends that would immediately suggest that the installation of wider edge lines affected a driver’s selection of speed, but it does appear that the installation of wider edge line markings in rural curves with small radii (≤ 700 ft) and higher speed limits (≥ 55 mi/h) may have impacted a driver’s selection of lateral position through horizontal curves with a slight shift toward the centerline once in the curve (see table 23). However, there were no mean changes of speed that exceeded 3 mi/h or mean changes in lateral position that exceeded 6 inches, which were established as the practical statistically significant differences during the sample size calculations.

Table 24. Sample size.


Curve

Code

Comparison/
Treatment

Speed

Radius
(ft)

Shoulders
(Y/N)

Day/
Night

Observations

Limit
(mi/h)

Advisory
(mi/h)

Before

After

1

1

Treatment

55

30

318

N

Day

849

752

Night

86

82

2

1

Treatment

55

35

539

N

Day

388

613

Night

44

84

3

1

Comparison

55

35

649

N

Day

804

828

Night

75

83

4

1

Comparison

55

40

663

N

Day

492

674

Night

66

135

5

2

Treatment

55

30

314

Y

Day

298

810

Night

76

100

6

2

Comparison

55

35

613

Y

Day

2,770

1,031

Night

199

274

7

3

Treatment

55

30

881

N

Day

871

916

Night

83

56

8

3

Comparison

55

40

1,857

N

Day

408

770

Night

43

99

9

4

Treatment

55

N

1,171

Y

Day

904

735

Night

84

86

10

4

Treatment

55

45

1,250

Y

Day

890

1,050

Night

60

97

11

4

Comparison

55

N

1,425

Y

Day

923

790

Night

83

94

12

5

Treatment

35

30

406

N

Day

891

914

Night

72

98

13

5

Treatment

50

40

672

N

Day

1,340

1,224

Night

117

102

14

5

Comparison

45

30

460

N

Day

1,291

686

Night

95

193

15

5

Comparison

35

30

511

N

Day

1,801

1,403

Night

116

261

16

7

Treatment

50

40

1,193

N

Day

279

1,083

Night

113

104

17

7

Comparison

45

30

860

N

Day

626

846

Night

354

211

18

8

Treatment

45

N

1,161

Y

Day

2,065

772

Night

247

143

19

8

Comparison

35

N

1,650

Y

Day

1,337

1,222

Night

129

169

Table 25. Speed data by location.


Curve

Code

Statistic

Speed by Location (ft/s)

U

W

PC

MC

Before

After

Before

After

Before

After

Before

After

1

1

Mean

71.6

71.5

71.9

75.4

67.4

67.6

60.2

60.1

Std. dev.

8.4

8.8

12.5

9.6

7.7

7.7

5.9

6.5

2

1*

Mean

69.7

69.8

69.9

 

 

69.1

62.8

64.5

Std. dev.

10.1

9.4

8.6

 

 

7.9

8.7

8.8

3

2

Mean

73.7

73.8

70.8

70.9

66.4

65.9

55.9

56.3

Std. dev.

8.1

8.2

8.1

8.2

7.6

7.4

6.0

5.8

4

2

Mean

73.1

72.1

78.8

77.1

78.3

76.1

74.9

73.7

Std. dev.

7.7

9.2

8.7

9.1

8.5

8.8

8.0

8.5

5

2*

Mean

75.0

73.3

73.3

 

68.0

 

65.0

65.7

Std. dev.

10.6

10.7

9.9

 

9.8

 

8.5

8.6

6

2*

Mean

75.4

74.2

76.5

76.4

75.2

75.7

72.8

72.8

Std. dev.

12.5

12.5

9.5

9.7

9.2

9.3

9.2

8.9

7

3

Mean

84.4

83.5

83.5

83.2

82.7

82.3

81.0

81.7

Std. dev.

7.2

7.2

7.3

7.2

7.5

7.4

7.7

7.4

8

3

Mean

75.9

74.3

 

82.1

82.0

81.5

80.7

80.6

Std. dev.

