Dial Keju (Research Assistant)
Department of Urban and Regional Planning · University of Hawaii at Manoa
E-mail: karlk@hawaii.edu
Website: www.durp.hawaii.edu
Phone: 808-956-7381
Fax: 808-956-6870
A probability-based design was used to produce estimates of seat belt use for the entire state. Site selection was based on three different objectives: (1) to ensure adequate counts to enable statewide, county, and district level estimates of seat belt use; (2) to ensure a mix of roadway types, volumes, and locations (urban, suburban, and rural); and (3) to allow for comparison across different time periods. Selection was based on random selection of tow types of roadways: high volume and low volume roads. Due to resource constraints, some of the extremely low volume sites were excluded from the final sample. The sites were distributed across each county proportionate to population. Site selection has remained relatively the same since 1986 to ensure compatibility of date across time.
Stratified Sample Estimation of Hawaii Seatbelt Use Rates. In designing stratified sample estimates, the model used was borrowed from Cochran (1977) and Kish (1965). The main stratum used was by island (County) because observation teams required air travel to neighboring islands in order to capture vehicles in other areas of the state. The sampling unit in this study was defined in terms of the total number of registered vehicles in each county. Table 1 provides the number of units, the number selected, and the respective probabilities of selection:
| County | Registered Autos | Sample Size (nh) | Probability of Selection |
| Oahu | 484,761 | 23,670 | .0484 |
| Maui | 90,573 | 9,113 | .100 |
| Hawaii | 90,281 | 6,078 | .067 |
| Kauai | 39,078 | 5,208 | .133 |
The unbiased estimator of the population proportion, p
where the county stratum weight,
Wh=county registration
state registration
and where ph is the proportion of belt users in the stratum,
where f, the sampling fraction, is approximately 44,069/704,693–.06, and
Sampling Weights. Weights for the counties and the weighted state proportion of seatbelt users are given in Table 2. A complete table is provided in appendix 8 of this report.
| County | Stratum Weight | Stratum p | Contribution to State p [wt*p(usage)] | Sample size(nh) | Stratum Variance Component [(p)(1-p)]/nh/(nh-1) | Contribution to State Variance |
| Honolulu | 0.652905 | 0.8060 | 0.5262 | 23,670 | 0.1564 | 0.1021 |
| Maui | 0.130700 | 0.7554 | 0.0987 | 9,113 | 0.1848 | 0.0242 |
| Hawaii | 0.145133 | 0.8032 | 0.1166 | 6,078 | 0.1581 | 0.0229 |
| Kauai | 0.071260 | 0.8750 | 0.0624 | 5,206 | 0.1094 | 0.0078 |
| State Total | 0.8039 | 44,069 | ||||
| Overall Sampling Fraction= | State Variance | 0.0000035427 | ||||
| f/n = | 2.25667E-05 | Standard Deviation | 0.001882209 | |||
| Total Sample (N)= | 44,069 | Confidence Interval | 0.003689129 | |||
| Confidence Interval= | ±0.37% |
The overall weighted state proportion is 0.8039. Each county contributes stratum variance between 0.1094 and 0.1848 and the variance of the proportion estimated form the total state sample is less than 3.54 X 10 -6.
Observation requirements. Observation team spent 40 minutes during daylight hours at each site. This was enough time to capture the necessary cases for statistical analysis at low volume sites. Well documented seat belt use rate was obtained by two trained observing team members; one to observe all front-seat occupants while a second member recorded the data. Observers were instructed to record shoulder belt use of drivers and front seat occupants for all incoming traffic. Further, observers were instructed to only record vehicles for which they were certain of whether or not the occupants used seat belt. When observation became difficult due to heavy traffic volume, every nearest fifth vehicle was observed. Further, instructions were given to each observation team that specifies the exact site locations (by providing the main street with a reference street) and the direction of traffic to be observed. Team members were required to exercise extreme caution in parking vehicles, in conducting observations, and avoiding potentially dangerous situations. The risks of error and mistake were minimized by thorough preparation before, during, and after conducting observations. After scoring the results of observation on paper forms, data were then entered into the database for analysis at DURP using SAS, a statistical software package.
