Content Mapping: the importance of hazard impact information
Updated: 6 days ago
For more than five decades, warning researchers have been saying alerts must include five key contents:
Source
Hazard
Location
Time
Guidance
These contents, found in the Warning Response Model, have been identified as being the ones that motivate action and reduce delay.
Around 2016, Dennis Mileti, the godfather of warning research, said you need to include hazard impact - that is, what the hazard is doing, or could do - along with the hazard name. This was echoed by researchers who studied "impact based warnings" (IBW) and demonstrated that knowing the impact information can help to improve message understanding.
IBW research did not include "calls to action" and therefore did not measure the effects of hazard impact + guidance information in a message. We know that people need to feel efficacy in order to be able to take action, and efficacy is affected by knowing what to do. So it makes sense that IBW researchers found that including IBW increased understanding, but not action.
More recently, Sutton et al. set out to study the impact of each of those 5 contents on cognitive outcomes (understanding, believing, personalizing, deciding) and intention to search for more information. We conducted an experiment comparing a complete message (containing all of the recommended contents), with messages that lack one piece of content in each message. Importantly, we chose to specific hazard (name) and hazard impact as two different contents, creating 6 comparisons against the complete message.
What we found was quite surprising.
We found that across all of the outcomes measured, hazard impact had the greatest effect. It was even more important than hazard name and protective action guidance (when compared to a complete message).
This finding demonstrated the importance of hazard impact information and also clearly shows that it is a type of content that should be included in every message.
The bottom line: if you want to increase all of the outcomes in response to a message, include the hazard impact information.
When we apply this finding to WEAs, we start to see new and important aspects that can help to improve messages. Let's take this message below.
If we base our evaluation of this message using the 5-contents from the Warning Response Model, we would say "this is a pretty good message!" They've included the source, a sort of hazard name (law enforcement incident), location, and guidance. They use CAPS to emphasize SHELTER IN PLACE, and they explain what it means in plain language. They even added URLs for more information. It is clear and actionable at the time of message receipt.
[In writing this blog, we looked at those URLs, and found that one of the URLs was incomplete (the facebook link is cut off), and the active alert information was removed before we could learn more about the incident.]
But, if we look at the message for what is missing (hazard impact), we can see where it might be confusing to recipients or raise concerns. While the message says "law enforcement incident" and tells people to "go indoors, close and lock windows and doors" we are left with a lot of questions. What exactly is the danger we are locking out? What is lurking about in the area? Is it a person we should be alerted to? Do they have a weapon? Have they injured someone? Is this for our protection against this incident, or is it to control the scene for the law enforcement response (perhaps both).
Each of these questions are the type that could be addressed by adding an impact statement to the message. A clear statement about the hazard impact would look something like this: "law enforcement are responding to an [active shooter/domestic violence/bomb threat/another hazard type]. If you are in the area, you could be injured. Go indoors now or stay away." By phrasing it this way, organizations are also telling us what they are doing (responding to the incident) and placing the emphasis on what people can do to be safe rather than ordering them to do something (because the Sheriff's Office told them to).
I don't know how many hits the website or facebook site might have gotten as a result of this WEA, but we can bet that there was a lot of speculation among those who received the message.
As a wrap up, we found that the message was cancelled some time later. Unfortunately, a WEA was not issued to cancel the shelter in place order. This is especially concerning when the agency clearly says "do not come out until we tell you it is safe." If you never tell people, how would you expect them to know?
For more assistance on your emergency alerts, be sure to read the blogs found on The Warn Room. Or download the Warning Lexicon. Subject matter expert vetted hazard impact statements are included for 48 hazards, making it easy to answer that question that people need an answer to.