17.9

17.7

 

9.9

9.6

9.1

9.3

9.0

9

3*

Mean

88.2

87.0

86.6

86.7

84.6

84.4

84.9

84.7

Std. dev.

7.6

7.6

7.6

7.8

7.8

8.0

7.5

7.4

10

4

Mean

69.1

69.2

69.2

68.8

68.9

68.5

68.1

68.0

Std. dev.

7.5

7.4

8.2

8.4

8.2

8.4

8.4

8.7

11

4*

Mean

77.8

76.1

83.1

81.7

80.4

80.6

82.3

 

Std. dev.

11.3

12.7

9.8

10.6

9.2

9.8

8.8

 

12

6

Mean

62.4

 

64.5

64.8

61.4

54.7

55.0

54.2

Std. dev.

9.3

 

8.0

9.1

7.4

6.8

5.9

6.1

13

6

Mean

75.3

74.9

77.8

77.9

74.3

76.0

72.7

72.6

Std. dev.

12.0

14.0

9.5

9.3

9.0

10.1

9.1

8.8

14

6*

Mean

75.4

76.1

72.3

76.2

71.3

73.2

64.3

65.7

Std. dev.

8.7

8.9

7.8

7.9

6.6

6.8

6.3

6.4

15

6*

Mean

73.1

72.5

69.9

69.4

 

63.5

60.5

59.4

Std. dev.

8.6

9.1

8.3

8.6

 

8.3

7.7

8.5

16

7

Mean

79.3

77.0

80.1

77.9

80.2

78.0

73.8

72.3

Std. dev.

8.6

8.7

9.0

8.7

9.0

8.6

8.5

8.1

17

7*

Mean

60.6

60.8

72.3

71.3

70.3

70.2

68.3

 

Std. dev.

20.4

20.0

9.7

8.9

9.3

9.2

8.7

 

18

8

Mean

73.1

77.1

78.5

79.6

76.0

76.7

74.6

75.6

Std. dev.

8.8

8.9

9.7

8.8

9.5

8.7

9.2

8.1

19

8*

Mean

74.7

74.8

76.4

75.9

72.3

71.9

70.7

70.4

Std. dev.

7.2

7.7

7.9

7.8

7.0

6.9

6.8

6.7

*Indicates a comparison study site.
Note: Blank cells indicate no data were available.

Table 26. Change in speed data by location.


Curve

Code

Statistic

Change in Speed by Location (ft/s)

W - U

PC - W

MC - PC

Before

After

Before

After

Before

After

1

1

Mean

-1.1

3.9

-4.9

-7.8

-7.2

-7.6

Std. dev.

9.2

5.9

12.0

5.8

4.5

4.9

2

1*

Mean

.2

 

 

 

 

-4.6

Std. dev.

6.7

 

 

 

 

5.0

3

2

Mean

-2.9

-2.9

-4.4

-5.0

-10.5

-9.6

Std. dev.

4.1

4.3

2.1

2.3

4.4

4.8

4

2

Mean

5.7

5.0

-0.5

-1.0

-3.4

-2.5

Std. dev.

3.1

5.3

2.0

2.1

2.6

2.5

5

2*

Mean

-1.7

 

-5.3

 

-2.9

 

Std. dev.

7.3

 

5.2

 

6.4

 

6

2*

Mean

1.1

2.3

-1.2

-0.7

-2.5

-2.9

Std. dev.

7.9

15.8

3.5

13.2

3.3

12.0

7

3

Mean

-0.9

-0.3

-0.8

0.9

-1.6

-0.5

Std. dev.

2.8

3.1

2.5

2.7

2.3

2.1

8

3

Mean

 

8.0

 

-0.8

-1.0

-0.9

Std. dev.

 

13.1

 

3.4

2.9

2.3

9

3*

Mean

-1.6

-0.3

-2.0

-2.3

0.2

0.2

Std. dev.

3.2

3.7

3.1

3.7

2.2

3.0

10

4

Mean

0.1

-0.4

-0.4

-0.3

-0.8

-0.5

Std. dev.

