0 a result of a scientific test that appears to show something exists or is present, when this is not correct:
The anthrax test is unreliable because it produces too many false positives.
A sensitivity analysis of the 1881-1902 dataset was calculated at the 50 % false positive/false negative rate.
There is also no discussion concerning the false positive rate.
In the best case, the misidentifications of the two models would coincide, and the false positive rate would be max(0.03, 0.068) = 0.068.
Decreased false positive responses indicates improvement from baseline to retest.
The greatest concern is probably the high likely proportion of false positive results.
The variability of errors and false positive or negative responses is too small to differentiate the 5th and 95th percentiles.
Matching the cutoffs for all three events does, of course, reduce the false positive error; but the price is to increase the false negative error.
False positive rates were also published and varied significantly.