(1) differences by island;
(1) differences by factor;
(1) differences between day and night time period.
Table 3 presents a more detailed breakdown of the seat belt use rate by island (complete tables are found in appendix 1 through 5 of this report). It shows that drivers (81.6%) continue to have higher seat belt use percentage than front seat passengers (76.9%). Oahu continues to show the greatest difference in seat belt use between drivers (81.9%) and passengers (76.6%). On Maui, 76.9% of drivers and 72.4% of passengers were belted. Hawaii had 81.9% of drivers and 77.0% of passengers wearing seat belts. Kauai, which had the smallest difference among the four islands, had an 33.0% use rate for drivers and 86.2% for passengers. Site specific seat belt use rates and locations for all islands are found on maps located in appendix 9 of this report.
| ISLAND | DRIVER | PASSENGER | TOTAL | ||||
| . | Total | %Drivers Belted | Total | %Passenger Belted | Total Belted | Total Observed | %Total Belted |
| Oahu | 17,549 | 81.9% | 6,121 | 76.6% | 19,079 | 23,670 | 80.6% |
| Maui | 6260 | 76.9% | 2853 | 72.4% | 6,884 | 9113 | 75.5% |
| Hawaii | 4,045 | 81.9% | 2,033 | 77.0% | 4,882 | 6,078 | 80.3% |
| Kauai | 3,633 | 88.0% | 1,575 | 86.2% | 4,557 | 5,208 | 87.5% |
| Total | 31,487 | 81.6% | 12,582 | 76.9% | 35,402 | 44,069 | 80.3% |
Figures 3 through 6 show the overall seat belt use by year on the four major Hawaiian islands. Recent trends in seat belt use suggest an increase on Oahu, where the seat belt use rate has increased slightly from 79.8%(1998) to 80.6%(1999). The peak rate for Oahu, observed in 1991 was 87.7%(Figure 3). This was the highest use rate ever recorded in Hawaii.
For 1999, Maui’s use rate has declined to 75.5%. The peak year was in 1993 when occupants had a use rate of 81.7%(Figure 4). Since the survey began, Maui has consistently ranked the lowest in seat belt use.
The island of Hawaii, like Maui, has experienced a decrease in seat belt use in recent years. Its peak use rate was recorded at 83.2% in 1993(Figure 5). Currently, its seat belt use rate is 80.3%.
Kauai, however, has shown consistent gains in belt use rates throughout the survey period. Currently, 87.5% of vehicle occupants were belted, slightly less than the 87.7% previously reported in 1998(Figure 6).
The following section examines various factors that influence seat belt use in Hawaii. Figure 7 contains the use rates for drivers and front-seat passengers by vehicle type. Drivers and passengers in cars(83.2%) and sports utility vehicles (82.4%) were the most likely to be belted. These are followed by occupants of vans (80.5%). Truck occupants show a lowest seat belt use rates (69.2%). These results are consistent with previous surveys.
Figure 8 shows the difference in seat belt use rate by traffic volume. Seat belt use in low volume areas was 74.4%. In higher volume areas, the use rate increased to 80.9%.
Figure 9 contains information on seat belt use under different weather conditions. The belt use rate under sunny conditions was 80.0%. With rainfall, the use rate increased to 84.7%.
Figure 10 shows that occupants are more likely to wear seat belts as the speed limit increases.
Figure 11 shows that as the number of lanes increased, vehicle occupants were more likely to wear seat belts.
Figure 12 shows no significant difference in seat belt use between weekdays and weekends. Previous reports, however, have shown that occupants were more likely to use seat belts on weekdays than on weekends.
