Why people don't respond to alarm signals

People often fail to respond quickly to fire alarm signals. A common example: In a hotel, a fire alarm sounds late at night. Only a handful of people actually dress and leave (often using the elevators). A few people call the front desk. Many people periodically poke their heads out their doors to see what's going on. But mostly, people simply wait for the alarm signal to stop. Eventually, the alarm either shuts off, or someone (hopefully) comes to tell them that there is a real emergency and they must leave.

This problem has aggravated and perplexed fire safety professionals over the years. Some denounce the public for their stupidity in failing to recognize the potential danger indicated by a fire alarm signal. Their concern is valid: in many fire emergencies, a rapid response is critical to survival. But attributing the problem to public stupidity is inaccurate and of no value in correcting the problem. In reality, people are simply exhibiting natural tendencies.

Why don't people respond to alarm systems? There are several reasons. In this essay, these reasons are divided into two main categories:

Alarm signals as sources of information.
Other reasons why people do and don't respond apart from the information-value of alarm signals.
Alarm signals as a source of information. In order to understand why people fail to respond to alarm signals, we need to look at the information-value of alarm signals. By information-value, we mean the degree to which an alarm signal is useful in reducing uncertainty. Stated differently, how useful is an alarm signal in helping people achieve good situation awareness?

One problem is that people may fail to recognize that a signal indicates the need to immediately evacuate a building. Recent code revisions address this problem by mandating a "temporal code three" signal. Regardless of the notification device used to create the sound, building occupants hear a pattern of three sounds followed by a longer interval of silence. Because codes generally do not require the replacement of older alarm systems, many years will elapse before we can depend on hearing these coded signals. Standardizing an evacuation signal is a valuable step towards reducing the uncertainty associated with fire alarm signals, but there are many other aspects of the alarm signal information problem that a standardized coded signal does not address.

Limited information-value of a simple evacuation signal. Assuming that a building occupant recognizes a signal as indicating a request to evacuate, the signal, by itself, does little to reduce uncertainty from the building occupant's perspective. The signal does not indicate the location and severity of a fire, and, most importantly, whether an emergency even exists. This information is important because most adults instinctually want to delay their response to a threatening situation until the danger is well understood.

Experiences with alarm signals that increase ambiguity. We ask people to take the alarm signal as an indication of real danger, but most people's experiences lead them to believe the opposite. People typically experience alarm signals that are associated with system tests, surprise drills and false and nuisance alarms, not real fires. The alarm signal that indicates a real fire is relatively rare. The alarm signal that indicates a truly dangerous situation is very rare indeed. How can the information-value of fire alarm signals be improved? The following is a list of ways to increase the information-value of alarm signals.

Use vocal alarm systems for greater information-value. Vocal alarm systems (also called "voice/alarm signaling service") broadcast messages to building occupants using a dedicated or prioritized public address system. Vocal alarm signals have been demonstrated to be much more effective in inducing building occupants to action, because their information-value is much greater. Used correctly, they can both (1) define the nature and location of a threat; and (2) recommend an effective coping response. But vocal alarm systems must be used carefully. When a vocal alarm system is used to provide inaccurate information, its effectiveness is compromised. Building occupants may learn that the information provided is unreliable. Most importantly, the danger to building occupants can be increased when the information is associated with an inaccurate awareness of the situation. As an example, building occupants can be instructed to evacuate using smoke-contaminated stairs when they would have been much safer remaining in their rooms.

Minimize system tests and false and nuisance alarms. This has been a major problem. Fortunately, new smoke detector technology is rapidly decreasing the incidence of nuisance and false alarms in newer installations.

Use alarm verification features, positive alarm sequences, and presignal alarm systems to minimize the impact of false and nuisance alarms. These are measures that require, to various degrees, the interaction of trained personnel with a fire alarm system for the purpose of avoiding false alarms. When used correctly, these measures can greatly decrease the likelihood of false and nuisance alarms, thereby improving the responses of building occupants when real fires are present. But when poorly planned or executed, these measures can increase the danger to building occupants by delaying a general alarm when real fires are present.

Minimize the use of surprise fire alarms drills. Surprise fire alarm drills are a useful way of evaluating preparedness, but they also reduce the building occupant's perception that alarm signals indicate real emergencies. For this reason, surprise drills can be counterproductive. Fire safety codes have been changed to allow greater use of announced drills in recognition of that they often just as effective for training purposes. In many settings, surprise drills are best reserved for those occasions when the overall emergency response must be evaluated.

Provide information about ALL alarm signals. Provide information about the origin of the alarm and management's efforts to avoid repeats. More often than not, building managers fail to offer any explanation at all. And when they do tell occupants that an alarm signal was false or a systems test, they don't tell building occupants whether measures are being undertaken to prevent recurrences of the same problem. Failing to provide this type of information encourages building occupants to believe (sometimes accurately) that the next alarm signal is likely to be a repeat of the same problem. Research on reducing false alarm effects has demonstrated that providing information about false alarm sources and corrections is effective in increasing appropriate responses to future alarms signals.

Supplement the low information-value of simple alarm signals with information from emergency team members. Good emergency planning has proven to be very effective in getting people to respond when onsite emergency response team members (e.g., floor wardens, fire brigade member) provide the information that is lacking from an alarm signal. Ideally, such persons should provide information about the nature of a threat ("there is a small fire in the basement"), the appropriate response ("the building manager has ordered all occupants to evacuate the building and wait in the parking lot"), and the reason ("because we can't be certain that the fire will be easily controlled").

Reasons for evacuating other than the information-value of alarm signals.

The information-value of alarm signals is not the only factor in determining whether people respond to alarm signals. Here are some other important considerations.

Task persistence. People do not like to be interrupted. The remote possibility that there could be a fire is often not a good enough reason.

Denial and avoiding anxiety. People want to avoid the feelings of anxiety that the danger of a fire evoke. Therefore, they may tend to avoid interpreting an alarm signal as an indication of real danger.

Social roles. People often respond to fire alarm in the total absence of any sense of danger, because other people expect this behavior from them. A familiar example: young children can be easily taught to immediately evacuate in a school setting. But with adults in most settings, it is important to recognize that effectiveness of social roles depends on complicated cultural and organizational contexts. The willingness of building occupants to cooperate with fire drills typically depends on how effective building managers are in their roles. However, Even the best building manager will find it difficult to convince people that every alarm signal should indicate danger when occupants are frequently inconvenienced by systems tests, surprise drills and false alarms. It is also useful to remember that building managers' roles require them to retain tenants, so they will understandably reluctant to alienate tenants by intimidating them into conforming to fire safety regulations.

Risk perceptions. The greater the perceived risk when a fire alarm signal is detected, the more likely a person is to respond. Mental models. Related to risk perceptions, people often have a faulty mental model about how quickly a minor fire can evolve into a life-threatening situation.

Copyright © 2000. Norman E. Groner. All rights reserved