Effect of signal word and source attribution on judgments of warning credibility and compliance likelihood

Social-communication models frequently include message source as an important factor in persuasion. However, research on the contribution of source characteristics to warning e ! ectiveness is virtually non-existent. The present research involved two experiments. Experiment 1 examined the e ! ects of the presence (vs. absence) of the signal word WARNING, supplementing it with the source-related term GOVERNMENT to the signal word, and the addition of more speci " c terms (i.e., US and FEDERAL) on ratings of credibility and compliance likelihood for alcohol, cigarette, and iron supplement warnings. Higher ratings were produced with the signal word ' s presence than its absence and adding more speci " city (and length) to the source. The highest ratings accompanied the longest, most speci " c pre " x: US FEDERAL GOVERNMENT WARNING. Experiment 2 investigated three types of sources on credibility and compliance likelihood: (1) speci " c regulatory governmental agencies (e.g., US FOOD AND DRUG ADMINISTRA- TION), (2) speci " c scienti " c professional groups (e.g., AMERICAN MEDICAL ASSOCIATION), and (3) general statements in which an explicit source is not mentioned (e.g., Important Health Warning). The inclusion of speci " c sources produced higher ratings compared to a signal word (WARNING) alone. Implications for warning design and further research are discussed. Relevance to industry Warning messages are used to communicate information about potential hazards and how to avoid injury and property damage. This research shows that the presence of a signal word and adding speci " c source information (telling who is giving the message) was found to increase credibility judgments and compliance intentions. Implications for enhancing warning e ! ectiveness are discussed. ( 1999 Elsevier Science B.V. All rights reserved.