SAFETY BELT SURVEY DESIGN IN ILLINOIS Mehdi Nassirpour, Ph.D. Illinois Department of Transportation
Mar 27, 2015
SAFETY BELT SURVEY DESIGN IN ILLINOIS
Mehdi Nassirpour, Ph.D.Illinois Department of Transportation
OBJECTIVE OF THE SURVEY
•To design a multi-stage cluster random sample in order to estimate the statewide safety belt usage rate in Illinois.
Division of Traffic Safety at IDOT
Census Tracts Census Tracts Census TractsCensus Tracts
Road TypeRoad TypeRoad Type
HVR HVR LVRHVR LVR HVR LVRLVR
City of Chicago77 Communities
Cook CNo Chicago
30 Townships
Other Counties101 Counties
Urban Rural
Road Type
SAMPLE DESIGN SUMMARY IN ILLINOIS
Stage 1
Stage 2
Stage 3
ALLOCATION OF SAMPLE SITES
Total Sample (258 Sites
12 Counties (166 Sites)
Cook County (92 Sites)
Cook CountyNo Chicago (41 Sites)
City of Chicago (51 Sites)
STAGE 1 - PSU SELECTION
• Sampling Units - 20 PSU– 5 Community Areas in the City of Chicago– 3Townships in Cook County– 11 Counties (DuPage County was selected twice)– 4 out of 11 counties were rural counties. – 47 counties were eliminated due to a small number of
VMTs and population.– 54 counties contains about 90 percent of VMT and
population• Selection - Probability Proportional to Vehicle Miles of
Travel and Population.• Selection is systematic (1/Ns)
– After Random Start
SAFETY BELT LOCATIONS IN IL
MCHENRY COUNTYLOCATIONS (6)
LAKE COUNTYLOCATIONS (31)
KANE COUNTYLOCATIONS (14)
WILL COUNTYLOCATIONS (18)
CHICAGO LOCATIONS (49)
COOK COUNTYLOCATIONS (40)
DUPAGE COUNTYLOCATIONS (42)
CHAMPAIGN COUNTYLOCATIONS (9)
ROCK ISLANDCO. LOCATIONS (7)
PEORIA COUNTYLOCATIONS (9)
MONTGOMERGY COUNTYLOCATIONS (2)
ST. CLAIR COUNTYLOCATIONS (14)
Division of Traffic Safety at IDOT
STAGE 2 - CENSUS TRACT SELECTION
• Census Tracts – A total of 96 census tracts (10-12
percent of total tracts) within those selected counties were selected randomly.
• Selection - Probability Proportional to Population.
• Selection is systematic (1/Ns).– After Random Start
Division of Traffic Safety at IDOT
STAGE 3 - ROAD TYPE SELECTION
• Road Type – All roads (Interstates, freeways, US/IL
highways, and residential streets) within selected tracts were identified using tract maps.
– All Interstates and IL/US highways within these tracts were selected with certainty and other roads were selected randomly.
Division of Traffic Safety at IDOT
PROCEDURE USED TO ASSIGN OBSERVATION SITES TO OBSERVATION TIME PERIOD
• Selected roads randomly assigned to time of day and day of week.
• To minimize travel time and distance required to conduct the observations, the sampled sites were grouped into 29 geographic clusters, with each cluster containing about between 6 and 12 road sites.
• Each cluster is randomly (without replacement) assigned to a day of the week. Each day was divided into eight one-hour time periods between 7:00am to 6:30pm when the light was adequate for observation.
Division of Traffic Safety at IDOT
Weighting Procedures
Wctr = 1/Pctr
Pctr = Pc . Pt/c . Pr/ct
c identifies a sample of countyt identifies a sample of tractr identifies a sample of road sitePc probability of selecting county cPt/c probability of selecting tract t within
selected county cPr/ct probability of selecting road r, conditional
to the tract and county
Division of Traffic Safety at IDOT
FORMULA FOR ESTIMATING SAFETY BELT USAGE RATE
ctr
ctr
YctrWctr
XctrWctrR
.
.
R = Safety belt usage rateWctr = weight associated with road r in tract t in county c;Xctr = number of safety-belted front seat occupants of passenger cars and pick-up trucks observed in county c, tract t, and road site r; andYctr = number of front seat occupants observed in county c, tract t, and road site r.
Division of Traffic Safety at IDOT
FORMULA FOR ESTIMATING SAFETY BELT USAGE RATE
.
Octr
LctrFctr
FctrWctrctrW
YctrctrW
XctrctrWR
CTR
CTR
.'
.'
.'Fctr = adjustment for lanes not observed in road site r, tract t in county c;
Lctr = total number of lanes in road site r, tract t in county c; and
Octr = total number of observed lanes in road site r within tract t in county c
VARIANCE ESTIMATION
.
Estimated safety belt usage rate is subject to sampling error since a small number of front seat occupants were observed. These sampling errors are expressed as coefficients of variation (relative error), that are the ratios of the relative standard errors of estimates to the estimates themselves.
Division of Traffic Safety at IDOT
VARIANCE ESTIMATION PACKAGE
.
