give a good research talk - university of georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • a...

42
Give a Give a Good Good Research Talk Research Talk with a bit of bioinformatics flavor with a bit of bioinformatics flavor Ying Ying Xu Xu Biochemistry and Molecular Biology Department, and Biochemistry and Molecular Biology Department, and Institute of Bioinformatics Institute of Bioinformatics University of Georgia University of Georgia http:// http:// csbl.bmb.uga.edu csbl.bmb.uga.edu

Upload: others

Post on 20-May-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Give a Give a ““GoodGood”” Research TalkResearch Talkwith a bit of bioinformatics flavorwith a bit of bioinformatics flavor

Ying Ying XuXuBiochemistry and Molecular Biology Department, and Biochemistry and Molecular Biology Department, and

Institute of Bioinformatics Institute of Bioinformatics University of GeorgiaUniversity of Georgia

http://http://csbl.bmb.uga.educsbl.bmb.uga.edu

Page 2: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

•• A research talk should include the following componentsA research talk should include the following components–– define the problem you want to solvedefine the problem you want to solve–– provide motivation of why to solve itprovide motivation of why to solve it–– present your solutionpresent your solution–– convince the audience that your solution is goodconvince the audience that your solution is good–– summarize your contributionsummarize your contribution

•• …….. why some people give good talks while others give .. why some people give good talks while others give boring or even bad talks?boring or even bad talks?

••

Give a Good Research TalkGive a Good Research Talk

Page 3: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Give a Good Research TalkGive a Good Research Talk

•• Giving a good talk is an art; it takes practice but there Giving a good talk is an art; it takes practice but there are some rules we can followare some rules we can follow–– to accomplish what we intend to achieve to accomplish what we intend to achieve –– to keep our audience interestedto keep our audience interested

Page 4: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Goal of This TalkGoal of This Talk

•• Help young scientists to better prepare and present Help young scientists to better prepare and present research talks byresearch talks by–– explaining some of the explaining some of the ““rulesrules”” through examplesthrough examples

Page 5: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Outline of the TalkOutline of the Talk•• Set goals you want to accomplish through your talkSet goals you want to accomplish through your talk

•• Know who your audience isKnow who your audience is

•• Design your talk to accomplish your goalsDesign your talk to accomplish your goals

•• Present your talkPresent your talk

•• Things you should or should not doThings you should or should not do

•• Answer questionsAnswer questions

Page 6: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Set Goals for Your TalkSet Goals for Your Talk•• Before you start preparing, ask yourself the questionBefore you start preparing, ask yourself the question

–– Just want to present a clear talk about what I have doneJust want to present a clear talk about what I have done

–– Show that I have solved a challenging open problem Show that I have solved a challenging open problem …….. .. so I so I am technically strongam technically strong

–– Show that I have methodically solved a messy and complex Show that I have methodically solved a messy and complex problem problem ………… so I am good at handling large projectsso I am good at handling large projects

–– Show that I have drilled deep into a problem and published ten Show that I have drilled deep into a problem and published ten papers papers …………. . so I am a deep thinker and productiveso I am a deep thinker and productive

““What do I hope to accomplish through this talk?What do I hope to accomplish through this talk?””

Page 7: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Set Goals for Your TalkSet Goals for Your Talk

•• There are different types of talks so prepare accordinglyThere are different types of talks so prepare accordingly

–– interview talkinterview talk

–– conference talkconference talk

–– project report talkproject report talk

–– journal club talkjournal club talk

–– ……....

–– introductory bioinformatics talk to laypeople introductory bioinformatics talk to laypeople

Page 8: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Interview TalkInterview Talk•• The goal is to convince the employer to hire you by showing yourThe goal is to convince the employer to hire you by showing your

good features, including technical and personal onesgood features, including technical and personal ones

•• If interviewing for a faculty position, you want to show If interviewing for a faculty position, you want to show

–– you do good science, hopefully in an area that is fundableyou do good science, hopefully in an area that is fundable

–– you have both breadth and depth technicallyyou have both breadth and depth technically

–– you are productive you are productive

–– you know the big picture of the field with a visionyou know the big picture of the field with a vision

–– you present things well: clear, convincing and easy to followyou present things well: clear, convincing and easy to follow

–– you have experience in writing papers, grants, reviewsyou have experience in writing papers, grants, reviews

Page 9: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Conference TalkConference Talk

•• The goal is to showThe goal is to show

–– you have made a great contribution to science by having solved you have made a great contribution to science by having solved an interesting and challenging probleman interesting and challenging problem

–– your solution is correct, and can benefit othersyour solution is correct, and can benefit others’’ researchresearch

–– interested in collaborating with othersinterested in collaborating with others

–– ……....

