Using a data sheet, the physicians answered the question , Can you definitely see a blue color? (same as experiment 1 ). The physicians were not instructed how many samples,if any, would be positive. None of these physicians participated in experiment 1 , so they were highly unlikely to have seen a positive Meixner reaction recently. Each observer was shown the papers without any input from the other observer. The number of emergency physicians (5) and the number of different amounts of each substance (2) resulted in a total of 10 readings for each substance. This amount of data was believed to be sufficient for the comparison of percentage of agreement between analytes and to determine any statistically significant difference in rate of positive Meixner interpretation between the analyte groups.