Communicating coverage data with non-literate communities: beans, sticks or pie charts?

Abstract

A study was undertaken by staff of a health project in Malawi to identify the optimum strategy for reporting information to beneficiary communities regarding project coverage and effectiveness. 3 candidate methods for presentation of coverage data to this nonliterate or semiliterate population were selected based on preliminary qualitative research: 1) a 2-color rotating pie chart; 2) a set of sticks cut in varying lengths relative to a full length (100%) standard stick; and 3) measures of beans with 10 measures representing a full (100%) measure of beans. Data were presented to 60 individual participants using 2 examples for each of the 3 methods. Questions requiring interpretation of the data were then asked of each of the 60 individual participants using 2 examples for each of the 3 methods. Questions requiring interpretation of the data were then asked of each of 60 participants and responses recorded as correct or incorrect. All 3 methods were more than 80% effective in communicating concepts of proportion as measured by the percentage of correct responses. The pie chart method appeared least effective and the beans method most effective in presentation of the information. The effectiveness of the beans method was significantly greater than that of the pie chart method (p = 0.02). This study provides information which may be of use to program managers in selecting strategies for communication of health service data to the communities in which they work. It also demonstrates a simple methodology for identifying and assessing the relative effectiveness of presentation methods. (authors)