Learning to test one's hypothesis and being able to find its flaws is a central idea behind Stage 3.  Specific learning outcomes will depend on the approach to (dis)proving the hypotheses in a given discipline. In case of the empirical approach to testing, the learner develops an algorithm for dealing with the chosen problem.

By the end of Stage 3, the learner is able to do the following:

• Show the data collected as a result of following the hypothesis
• Select the data where the hypothesis “failed” and choose those examples that will help to further improve the hypothesis
• Demonstrate how the proposed algorithm helps to resolve the problem with previously selected data

It is important to add that the learner should be disposed to treat own hypotheses as models that have drawbacks and applicability limits.