Negative Testing

I am currently working on some test cases to get a proper package reference for the XML Gateway (more specifically for company incorporations). In the Technical Interface Specification, the testing section says on page 52:

Developers will be expected to make extensive use of the test service, covering all services and transaction types available, with both positive and negative testing included.

Page 52 also says:

Developers that fail any tests when seeking approval will be asked to cease approval testing for a period to allow for proper testing to occur.

Given that negative testing implies we need to send some invalid data to test how our system handles it, how does CH tell the difference between Negative testing and failing a test?

I think the wording is a bit misleading here Dan - what this is basically saying is that if your tests fail and return a validation error, you should resolve the issue before submitting fresh ones.