The key to a successful thesis defence is to keep your committee as involved as possible throughout the entire process. Your committee members are there to provide specific expertise and make sure your thesis is as thorough as possible.

Here are a couple of pointers for empirical theses:

An Ounce of Prevention is Worth a Pound of Cure

When designing your experiments, anticipating possible issues is paramount. Have a clear thesis and keep your methodology as simple as possible. Here of some issues that might crop up:

1. Number of Factors: Have no more than 2-3 independent variables. You cannot really process more than a 3-way interaction so don't set yourself up to fail.
2. Replication: Make sure you have some replication in your trials. Run everyone through each stimulus at least 3 times if possible. This way, in a repeated-measures design you can be assured of not having empty-cells when it comes time to analyze the data. It also makes it easer (possible) to strip out dropped-trials if you have issues with recording (E-Prime has a tendency to lose the odd trial, so replication is 100% necessary).
3. Confounds: If you're working with visual stimulus, can a participant move his/her head? Is light-level important? Figuring out covariates and possible confounds means that you won't have to re-run an experiment later.
4. Balancing Experiments: Make sure every participant goes through every condition (in repeated-measures).
5. Early Stats: Totally write up one experiment before continuing onto the next, that way if there is an issue, you can address it in future work. Doing all your experiments up front means that any mistakes may be multiplied throughout your study, possibly invalidating your findings.

Work in Sections

This idea is quite simple: hand in chapters at a time. It gives you more feedback and allows you to catch errors early.

1. Write your method first: This is just a suggestion, but it's good to have an idea of what your method and preliminary results are before fleshing out your introduction. Otherwise, you might get caught on tangents or missing out on crucial information  you'll 'discover' in later experiments.
2. Keep your introduction focussed: In your prospectus the introduction is really a big literature review. For the thesis, you should have this review focussed and condensed to address your particular question and region of interest.
3. Keep your commitee involved: Each committee member has a particular gift, whether it be indepth topical knowledge, stats, or just knows the right person to answer a given question. Use these people to your advantage or risk them blasting you at the defence!

Knowledgeable Supervision

It is essential that SOMEONE on your committee have strong knowledge of your topic. Cognitive Science is so broad that your immediate supervisor might only have skills and knowledge tangental to your particular field, but make sure someone else on your committee does have that knowledge. Otherwise, you enter your defence without being sure that an expert already has knowedge or has done something similar but is only published in an obscure location.