Request for Approval of Information Collection Under Generic Clearance
OMB No. 2127-0646
2127-0646 Category: Demonstration Surveys
2127-0646 Sub-Category: High Visibility Alcohol Enforcement Demonstration Projects
Project Title: Improving the General Deterrence Effects of Sobriety Checkpoints
Project Contact: Alan Block, [email protected], 202-366-6401
Project Summary:
The purpose of this project is to assess an approach to increasing the visibility of sobriety checkpoints. It stems from complaints that sobriety checkpoints are not as effective as they should be because most people do not see them (i.e., they typically are conducted during late-night hours when traffic flow is light). This project will conduct sobriety checkpoints at two intervention sites during earlier hours (6:00 pm – 9:00 pm) than the traditional 10:00 pm to 2:00 am time frame. The project will also assess whether strategically located signs alerting the public to the checkpoints enhances public awareness of the checkpoint activity. Westminster Colorado and Jackson County Missouri have been selected as the two intervention sites. They will conduct the intervention at two different times separated by 5 or more months, with the field period for each of the two intervention waves being 2 weeks (3 evenings of checkpoint activity during two successive weeks). The use of signs will be rotated, with Westminster using the signs during the first intervention wave, and Jackson County using the signs during the second intervention wave. There will also be a comparison site in both Colorado (Fort Collins) and Missouri (St. Charles County) that will not be conducting the interventions.
Data to assess public awareness of the sobriety checkpoint activity, and perceived risk of an alcohol-impaired driver being stopped by police, will be obtained through telephone interviews, both with persons on landline and cell phones (this addresses a concern expressed in the terms of clearance regarding cell phone only respondents). A pre/post design will be utilized whereby interviews are conducted prior to, and immediately following, each intervention wave at each intervention site. Each comparison site will be surveyed at the same time as its matched intervention site.
Items To Address Per The Terms Of Clearance:
The exact questionnaire to be used. Included as Attachment 1.
Designated survey population and sampling design. The survey population will be drivers age 18 and older currently residing in the demonstration and comparison communities. Ages 18 through 39 will be over-sampled because of the greater contribution of that age group to the drinking and driving problem. A geographic area will be defined for each project site by zip code or area name, which will then be used to identify associated phone exchanges in both landline phone banks and cell phone banks. Systematic random selection of sample will be made from those exchanges. Fifty percent of the sample will be drawn from landline phone banks, and the remaining 50% from cell phone banks. For both the landline and cell phone samples, 50% will be administered the over-sample screener that will limit eligibility for participation to ages 18 through 39. The remaining 50% will be screened for 18 and older. As indicated in #3 below, the goal of the over-sampling is a 40%/60% split between those 18-39 and those 40 and older, respectively. Raising the over-sampling percentage to 50% is intended to compensate for traditionally lower response rates among younger age groups.
Completed interviews from the landline sample will come from persons who are landline only as well as persons who have both a landline phone and a cell phone. Completed interviews from the cell phone sample will come from persons who are cell phone only as well as persons who have both a cell phone and a landline phone in their place of residence. Since exchanges for cell phone numbers may radically differ from the cell phone owner’s place of residence, the cell phone sample will be screened to make sure they reside in the geographic area defined for the site. Those reporting they currently live outside the defined geographic area will be ruled ineligible to participate in the survey.
Whether or not the sample will be stratified. Over-sampling of ages 18-to-39 will be conducted to achieve a 40%/60% split in the percentage of the sample that is 18-to-39 and 40-and-older, respectively. The over-sampling will be achieved by having each sample composed of a systematically selected cross-sectional sample, plus an independently drawn over-sample screened to be composed entirely of persons 18-to-39. The independently drawn samples will be administered different age screening questions. The screener for the cross-sectional sample is included with the questionnaire in Attachment 1. The screener for the 18-39 oversample is included as Attachment 2.
Specific sample sizes. The sample size will be 500 completed interviews per site per survey wave. The 500 sample size was selected as a balance between cost and precision. The core of the analysis will be to compare differences between survey waves on individual questions to detect whether the intervention corresponded with a statistically significant increase in awareness of checkpoints and perceived risk of impaired drivers being stopped by police. If the design was simple random sampling, the pre/post differences would have to be 6.2 percentage points to be statistically significant. Since the proposed sampling involves a complex design, a somewhat larger difference would be required to achieve statistical significance. Past experience on NHTSA surveys suggests a modest design effect (e.g., an average difference in confidence intervals of 17% on selected items in the 2007 MVOSS). This intervention is not intended to achieve a difference smaller than several percentage points because the level of effort needed to sustain the intervention would be too costly for such a small gain. Conversely, pre/post differences to achieve statistical significance can’t be set overly high due to limitations on the extent to which a study strictly focused on checkpoint visibility can penetrate public awareness. Therefore, pre/post differences slightly in excess of 6 percentage points to reach statistical significance are considered acceptable for this project. Each site will be administered a survey wave prior to, and immediately following, each of their two intervention periods for a total of four survey waves. The total number of completed interviews will be: 4 Sites x 4 Survey Waves x 500 Interviews = 8,000 Completed Interviews. Average interview time will be 10 minutes or less. Eight thousand interviews with average duration of 10 minutes equates to a burden of 1,333 hours to the public over the course of a year.
