Effectiveness of Safety and Public Service Announcement Messages on Dynamic Message Signs

Executive Summary

The number of transportation agencies that use dynamic message signs (DMS) to provide traffic information to motorists has increased dramatically over the past four decades. This growing trend of DMS deployment is a reflection of the public interest in more information about travel conditions and the importance of traveler information for transportation system operations.

There has been extensive research conducted on traffic-related messages. However, policies regarding the display and type of non-traffic-related messages vary greatly among states. Thus, it is unclear how effective these non-traffic-related messages are in modifying driver or travel behavior. It is also unclear whether these messages are acceptable to motorists and whether they have a positive or negative impact on their driving behavior.

The purpose of this study was to identify how safety and public service announcement (PSA) messages influence driver behavior and to ultimately assist the U.S. Department of Transportation (USDOT), transportation management centers (TMCs), state agencies, and local transportation partners in optimizing the utility of safety and PSA messages on DMS. Four urban areas in the United States were selected as study sites—Chicago, IL; Houston, TX; Orlando, FL; and Philadelphia, PA. Surveys were designed to specifically address the types of safety and PSA messages for each respective city. The goal was to collect approximately 500 survey responses per city. In total, 2,088 responses were received.

The data from each survey was summarized by city and analyzed in terms of usefulness and effectiveness. That is, summary statistics were compiled for the survey questions, and inferential statistics were used to examine the usefulness and effectiveness of safety and PSA messages. Usefulness was defined as the practical and functional application of PSAs on DMS, and examined within each city. For this study, effectiveness was defined as the ability of DMS to positively impact driver behavior by displaying safety and PSA messages. Usefulness and effectiveness are clearly subjective measures since they are based on individual perceptions. For the inferential model, only those variables that were shown to have a significant impact on the respective outcome (usefulness or effectiveness) for DMS remained in the final model. If a variable was not included, there was no significant impact on the usefulness or effectiveness of having safety and PSA messages on DMS.

Most respondents reported that they do see safety and PSA messages on DMS while driving, at least sometimes. The majority also noted that safety and PSA messages on DMS are useful, with some even noting that those messages are more effective on DMS as compared to other media (such as television). Respondents in each location were also asked to interpret the meanings of several common safety and PSA messages (e.g., “Don’t drive impaired” and “Eyes on road, hands on wheel”). In general, most respondents had a fairly good understanding of these messages.

Furthermore, the survey asked about safety and PSA messages that have a more threatening connotation. The majority of respondents indicated that such messages like “Click it or ticket or get $100 fine” or “100 deaths this year on Texas road,” would impact their driving behavior. In Chicago, respondents indicated that they would change their driving behavior for messages that relate to slowing down for emergency vehicles and in construction zones. Houston respondents indicated that messages that would most affect their driving are those with more assertive language (e.g., “Drunk driving, over the limit, under arrest”). Orlando motorists were most likely to change their driving behaviors for all safety and PSA messages; while, Philadelphia motorists were neutral on the effects of safety and PSA messages on their driving.

In addition, the perceived usefulness of safety and PSA messages was examined in a binary logit model. The model showed that the messages were considered useful if the driver encountered them often. Respondents also thought that those messages could be effective in changing behavior. Hence, greater exposure to such messages impacts the perceptions of the messages.

In three of the locations, drivers that considered driving under the influence (DUI) messages (Philadelphia, Chicago, Houston) and speeding messages (Orlando, Chicago, Houston) important also found safety and PSA messages to be useful. In two of the locations, individuals younger than 30 years old did not think that safety and PSA messages were useful. More factors that affect perceived usefulness were also mentioned, but they were on an individual city basis and are further discussed in Chapter 6: Findings and Recommendations/Guidance.

Similarly, the perceived effectiveness was analyzed in a binary logit model, where all responses were aggregated together, ignoring locational factors. Safety and PSA messages were considered more effective when they were encountered often. They were considered useful for respondents older than 60 years old and respondents who had some graduate school or a post-graduate degree. Males with an income less than $25,000 per year and individuals younger than 30 years old did not perceive safety and PSA messages on DMS to be effective.

There are limitations associated with this study as there are always sampling biases with surveys. To minimize such biases, a pilot test was conducted, a representative sample of the driving population was captured, and surveys were collected via face-to-face communication. A website was also set up to offer another mechanism for participants to answer the survey. Possible differences were also examined between those who lived in the general area of the city and those who did not.

The general recommendation based on the surveys was that safety and PSA messages need to be considered useful and effective to maximize their influence on driver behaviors. Certain socioeconomic characteristics influence drivers’ perceptions of these messages. For instance, younger respondents were less likely to consider the messages effective, and future efforts should be focused on promoting awareness targeted toward this group. Respondents also indicated that they took assertive safety and PSA messages seriously, but further examination should consider the magnitude of this impact.

Finally, drivers’ stated preferences usually differ from their revealed preferences. Thus, an on-road impact assessment of safety and PSA messages on DMS is needed to confirm the findings of this survey-based perceptional study.