What Makes a Good Format: Frameworks for Evaluating the Effect of Graphic Risk Formats on Consumers’ Risk-Related Behavior

Abstract

The format in which risk information is presented has a fundamental influence on the quality of risk communication with health consumers. Graphic formats might enhance the risk communication process, but we know little about how these formats affect risk-related behavior. Based on a review of literature within and outside the medical domain, we present four useful frameworks for evaluating the effects of graphic format on risk-related behavior.

Well-informed consumers have the potential to improve the quality of their health care and reduce health care costs. Sound understanding of health risks can help consumers achieve these ends through active participation. Extensive evidence indicates that variation in traditional numerical and textual risk presentation formats can profoundly affect risk-related behavior (e.g. risk perception, risk comprehension, and decision-making). Graphical formats are commonly recommended to enhance risk communication,¹ but research has not demonstrated their clear benefit over traditional formats.² Further, this research has been criticized for lacking theoretical bases as well as attention to the role of task.² We present four interpretive frameworks that could enhance our understanding of the effect of graphic presentation format on risk-related behavior.

The Format Preference Framework asserts that optimal graphic formats are those people most prefer.¹ Preferred formats are more likely to engage attention and be deemed more useful and meaningful than non-preferred formats. Format preferences have been elicited through a number of approaches including format rankings, Likert scales, and interviews. Preferred formats might not necessarily lead to the most accurate understanding of risk information. The interaction between format preference and risk comprehension is an important area of future research.

The Information Salience Framework asserts that optimal graphic formats highlight the most relevant information. Salient information is more likely to draw attention, be given more consideration, and have a stronger effect on risk-related behavior than less salient information. The Foreground-Background Salience model³ differentiates between foreground and background risk information. Given a risk represented as the proportion of people harmed out of people at risk of harm, the numerator represents foreground information, and the denominator background. Foreground salient formats (e.g. stick figures depicting number of people harmed) induce heightened risk perception and greater risk avoidance than formats with both foreground and background salience (e.g. pie chart depicting number of people harmed out of number of people at risk).³ If our goal is to enhance understanding of risks rather than persuade, the Salience framework offers clues about how to manipulate risk perception and raises ethical issues.

The Task Consistency Framework asserts that optimal graphic formats make accessible the information required to complete a given task. Cognitive fit theory⁴ differentiates between formats and tasks that are symbolic or spatial in type. Symbolic formats emphasize discrete data values (e.g. text, table). Symbolic tasks require extraction of specific values and rely on analytic processes (e.g. “what is the probability of treatment success?”). Spatial formats emphasize holistic patterns, trends, and relationships in data (e.g. graphs). Spatial tasks require making associations, perceiving relationships, and rely on perceptual processes (e.g. “Does risk A or risk B pose the greatest risk?”). Problem solving is most effective and efficient when the format and task types match.⁴ Although this framework incorporates the influential role of tasks, it could oversimplify the complexity of real-life risk communication contexts.

The Rational Expectation Framework asserts that optimal graphic formats result in behavior best aligned with how a rational agent would make decisions. Expected utility theory⁵ serves as the rational gold-standard for decision-making under risk. This theory adheres to the principle of invariance: preferences for risky alternatives should not be affected by logically equivalent ways of framing the risky choice. However, two health risks framed in terms of mortality (losses) lead to different risk preferences than when framed in terms of survival (gains).⁶ Exploring framing effects and other distortions in interpretation can provide research with greater theoretical basis. Each of these frameworks provides a unique perspective and could make opposing recommendations for formats that enhance each risk-related behavior. Preferred and salient formats engage attention, salient and task-consistent formats reduce cognitive load, and rationally aligned formats minimize irrational choices informed by risk information. A broadly informative approach is comparison of graphic effects among a number of interpretive frameworks.