III. EVALUATION METHOD

All states committed resources to evaluate their individual effort. National coordination facilitated shared data collection procedures among evaluators.

Observational Surveys of Belt Use
Nearly every state conducted and reported statewide surveys of belt use immediately following the period of stepped up enforcement. Most of the surveys were completed within the month of June 2003. These surveys generally followed NHTSA guidelines for conducting statewide surveys. NHTSA guidelines require that:

  • states have a probability-based survey design;
  • data be collected from direct observation of safety belt use;
  • the relative error of the safety belt use estimate not exceed five percent;
  • counties or other primary sampling units totaling at least 85 percent of the State’s population be eligible for inclusion in the sample;
  • all daylight hours for all days of week be eligible for inclusion in the sample.

NHTSA guidelines also require that the determination of a safety belt use rate be based on all types of passenger motor vehicles including passenger cars, pickup trucks, vans, minivans, and sport utility vehicles and that surveys include observation of both drivers and front seat outboard passengers and both in-state and out-of-state vehicles.

Observational surveys of statewide safety belt use from the previous year were compared with state rates reported for June 2003.

The national safety belt use rate was determined by the National Occupant Protection Use Survey (NOPUS). NOPUS is an observational survey of safety belt use that began in 1994 and has been used by NHTSA to measure the nation’s belt use. NOPUS observes actual use on the roads and provides a reliable estimate of safety belt use nationwide. The 2003 NOPUS survey observed 213,195 occupants in 162,195 vehicles in 1,972 observation sites across the country in June (Glassbrenner, 2003).

Pre/Post Telephone Survey
Random dial telephone surveys were conducted before announcing the enforcement program to the public (May 2003) and after the period of enforcement ended (June 2003). The survey instrument used was a NHTSA developed instrument, designed to measure drivers’ knowledge and awareness related to safety belts, laws governing their use, and exposure to safety belt enforcement programs (see Solomon, 2002). The survey instrument did not change between the two survey waves.

Survey samples were designed to represent the national population using approximately 1,200 respondents with an approximate even split between males and females.

Results from the May/June 2003 surveys were compared with previous pre/post national surveys, conducted May/June 2002 and November/December 2002. Survey methodology was consistent across these surveys.

Driver Licensing Office Survey
A small number of states (10) reported results from surveys of motorists coming into Driver Licensing Offices (DL) for license services. These surveys were conducted before and immediately after the National Mobilization’s publicity and enforcement.

The one-page questionnaire was used to assess public knowledge and awareness, changes motorists may have made in their safety belt use behaviors, how vigorously they felt their police agencies enforce the law and the likelihood police would stop them. The survey form used in each state, by and large, was the same with only minor modifications to names of states, type of law, and names of law enforcement agencies as they appear on the questionnaire (see example questionnaires in Appendix A).

The questionnaire remained unaltered between pre and post waves in order to measure change as the campaign progressed. The first wave of surveys provided baseline information. DL conducted that wave over the two-week period leading up to the announcement of the enforcement program. A second survey wave measured program effects immediately after the enforcement period ended.