I've been asked this question a few times recently, "how do you (or should you) include negative results in your thesis?"

This one is really easy to answer, but for some it is not an easy answer to hear.

If, based on prevailing theory, you have every reason to expect a particular result but observation proves otherwise, then this may be of interest to the field and worth reporting if - and only if - you have conducted the work competently.

If, on the other hand, you just screwed up, this is not of interest. The only thing you can do is repeat the research.

Unfortunately, no matter how hard you have worked, if the execution of your research is poor, there is no honest way to raise it to the required standard through clever writing.

This is why it's so very important to analyse your data early, so you can spot mistakes while you still have the time to rectify them.