Diagnostic tests are widely used in medicine and many other fields. For electric power infrastructure testing is done on many systems including power transformers and underground cables.

However there is a problem - tests rarely provide perfect predictions. Given this fact, central to diagnostic testing is how to interpret test results.  A rigorous framework for probabilistic interpretation is provided here. The framework is implemented in a computer program.

Two Key Concepts:

1. The pre-test probability of the condition is the likelihood of the condition based on your assessment prior to test.
2. A test creates two groups - those who test positive and those who test negative -- The KEY questions are what is the likelihood that each of these groups has the conditiom

Formal probabilistic statements about positive and negative test results require application of Bayes' Theorem. This requires the specification of three parameters. The input parameters are:

• Your pre-test assessment of the likelihood that an object or group of objects has a specific condition.
• For the specific test you must specify two additional parameters: the probabily that the test will return a “false positive&rdquo result, and the probability that the test will return a “false negative” result.

With these parameters and Bayes' Theorem you can make formal probabilistic statements about the likelihood that an object has a condition for both positive and negative test outcomes.

Click the Run Model button and you will be able to apply the framework.

• Enter input parameters or leave the default value.
• Click the Update Assessment button to compute updated likelihoods of the condition.
• Click the Explain Results button for an explanation of the numbers behind the updated probabilities.