Scheduling of each survey instance. The requested survey field periods are:
Westminster Colorado and its Comparison Site
June 16 to July 1 2010 (Pre-Intervention I Survey)
August 1 to August 18, 2010 (Post-Intervention I Survey)
March 7 to April 4, 2011 (Pre-Intervention II Survey)
April 17 to May 16, 2011 (Post-Intervention II Survey)
Jackson County Missouri and its Comparison Site
September 28 to October 18, 2010 (Pre-Intervention I Survey)
November 1 to November 24, 2010 (Post-Intervention I Survey)
March 7 to April 4, 2011 (Pre-Intervention II Survey)
April 17 to May 16, 2011 (Post-Intervention II Survey)
The project requirements were reviewed to determine if all interviewing could be scheduled later than the Census data collection period ending August 31, 2010. Among the project requirements was that there not be any national alcohol or seat belt campaign activity for at least a few weeks prior to the start of any survey data collection wave in order not to contaminate the data with activity outside the project. The survey field period itself would then take a minimum of about 3 weeks, which meant that the early evening checkpoints could not be implemented any sooner than 6-7 weeks after the end of a national Crackdown or Mobilization. It was determined that the Intervention I survey waves could be conducted later than August 31 in Jackson County Missouri and its comparison site. However, our contacts in Westminster Colorado indicated that the timing of the Wave 1 sobriety checkpoint interventions in Jackson County (Oct. 21, 22, 23, 28, 29, 30) would not be suitable for them because the weather conditions would be too unpredictable that late in the year. Therefore, permission is requested to conduct the 1,000 Pre-Intervention I interviews in Colorado during the second half of June 2010, and the 1,000 Post-Intervention I interviews during August. That timing would leave a gap of 2-3 weeks between the end of the national Click It or Ticket Campaign and the beginning of the first survey wave, and complete the intervention and survey waves before the start of the National Alcohol Crackdown. The timing also addresses contamination with likely anti-DWI activities at the sites during the July 4 weekend (i.e., completing most if not all the baseline interviewing before July 4, and waiting a few weeks after July 4 before conducting the intervention). This was the best schedule that could be worked out given the preponderance of traffic safety activity.
6. Goals of each survey instance. Survey waves will be conducted prior to, and following, each intervention in order to identify if the intervention corresponded with an increase in reported sightings of sobriety checkpoints, as well as any change in perceived risk of impaired drivers being stopped by police. However, if the surveys were conducted only at intervention sites, one could not say for sure whether a detected change was due to the intervention, or was due to some factor independent of the intervention. Therefore, surveys at matched times will also be conducted at comparison sites that are not conducting the intervention in order to provide a standard against which to compare the intervention site findings. Each comparison site will be in the same State as an intervention site, therefore being subject to the same drinking and driving laws. The decision to have 2 intervention sites and 2 intervention waves is based on the signage conditions we want to impose, whereby each site conducts one intervention wave augmented by signage and one intervention wave without, with the order of the conditions at each intervention site being the reverse of the order at the other intervention site.
Call attempts will be staggered across weekday, weeknight, and weekend hours, with no calls made before 9 am or after 9 pm. A maximum of 10-dialing attempts will be made to reach and complete an interview. The only time this 10-dialing attempt ceiling will be exceeded is if the respondent requests to be contacted again at a specific time. In the event that someone refuses participation, that refusal will be treated as final and no call back will be made to the respondent at some later time. That said, if a respondent goes part-way through a survey and states that he/she cannot complete the survey but is willing to do so at a later time/date, that callback request will be honored. If they go part way through a survey, but then refuse to finish, that refusal will be treated as final and not a complete – despite the fact there will be partial data. Lastly, in instances where the call attempt reaches an answering machine, a brief (<10sec) message will be left giving the reason for the call and saying that we will call back
7. How the data will be analyzed. The core of the analysis will be comparison of response percentages to identical questionnaire items between the paired pre- and post-survey waves surrounding each intervention period to identify statistically significant differences. Data from comparison sites will be similarly analyzed, and their results compared to the intervention sites. To assess the signage conditions, the response percentages at each site associated with the first intervention wave will be compared to the response percentages associated with the second intervention wave at the same site.
