Field testing of consumer safety signs: the comprehension of pictorially presented messages

This paper describes one of a series of studies on the investigation of desirable characteristics of a labelling system for use on potentially dangerous household products. The suitability of a range of signs based on symbolic coding of hazards has been investigated. The hazards investigated were: Fire, Poison, Caustic, Electrical and General. The first stage of the programme involved a survey of all known symbols used for these and similar applications, followed by some preliminary testing to determine a range of symbols for more detailed testing. This subsequent testing of the symbols involved their incorporation into a range of sign variants, with differing colours of image and background together with variations in the shape and colour of enclosure containing symbol. These symbolically coded sign variants were tested in the field using a structured national random sample of 4000 respondents. The data were analyzed using a variable, but systematically controlled, criterion for correct comprehension, which allowed reliable discrimination between certain of the sign variants in terms of comprehension performance. In absolute terms, comprehension of even the best of the signs was not markedly high, ranging from around 20% with s stringent criterion up to 50% if a less stringent criterion of correct comprehension was used. Some of the poorer signs tested failed to achieve better than 5% correct comprehension rate even with a lax criterion for comprehension. Prior experience of the sign was found to enhance comprehension, increasing the correct comprehension rate by a factor of 1.5-2.0 times the correct comprehension rates obtained with those respondents who affirm that they have not seen the sign before.