At UQ all physics honours students (4th year undergrad) have to give two 15 minute talks about their year long research project. The first is a progress report at the end of the first semester and the second at the end of the project. No grades are given for these presentations but they are attended by the 3 thesis examiners [supervisor, expert, and non-expert] and so may influence the grade for the thesis.
I think these presentations are very challenging for the students and I am sometimes impressed at the quality of the talks. This is a great opportunity for students to develop and improve their communication skills. When I was an undergrad we never had opportunities like this. Most of us also had very little public speaking experience. Students today are quite different and much more confident and polished.
Here is my advice to students.
First, review general material on giving scientific talks such as Garland's Advice to Beginning Physics Speakers or Wilkins' one page or Geroch's suggestions or Mermin's. Don't think you know better than these old timers.
Second, decide on your audience [your friends, other students, faculty, your research group, the examiners?]. I hate to say it but the examiners is the correct answer.
Taylor the talk accordingly.
Third, decide on your real goal [impress others, show how much work you have done, show off how much jargon you have learnt, be entertaining, talk about how great your research field is, make excuses for your lack of results....?].
Taylor the talk accordingly.
The goals of the progress seminar are simple.
Show you have a well defined and realistic project.
Show you have a clear plan.
Show you have started to make some progress.
The goal of the final seminar is simple.
Show you have achieved something concrete and worthwhile.
Anything else is subsidiary.
Be realistic about how much you can achieve in 15 minutes.
Some background is crucial but don't let it dominate your talk.
Don't spend more than a minute about why the research field is important and interesting.
Don't spend more than a minute on the history of the field.
Don't think you can explain how Shor's algorithm works, the subtleties of the quantum measurement problem, or the microscopic basis for Landau's Fermi liquid theory, ...
Most of the talk should be about what you have done and why it is significant.
Yet, if you can teach people one small thing they will be very appreciative.
Clearly distinguish between the contributions of founders of the field, those of your supervisor, and yourself.
Include relevant references on your slides.
Avoid irritants: being late, having problems with the technology, small fonts, endless jargon, hype, lavish Powerpoint animations, .....
Practise. Practise. Practise.
Consider writing out explicitly what you are going to say.
Start preparations early. Get feedback.
How you answer the questions is important.
Listen carefully. Don't cut off the questioner.
Don't bluff an answer. Saying you don't know is o.k.
Relax. The audience knows that this is a stressful experience, particularly for the inexperienced, and does not expect a perfect talk.