5.3

4.9

2.3

2.1

3.2

2.2

11

4*

Mean

5.4

5.6

-2.7

-1.1

1.9

 

Std. dev.

7.3

8.0

3.9

3.7

3.5

 

12

6

Mean

2.1

 

-3.1

-10.0

-6.4

-0.5

Std. dev.

10.6

 

3.2

8.0

8.7

3.3

13

6

Mean

2.5

3.1

-3.6

-3.7

-1.5

-0.8

Std. dev.

7.7

10.2

2.8

2.5

1.8

1.7

14

6*

Mean

-3.2

0.1

-0.9

-2.9

-7.0

-7.5

Std. dev.

6.5

5.5

3.5

3.0

4.4

4.5

15

6*

Mean

-3.3

-3.0

 

-5.9

 

-4.0

Std. dev.

4.1

8.4

 

9.6

 

8.5

16

7

Mean

0.8

0.9

0.1

0.1

-6.4

-5.9

Std. dev.

4.9

4.9

2.8

2.6

4.3

3.6

17

7*

Mean

11.8

10.5

-2.0

-1.2

-2.1

 

Std. dev.

18.3

17.5

3.5

3.6

3.2

 

18

8

Mean

4.3

2.6

-2.7

-3.0

-1.3

-1.2

Std. dev.

4.8

4.6

3.7

3.6

3.9

3.1

19

8*

Mean

1.6

1.1

-4.1

-4.1

-1.6

-1.5

Std. dev.

3.6

4.6

4.1

3.8

2.1

2.2

*Indicates a comparison study site.
Note: Blank cells indicate no data were available.

Table 27. Lateral position data by location.


Curve

Code

Statistic

Lateral Position by Location (inches)

U

W

PC

MC

Before

After

Before

After

Before

After

Before

After

1

1

Mean

42.2

40.4

34.5

32.5

22.1

30.1

46.2

50.2

Std. dev.

13.5

13.9

16.5

9.5

10.9

9.2

17.4

15.5

2

1*

Mean

27.7

27.4

24.7

 

 

26.5

47.6

48.3

Std. dev.

10.3

11.1

9.5

 

 

12.0

13.8

14.1

3

2

Mean

35.8

43.4

28.1

26.1

22.3

21.7

36.2

39.1

Std. dev.

13.6

11.5

10.1

9.2

10.5

9.1

14.5

13.4

4

2

Mean

23.7

 

25.1

27.6

30.9

32.9

40.5

44.2

Std. dev.

12.3

 

11.4

11.3

10.9

11.8

14.9

15.7

5

2*

Mean

40.7

38.8

30.2

 

30.7

 

53.7

55.2

Std. dev.

11.3

11.7

9.8

 

11.7

 

16.0

15.6

6

2*

Mean

33.4

 

27.5

25.4

17.1

20.5

34.0

34.3

Std. dev.

13.2

 

11.6

10.9

10.9

12.0

15.0

15.3

7

3

Mean

31.1

31.0

39.6

35.9

37.3

38.9

43.6

43.0

Std. dev.

13.3

13.6

11.4

11.6

13.0

13.9

13.9

14.2

8

3

Mean

30.7

30.1

 

40.6

46.7

41.5

56.2

54.6

Std. dev.

13.5

13.0

 

11.0

14.6

13.6

15.5

16.2

9

3*

Mean

43.3

41.1

42.9

 

37.5

34.6

49.2

45.2

Std. dev.

10.8

10.9

11.4

 

11.8

13.2

13.7

14.2

10

4

Mean

31.6

30.5

28.9

 

28.1

25.8

29.3

28.9

Std. dev.

12.0

13.1

11.1

 

10.7

12.3

13.3

12.9

11

4*

Mean

37.9

36.4

33.9

30.2

26.5

32.9

34.1

 

Std. dev.

13.8

15.2

10.8

11.9

11.2

12.2

12.0

 

12

6

Mean

37.5

 

39.3

35.2

18.0

26.0

43.5

51.7

Std. dev.

