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-04-138
Date: December 2005

Enhanced Night Visibility Series, Volume VII: Phase II—Study 5: Evaluation of Discomfort Glare During Nighttime Driving in Clear Weather

PDF Version (574 KB)

PDF files can be viewed with the Acrobat® Reader®

CHAPTER 3—RESULTS

ANALYSIS OF VARIANCE

An analysis of variance (ANOVA) was conducted on the 11 (VES) by 3 (Age) model for both near and far ratings of the discomfort glare. Recall that two glare ratings (i.e., near and far) were given for each VES by each participant using the deBoer scale. The deBoer scale ranges from 1, “Unbearable,” to 9, “Just noticeable.” Therefore, if a participant experiences more glare, this will result in a lower rating. The results of the ANOVA discomfort ratings for the VES by Age model are described in the following paragraphs. The results of the ANOVA are summarized in table 4 and table 5.

Table 4. Summary of far rating ANOVA.
Source DF SS MS F Value p Value
VES 10 117.6 11.8 6.93 <.0001
Age 2 14.3 7.2 0.79 0.4604
Age by VES 20 43.4 2.2 1.28 0.1951
Error 240 407.6 1.7    
Total 272        


Table 5. Summary of near rating ANOVA.
Source DF SS MS F Value p Value
VES 10 415.1 41.5 37.5 <.0001
Age 2 22.7 11.4 0.82 0.4454
Age by VES 20 53.4 2.7 2.41 0.0009
Error 240 265.7 1.1    
Total 272        

For the far discomfort rating (table 4), the effect of VES was statistically significant (p < 0.0001). The effect of age was not statistically significant (p = 0.46), with younger participants reporting a mean discomfort of 5.5, middle-aged participants reporting a mean discomfort of 5.7, and older participants reporting a mean discomfort of 5.2. There was also no significant interaction between VES configuration and age (p = 0.2).

The effect of VES was also statistically significant (p < 0.0001) for the near discomfort rating (table 5). The effect of age was not statistically significant (p = 0.45), with younger participants reporting a mean discomfort of 3.7, middle-aged participants reporting a mean discomfort of 4.3, and older participants reporting a mean discomfort of 3.8; however, there was a significant interaction between VES configuration and age (p = 0.0009) for the near discomfort rating.


VES MAIN EFFECT

Because VES had a statistically significant main effect for both the near and far discomfort ratings, a post hoc analysis was conducted for this variable using Student-Newman-Keuls (SNK) tests. These tests provide a way to compare means and determine significant differences between each configuration. The SNK tests were used because they do not produce overly conservative results when many levels of a single independent variable are compared. The results of the SNK tests are shown in table 6 and table 7, which list the VESs and their mean discomfort ratings in descending order. In other words, the VESs are ordered from the least amount of discomfort (therefore receiving the highest rating) to the most discomfort (therefore receiving the lowest rating). Table 6 and table 7 also show the SNK result grouping. In SNK tests, means that are assigned the same letter are not significantly different from each other.

For the far rating, the range of mean discomfort ratings was small, varying from 6.6 for the three UV–A + HID to 4.7 for the three UV–A + HLB (table 6). Accordingly, the SNK groupings were rather large, indicating that when the distance between the participant and the opposing vehicle was 396.2 to 304.8 m (1,300 to 1,000 ft), there were not many significant differences between the different VESs tested. The three VESs rated most discomforting were the three UV–A + HLB, five UV–A + HLB, and HOH; they all scored below a 5 (“Just acceptable”) on deBoer’s scale. The significant difference shown between the three UV–A + HID and the HID alone is surprising because adding the UV–A systems to the base lamps should not add enough visible light to affect driver discomfort and should certainly not reduce it.

Table 6. Mean far discomfort ratings in descending order.
VES Mean SNK Grouping
Three UV–A + HID 6.6 A
HLB–LP 6.3 A, B
Five UV–A + HID 6.0 A, B, C
HHB 5.8 A, B, C, D
HLB 5.5 B, C, D, E
HID 5.5 B, C, D, E
Hybrid UV–A + HID 5.5 B, C, D, E
Hybrid UV–A + HLB 5.0 C, D, E
Five UV–A + HLB 4.9 D, E
HOH 4.8 E
Three UV–A + HLB 4.7 E

For the near rating, the mean discomfort ratings ranged from 5.9 for the three UV–A + HID to 2.6 for the HHB (table 7). The three UV–A + HID, five UV–A + HID, and HID provided the least amount of discomfort when there was 137.2 to 45.7 m (450 to 150 ft) of separation between the driver and the opposing car. All three of these lighting configurations were classified above the “Just acceptable” boundary of deBoer’s scale. The ratings for HLB–LP indicated significantly more glare discomfort than the three UV–A + HID and five UV–A + HID.

The HHB headlamps caused a “Disturbing” amount of discomfort to the drivers at the near separation distance from the glare source, with a significantly lower mean deBoer rating (higher discomfort) than the HLB; however, the HHB did not result in significantly more discomfort than the HOH or any of the HLB and UV–A combinations. The HLB did not cause significantly less discomfort than the HOH or any of the HLB and UV–A combinations.

Table 7. Mean near discomfort ratings in descending order.
VES Mean SNK Grouping
Three UV–A + HID 5.9 A
Five UV–A + HID 5.5 A, B
HID 5.1 B, C
HLB–LP 4.8 C
Hybrid UV–A + HID 4.1 D
HLB 3.5 E
Three UV–A + HLB 3.1 E, F
Five UV–A + HLB 3.0 E, F
HOH 3.0 E, F
Hybrid UV–A + HLB 2.9 E, F
HHB 2.6 F

Table 8 shows a comparison of far rating and near rating for each VES. As expected, a higher level of discomfort was always reported for the near rating. The largest differences are observed for the halogen configurations, with the HHB causing a reported level of discomfort three deBoer scale levels lower at the near distance than at the far distance. HID alone and with nonhybrid UV–A configurations showed the smallest differences of less than one deBoer scale level.

Table 8. Differences between near and far discomfort ratings.
VES Near Rating − Far Rating
HHB 3.2
Hybrid UV–A + HLB 2.1
HLB 2.0
Five UV–A + HLB 1.9
HOH 1.8
Three UV–A + HLB 1.6
HLB–LP 1.5
Hybrid UV–A + HID 1.4
Three UV–A + HID 0.7
Five UV–A + HID 0.5
HID 0.4

VES CONFIGURATION BY AGE INTERACTION

Results of the ANOVA showed a significant interaction between VES and age for the near discomfort ratings (p = 0.0009). Figure 7 shows a plot of near discomfort rating versus VES for each age group. This figure illustrates three differences in VES condition means: (1) for the three UV–A + HID, the older and middle-aged participants reported less discomfort than the younger participants; (2) for the HID headlamps, the middle-aged participants experienced less discomfort than the younger participants; and (3) for the hybrid UV–A + HLB, the middle-aged participants reported less discomfort than the younger participants.

Graph. Near discomfort rating versus VES for each age group. Click here for more detail.

Figure 7. Graph. Near discomfort rating versus VES for each age group.

 

Previous    Table of Contents    Next
ResearchFHWA
FHWA
United States Department of Transportation - Federal Highway Administration