Effects of a Role Model and Fear in Warning Label on Perceptions of Safety and Safety Behavior

Mark A. deTurck, State University of New York, Buffalo
Robert A. Rachlin, State University of New York, Buffalo
Melissa J. Young, State University of New York, Buffalo
ABSTRACT - Subjects examined an oven cleaner containing either a low or high fear appeal in the warning label. Subjects examined the cleaner either alone, or were paired with a confederate (role model). After examining the label, subjects paired with a confederate observed the confederate test the cleaner prior to testing the cleaner themselves. The confederates tested the cleaner in one of three conditions: 1) not wearing safety gloves and getting some of the product on themselves and experiencing a chemical burn, 2) not wearing safety gloves and not getting any cleaner one them, and 3) wearing the safety gloves and not experiencing a chemical burn. Although subjects perceived the oven cleaner to be more hazardous when they read the high fear appeal label, they recalled more information from the low fear appeal warning label. Moreover, the level of fear in the warning did not influence subjects' safety behavior. As predicted, subjects were most likely to wear the safety gloves when they observed the role model wear the safety gloves. Observing the role model suffer a chemical burn when not wearing the safety gloves had almost no effect on subjects' likelihood of wearing the safety gloves.
[ to cite ]:
Mark A. deTurck, Robert A. Rachlin, and Melissa J. Young (1994) ,"Effects of a Role Model and Fear in Warning Label on Perceptions of Safety and Safety Behavior", in NA - Advances in Consumer Research Volume 21, eds. Chris T. Allen and Deborah Roedder John, Provo, UT : Association for Consumer Research, Pages: 208-212.

Advances in Consumer Research Volume 21, 1994      Pages 208-212

EFFECTS OF A ROLE MODEL AND FEAR IN WARNING LABEL ON PERCEPTIONS OF SAFETY AND SAFETY BEHAVIOR

Mark A. deTurck, State University of New York, Buffalo

Robert A. Rachlin, State University of New York, Buffalo

Melissa J. Young, State University of New York, Buffalo

ABSTRACT -

Subjects examined an oven cleaner containing either a low or high fear appeal in the warning label. Subjects examined the cleaner either alone, or were paired with a confederate (role model). After examining the label, subjects paired with a confederate observed the confederate test the cleaner prior to testing the cleaner themselves. The confederates tested the cleaner in one of three conditions: 1) not wearing safety gloves and getting some of the product on themselves and experiencing a chemical burn, 2) not wearing safety gloves and not getting any cleaner one them, and 3) wearing the safety gloves and not experiencing a chemical burn. Although subjects perceived the oven cleaner to be more hazardous when they read the high fear appeal label, they recalled more information from the low fear appeal warning label. Moreover, the level of fear in the warning did not influence subjects' safety behavior. As predicted, subjects were most likely to wear the safety gloves when they observed the role model wear the safety gloves. Observing the role model suffer a chemical burn when not wearing the safety gloves had almost no effect on subjects' likelihood of wearing the safety gloves.

Students of product safety information have devoted considerable effort toward identifying the message factors that enhance consumers' compliance with product safety recommendations. Despite their efforts, however, the growing literature on the effects of product safety information indicates that product warning labels exert very little influence on consumers' safety behavior (see Lehto & Miller, 1986 for a review). One reason for the apparent lack of effectiveness in product warning labels is the assumption that there is a direct relationship between exposure to a product safety message and consumers' decision to comply with the message's recommendations. Models of consumer information processing of safety information (deTurck, Goldhaber & Richetto, 1992a; McGuire, 1980) suggest that a number of mediating cognitive factors (e.g., comprehension, perception of personal hazard, etc.) may dilute the effects of an otherwise potent warning message.

Situational cues may help determine consumers' decision to comply with a product warning hazard. One salient cue that consumers may rely on is whether others in the situation comply with safety recommendations. The purpose of the current research was to determine whether a product warning label communicating low versus high hazard or the presence of a compliant or noncompliant role model in the situation affected: 1) perceptions of hazard associated with the product, and 2) compliance with the product warning label.

A MODEL OF SAFETY INFORMATION PROCESSING

Consumers possess varying levels of knowledge regarding the safe use of products. Manufacturers draft warning labels to provide consumers with information they may need so as to use a product safely. Whereas some consumers may heed the instructions on a warning label, others may deliberately fail to read it or they may decide not to comply if they do read it. Understanding the circumstances underlying consumers' decision to comply or not comply with product safety messages would contribute significantly to enhancing the quality of safety programs.