PC CARP from the Iowa state University was used to estimate the coefficient of variation for a ratio from a stratified cluster sample.
PC CARP is a menu driven program and is designed to construct estimates and standard errors for totals, means, ratios, differences of ratios, and entities in two-way tables for multi-stage stratified sample. The basic input for the program is the stratum identification, cluster identification, weight, and data vector for a set of observations.
Division of Traffic Safety at IDOT
VARIANCE ESTIMATION FORMULA
.
(Yijkl, Xijkl, Yijk2, Xijk2, ...., Yijkp, Xijkp)
where i= 1, 2, ....l j = 1, 2, ...,nj; k = 1, 2,...., mij; i is the stratum identification, j is the cluster identification, k is the element-within-cluster identification.
The above vector is used in the analysis and estimator is
.....3,2,1,.
.
1 1 1
1 1 1 psXijksWij
YijksWij
Xs
YsRs L
i
nij
j
mij
k
L
i
nij
j
mij
k
95 Percent ConfidenceInterval
EstimatedUsage Rate
Coefficientof Variance
StandardError
Lower Limit%
Upper Limit%
Statewide (258) 76.2 4 3 70 82
RegionsCity of Chicago (46) 69.1 8 5 59 79Cook County (40) 70.9 5 4 63 79Other Counties (172) 79 5 3 73 85
Road TypeResidential (190) 71.8 5 3 66 78U.S./Illinois Highways(40)
76.3 2 3 70 82
Interstate Highways (28) 87.3 4 3 81 93
Time of DayMorning Rush Hours (55) 75.7 5 4 68 84Noon Rush Hours (45) 74.8 5 3 69 81Evening Rush Hours (23) 80.9 7 5 71 91Non-Rush Hours (135) 76.2 2 1 74 78
Day of WeekWeekends (115) 79.1 4 3 73 85Weekdays (143) 73.5 5 3 68 79
Safety Belt Usage Rates in Illinois, July 2003
PUBLIC OPINION SURVEY OF “CLICK IT
OR TICKET” CAMPAIGN
Mehdi NassirpourIllinois Department of Transportation
SURVEY RESEARCH CENTER
• In the Spring of 2003, the Illinois Department of Transportation contracted with the Survey Research Office at the University of Illinois at Springfield to conduct two telephone surveys. Two of the surveys were conducted before and after a major seat belt initiative (that involved both media and enforcement-related activities that occurred during a time period surrounding Memorial Day Weekend, 2003. Interviewing for the “pre-test” (before) survey was conducted from April 25 to May 4, 2003, and interviewing for the “post-test” (after) survey was conducted from June 3 to 18, 2003.
Division of Traffic Safety at IDOT
OBJECTIVES
• To collect statewide data and information on public awareness and acceptance of the safety belt program during the “Click it or Ticket” program in Illinois at five points (before, during and after the campaign).
• To describe motorists’ opinions on safety belt use and related issues, controlling for demographics during the baseline and follow-up data collection periods.
• To identify key factors to assist the Division of Traffic Safety to measure effectiveness of the overall “Click it or Ticket” program and services in Illinois using the pre and post approach.
Division of Traffic Safety at IDOT
CLICK IT OR TICKET CAMPAIGN
• “Click It or Ticket” is a high visibility, massive enforcement effort designed to detect violators of Illinois traffic laws with special emphasis on occupant protection. An intense public information and education campaign was run concurrently with the enforcement blitz to inform the motoring public of the benefits of seat belt use and of issuing tickets for seat belt violations. The goal of the “Click It or Ticket” campaign is to save lives and reduce injuries resulting from motor vehicle crashes by increasing the safety belt usage rate in Illinois by at least 3-5
percentage points.
Division of Traffic Safety at IDOT
MEDIA/ENFORCEMENT
Several media events were held throughout the state on April 28, 2003 announcing the enforcement efforts planned for May 15 - June 1, 2003.
A total of 421 local agencies as well as Illinois State Police and Secretary of State Police agencies participated in the enforcement campaign and issued 46,434 citations. Over 26,085 man-hours were conducted a variety of enforcement efforts.
Division of Traffic Safety at IDOT
STRATIFIED RANDOM SAMPLE
Chicago Metro Downstate
City of Chicago
Chicago Area
Suburbs
North and
Central Illinois
Southern Illinois
Total Sample
Division of Traffic Safety at IDOT
SAMPLING METHODOLOGY
• Random samples of telephone numbers were purchased for each of the four stratification areas (City of Chicago, Chicago Suburban Counties, North/Central Illinois, and Southern Illinois). Each telephone number in the samples was called a maximum of six times, at differing times of the week and day. Within households, interviewers asked for the youngest licensed driver 75 percent of the time, because earlier experience showed that we under-represent younger drivers. In the other 25 percent of the time, interviewers asked for a licensed driver who was male/female (varying at random) and who had the next birthday. Replacements were accepted if that designated household member was not available.