–– plus a plus a hiddenhidden agenda: to impress the editoragenda: to impress the editor--inin--chief of Journal X in chief of Journal X in the audience so my next paper the audience so my next paper …………. .

Page 10: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Introductory Bioinformatics Talk Introductory Bioinformatics Talk •• When giving an introductory bioinformatics talk to, say, When giving an introductory bioinformatics talk to, say,

college freshmen, you want to college freshmen, you want to

–– To explain to students an emerging field by To explain to students an emerging field by

•• giving an example to show how computation can help to solve giving an example to show how computation can help to solve biology problem biology problem (gene finding is my favorite),(gene finding is my favorite), using simple languageusing simple language

•• extrapolating the idea to other examples to illustrate some geneextrapolating the idea to other examples to illustrate some general ral principles, principles, e.g.,e.g., computation can help speed up biology research computation can help speed up biology research

•• showing real examples of what computation has done to biology, showing real examples of what computation has done to biology, e.g.e.g.identifying a gene responsible for a diseaseidentifying a gene responsible for a disease

–– The The hiddenhidden goalgoal: attract them to study bioinformatics in your lab: attract them to study bioinformatics in your lab

Page 11: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Set Goals for Your TalkSet Goals for Your Talk

•• Define clearly what you want to accomplish Define clearly what you want to accomplish –– keep the audience awake and get your technical points across, keep the audience awake and get your technical points across,

as the minimum goalas the minimum goal–– additional goals can be achieved through careful designadditional goals can be achieved through careful design

•• Design your talk to best accomplish your goalsDesign your talk to best accomplish your goals

Page 12: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Know Who Your Audience IsKnow Who Your Audience Is•• You will lose the audience quickly if did not prepare the You will lose the audience quickly if did not prepare the

talk according to their knowledge leveltalk according to their knowledge level–– avoid using terminology the audience does not understand avoid using terminology the audience does not understand

–– avoid repeating things your audience already knewavoid repeating things your audience already knew

–– avoid going too deep too quickly for them to followavoid going too deep too quickly for them to follow

•• Example I: Example I: PoincarPoincaréé conjectureconjecture–– In mathematical termIn mathematical term: : every every simply connected closedsimply connected closed threethree--manifold is manifold is

hemeomorphichemeomorphic to the threeto the three--sphere sphere

–– Using laymanUsing layman’’s languages language: for : for a threea three--dimensional objectdimensional object, if every loop , if every loop on its surface on its surface can be shrunken to can be shrunken to a a point, then it is a point, then it is a ((deformeddeformed)) threethree--D D spheresphere

Page 13: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Know Who Your Audience IsKnow Who Your Audience Is•• Example II: protein threadingExample II: protein threading

–– To a structural biologistTo a structural biologist: : find a sequencefind a sequence--structure alignmentstructure alignment between a between a target sequence and its nativetarget sequence and its native--like fold in PDB to minimize energy: like fold in PDB to minimize energy: E_p + E_s + E_g

– An advanced version: threading with side-chains – find a sequence-structure alignment and placement of side-chains to minimize energy: E_p + E_s + E_g + E_sd

MTYKLILN …. NGVDGEWTYTE`

Page 14: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Know Who Your Audience IsKnow Who Your Audience Is–– To a laypersonTo a layperson: put a row of people on different stairs of a : put a row of people on different stairs of a

stairway, stairway, keeping the sequential order and allowing gapskeeping the sequential order and allowing gaps; some people like to ; some people like to be close to each other while others dislike thatbe close to each other while others dislike that

•• measure the degree of (measure the degree of (dis)likingdis)liking between each pair of people using between each pair of people using a a disatncedisatnce--dependent functiondependent function

•• put people on the stairs so the total degree of disliking is minput people on the stairs so the total degree of disliking is minimizedimized

Page 15: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Know Who Your Audience IsKnow Who Your Audience Is

–– two versions of the problemtwo versions of the problem•• a persona person’’s arms cannot moves arms cannot move•• a persona person’’s arms can be in any positions arms can be in any position

Page 16: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Know Who Your Audience IsKnow Who Your Audience Is•• Consider adjusting the focus of your talk, depending who Consider adjusting the focus of your talk, depending who

your audience isyour audience is

•• Example: giving a bioinformatics talk toExample: giving a bioinformatics talk to–– A. computational scientists, you want to focus onA. computational scientists, you want to focus on