Figure 1 depicts a model of consumer information processing related to safety information. According to the model, there is not a direct relationship between exposure to a product safety message and the decision to comply. If we were to calculate the effect of exposure to a warning message and compliance, excluding for the moment psychological and situational cues, based on a .6 correlation (a large effect by social science standards) between adjacent components of the model, the resulting effect of exposure on compliance would be .13 (a relatively small effect).

Two factors which have not received a great deal of empirical attention in previous models of safety behavior are consumer's psychological dynamics and situational cues. At least two psychological factors that have received researchers' empirical attention are product users' information processing objectives and product users' familiarity with a product. Product users' information processing objectivesCtheir reason for attending to product information (form an impression versus memorize)Cmediate their attention to product labels, recall of safety information, perception of hazard, and compliance with product safety recommendations (deTurck & Goldhaber, 1989b; deTurck, Goldhaber & Richetto, 1992b).

A consumer's familiarity with a product refers to the extent to which he/she is knowledgeable about the safe use of a product. Consumers' knowledge may come from direct experience with a product, or from exposure to other sources, such as manuals, friends, family members, or the media. Research on familiarity indicates that consumers who are more familiar with a product tend to perceive less hazard associated with a product than their less informed counterparts (deTurck & Goldhaber, 1989a, deTurck, Goldhaber & Richetto, 1992a, 1992b; Godfrey & Laughery, 1984).

By contrast, very little is known about how consumers' safety behavior is affected by situational cues present when they use a product. A consumer's decision to comply with safety recommendations may be based to a lesser extent on safety information, and more on contextual cues when he/she uses a product. One situational cue found to affect individuals' compliance with safety messages is the amount of effort they must exert to comply with the safety recommendations. Wogalter, Godfrey, Fontenelle, Desaulniers & Rothstein (1987) found that people essentially ignored safety recommendations when they were required to exert a substantial amount of effort (walk down a hall 50 feet to use another exit, or to return down the hall they came and use another exit) compared with a less safe alternative requiring little effort (using the door next to the broken one).

Another cue in situations that may affect product users' compliance with safety recommendations is the safety related-behavior of a role model. Consumers do not use products in a social vacuum. Children grow up observing their parents and older peers using products that may pose potential hazards. Family members, friends, or coworkers also collaborate on many projects that may entail using potentially hazardous products and/or engaging in hazardous circumstances. In addition, there are many hazardous situations in our daily routines that involve interacting with others away from formal work projects (e.g., driving in a car).

Despite the fact people may typically avoid using certain hazardous products, or use potentially hazardous products with a great deal of care, they may fail to use hazardous products safely when they observe others failing to take proper precautions. A preteen, for example, may believe that smoking is extremely harmful, but decide to smoke because he/she observes some friends smoke, or from observing his/her parents smoke. Similarly, adults may use their safety harnesses when driving their own automobiles, but fail to wear a safety harness when they are a passenger with a driver who does not wear his/her safety harness. Stated differently, product users may rely on role models in a given context as a guide for their own safety behavior rather than their own personal knowledge of a product's potential hazard.

FIGURE 1

A MODEL OF THE EFFECTS OF PRODUCT WARNING LABELS

The current model is consistent with models of the attitude-behavior relationship (Fazio, 1986; Fishbein & Ajzen, 1975). These models indicate that the social context exerts a substantial influence on a person's behavior. Although product users may perceive a product or situation to be hazardous, others' behavior in the situation may override their own perception of product hazard, and as a result, they may fail to act prudently when using a product.

Research on helping in hazardous situations suggests that people rely on others' safety behavior when deciding on the degree of hazard posed by certain situations. Latane and Darley (1968) found that subjects working with a confederate would remain in an unsafe environment (smoke-filled room) so long as the confederate remained. In fact, 90% of the subjects remained in the smoke-filled room rubbing their eyes, coughing, and waving the fumes away as they worked. However, when subjects worked alone and the room began to fill with smoke, 75% of them left the room to report that there was smoke. Subjects paired with a confederate obviously relied on the confederate's behavior to determine the hazardous nature of the situation. Similarly, Bryan and Test (1967) obtained results in four experiments indicating that people rely on the conduct of a role model when deciding when to render help to another person. The presence of a helping role model significantly increased incidents of subjects' willingness to help.

Indeed, the social learning literature (Bandura, 1962, 1977) indicates that people often rely on role models' behavior when deciding how they should behave. A great deal of product users' safety behavior may be determined by their observations of how others use a product. Much of the information product users rely on for their own behavior may be a result of vicarious learning rather than direct experience or from product safety information.