Division of Traffic Safety at IDOT
DISTRIBUTION OF THE SAMPLES (Pre=560 and Post=623)
Pre-Test Post-Test
Chicago 113 160
ChicagoSuburbanCounty
228 177
DownstateCounties
219 286
North/CentralIllinois
137 166
SouthernIllinois
83 120
Division of Traffic Safety at IDOT
SAMPLING ERROR
• The sampling error for the statewide results is under +/- 5 percent (at the 95th confidence level) for both the pre-test survey (+/- 4.2%) and the post-test survey (+/- 4.0%). The error for subgroups is, of course, larger.
Division of Traffic Safety at IDOT
SAMPLE WEIGHTS
• The statewide results for both surveys have been weighted to arrive at a proper distribution by region and by gender.
Division of Traffic Safety at IDOT
The largest difference appears to be that for employment status, where the post-test survey has fewer full-time employees (37%, compared to 45% in the pre-test). The post-test respondents are also somewhat younger and have somewhat more minority respondents. The post-test sample also has somewhat more households with more than one person of driving age in the household.
DEMOGRAPHIC CHARACTERISTICS OF PRE AND POST SAMPLES
Division of Traffic Safety at IDOT
When Driving, how often do you wear your seat belt (composite of shoulder and lap)?
81.6%
9.5%4.6% 1.8% 2.5%
82.4%
9.1%4.7% 3.0% 0.8%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
All of the Time Most of theTime
Some of theTime
Rarely Never
Pre-Mobilization Post Mobilization
Data Source: Telephone Survey
N=528 for Pre Mobilization
N=576 for Post Mobilization
When was the last time you did not wear your seat belt when driving?
11.5% 11.4%5.9%
3.5%
59.7%
8.0%5.8% 4.7%
8.2%11.7%
8.4%
61.1%
0%
10%
20%
30%
40%
50%
60%
70%
Within the lastday
Within the pastweek
Within the pastmonth
Within the pastyear
More than ayear ago
(always wear)
DK/NA
Pre-Mobilization Post Mobilization
Data Source: Telephone Survey
In the past thirty days, has your use of seat belts when driving increased, decreased, or stayed the same?
5.4%0.5% 0.6%
10.4%
0.3%
88.8%
0.6%
93.5%
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
70.0%
80.0%
90.0%
100.0%
Increased Decreased Stayed the same DK/NA (Includes "newdriver")
Pre-Mobilization Post Mobilization
Data Source: Telephone Survey
According to Illinois state law, can police stop a vehicle if they observe a seat belt violation or do they have to observe
some other offense in order to stop the vehicle?
69.1%
19.6%
11.3%
70.9%
20.6%
8.4%
0%
10%
20%
30%
40%
50%
60%
70%
80%
Can stop just for seatbelt violation
Must see other offensefirst
DK/NA
Mobilization
Pe
rce
nt
Pre-Mobilization Post Mobilization
Data Source: Telephone Survey
In the past thirty days, have you seen or heard anything about the police setting up roadside safety checks where
they stop to check drivers and vehicles?
20.0%
77.4%
2.6%
44.7%
53.2%
2.1%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
Yes, have seen/heard No, have not DK/NA
Pre-Mobilization Post Mobilization
Data Source: Telephone Survey
In the past thirty days, have you seen or heard of any special effort by police to ticket drivers in your community for seat
belt violations?
12.2%4.3%
48.9% 48.1%
2.9%
83.5%
0%
10%
20%30%
40%
50%
60%70%
80%
90%
Yes, have seen orheard
No, have not DK/NA
Mobilization
Pe
rce
nt
Pre-Mobilization Post Mobilization
Data Source: Telephone Survey
Figure 21: In the past thirty days, have you seen or heard any messages that encourages people to wear their seat belts?
63.2%
35.7%
1.0%
82.2%
17.0%
0.8%0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
Yes, have seen/heard ofsuch
No, have not DK/NA
Pre-Mobilization Post Mobilization
Data Source: Telephone Survey
Do You Recall Hearing, or Seeing, the Following Slogans in the PastThirty Days?
Pre-Mobilization Post Mobilization Percent Change
Police in Illinois Arrest Drunk Drivers. 29.3% 23.6% -5.7%
Wanna drink and drive? Police in Illinois will show you thebars.
29.3% 48.0% 18.7%
Smart motorists always respect trucks. 10.9% 11.3% 0.4%
Help promote wreck-less driving. Don't hang out in the NoZone.
9.6% 9.0% -0.6%
Friends don't let friends drive drunk NA 89.0% NA
Stupid Driving Tricks /or/ Don't Pull a Stupid Driving Trick 5.0% 6.7% 1.7%
Click It or Ticket 66.7% 84.6% 17.9%
Buckle Up America 47.6% 53.2% 5.6%
Drive Hammered…Get Nailed! NA 30.1% NA
Children in Back 21.8% 23.6% 1.8%Cell phones save lives. Pull over and report a drunk driver. 39.2% 46.0% 6.8%
Drive smart, drive sober. 64.9% 66.9% 2.0%
Operation A-B-C 6.6% 5.1% -1.5%
You Drink, You Drive, You Lose NA 55.2% NA
Drink and Drive? Police in Illinois have your number. NA 22.3% NA
Data Source: Telephone Survey