The basic approach to weighting the data will be to adjust the achieved samples so that they approximate key demographics of the selected sites according to Census figures (sex, age, race/ethnicity). Since the study will be unable to obtain this demographic data for drivers only, the study will use the general population figures that are available. If there are indications of substantial differences between the breakout of the samples by phone status versus what would be expected within the current environment, then further adjustment may be applied. However, there are no data to provide a precise barometer of what the phone status distribution should be at the community level, and available State figures are likely too old to use for weighting purposes and may not be representative of the selected communities anyway. Therefore, it would take an obvious discrepancy with State and national figures for this study to consider weighting on this factor.
Complexities in how the public are mixing the different technologies, and the implications for respondent contact and entry into the sample, recommend extreme caution regarding any further weighting adjustments. While this study is willing to consider additional steps in the weighting process, the goal will be to limit weighting adjustments as much as possible. Because the objective of the study is to determine if there is a difference between independently selected samples in responses to identical questions, the most important aspects of the data collection are coverage within each of the survey waves, and the use of identical procedures and data collection instruments during each survey wave. Coverage in each of the survey waves will be persons of a specified age range living in households who have phone service, which will exclude the approximately 2% of the population not having phone service. While some population characteristics will be under-represented or over-represented in the obtained samples, that bias should be fairly consistent across the different survey wave samples due to the consistency of the data collection methods.
There is little to provide guidance as to expected response rates for community-level survey activity. We do expect the response rates for the landline samples to be higher than the response rates for the cell phone samples based on what others have found. In addition, NHTSA’s National Alcohol Crackdown surveys, which are of similar length and content and conducted entirely on landline, obtained response rates of 30-35% last year, with young adults being closer to 25%. Given the contracted field period with which we’ll need to operate, the combined cell/landline response rate over the course of the study may be between 20% and 30%. But we may instead find that the communities that we will be surveying are different from the norm. Of import to the study is that we incorporate procedures to encourage higher response rates, such as well-designed call schedules.
ATTACHMENT 1
QUESTIONNAIRE WITH CROSS-SECTIONAL SAMPLE SCREENER
OMB control No. 2127-0646
Expiration Date: 8/31/2011
Hello, my name is _______ and I am calling on behalf of the U.S. Department of Transportation. We are conducting a study about Americans’ driving habits and opinions. The interview is voluntary, and will take less than 10 minutes to complete. We will not collect any personal information that would allow anyone to identify you. Please note that an agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a current valid OMB control number. The OMB control number for this study is 2127-0646. Could we begin now?
(IF CALLING FROM CELL SAMPLE) SCREENING QUESTIONS
Cell Screen Q1. Are you driving or in any other situation that would make it unsafe to talk at this time?
IF YES, TERMINATE IMMEDIATELY AND CALL BACK AT A FUTURE TIME.
Cell Screen Q2. Are you 18 years of age or older?
Yes
No IF NO TERMINATE THE INTERVIEW
Cell Screen Q3. Did I reach you on a landline, or on a cell phone?
Cell Phone
Landline
Refused IF REFUSED TERMINATE INTERVIEW
Cell Screen Q4. Do you currently reside in the [NAME OF DEMONSTRATION /COMPARISON COMMUNITY] area?
Yes
No IF NO TERMINATE THE INTERVIEW
Cell Screen Q5. Do you use both a cell phone and a landline phone for your home telephone service, or do you exclusively use a (INSERT CSQ3 - cell phone/landline phone)?
Use Landline only
Use Cell phone only
Use Both landline and cell phone
Don’t know
Refused
(IF CALLING FROM LANDLINE SAMPLE) SCREENING QUESTIONS
Landline Screen Q1. Are you driving or in any other situation that would make it unsafe to talk at this time?
IF YES, TERMINATE IMMEDIATELY AND CALL BACK AT A FUTURE TIME.
Landline Screen Q2. May I please speak with the person in your household who had the most recent birthday AND is at least 18 years old?
Yes – The same person who answered the phone
Yes – Eligible respondent summoned by person who answered the phone
Not available - Eligible respondent not presently available IF NOT AVAILABLE, CALL BACK
No IF NO TERMINATE THE INTERVIEW
Landline Screen Q3. Did I reach you on a landline, or on a cell phone?
