On the Prediction of Pictorial Comprehension

Abstract

Pictorial development and testing can be a costly and inefficient process. The process of designing and testing pictorial symbols could benefit from a precursor test to determine the likelihood that a concept will permit the design of a successful symbol (according to subsequent comprehension testing). This study examines whether ratings of the concepts of to-be-designed symbols could be useful in the prediction of comprehension of the ensuing symbols for those concepts. Participants rated 50 text descriptions (referent plus further verbal context) on: (1) how concrete is this concept?(2) how easy is it to visualize this concept?; and (3) how effective would a simple picture be in conveying this concept? These ratings served as predictors of the population estimates and openended comprehension scores obtained from previous research by Young and Wogalter (2001). Results showed that there was a high correlation between the measures used in this study with both population estimates and open-ended comprehension scores. Ease of visualization of a concept had the highest predictive value with concreteness, being the second highest predictor. Measures of the ease of visualization or imaging a concept may be useful preliminary tools for pictorial designers.

Refining teratogen warning symbols for diverse populations

“…Although Symbols 4 and 5 appear to be viable alternate symbols, it should be reiterated at this time that none of the symbols surpassed the ANSI 85% correct comprehension criteria, although several did surpass the International Organization for Standardization standard of 67%, and, moreover, none of the symbols was effective in communicating the concept “Do not take if you may become pregnant.” Based on the earlier results and development work of Goldsworthy and Kaplan (2006b), Symbols 4 and 5 used a dual‐ panel design to include two illustrations on a medication label in an attempt to address incorporate both the prohibition and the consequence messages. Because previous research has indicated that abstract concepts such as the passage of time are often difficult to convey pictorially (Hicks et al, 2003; Leonard et al, 1999; McDougall and Curry, 2000), it was hoped that the multiple panel design of Symbols 4 and 5 (where one image [left panel] illustrates the prohibitive “do not take” action paired with another image [right panel] to illustrate a consequence that indicates “harm to the fetus”) would impart not only the dual components of the warning but also carry the relevance of the conditional state of possibly becoming pregnant. The current data indicate that this goal remains elusive and ripe for further investigation.…”

“…In addition to issues of efficacy brought about by suboptimal pictorial design, there are limitations to the concepts that can be captured by a pictorial representation, especially in the absence of other instruction and in a limited space. Abstract concepts that can only be visualized sequentially, such as the passage of time or those that are complex, such as conditional states, are often difficult to convey pictorially (Hicks et al, 2003; Leonard et al, 1999; McDougall and Curry, 2000). Thus, the ability to communicate danger to women who may become, but are not already, pregnant seems particularly difficult for pictorial warnings, especially those that require a high degree of efficiency in an extremely limited space, as is the case with medication warnings.…”

Alternative Information Signs: Evaluation of Driver Comprehension and Visual Attention

“…3. Sign design (Dewar et al 1974, Hicks et al 2003, Mackett-Stout et al 1981. On a similar note, research has also developed a number of design factors for symbol signs (Dewar 1988, Zwaga et al 1998).…”

Alternative questionnaire formats in usability evaluation

“…As Table 1 indicates, comprehension rates are high for most of the items. Only PUI item 2 obtained a value lower than the minimal comprehension rate cut-off value of 67% required in ISO 3864 (see Hicks et al, 2003). Since these pictorial items are not used in a safety-critical environment, we considered these results satisfactory.…”