Current Study

Subjects were paired with a confederate to test a new product as part of a marketing study. Each confederate-subject pair was given a product (oven cleaner) to examine. The label on the container communicated that there was a either a low or high level of hazard associated with using the product. After examining the product they were asked to test the product to determine its effectiveness. The confederate always tried the product first.

In one condition, the confederate tested the product on a grill without wearing safety gloves. In a second condition, the confederate tested the product while wearing safety gloves. In the third condition, the confederate tested the product without wearing the safety gloves and he/she got some of the cleaner on his/her hand and acted as though he/she got a chemical burnCsuffered some discomfort. There were two control conditions (low versus high hazard label) where there was not a confederate present.

It is predicted that the warning label will exert an effect on subjects' perception of hazard associated with the product. More specifically, it is expected that subjects exposed to the higher hazard warning label will perceive there is a greater level of hazard associated with the product than subjects exposed to the lower hazard warning label. However, subjects' compliance with safety recommendationsCwearing the safety glovesCwill not be affected by the degree of hazard communicated in the warning label but by whether or not the confederate complies with the safety instructions in the label.

Results obtained by deTurck, Goldhaber, Richetto, and Young (1992) indicate that high fear appeal warning labels produced greater perceptions of hazard, but that the high fear warning labels were not perceived to be as credible as moderate or low fear appeal labels. To the extent high fear appeal warning labels are not perceived to be as credible as other relevant safety information in a situation, product users should be more likely to comply with other more credible safety cues that are present in the situation such as others' safety behavior. It is assumed that a role model's safety behavior will serve as a credible situational cue that product users will rely on more than safety information in a product label. More specifically, it is predicted that subjects will be more likely to wear the safety gloves when the confederate wears the gloves.

METHOD

Sample

One hundred and twenty-six students (62 females, 49 males, and 15 which did not report gender) from an introductory communication class at the State University of New York at Buffalo were solicited to participate in this study in partial fulfillment of a course requirement.

Procedure

After subjects arrived in the testing site, they were greeted by a research assistant who introduced them to the confederate who was seated in a waiting area and both were escorted to the research laboratory. The participants were then given a written prebriefing statement which told them that the department had been asked by a major manufacturer (no company name was given) to test a new oven cleaning product. After the subject and confederate finished reading the prebriefing statement and had signed the adjoining consent form, they were told that they would have an opportunity to examine and test the effectiveness of this new oven cleaning product. Only one subject participated with a confederate during each session.

On the other side of the room, a container of oven cleaner was set out on a table with a pair of safety gloves along side of it. Additionally, there was a brush (used to apply the product), a wash basin, a roll of paper towels, and a sponge placed nearby on the table. In the corner of the same table, there was a portable hibachi set up on top of a protective covering to protect the tabletop.

Subjects were randomly selected to receive one of two types of warning labels. Two product labels were generated which were identical in every regard except the wording on the warning label. The label was black and white, and contained information regarding the proper use of the product as well as detailed instructions for using the product. Both labels contained a safety warning printed in reverse (white on black) so that it was more salient than the other information on the label. Subjects were randomly assigned to receive either a low or high fear warning label.

Low fear. The low fear warning label read: "Caution: This product may cause minor skin irritation".

High fear. The high fear warning label read: "Danger: This product will cause severe chemical burns to skin".

Both warning labels contained identical additional information regarding what to do if the product came into contact with skin or eyes.

Safety Behavior

The confederate was selected by the research assistant to be the first of the pair to examine and use the product. In all conditions, the confederate examined the jar, and read the product label for approximately one minute. Confederates were instructed to perform one of three safety behaviors.

No Gloves-No Burn. In this condition, the confederate applied a small amount of the oven cleaner to the portable hibachi without wearing the safety gloves as stated on the product label. When testing the oven cleaner, the confederate did not get any of the cleaner on his/her skin and experienced no chemical burn.

No Gloves-Burn. In this condition, the confederate did not wear the safety gloves, and got a small amount of the cleaner on his/her hand while applying it to the grill. The confederate then exclaimed, "Ouch, that burns" and quickly wiped the oven cleaner off using one of the paper towels setting next to the hibachi. The confederate was then instructed to wash his/her hands in the wash basin.

Gloves. In this condition, the confederate wore the safety gloves while applying the oven cleaner to the hibachi and did not get any on his/her skin.

No Confederate (Control). In this condition, subjects examined and tested the oven cleaner without a confederate present.

