The University Record, November 16, 1992

Product warning labels: how best to get the message across

By Sally Pobojewski
News and Information Services

Product warning labels—meant to protect you from the potential dangers of consumer products like pesticides and drain cleaners—may be hazardous to your health, according to a study by a U-M doctoral student in engineering.

“The current practice of separating warnings from the directions for use increases the chances that people will filter out the safety messages printed on the label,” says J. Paul Frantz, a consulting engineer who recently completed the requirements for a Ph.D. degree in industrial and operations engineering.

In a series of experiments—believed to be among the first to systematically analyze how people use product labels—Frantz studied how 80 undergraduate students used two household products, a drain cleaner and a wood sealant, which can be hazardous if used incorrectly.

Frantz discovered that when safety warnings were integrated with directions for use—instead of the current practice of printing them in a separate warning section—the percentage of subjects in his experiment who read and followed safety procedures went up dramatically.

Frantz says current warning labels are typically designed with the best motive possible: Manufacturers want people to be aware of product hazards, so they try to make the warning stand out. Safety warnings are often printed in a separate location on the label, frequently in a box in prominent, colored type.

“Most product manufacturers follow government regulations, trade association guidelines and attorney recommendations when they write and design product labels,” Frantz explains. “With respect to the location of warnings, all these guidelines are based on the presumption that people are looking for safety information.

“When I studied how people actually use product labels, however, I found that they were highly goal-directed,” Frantz adds. “They tend to scan the label, focusing on the portion that helps them complete the task, and [they] filter out other information.”

To test different ways of presenting product safety information, Frantz developed four sets of labels for the “mock” drain cleaner and wood sealant used in his experiments. In each case, one label was modeled after the label on the product as sold to the consumer. The wording of the other three labels was changed to test the location of safety information, the explicitness of warnings, and whether instructions and warnings appeared as a paragraph or as a numbered list.

Experimental subjects were told the study was a test of the effects of background music on the performance of household tasks. Subjects were videotaped with a hidden camera while they used the drain cleaner to unclog a plugged sink and applied wood sealant to a plant stand.

When Frantz analyzed the results of his experiments, he found that:

—Moving any type of information from the safety warning section into the directions for use section more than doubled the number of subjects who read the information, from 48 percent to 82 percent.

—Stating safety precautions explicitly (for example, “open a window,” instead of “use in a well-ventilated area”) more than doubled the number of experimental subjects who complied with the precautions.

—Including explicit safety precautions within the directions for use produced rates of compliance six times higher than the rates of subjects who read the original label on the product as sold to the consumer.

—Presenting the directions for use as a numbered list, instead of a paragraph of text, had no significant effect on label effectiveness.

“This research is encouraging, because it shows it is possible to develop more effective labeling that can potentially reduce consumer injuries and product liability lawsuits,” Frantz says. “Manufacturers can use the techniques developed in this research to improve existing product labels.”

Financial assistance and experimental facilities for the research project were provided by J.M. Miller Engineering Inc. of Ann Arbor.