Comprehension and perceived quality of warning pictorials

The present study assessed the comprehensibility and quality of warning pictorials in the presence and absence of explicit context Context was provided by a photograph and a verbal description of an environmental scene in which the pictorial might appear. A total of 248 individuals performed a comprehension test on a randomly-assigned pictorial from each of three referent categories (Keep Out, Electrical Shock. and Do Not Dig). Following this task. 185 participants were shown five pictorials (four others plus the one they had seen) associated with each of the three referent categories and then rated and ranked them on their quality to convey the referent message effectively. Results indicated that the context manipulation enhanced comprehension for pictorials two out of three referent categories. Confidence intervals indicated that comprehension levels of all the Electrical Shock symbols would fall within the ranges specified by ISO's 67% and ANSI's 85% comprehension criteria. Three of the Do Not Dig pictorials and none of the Keep Out pictorials fell within the acceptable ISO and ANSI comprehension criteria. Statistically significant average point biserial correlations were obtained between the comprehension and quality scores for each referent category. Implications for warning pictorial test and design are discussed. IN1RODUCI10N Warning designers have increasingly made greater use of pictorials in hazard communications. The potential benefits of pictorials include being able to see them at greater distances (and in smaller sizes) than words (Jacobs, Johnston, and Cole, 1975), as a way of reaching persons unable to read the language (e.g., foreign speakers, children), and as a way of increasing the salience of the message (Young, 1991). Although there are several criteria for the design of good pictorials, understandability (i.e., comprehension) is usually considered the most important criterion (Dewar, 1994). Without empirical evidence on comprehension rates, it is not possible to tell whether existing pictorials are adequate on this dimension. The present study examined the comprehension rates of several pictorials intended to convey the concepts of Keep Out, Electrical Shock. and Do Not Dig. A second purpose of the study was to determine whether the presence of context facilitates icon comprehension compared to its absence. Pictorials are normally viewed in locales that assist understanding of their meaning. However, when pictorials are tested for comprehension, they are usually evaluated without any explicit contexL In other words, the test lacks external validity. Without context, the viewer has to generate a mentally-derived locale for the pictorial. Fa- example, a pictorial of a boot could be interpreted in different ways depending on where a sign with this symbol is posted. At the entrance of a shopping mall, it could mean that shoes are required (no bare feet). At a construction site, it could mean that steel-toed boots are required. In other situations, it could designate a shoe store or shoe repair facility. The issue of combining pictorials with an appropriate context during testing is important because according to international and U.S. standards pictorials need to meet certain levels of comprehension for acceptability.