The purpose of the present study was to investigate how pre-recorded fire evacuation messages should be designed and presented in order to be effective. Five different messages were tested in the study. Visitors at an IKEA store in Malmö, Sweden, evaluated all the messages. The visitors listened to one message each and were interviewed about their associations. They also filled out a questionnaire about the message. Three messages were then tested in unannounced evacuation experiments at Lund University. The participants were first year computer science and electro engineering students and the experiments were performed during exercises in mathematics. All experiments were filmed and the students filled out a questionnaire about the tested message. Based on the study it is recommended that pre-recorded fire evacuation messages should not contain more than five pieces of information, namely more than five phrases. The results also suggest that it is preferable to mention fire, since people will remember the message more accurately. Based on the experiments no difference could be observed between messages that were recorded with a synthetic female voice and with a human male voice.
- Annan samhällsbyggnadsteknik