11.2

 

11.1

11.7

5.2

11.1

16.5

16.3

13

6

Mean

34.8

41.6

32.0

30.0

30.8

30.6

40.3

37.7

Std. dev.

12.7

15.2

10.8

11.4

11.9

11.8

12.8

13.1

14

6*

Mean

29.7

29.6

44.1

37.5

39.1

38.0

53.8

52.4

Std. dev.

9.0

8.9

10.5

10.9

10.3

10.8

16.0

17.2

15

6*

Mean

32.0

6.6

45.1

23.5

 

41.6

38.6

43.2

Std. dev.

12.1

11.6

14.9

13.2

 

12.3

13.1

15.6

16

7

Mean

38.0

37.5

31.2

34.1

23.7

23.0

38.8

38.8

Std. dev.

13.2

12.7

12.7

11.9

11.1

10.9

12.6

13.8

17

7*

Mean

40.6

39.0

23.6

24.7

28.9

30.2

34.6

 

Std. dev.

16.9

16.0

10.5

10.1

12.2

10.7

16.8

 

18

8

Mean

57.7

33.9

30.7

29.7

24.8

22.3

38.1

36.6

Std. dev.

22.5

11.3

11.7

9.7

11.1

9.7

13.1

12.0

19

8*

Mean

31.8

39.4

38.7

38.3

32.1

34.1

51.6

53.9

Std. dev.

11.4

10.0

10.2

11.1

10.8

10.3

13.6

14.8

*Indicates a comparison study site.
Note: Blank cells indicate no data were available.

Table 28. Change in lateral position data by location.


Curve

Code

Statistic

Change in Lateral Position by Location (inches)

W - U

PC - W

MC - PC

Before

After

Before

After

Before

After

1

1

Mean

-5.8

-7.9

-10.2

-2.4

24.2

20.1

Std. dev.

20.4

14.5

16.8

10.9

17.5

15.6

2

1*

Mean

-3.0

 

 

 

 

21.9

Std. dev.

11.7

 

 

 

 

15.1

3

2

Mean

-7.7

-17.3

-5.8

-4.4

13.9

17.3

Std. dev.

14.1

12.1

6.8

6.3

14.4

13.5

4

2

Mean

1.4

 

5.9

5.3

9.6

11.3

Std. dev.

12.7

 

10.7

11.2

13.3

14.4

5

2*

Mean

-10.5

 

0.5

 

23.1

 

Std. dev.

12.6

 

10.9

 

15.7

 

6

2*

Mean

-5.9

 

-10.4

-5.0

16.9

13.8

Std. dev.

14.4

 

12.8

15.6

13.2

18.6

7

3

Mean

8.5

4.9

-2.3

3.0

6.3

4.0

Std. dev.

14.1

13.6

12.8

13.6

13.3

14.1

8

3

Mean

 

10.4

 

1.0

9.4

13.2

Std. dev.

 

14.1

 

12.5

17.3

16.7

9

3*

Mean

-0.3

 

-5.5

 

11.7

10.5

Std. dev.

11.5

 

11.8

 

13.2

14.8

10

4

Mean

-2.7

 

-0.8

 

1.2

3.2

Std. dev.

15.0

 

8.9

 

12.4

12.3

11

4*

Mean

-4.0

-6.2

-7.4

2.7

7.6

 

Std. dev.

14.5

16.0

11.1

11.7

12.3

 

12

6

Mean

1.8

 

-21.3

-9.2

25.5

25.8

Std. dev.

14.6

 

9.8

13.3

17.5

16.1

13

6

Mean

-2.9

-11.5

-1.2

-2.0

9.5

11.2

Std. dev.

13.6

16.8

10.5

9.5

10.8

11.6

14

6*

Mean

14.3

7.9

-5.0

0.6

14.7

14.5

Std. dev.

11.9

11.3

12.4

12.3

15.8

16.4

15

6*

Mean

13.0

17.4

 

18.1

 

1.6

Std. dev.