•• biology background so the audience knows why your problem is impbiology background so the audience knows why your problem is importantortant

•• how you have formulated a biology problem as a computational onehow you have formulated a biology problem as a computational one

•• how your methods have solved an important biology problem, bettehow your methods have solved an important biology problem, better than r than previous onesprevious ones

•• how you have avoided highhow you have avoided high--computational complexity, by taking advantage computational complexity, by taking advantage of special biological insights you haveof special biological insights you have

Page 17: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Know Who Your Audience IsKnow Who Your Audience Is

Avoid a situation: a computer scientist shouts in audience “I can solve the same problem” …

only if he/she knew the biology – your talk should let him/her understand that clearly

Page 18: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Know Who Your Audience IsKnow Who Your Audience Is–– B. biologists, you want to focus onB. biologists, you want to focus on

•• the biology problem you try to solve and the background the biology problem you try to solve and the background –– make sure that make sure that your audience is convinced that you know biologyyour audience is convinced that you know biology

•• how previous methods are inadequate for solving the problem how previous methods are inadequate for solving the problem

•• the technique you used, at a conceptual levelthe technique you used, at a conceptual level

•• why your advanced technique can solve the problem better than uswhy your advanced technique can solve the problem better than using ing simple methods simple methods

•• your method can be used to solve other problems, particularly thyour method can be used to solve other problems, particularly the ones e ones relevant to the audiencerelevant to the audience

You want to maximize the positive contrast between your strength and that of the audience, without losing them

Page 19: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Know Who Your Audience IsKnow Who Your Audience Is

•• Should make up for the deficiency of audience Should make up for the deficiency of audience bybyproviding basics so they can follow the idea of your talkproviding basics so they can follow the idea of your talk

•• To computational scientistsTo computational scientists–– cellcell–– genomegenome–– genesgenes–– proteins proteins –– pathwayspathways–– ……

•• To biologistsTo biologists–– use examples to illustrate key use examples to illustrate key

ideas of your algorithmideas of your algorithm•• branchbranch--andand--boundbound•• linear programminglinear programming•• support vector machinesupport vector machine•• …………..

Page 20: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Prepare Your TalkPrepare Your Talk

Now I know what to talk about and at what technical level. Tell me how to put the talk together, or even better do the talk for me

Page 21: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Prepare Your TalkPrepare Your Talk•• Giving a talk is like telling a story; should design the structuGiving a talk is like telling a story; should design the structure of the re of the

story that is most interesting to the audiencestory that is most interesting to the audience

•• Example: Example: telling a story about mountain climbing in two waystelling a story about mountain climbing in two ways•• One wayOne way

–– goal is to climb to top of mountain Xgoal is to climb to top of mountain X–– here is the planned routehere is the planned route–– at point A, used strategy at point A, used strategy aa and tool kit and tool kit

{{AaAa} } –– at point B, used strategy at point B, used strategy bb and tool kit and tool kit

{{BbBb} } –– ……–– made it to top five minutes faster than made it to top five minutes faster than

the best time beforethe best time before

Page 22: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Prepare Your TalkPrepare Your Talk•• Another wayAnother way

–– the goal is to see the view on top the goal is to see the view on top of mountain Xof mountain X

–– have selected a particular route, have selected a particular route, because of a, b, cbecause of a, b, c

–– along the way, need to overcome along the way, need to overcome five major obstaclesfive major obstacles

–– e.g. have to use technique Y to e.g. have to use technique Y to overcome Aovercome A

–– the view at A is great: the view at A is great: the scenery, the scenery, the monkey, the airplanethe monkey, the airplane

–– at top, the view is spectacular; at top, the view is spectacular; and we got there faster than other and we got there faster than other folksfolks

tough places

Page 23: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Design Your TalkDesign Your Talk

• Don’t let your goals and brilliant ideas get buried in technical details so the audience could not see

Page 24: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Design Your TalkDesign Your Talk•• For any research talk, we need toFor any research talk, we need to

–– explain what we have doneexplain what we have done–– explain why we do itexplain why we do it–– convince our audience that we did a good jobconvince our audience that we did a good job–– keep our audience interested (or entertained)keep our audience interested (or entertained)

•• So need to design the structure of our talk accordingly!So need to design the structure of our talk accordingly!