Dependent Measures

After the confederate finished testing the oven cleaner, he/she was directed to another room to complete a questionnaire while the subject remained to examine and test the oven cleaner. The research assistant remained with the subject and recorded whether the subject wore the safety gloves before applying the oven cleaner to the hibachi.

When the subject finished applying the oven cleaner to the hibachi, he/she was asked to complete a questionnaire. The questionnaire asked the subject to indicate how safe they perceived the product to be on an eight-point scale ranging from 1 (Safe) to 8 (Unsafe). In addition, subjects indicated on an eight point scale from 1 (Certain to Use Safely) to 8 (Uncertain to Use Safely) how certain they would be to know how use the product safely in the future. Questions tested the subjects knowledge of safety-related information regarding the oven cleaner (e.g., what to do if the cleaner gets in eyes, how to apply the product, what safety equipment should be worn while using the product). The final section of the questionnaire contained several demographic questions.

After subjects completed the questionnaire, they were debriefed and given an opportunity to ask any questions. Subjects were thanked for their participation and dismissed.

RESULTS

Perceived Safety

A 2 (fear: low/high) X 4 (role model: none/no gloves-no burn/gloves/no gloves-burn) multivariate analysis of variance was conducted with perceived safety and certainty of safe use as the dependent variables. [No main or interaction effects were obtained for gender of confederate or subject; therefore data were collapsed across this variable for all analyses.] A main effect was obtained for fear, F(2,125)=4.25, p <.02. Due to the significant omnibus effect, the univariate ANOVAs were examined to determine the precise effect of fear communicated in the warning on perceived safety and certainty of safely using the oven cleaner. Subjects examining the low fear label reported the product as safer (M=3.49) than those subjects who examined the high fear label (M=4.31), F(1,126)=5.26, p<.03. Similarly, subjects in the low fear condition reported feeling more certain about how to use the product safely (M=2.59) than subjects in the high fear condition (M=3.31), F(1,126)=5.22, p<.03.

Recall

A main effect for fear was obtained: Subjects in the low fear condition correctly recalled more facts about the product (M=7.72) than subjects in the high fear appeal condition (M=6.68), F(1,108)=7.64, p<.01.

Safety Behavior

As predicted, results from a chi-square analysis revealed that subjects who observed the confederate wear the safety gloves when using the oven cleaner were more likely to wear gloves than in any other condition, even subjects who observed confederates apply the oven cleaner without safety gloves and suffering a chemical burn, Chi2 (3),=34.30, p<.00 (Figure 2). The contingency coefficient associated with this effect was .46.

DISCUSSION

As its primary goal, the current study sought to determine the effects of a role model's safety behavior on a naive bystander. Results indicate that although perceptions of a product's hazard may be affected by the product's warning label, compliance with recommended safety behavior is affected by another person's safety behavior in the context and not the warning label. Clearly, subjects followed the confederate's example despite the recommendations in the product warning. These differences cannot be attributed to differential learning from the warning label.

Perhaps the most interesting finding was that observing the role model improperly use the product and get hurt had almost no effect on subjects' subsequent safety behavior. Although research in social learning indicates that when a role model's behavior is punished observers are less likely to engage in the target behavior (Walters, Leat & Mezei, 1963; Walters & Parke, 1964), subjects in the current study were not affected by observing the role model get hurt. It may be the case that the current subjects assumed that they would not be as careless as the role model when they tried the product.

FIGURE 2

PERCENTAGE OF SUBJECTS WEARING GLOVES AS A FUNCTION OF ROLE MODEL'S SAFETY BEHAVIOR

It would be useful to extend the current design to determine if well liked role models, e.g., friends or family members, exert more influence on observers' safety behavior than less familiar role models. Well liked role models should exert greater influence over safety behavior than strangers due to the greater affinity between role model and observer and because observers probably spend more time with well liked role models. It would be useful to replicate the current study pairing subjects with a best friend posing as the confederate or a stranger.

In addition, it would be useful to determine if the current effects are robust across a variety of products and/or situations. It is possible that the current effects are limited to products or situations that are somewhat novel to product users. Product users may not be so readily influenced by a careless role model when they are very familiar with a product or situation, although the research reported earlier on familiarity (deTurck & Goldhaber, 1989b, deTurck, Goldhaber & Richetto, 1992a, 1992b; Godfrey & Laughery, 1984) and Latane and Darley's (1968) results from the smoke-filled room study suggest that even with highly familiar hazardous circumstances people are not likely to take proper safety precautions.