15.9

15.3

 

16.3

 

17.8

16

7

Mean

-6.7

-3.5

-7.6

-11.1

15.1

15.8

Std. dev.

15.5

14.6

11.8

12.2

14.0

14.1

17

7*

Mean

-17.0

-14.4

5.3

5.5

5.8

 

Std. dev.

17.3

15.3

12.2

9.7

16.0

 

18

8

Mean

-27.5

-4.2

-6.0

-7.4

13.4

14.3

Std. dev.

21.5

11.7

12.0

10.1

13.4

12.8

19

8*

Mean

6.8

-1.1

-6.6

-4.3

19.6

19.9

Std. dev.

15.1

10.8

14.9

11.1

13.0

13.9

*Indicates a comparison study site.
Note: Blank cells indicate no data were available.

Enhanced Statistical Analysis

The research team also conducted an enhanced statistical analysis of the lateral position data at MC and the speed change from PC to MC. There were a total of 40,673 measurements of lateral position at MC and 33,458 values on the speed change from PC to MC (MC - PC speed) before and after installation of wider edge lines, along with curve characteristic variables such as shoulder width, speed limit, and radius. Note that 40,673 and 33,458 correspond to the number of vehicles for which the measurements (for lateral position at MC and speeds at both MC and PC) were taken.

The lateral position data at MC were analyzed by employing a split-plot analysis having curve as a random effect and before-after, day/night, presence of paved shoulder, speed limit (high = ≥ 55 mi/h and low = ≤ 50 mi/h), radius (large = ≥ 800 ft and small = ≤ 700 ft), site type (trt = treatment and comp = comparison) and interaction effects among them as fixed effects. Note that the before-after variable for treatment sites corresponds to both the passage of time and edge line width (before = 4 inches and after = 6 inches), while the same variable for comparison sites corresponds to the passage of time (edge line width is 4 inches for both before and after periods). The initial analysis, including all main effects, two-way interactions, and some three-way interactions of interest, revealed that there were statistically significant three-way interactions such as before/after x day/night x site type, before/after x shoulder x site type, and before/after x radius x site type. This implies that the effect of edge line width on the lateral position at MC may change with the level of other variables such as day/night and/or radius.

Researchers conducted separate analyses for each combination of day/night and large/small radius. The results of the split-plot analyses were obtained by the restricted maximum likelihood (REML) method.

For curves with a small radius (≤ 700 ft), the mean lateral placement at MC at the treatment sites increased after installation of wider edge lines. The change was larger for the nighttime data (4.9 inches during the nighttime versus 2.4 inches during the daytime). No significant changes were observed for the comparison sites or curves with a large radius (≥ 800 ft).

Next, the speed change from PC to MC (MC - PC speed) was used as a dependent variable. The initial analysis, including all main effects, two-way interactions, and three-way interactions of interest, revealed that two three-way interactions, before/after x speed limit x site type and before/after x radius x site type, produced statistically significant results. This implies that the effect of edge line width on speed change (MC - PC speed) may be different for various levels of variables such as speed limit and/or radius. Researchers conducted separate analyses for each combination of speed limit (high/low) and radius (large/small). The results of the split-plot analyses were obtained using the REML method implemented using a statistical package (Statistical Analysis Software (SAS®) product).

For the treatment sites, the mean speed change (MC - PC speed) during the after period (6-inch edge lines) was consistently smaller (in magnitude) than the mean speed change during the before period (4-inch edge lines), which suggests that drivers decelerated (from the beginning of the curve to MC) less after installation of wider edge lines. However, this change seems to be significant only for the curves with a small radius (≤ 700 ft) and low speed limit (≤ 50 mi/h).

SUMMARY

Previous studies on the operational effects of pavement markings and wider edge lines showed mixed results regarding lateral placement and vehicle speed. The study reported herein produced similar findings to previous research. While some particular instances of either lateral placement and/or change in speed were found to be statistically significant, the findings were not consistent, and the magnitude of the change was not deemed practical. For the conditions studied in this report, it appears that wider edge line had a net zero impact in terms of vehicle lateral placement and speed.

 

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