If your data represent evaluation of a diagnostic test, Prism reports the results in five ways:
Term |
Meaning |
Sensitivity |
The fraction of those with the disease correctly identified as positive by the test. |
Specificity |
The fraction of those without the disease correctly identified as negative by the test. |
Positive predictive value |
The fraction of people with positive tests who actually have the condition. |
Negative predictive value |
The fraction of people with negative tests who actually don't have the condition. |
Likelihood ratio |
If you have a positive test, how many times more likely are you to have the disease? If the likelihood ratio equals 6.0, then someone with a positive test is six times more likely to have the disease than someone with a negative test. The likelihood ratio equals sensitivity/(1.0-specificity). |
The sensitivity, specificity and likelihood ratios are properties of the test.
The positive and negative predictive values are properties of both the test and the population you test. If you use a test in two populations with different disease prevalence, the predictive values will be different. A test that is very useful in a clinical setting (high predictive values) may be almost worthless as a screening test. In a screening test, the prevalence of the disease is much lower so the predictive value of a positive test will also be lower.
Prism computes confidence intervals for all these values using a method you choose on the Options tab for computing the confidence interval of a proportion. Prism offers three methods. We recommend the hybrid Wilson/Brown method.