Page 25: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Design Your TalkDesign Your Talk•• First let your audience know briefly what you are going to tell First let your audience know briefly what you are going to tell themthem

•• Set the background so your audience know the big picture in whicSet the background so your audience know the big picture in which h your specific accomplishment is made (why)your specific accomplishment is made (why)

•• Define your specific goals clearly Define your specific goals clearly

•• Lay out technical challenges you need to overcomeLay out technical challenges you need to overcome

•• Present your overall idea for solving the problemPresent your overall idea for solving the problem

•• Highlight some techniques used in your solutionHighlight some techniques used in your solution

Page 26: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Design Your TalkDesign Your Talk

•• Use Use examplesexamples to explain complex ideas/techniques so your audience to explain complex ideas/techniques so your audience can follow at conceptual levelcan follow at conceptual level

•• Only give technical description after you show basic ideas at Only give technical description after you show basic ideas at conceptual level and when you have toconceptual level and when you have to

•• Present validation of your solution to convince your audience Present validation of your solution to convince your audience

•• Demonstrate the usefulness of your workDemonstrate the usefulness of your work

•• Conclude with thoughts that can provoke your audience to thinkConclude with thoughts that can provoke your audience to think

•• Never forget to acknowledge where acknowledgment is dueNever forget to acknowledge where acknowledgment is due

Page 27: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Prepare Your TalkPrepare Your Talk•• A design templateA design template

problem definition and background

your contribution and methods conclusion and future research

conceptual level

Page 28: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Things You Want to Pay AttentionThings You Want to Pay Attention

•• Conceptual clarityConceptual clarity–– try to decompose complex concepts/issues try to decompose complex concepts/issues

•• e.g., presenting a complex algorithm using a two level presentate.g., presenting a complex algorithm using a two level presentation, ion, one at high level and one at detailed levelone at high level and one at detailed level

–– avoid mixing things that do not belong togetheravoid mixing things that do not belong together•• a common mistakea common mistake: mixing : mixing ““howhow”” and and ““whywhy””•• e.g., here is a procedure for solving problem X; in the middle oe.g., here is a procedure for solving problem X; in the middle of the f the

procedure, discus why you do Yprocedure, discus why you do Y

Page 29: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Things You Want to Pay AttentionThings You Want to Pay Attention

•• Big picture Big picture versusversus technical detailstechnical details–– avoid putting too much emphasis on technical detailsavoid putting too much emphasis on technical details

•• talks are for providing ideas while papers give detailstalks are for providing ideas while papers give details

–– a common mistakea common mistake: don: don’’t show (see) the forest for the treest show (see) the forest for the trees

Page 30: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Things You Want to Pay AttentionThings You Want to Pay Attention

•• Minimize text presentation and use visuals when you canMinimize text presentation and use visuals when you can(a picture is worth 1,000 words; and (a picture is worth 1,000 words; and examples are your best weapon!)examples are your best weapon!)

•• Minimize things that the audience could not see, Minimize things that the audience could not see, mathematical notations, tablesmathematical notations, tables

•• Give your presentation a smooth flowGive your presentation a smooth flow–– use good transition slidesuse good transition slides

•• State things that can provoke your audience to thinkState things that can provoke your audience to think

•• Give audience a break, by having humorous slides, once Give audience a break, by having humorous slides, once in a whilein a while

Page 31: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Polish Your DesignPolish Your Design•• Ask yourself if your audience can follow and stay interested in Ask yourself if your audience can follow and stay interested in your your

talk, after initial design; revise if neededtalk, after initial design; revise if needed

•• Trim your slides to make the presentation as succinct as possiblTrim your slides to make the presentation as succinct as possiblee–– rruthlessly prune material that is uthlessly prune material that is not essentialnot essential to your goalsto your goals

•• Avoid having busy slides Avoid having busy slides

•• Refine your presentation so the overall idea is crystal clearRefine your presentation so the overall idea is crystal clear–– even at the expense of cutting some interesting but messy detaileven at the expense of cutting some interesting but messy detailss

Page 32: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Refine Your Presentation Refine Your Presentation Slide #972Slide #972•• Consider two genes Consider two genes ggii and and ggii, we use the following to measure the , we use the following to measure the

functional relatedness between functional relatedness between ggii and and ggii

( ) ( )( )jkikkji ppGggP −−= 11,,

( ) ( ) jkikkji ppGggP −= 1,,

( ) ( )jkikkji ppGggP −= 1,,

( ) ( )∑= ⎭

⎬⎫

⎩⎨⎧ −−+

⎟⎟⎠

⎞⎜⎜⎝

⎛=≤−

L

l kl

klkl

k

kljkikkjiji N

ddNNNNppGdggggP

12

22

1,12min,,,

( )2

11,,, ∑

=⎟⎟⎠

⎞⎜⎜⎝

⎛−=

L

l k

klkji N

NGNAggP

where P is …….; G is ……; N is ……; || is ……..; N is ……; d is ….., NA is …..; ………

If P is within (a, b], we call the two genes very related; if P is in [c, d], we call them kind of related; otherwise not related

