1 - Remember your audience.
For most talks -- say, colloquia -- about 50% of your audience will be graduate students. And some of those will be entry level graduate students. You can safely assume they know about Newton's Laws, Special Relativity, Thermodynamics, Nonrelativistic Quantum Mechanics, and some electricity and magnetism. Maybe some solid state. And that's about it. Anything else should be introduced and explained.
Instead, many talks seemed geared towards the two or three professors in the audience that are in the speaker's field. That not only cuts out the graduate students, but I imagine much of the faculty, too.
I don't know how people that give colloquia are invited to do it, but I imagine a faculty member invites them. It might be a good idea to have some back and forth to find out what the audience might be likely to know.
For instance, we have had so many astronomy talks here that I'm starting to pick up on some things. I at least know about things like the main sequence and the Hertzsprung-Russell diagram, even though outside of colloquia, my astronomy experience is zilch.
And of course, at professional meetings, as opposed to colloquia the standards change, since presumably there there might be more familiarity with the basics.
- Have some structure to your talk. A beginning, a middle, and an end.
I've seen talks that fall apart at all or any of those stages.
A lot of talks just end with data after data after data. "This is a spectrum of this, this is a spectrum of that." What does it mean? Who knows, but they took the data, so now you have to sit through it. Even worse are talks that are just data from beginning to end.
Introduction: Is your talk about a field a student is familiar with? If not, introduce it. Do you have a specific problem you're working on? If not, narrow your talk down. Introduce the problem, too.
And have transitions. How will this technique solve that problem?
I've seen some talks that had interesting introductions that went no where. One that comes to mind was about "quantum computing." That's a hot topic now. Well, when it introduced quantum computing, it was interesting. I guess it had to be; that was probably his sell that he used to get funding.
When the talk went into act 2, it was about decays of states that have been excited by light. Uh, what does that have to do with quantum computing? There was nothing in the introduction that tied into his specific work, hence the transition made no sense, and it wasn't possible to follow the rest of the talk because he didn't explain why he was even doing what he was doing.
-Have some enthusiasm.
There is no surer way to be bored by a talk than if the speaker him or herself is bored by it.
Come on. You wrote the talk. It might even be your research. If you can't be excited about it, who could?
For most talks -- say, colloquia -- about 50% of your audience will be graduate students. And some of those will be entry level graduate students. You can safely assume they know about Newton's Laws, Special Relativity, Thermodynamics, Nonrelativistic Quantum Mechanics, and some electricity and magnetism. Maybe some solid state. And that's about it. Anything else should be introduced and explained.
Instead, many talks seemed geared towards the two or three professors in the audience that are in the speaker's field. That not only cuts out the graduate students, but I imagine much of the faculty, too.
I don't know how people that give colloquia are invited to do it, but I imagine a faculty member invites them. It might be a good idea to have some back and forth to find out what the audience might be likely to know.
For instance, we have had so many astronomy talks here that I'm starting to pick up on some things. I at least know about things like the main sequence and the Hertzsprung-Russell diagram, even though outside of colloquia, my astronomy experience is zilch.
And of course, at professional meetings, as opposed to colloquia the standards change, since presumably there there might be more familiarity with the basics.
- Have some structure to your talk. A beginning, a middle, and an end.
I've seen talks that fall apart at all or any of those stages.
A lot of talks just end with data after data after data. "This is a spectrum of this, this is a spectrum of that." What does it mean? Who knows, but they took the data, so now you have to sit through it. Even worse are talks that are just data from beginning to end.
Introduction: Is your talk about a field a student is familiar with? If not, introduce it. Do you have a specific problem you're working on? If not, narrow your talk down. Introduce the problem, too.
And have transitions. How will this technique solve that problem?
I've seen some talks that had interesting introductions that went no where. One that comes to mind was about "quantum computing." That's a hot topic now. Well, when it introduced quantum computing, it was interesting. I guess it had to be; that was probably his sell that he used to get funding.
When the talk went into act 2, it was about decays of states that have been excited by light. Uh, what does that have to do with quantum computing? There was nothing in the introduction that tied into his specific work, hence the transition made no sense, and it wasn't possible to follow the rest of the talk because he didn't explain why he was even doing what he was doing.
-Have some enthusiasm.
There is no surer way to be bored by a talk than if the speaker him or herself is bored by it.
Come on. You wrote the talk. It might even be your research. If you can't be excited about it, who could?
