Purpose: Mathematically, precision is a function of error: the smaller the potential error, the more precise the estimate. Yet, there is some evidence that people may think about precision in this way. The purpose of this study was to determine whether, and to what extent, people’s preferences for precise information are different for weather forecasts versus genetic tests for pancreatic cancer.
Method: 1599 participants in a demographically-diverse online survey read two scenarios: a health scenario about two different genetic tests for pancreatic cancer (one with a narrow range and one with a wide range of predicted risks) and a weather scenario about two different meteorologists’ predictions of rain (also with narrow and wide ranges of precipitation likelihoods). Participants chose their preferred genetic test or meteorologists and then rated test/meteorologist was most precise and most accurate.
Result: Participants were more likely to identify the meteorologist with a narrow range as being precise than they were to identify the genetic test with the narrow range as having those characteristics (Precise: 56% vs. 43%). Similar patterns were observed for assessments of accuracy. Participants with high numeracy skills were significantly more likely to choose the test with a narrow range (69% vs. 62%), as well as state that the test or meteorologist with the narrow range was more precise (Test: 50% vs. 40%; Meteorologist: 62% vs. 50%). While most participants preferred the option with the narrow range in both the health and the weather scenarios, in the health context those with high subjective numeracy skills were significantly more likely to choose the narrow range.
Conclusion: Numeracy skills are important determinants of whether people are able to correctly identify a more precise test, but context matters. It appears that it is easier for most people to interpret information about precision in the context of weather forecasts than in the context of genetic tests, even when the information in the two contexts is technically very similar.
See more of: The 32nd Annual Meeting of the Society for Medical Decision Making