What kind of pictographs are suitable to be used for natural disaster alerting? To find out, we were addressing marginalized communities with reading and writing problems in Sri Lanka and the Philippines last month which he have reported on before. Before we go back for a thorough comprehension study in May, we used the visits to get a clearer picture of the needs ands preferences of target audiences. Many sets of disaster symbols are around – most of them look great. But which ones are really working? And why?

We took icons from different sets, including the UN-endorsed “OCHA Humanitarian Icons” set and the Guemil project, and hand-picked other icons. What we can say after this study: Most of the pictographs can be understood, as long as a symbol is present at all. The OCHA set, for example, does not include a symbol for heavy rains, GUEMIL lacks specific icon for storm. Some symbols seem impossible to comprehend, such as the OCHA symbol for cyclone (see above). But in general, the sets work out well to identify hazards, at least if added on and modified here and there.

We found out that for representing hazard events most pictographs can be seen as suitable, but some had been preferred and frequently used when we asked participants to choose pictographs. One general rule can be seen: if a pictograph shows context (such as the car next to the house in the figure above), it is used more than more unspecific symbols.

Generally more difficult are “response actions” for actions people are advised to take, like symbols for “evacuate” or “observe”. Without giving context, the comprehension was generally rather low, but when used in the reporting exercise, this changed. So also here, it is a matter of context. Especially we found that feature-rich context-driven symbols were preferred in the reporting exercises – if a symbol showed both the hazard and the required action, it was generally preferred. Whether context-driven symbols are not only popular but also help in overall understanding hazard warnings we will need to verify in the next round of the study.