Although warning labels may influence product users' perception of a product's hazard, the current findings, as well as previous research results, indicate that it is unreasonable to assume a warning label will directly affect safety behavior. Rather than channeling more resources into increased labeling, greater effort is needed in face-to-face educational programs so as to provide role models for safety behavior. Research on smoking (McAlister, 1989), for example, indicates that educational programs with role models are an effective strategy for deterring smoking. This is not to argue that we should abandon product safety labeling. It is apparent, however, that factors other than the warning labels exert greater influence on safety behavior. It would be more productive if safety researchers did not test the seemingly infinite permutations of passive warning message strategies, but instead, directed their energies toward studying more intrusive warning procedures.

REFERENCES

Bandura, A. (1962). Social learning through imitation. In M.R. Jones (Ed.), Nebraska symposium on motivation (pp. 211-269). Lincoln, NE: University of Nebraska Press.

Bandura, A. (1977). Social learning theory. Englewood Cliffs, N.J.: Prentice-Hall.

Bryan, J.H., & Test, M.A. (1967). Models and helping: Naturalistic studies in aiding behavior. Journal of Personality and Social Psychology, 6, 400-407.

deTurck, M.A., & Goldhaber, G.M. (1989a). Effectiveness of signal words in warnings: Effects of familiarity and gender. Journal of Products Liability, 12, 104-114.

deTurck, M.A., & Goldhaber, G.M. (1989b). Effectiveness of product warning labels: Effects of consumers' information processing objectives. Journal of Consumer Affairs, 23, 111-126.

deTurck, M.A., & Goldhaber, G.M., & Richetto, G.M. (1989). Effectiveness of product warnings: Effects of language valence, redundancy, and color. Journal of Products Liability, 12, 93-103.

deTurck, M.A., Goldhaber, G.M., & Richetto, G.M. (1991). Uncertainty reduction in product warnings: Effects of fear in signal word and hazard statement. Journal of Products Liability, 13, 329-338.

deTurck, M.A., Goldhaber, G.M., & Richetto, G.M. (1991). Uncertainty reduction in product warnings: Effects of fear and color. Journal of Products Liability, 13, 339-346.

deTurck, M.A., Goldhaber, G.M., & Richetto, G.M. (1992a). Familiarity and awareness: Effects of conscious and nonconscious safety information. Journal of Products and Toxic Liability, 14, 341-350.

deTurck, M.A., Goldhaber, G.M., & Richetto, G.M. (1992b). Effectiveness of warnings on alcoholic beverages: Consumers' information processing objectives and level of fear. Journal of Products and Toxic Liability, 14, 329-339.

deTurck, M.A., Goldhaber, G.M., Richetto, G.M., & Young, M.J. (1992). Effects of fear-arousing warning messages. Journal of Products Liability, 14, 217-223.

Fazio, R. H. (1986). How do attitudes guide behavior? In R.M. Sorrentino & E. Tory Higgins (Eds.), Motivation & cognition: Foundations of social behavior (pp. 204-243). New York: Guilford Press.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley.

Godfrey, S., & Laughery, K. (1984). The biasing effects of product familiarity on consumers' awareness of hazard. Proceedings of the Human Factors Society, pp. 483-486.

Latane, B., & Darley, J.M. (1968). Group inhibition of bystander intervention in emergencies. Journal of Personality and Social Psychology, 10, 215-221.

Lehto, M.R., & Miller, J.M. (1986). Warnings: Fundamentals, design, and evaluation methodologies (Vol. 1). Ann Arbor, MI: Fuller Technical Publications.

McAlister, A., Ramirez, A.G., Galavotti, C., & Gallion, K.J., (1989). Antismoking campaigns: Progress in the application of Social Learning Theory. In R.E. Rice & C.K. Atkin (Eds.), Public communication campaigns (2nd ed., pp. 273-290). Newbury Park, CA.: Sage.

McGuire, W. (1980). The communication-persuasion model and health-risk labeling. In L.A. Morris, M.B. Mazis & I. Barofsky (Eds.), Product labeling and health risks, (pp. 99-122). Banbury Report 6, Cold Spring Harbor Laboratory.

Walters, R.H., Leat, M., & Mezei, L. (1963). Inhibition and disinhibition of response through empathic learning. Canadian Journal of Psychology, 17, 235-243.

Walters, R.H., & Parke, R.D. (1964). Influence of response consequences to a social model on resistance to deviation. Journal of Experimental Child Psychology, 1, 269-280.

Wogalter, M.S., Godfrey, S.S., Fontenelle, G.A., Desaulniers, D.R., & Rothstein, P.R. (1987). Effectiveness of warnings. Proceedings of the Human Factors Society, 29, 599-612.

----------------------------------------