Landline
Cell Phone
Refused IF REFUSED TERMINATE INTERVIEW
Landline Screen Q4. Do you use both a landline phone and a cell phone for your home telephone service, or do you exclusively use a (INSERT LLSQ3 - landline phone/cell phone)?
Use Landline only
Use Cell phone only
Both landline and cell phone
Don’t know
Refused
MAIN SURVEY QUESTIONS
Q1. How often do you drive a motor vehicle? Is that…. (Read list, enter one response)
Every day
Almost every day
A few days a week
A few days a month
Never IF NEVER TERMINATE INTERVIEW
Don’t know
Refused
Q2. In the past 30 days, have you seen a DUI/DWI checkpoint in [NAME OF DEMONSTRATION/COMPARISON COMMUNITY] where drivers were briefly stopped by police at a set point on a roadway to check for drunk driving or driving under the influence? DUI stands for driving under the influence/DWI stands for driving while intoxicated. (Note: use DUI in Colorado and DWI in Missouri).
Yes
No IF NO GO TO Q6
Don’t Know IF DON’T KNOW GO TO Q6
Refused IF REFUSED GO TO Q6
Q3. Where did you see the DUI/DWI checkpoint? [Intervention site surveys would have exact locations that interviewers could check off] (Note: use DUI in Colorado, and DWI in Missouri)
Q4. What time of day or night did you see the checkpoint? (read the categories; allow for multiple responses)
Morning (6 am – 11 am)
Afternoon (Noon – 5 pm)
Early Evening (6 pm – 9 pm)
Late night (10 pm – after Midnight)
Don’t Know
Refused
Q5. In addition to seeing the checkpoint, did you read or hear about DUI/DWI checkpoints being conducted in [NAME OF DEMONSTRATION/COMPARISON COMMUNITY]? (note: use DUI in Colorado, and DWI in Missouri)
Ye s IF YES GO TO Q7
No IF NO GO TO Q8
Don’t Know IF DON’T KNOW GO TO Q8
Refused IF REFUSED GO TO Q8
Q6. In the past 30 days have you read or heard about DUI/DWI checkpoints being conducted in [NAME OF DEMONSTRATION/COMPARISON COMMUNITY]? (note: use DUI in Colorado, and DWI in Missouri)
Yes
No IF NO GO TO Q8
Don’t Know IF Don’t Know GO TO Q8
Refused IF Refused GO TO Q8
Q7. Did you read or hear about the checkpoints from (categories to be read, allow multiple response):
Radio or TV
Newspapers
From someone who told you
Other (Volunteered)
Don’t Know
Refused
Q8. In the past 30 days, did you see any road signs that said DWI/DUI checkpoints were being conducted in [NAME OF DEMONSTRATION/COMPARISON COMMUNITY]? (note: use DUI in Colorado, and DWI in Missouri)
YES
NO IF NO, GO TO Q11
Don’t Know IF DON’T KNOW GO TO Q11
Refused IF REFUSED GO TO Q11
Q9. (only for respondents that indicated they saw message on road signs in Q8). Do you recall where you saw the road signs?
Provide road sign locations (main roads, not cross streets):
Q10. Do you recall what message was on the road signs? (we are looking for coders to indicate whether respondents have captured the gist of the sign message, not necessarily the exact message)
Provide the message here: (e.g., DUI Checkpoints in Westminster)
Q11. If people in (NAME OF DEMONSTRATION/COMPARISON COMMUNITY) drive when over the legal limit for alcohol, how likely are they to be stopped by a police officer? Would they be…
Extremely Likely
Very Likely
Moderately Likely
Slightly Likely or
Not At All Likely To Be Stopped
Don’t Know (Volunteered)
Refused (Volunteered)
Q12. Do you think the chances of being stopped by the police for violating drinking and driving laws in [NAME OF DEMONSTRATION/COMPARISON COMMUNITY] have changed in the past month? That is, compared to a month ago, do you think a driver who is over the legal limit for alcohol is more likely, less likely or about as likely to be stopped by the police?
More likely
Less likely
About as likely
Don’t Know (Volunteered)
Refused (Volunteered)
Q13. I would like to ask your opinion regarding the use of DUI/DWI checkpoints. (note: use DUI in Colorado, and DWI in Missouri) Do you… (read list)
Strongly favor IF FAVOR, GO TO Q15
Somewhat favor IF FAVOR, GO TO Q15
Somewhat oppose IF OPPOSE, GO TO Q14
or Strongly oppose the use of DUI/DWI checkpoints? IF OPPOSE, GO TO Q14
No opinion (Volunteered) IF NO OPINION, GO TO Q15
Don’t know (Volunteered) IF DON’T KNOW, GO TO Q15
Refused (Volunteered) IF REFUSE, GO TO Q15
Q14. (Only for those opposed to sobriety checkpoints in Q13) Why do you oppose using sobriety checkpoints?
