Academic journal article Human Factors

Comprehension of Pictorial Symbols: Effects of Context and Test Method

Academic journal article Human Factors

Comprehension of Pictorial Symbols: Effects of Context and Test Method

Article excerpt


With increasing attention to multiculturalism and worldwide trade, pictorial symbols are increasingly being used to convey important messages. Well-designed symbols have the ability to communicate large amounts of information at a glance. They can also be useful in conveying information to persons who cannot read a printed verbal message, either because they have vision problems (e.g., older adults), lower-level verbal skills, or inadequate knowledge of the language being used in the warning (Boersema & Zwaga, 1989; Collins, 1983; Laux, Mayer, & Thompson, 1989; Lerner & Collins, 1980; Zwaga & Easterby, 1984). Clear communication is particularly critical when a pictorial symbol conveys safety-related information, as the lack of understanding or misinterpretation could lead to injury.

Given pictorial symbols' potentially important role in communicating hazards, national and international standards have been established that describe how to evaluate their comprehensibility, such as the American National Standard Institute's ANSI Z535.3 (ANSI, 1991) and the Organization for International Standardization's ISO 3864 (ISO, 1984). ANSI and ISO advise that symbols must reach a criterion of at least 85% or 67% correct, respectively, in a comprehension test to be considered acceptable. Despite the existence of these standards, pictorial symbols are routinely placed on signs, labels, and other materials without any evaluation as to whether they communicate the intended concepts to the targeted audience. Moreover, very little research has evaluated the methods of testing comprehension. The present research examines some of the factors that might influence the results of symbol comprehension tests.

Study Goals

The present study had three major objectives. The first was to compare comprehension performance using two test methods commonly employed to assess symbol comprehensibility: the multiple-choice test and the open-ended test. In multiple-choice tests, respondents are asked to choose the answer that best expresses the symbol's meaning from several alternative answers. In open-ended tests, participants are shown a symbol and are asked to give its meaning in their own words. In the current version of the ANSI (1991) Z535.3 symbol standard, either kind of test is allowed, although preference is given to the openended test method.

Dewar (1994) and others have expressed concerns about multiple-choice tests by pointing out that the quality of the distractor alternatives (wrong answers) could greatly influence comprehension scores. An obvious example is a pedestrian crossing symbol in a test that includes distractors such as "keep refrigerated" or "no left turn." These distractor answers could be easily ruled out by respondents, enabling them to choose the correct answer and unfairly inflating the symbol's purported comprehension level compared with a test with more plausible distractors.

Although the lack of plausibility can be obvious (as in the previous example), distractor quality in actual multiple-choice tests can be subtle and difficult to spot and may influence test performance (Katz & Lautenschlager, 1994; Wogalter & Marwitz, 1987; Wogalter, Marwitz, & Leonard, 1992).

Thus because of the importance of having plausible distractor alternatives, multiple-choice tests might require considerable development work in the test construction stage. Openended tests are easier to develop because they require only the set of symbols to be placed on test sheets with blank spaces. Nevertheless, multiple-choice tests are generally much easier to score than open-ended tests. Indeed, because of its straightforward quantification (the simple counting of responses), multiple-choice tests might appear more scientific. The scoring of open-ended responses is more difficult and less clear-cut. One must establish criteria for the kinds of answers that will be counted as correct; there is usually (if not always) at least some subjective judgment of the correctness of participants' responses. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.