Color vision assessment by Farnsworth lantern: results using alternative pass-fail criteria

The Farnsworth lantern is used in the United States and Australia to assess pilot applicants who have deficient color vision (DCV). Its efficacy was questioned following a crash in July 2002 because the DCV pilot confused the red and white approach path signals, despite having passed the Farnsworth test. The Farnsworth lights are larger and brighter than many aviation signals and it has a higher pass rate than the lantern tests used in other countries. Moreover, applicants can pass by making no errors on 1 run of 18 lights, which is too small a sample. There were 3 runs of the Farnsworth lantern given to 100 male subjects with DCV. The effect on pass rate of changing the number of runs and the pass criterion was assessed. There were 20 subjects who passed the Farnsworth lantern test. Their average error rate over three runs was 3.9%; two had an error rate of 13% and five confused red and white signals, the colors used in approach path signals. One subject passed by having zero errors on run 1 but made 13% errors on the next two runs. If all subjects are given two runs after a practice run, 15% pass if the pass criterion is < or = 1 error and 11% pass if it is zero errors. No subject made red-white color confusions with the zero error pass criterion. A practice run and two test runs should always be given. The pass criterion should be < or = 1 error in total on the two test runs.