Run the developer tests. The procedure will vary, depending on whether the test is manual or automated and whether additional test components are necessary, such as drivers or stubs.

To run the tests, you need to make sure that you have initialized the test environment with all necessary elements, such as software, hardware, tools, data, and so on.

Automated tests will often update a test results which you can evaluate to determine where your tests went wrong.

Evaluate the test execution by analyzing the test run. 

Testing will complete either normally or abnormally.  For correctly implemented tests, a normal completion represents a passed test, though it could warrant additional examination of the test results log to ensure the test ran as expected.  Abnormal termination could be premature termination or just a test that does not complete as intended.

Review the test log to understand any reported failures, warnings, or unexpected results. The cause of the problem(s) might be that the implementation element being tested is faulty, a problem with the developer tests, or a problem with the environment.

Determine the appropriate corrective action to recover from a "failed" developer test run. If the implementation element under test is faulty, fix the problem if possible and rerun the tests. If the problem is serious and cannot be immediately addressed, report the defect. If the developer test is faulty fix the test and rerun the tests. If there was a problem with the environment, resolve it and then rerun the tests.

When the developer tests pass and no further work is need to complete the change set, promote the changes for integration test. If the passing of these tests represent completion of a requirement update the status of the work item.