Page 33: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Refine Your PresentationRefine Your PresentationRevised presentationRevised presentation

•• Two genes are Two genes are functionally relatedfunctionally related if they are in same genomic if they are in same genomic neighborhoods across multiple genomesneighborhoods across multiple genomes

•• The The ““degree of relatednessdegree of relatedness”” can be defined quantitativelycan be defined quantitatively

•• Intuitively, the tighter the neighborhood is and the more genomeIntuitively, the tighter the neighborhood is and the more genomes s are involved, the more related the two genes areare involved, the more related the two genes are

genomes

genes

Page 34: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Fun Time Starts Fun Time Starts …………. . •• Now you have prepared your talk and are ready for the Now you have prepared your talk and are ready for the

show time show time …….. ..

•• Go through the slides the night before so the whole Go through the slides the night before so the whole material is fresh in your headmaterial is fresh in your head

Page 35: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Present Your TalkPresent Your Talk•• Always thank your host or meeting chairAlways thank your host or meeting chair

•• Audience in general wants to see you succeed; so donAudience in general wants to see you succeed; so don’’t t be nervous!be nervous!

•• Always read your audience and react to itAlways read your audience and react to it–– skip, repeat or provide additional materialskip, repeat or provide additional material

•• Engage and provoke your audienceEngage and provoke your audience

Page 36: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Present Your TalkPresent Your Talk

•• DonDon’’t read your slides as they are for the audiencet read your slides as they are for the audience

•• Be enthusiasticBe enthusiastic –– if you are not, how can your audience be?if you are not, how can your audience be?

•• Tell a joke to wake up the audience or give them a Tell a joke to wake up the audience or give them a breakbreak

•• Never go overtime!Never go overtime!

Page 37: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Things You Should/Should Not DoThings You Should/Should Not Do

•• Should doShould do

–– do homework to learn about your audiencedo homework to learn about your audience

–– include examples that connect with the localsinclude examples that connect with the locals

–– attend many talks and study what goes well and what does notattend many talks and study what goes well and what does not

Page 38: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Things You Should/Should Not DoThings You Should/Should Not Do

•• Should avoidShould avoid

–– only tell your audience what you have done but not whyonly tell your audience what you have done but not why

–– cover too much material cover too much material –– less is in general betterless is in general better!!!!

–– rush through slides to get more material coveredrush through slides to get more material covered

–– use same level of voice throughout the talk use same level of voice throughout the talk –– it puts people into sleep it puts people into sleep quicklyquickly

–– show off show off –– it turns people off quicklyit turns people off quickly

–– look at the screen rather than the audiencelook at the screen rather than the audience

–– end your talk abruptlyend your talk abruptly

Page 39: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Answer QuestionsAnswer Questions•• There are different types of questions, and handle them There are different types of questions, and handle them

accordinglyaccordingly–– need clarificationneed clarification–– suggest something helpfulsuggest something helpful–– want to engage in an intellectual dialogwant to engage in an intellectual dialog–– show that he/she is better than you areshow that he/she is better than you are–– ……....

•• Repeat the questionRepeat the question–– so the whole audience hears the questionso the whole audience hears the question–– give you some additional time to thinkgive you some additional time to think

Page 40: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Answer QuestionsAnswer Questions

•• Should anticipate some questions and prepare additional Should anticipate some questions and prepare additional material in advancematerial in advance

•• Avoid long answers, and suggest further discussion after Avoid long answers, and suggest further discussion after the talkthe talk

•• Be thankful for each questionBe thankful for each question

Page 41: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

TakeTake--Home MessageHome Message•• Set your goals clearly and design accordinglySet your goals clearly and design accordingly

•• Know your audience and present at the right levelKnow your audience and present at the right level

•• Make your presentation succinct and simpleMake your presentation succinct and simple

•• Present things clearly at conceptual level and use visualsPresent things clearly at conceptual level and use visuals

•• Do not rush nor go overtimeDo not rush nor go overtime

•• Be enthusiastic, confident and gracious Be enthusiastic, confident and gracious

•• Your job is: explain, convince and entertainYour job is: explain, convince and entertain

Page 42: Give a Good Research Talk - University of Georgiacsbl.bmb.uga.edu/~xyn/talk-about-talk.pdf · • A research talk should include the following components – define the problem you

Have Fun Giving Your Next Talk!

THANK YOU!