Table 4 gives a detailed breakdown of the seat belt use rate (the complete table is found in appendix 1 through 5 of this report).
| . | Total Observed | %Drivers Belted | Total Observed | %Passenger Belted | Total Belted | Total Observed | %Total Belted |
| VEHICLE TYPE | . | . | . | ||||
| Car | 19,232 | 84.8% | 7,994 | 79.6% | 22,673 | 27,226 | 83.2% |
| Truck | 6,092 | 71.3% | 2,017 | 62.8% | 5,614 | 8,109 | 69.2% |
| Van | 3,295 | 81.6% | 1,476 | 78.2% | 3,845 | 4,771 | 80.5% |
| Sport Utility | 2,865 | 82.2% | 1,092 | 81.9% | 3,264 | 3,957 | 82.4% |
| VOLUME | . | . | . | ||||
| Low | 2,888 | 75.5% | 1,020 | 71.4% | 2,911 | 3,908 | 74.4% |
| High | 28,599 | 82.2% | 11,562 | 77.4% | 32,491 | 40,161 | 80.9% |
| TIME PERIOD | . | . | . | ||||
| 7:00AM-10:59AM | 6,945 | 82.1% | 2,466 | 78.5% | 7,642 | 9,411 | 81.2% |
| 11:00AM-2:59PM | 15,229 | 81.6% | 6,407 | 76.9% | 17,367 | 21,636 | 80.2% |
| 3:00PM-6:59PM | 9,313 | 81.3% | 3,709 | 75.9% | 10,393 | 13,022 | 79.8% |
| WEATHER | . | . | . | ||||
| Sunny | 18,837 | 81.1% | 7,367 | 76.9% | 20,963 | 26,204 | 80.0% |
| Overcast | 9,132 | 81.2% | 3,598 | 75.1% | 10,124 | 12,730 | 79.5% |
| Raining | 3,414 | 86.0% | 1,570 | 81.9% | 4,224 | 4,964 | 84.7% |
| SPEED | . | . | . | ||||
| Below 25 MPH | 2,715 | 73.8% | 1,051 | 72.2% | 2,765 | 3,766 | 73.4% |
| 25-34 MPH | 6,377 | 77.9% | 2,334 | 73.5% | 6,689 | 8,711 | 76.7% |
| 35-44 MPH | 11,542 | 82.5% | 5,133 | 77.2% | 13,488 | 16,675 | 80.8% |
| 45-54 MPH | 7,836 | 84.8% | 2,921 | 79.8% | 8,976 | 10,757 | 83.4% |
| 55 or more MPH | 2,665 | 85.7% | 1,008 | 81.1% | 3,102 | 3,673 | 84.4% |
| LANES | . | . | . | ||||
| One Lane | 13,136 | 79.0% | 5,874 | 75.7% | 14,832 | 19,010 | 78.0% |
| Two Lanes | 9,013 | 81.4% | 3,731 | 75.8% | 10,176 | 12,744 | 79.8% |
| Three Lanes | 7,954 | 85.4% | 2,569 | 80.0% | 8,853 | 10,523 | 84.1% |
| Four Lanes | 1,384 | 86.3% | 408 | 84.8% | 1,541 | 1,792 | 85.9% |
| WEEK | . | . | . | ||||
| Weekday | 18,763 | 81.8% | 6,438 | 76.6% | 20,286 | 25,201 | 80.5% |
| Weekend | 12,724 | 81.4% | 6,144 | 77.2% | 15,116 | 18,868 | 80.1% |
Figure 13 shows that the likelihood of wearing a seat belt during the day (86.6%) exceeds that of night (79.4%). This finding is consistent with past reports where seat belt use rates during daylight hours exceeded those of night. A complete table is found in appendix 6.
Table 5 shows the breakdown of the seat belt use rate between day and night time periods.
