A lot of traffic accidents are preventable. To prevent drivers to speed excessively and to drive unsafely, we often see traffic campaign signs along the road that call for safer driving. But do they work? Here, we use Predict to see if signs are likely to be seen.

Changing human behavior is always a challenge. Risky behaviors such as speeding and risky driving is one of the areas where we see an imminent need for intervention. After all, these types of risky behaviors are risky beyond the individual -- it affects everyone around that person.

Traffic accidents account for 1.35 million deaths annually, and excessive speed accounting for almost 20% of these accidents. No wonder that speeding and risky behaviors are among what we want to target for behavioral change in traffic. Among the initiatives we use are signs to remind drivers to reduce speed, drive safely, take notice of kids close to the road, and so on.

But do these signs really work?

To break down the effect, we need to ask ourselves:

• Are these signs seen?
• Do they lead to an emotional response of relevance?
• Are they understood?
• Will the signs be remembered?

In answering these questions, we sent out a few Neurons staffers into the traffic in a car. While our main purpose was to run Predict (known as NeuroVision at the point of recording) analyses, we also mounted eye-tracking glasses on those that were driving.

‍

‍

We then analyzed the video for all instances of signs that were seen along the trip. Some examples are shown below:

‍

‍

‍

‍

‍

‍

‍

‍

After running this recording, we drew Areas Of Interest (AOIs) around three types of signs:

• Speed signs -- these are signs that indicate the speed limit
• Traffic signs -- these signs indicate directions, distances etc.
• Informational signs -- these are the campaign signs that were currently positioned in Copenhagen

In analyzing the results we found that most of the signs were not successful in attracting much attention.

‍

‍

These results suggest that among most signs, attention is low. In itself, this means that signs that are meant to guide driving safety are rarely likely to be seen.

We then compared whether there would be any differences between the sign types. Here, we found that the campaign signs did show some more promise in drawing attention:

‍

‍

However, as you can see from this plot, this main effect was driven by two well scoring informational signs.

Also, the general level of these scores is less than 5, where the score ranges from 0-100. This suggests that all signs are not likely to attract much attention. This is indeed a problem, since the whole meaning of all sign types are to guide driving behavior. In particular, speed signs are made to let the driver adjust the car's speed, while informational signs are meant to lead to behavioral changes. But if the signs are not seen, there's no way they can lead to such changes.

Taken together, traffic signs that aim to improve driving behavior need improvement!

What better to do than to use Predict to boost the performance of such signs?