Open ended
Q15. Now I’m going to ask you a few questions about alcohol use. Answers to the following questions are anonymous.
About how often do you drink alcoholic beverages including beer, light beer, wine, wine coolers, liquor, or flavored malt beverages? Would you say you usually drink alcoholic beverages…
Every day
Nearly every day
Three or four days a week
One or two days a week
Two or three days a month
Once a month or less, OR
You never drink alcoholic beverages
Don’t know (Volunteered)
Refused (Volunteered)
Q16. During the past 30 days, have you driven a motor vehicle within two hours after drinking any alcoholic beverages, even if you had only a little?
Yes
No
Don’t Know
Refused
Male or Female? (Don’t ask)
Male
Female
Q17. How old were you on your last birthday?
Code respondent age
Don’t Know
Refused
Q18. Do you consider yourself to be Hispanic or Latino?
Yes
No
Not sure
Refused
Q19. Which of the following racial categories describes you? You may select more than one.
American Indian or Alaska Native
Asian
Black or African American
Native Hawaiian or other Pacific Islander
White
Other (Volunteered)
Refused
Q20. What is the highest grade or year of school you completed?
8th grade or less
9th grade
10th grade
11th grade
12th grade/GED
Some college
College graduate or higher
Q21. Please stop me when I reach the category that includes your household’s total annual income for last year, that is 2009:
Under $15,000
From $15,000 to less than $30,000
From $30,000 to less than $50,000
From $75,000 to less than $100,000
From $100,000 to less than $125,000
$125,000 or more
Don’t know
Refused
Thank you for participating in this important traffic safety survey
ATTACHMENT 2
AGE 18-39 OVERSAMPLE SCREENER
OMB control No. 2127-0646
Expiration Date: 8/31/2011
Hello, my name is _______ and I am calling on behalf of the U.S. Department of Transportation. We are conducting a study about Americans’ driving habits and opinions. The interview is voluntary, and will take less than 10 minutes to complete. We will not collect any personal information that would allow anyone to identify you. Please note that an agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a current valid OMB control number. The OMB control number for this study is 2127-0646. Could we begin now?
(IF CALLING FROM CELL SAMPLE) SCREENING QUESTIONS
Cell Screen Q1. Are you driving or in any other situation that would make it unsafe to talk at this time?
IF YES, TERMINATE IMMEDIATELY AND CALL BACK AT A FUTURE TIME.
Cell Screen Q2. Are you between the ages of 18 and 39 years?
Yes
No IF NO TERMINATE THE INTERVIEW
Cell Screen Q3. Did I reach you on a landline, or on a cell phone?
Cell Phone
Landline
Refused IF REFUSED TERMINATE INTERVIEW
Cell Screen Q4. Do you currently reside in the [NAME OF DEMONSTRATION /COMPARISON COMMUNITY] area?
Yes
No IF NO TERMINATE THE INTERVIEW
Cell Screen Q5. Do you use both a cell phone and a landline phone for your home telephone service, or do you exclusively use a (INSERT CSQ3 - cell phone/landline phone)?
Use Landline only
Use Cell phone only
Use Both landline and cell phone
Don’t know
Refused
(IF CALLING FROM LANDLINE SAMPLE) SCREENING QUESTIONS
Landline Screen Q1. Are you driving or in any other situation that would make it unsafe to talk at this time?
IF YES, TERMINATE IMMEDIATELY AND CALL BACK AT A FUTURE TIME.
Landline Screen Q2. May I please speak with the person in your household who had the most recent birthday AND is between the ages of 18 and 39 years?
Yes – The same person who answered the phone
Yes – Eligible respondent summoned by person who answered the phone
Not available - Eligible respondent not presently available IF NOT AVAILABLE, CALL BACK
No IF NO TERMINATE THE INTERVIEW
Landline Screen Q3. Did I reach you on a landline, or on a cell phone?
Landline
Cell Phone
Refused IF REFUSED TERMINATE INTERVIEW
Landline Screen Q4. Do you use both a landline phone and a cell phone for your home telephone service, or do you exclusively use a (INSERT LLSQ3 - landline phone/cell phone)?
Use Landline only
Use Cell phone only
Both landline and cell phone
Don’t know
Refused
| File Type | application/vnd.openxmlformats-officedocument.wordprocessingml.document |
| Author | alan.block |
| File Modified | 0000-00-00 |
| File Created | 2021-02-02 |