| . | Total Observed | %Drivers Belted | Total Observed | %Passenger Belted | Total Belted | Total Observed | %Total Belted |
| OVERALL BELT USE | . | . | . | ||||
| Oahu (Day) | 2,345 | 86.7% | 784 | 86.2% | 2711 | 3,129 | 86.6% |
| Oahu (Night) | 2,081 | 80.9% | 815 | 75.7% | 2,329 | 2,932 | 74.4% |
| VOLUME | . | . | . | ||||
| Low(Day) | 325 | 91.0% | 94 | 86.1% | 377 | 419 | 89.9% |
| High(Day) | 2,049 | 86.2% | 690 | 86.2% | 2,363 | 2,739 | 86.2% |
| High(Night) | 2081 | 80.9% | 851 | 75.7% | 2,329 | 2,932 | 79.4% |
| BY VEHICLE TYPE | . | . | . | ||||
| Car(Day) | 1,565 | 89.2% | 541 | 88.1% | 1,874 | 2,106 | 88.9% |
| Car(Night) | 1,411 | 84.0% | 594 | 75.2% | 1,633 | 2,005 | 81.4% |
| Van(Day) | 265 | 85.9% | 108 | 89.8% | 330 | 373 | 88.4% |
| Van(Night) | 248 | 79.8% | 105 | 79.0% | 281 | 353 | 79.6% |
| Truck(Day) | 315 | 75.8% | 77 | 68.8% | 292 | 392 | 74.4% |
| Truck(Night) | 236 | 66.9% | 81 | 71.6% | 216 | 317 | 68.1% |
| Sport Utility(Day) | 200 | 83.0% | 58 | 84.4% | 215 | 258 | 83.3% |
| Sport Utility(Night) | 186 | 76.3% | 71 | 80.2% | 199 | 257 | 77.4% |
| BY SPEED | . | . | . | ||||
| 25-34 MPH(Day) | 876 | 84.7% | 247 | 86.2% | 955 | 1,123 | 85.5% |
| 25-34 MPH(Night) | 735 | 76.3% | 405 | 73.0% | 857 | 1,140 | 75.1% |
| 35-44 MPH(Day) | 961 | 87.6% | 365 | 85.4% | 1,154 | 1,326 | 87.0% |
| 35-44 MPH(Night) | 835 | 83.9% | 252 | 75.4% | 891 | 1,087 | 81.9% |
| 45-54 MPH(Day) | 508 | 88.7% | 172 | 87.7% | 602 | 680 | 88.5% |
| 45-54 MPH(Night) | 511 | 82.5% | 194 | 81.9% | 581 | 705 | 82.4% |
| LANES | . | . | . | ||||
| Two Lanes(Day) | 1,314 | 86.9% | 380 | 87.8% | 1,476 | 1,694 | 87.1% |
| Two Lanes(Night) | 583 | 81.4% | 176 | 81.2% | 618 | 759 | 81.4% |
| Three Lanes(Day) | 1,031 | 86.6% | 404 | 84.6% | 1,235 | 1,435 | 86.0% |
| Three Lanes(Night) | 1,498 | 80.7% | 675 | 74.3% | 1,711 | 2,173 | 78.7% |
| WEEK | . | . | . | ||||
| Weekday(Day) | 1,818 | 85.7% | 563 | 85.4% | 2,040 | 2,381 | 85.6% |
| Weekday(Night) | 1,554 | 81.7% | 561 | 76.1% | 1,697 | 2,115 | 80.2% |
| Weekend(Day) | 527 | 90.3% | 221 | 88.2% | 671 | 748 | 89.7% |
| Weekend(Night) | 527 | 78.5% | 290 | 75.1% | 632 | 817 | 77.3% |
Hence, these findings suggest the following recommendations:(1) Further study is needed on locational differences. For instance, why does Kauai continue to have an increasing seat belt use rate since 1985, while other islands have not?
(2) As an outcome of current studies, area-specific education and enforcement policies may need to be formulated and implemented. For instance, the seat belt use rate in remote Waialua has stabilized at 67% in the past two years and should be targeted for special attention.
(3) Statewide programs of enforcement, public education, and public information should be developed to increase seat belt use. For seat belts to be effective in reducing injuries and fatalities, motorists need to use them.
(4) Further analysis is needed to find out why seat belt use rate during daylight hours exceeds that of night. Though on would expect seat belt use to be higher at night when visibility is lower, this was not the case. The study found that fewer seat belts were worn during night hours than during day hours. A possible explanation is that, disguised under low visibility conditions, people choose to take a chance that patrol officers would not notice lack of compliance with seat belt use regulations.
APPENDIX 1-9