How Point (Single-Probability) Tasks Are Affected by Probability Format, Part 1: A Making Numbers Meaningful...

Area of Science:

• Health communication research
• Human-computer interaction
• Data visualization

Background:

• The Making Numbers Meaningful project conducted a systematic review to guide health number communication.
• This review focused on the impact of data presentation formats on understanding single health probabilities.

Purpose of the Study:

• To present evidence on how data presentation formats affect identification, recall, contrast, categorization, and computation of health information.
• To synthesize findings from research on "point tasks" involving single probabilities.

Main Methods:

• Systematic review of experimental and quasi-experimental studies.
• Searched multiple databases (MEDLINE, Embase, CINAHL, Cochrane, PsycINFO, ERIC, ACM) and hand-searched journals.
• Extracted data on stimuli, tasks, and outcomes from 99 articles reporting 218 findings.

Main Results:

• Moderate to strong evidence indicates icon arrays with human figures improve recall over block arrays.
• Survival curves enhance identification of peak survival compared to mortality curves.
• Bar charts with data labels improve computation over numbers or graphics alone.
• Matching denominators improve rate computation; framing impacts risk choices.

Conclusions:

• Evidence on the effects of data presentation formats on identification, recall, contrast, categorization, and computation is often weak or insufficient.
• Few findings were comparable enough to establish strong evidence-based conclusions.
• Heterogeneous study designs limited the ability to generate robust evidence.