By Troy Campbell
There’s a lot advice out there on how to give the perfect conference presentation. People say to be engaging, have clean slides, have good graphics, explain your contribution, and champion its relevance. Sometimes sources of advice contradict each other when some say to be funny and others say to be dry and professional.
Yet, regardless of the content of the advice, the motive of the advice is always the same. The advice seeks to help you do one thing: look good and smart.
What is missing in the advice is how you can actually help advance science.
While looking good and advancing science are not mutually exclusive, there are a few things that are lacking in conference presentations that would greatly benefit of science.
The audience should leave your presentation with an idea of what the next step is for science, not just a memory of how awesome a finding was or how funny you were.
Below are three pieces of advice on how we can all better advance science through our presentations.
#1 Have a true and guided discussion of the limitation of the theoretical process.
Let’s be honest: there’s probably no chance that you’ve tested your theory in the absolute best way possible. Funding and constraints almost always prevent this.
Further, there’s also little chance that you’ve removed 100% of the potential confounds. And if you have any doubt about this, just submit your manuscript to any journal club and the feedback you receive will make it abundantly clear how much you might be missing.
Psychological research is imperfect—and as Anna Kirmani and Michelle Pham have recently argued, it is important we realize this.
So what does this mean for your conference presentations?
It means you need to take control over the discussion of your limitations. Maybe you are worried about a hidden ceiling effect in the Likert measure or an undetected mood effect; mention those things. Further, tell us how we as a field could better test your theory. Maybe it is with larger more diverse samples or maybe it is with different dependent variables.
Remember, our greatest skill as scientists is arguably designing tests of theories, so take a moment to show that skill. Make your presentation part of a larger scientific discussion, not a one-sided argument for the undeniable rightness of your four M-Turk studies.
#2 Have a true and guided discussion of the relevance of your findings.
When the first choice overload studies came out, the authors made it seem that lots of choice was almost always bad. Turns out that although choice overload does exist, the original idea as the author of The Paradox of Choice Barry Schwartz explains was overstated.
Instead of looking back at how awesome their choice overload work was, in their original presentations and papers, the choice researchers should have looked forward and said to their audiences, here’s how we can test the breadth of this choice overload work, rather than assume its ubiquity.
Recently, my colleagues and I have tried to use a humble approach by starting many presentations with the statement: “Today our only goal is to have you leave here thinking: ‘Hey, this phenomenon happens sometimes, isn’t that interesting, and shouldn’t we explore it more?’” We’ve tried to follow examples by indicating we are introducing new information that may be highly important, and present ways in which we can further test its relevance – in part because, at some point or another, we have all been guilty of overstating things ourselves.
#3 Don’t be a theoretical imperialist and always remember the motto: “Death to Dichotomy”
Many researchers seem to feel the need to explain why their theory explains everything. This can make one can look very good and lead to a type of prominence in one’s field and the popular press. Yet, rarely is one perspective as completely explanatory as it seems.
For instance in the field of motivated cognition, people often have dichotomous debates between the existence of a motivation explanation (e.g. self-deception) versus a completely non-motivational explanation (e.g. informational differences, self-presentation bias). Scientists often try to explain lab and complex real world phenomena within one psychological theory with nearly unqualified general claims.
This leads to a theory versus theory or “dichotomous” view of reality. When as Michelle Pham argues, this is often far from the truth.
The majority of the most interesting phenomena are determined by multiple factors. For example, in the case of motivated cognition, there is synthesis of factors in self-deception that can occur due to concerns over self-presentation.
Again this dichotomous type of thinking does not lead to our goal of optimally advancing science.
Instead of pitting psychological mechanisms against each other in absolute terms, we should develop models that allow for a multiple psychological mechanisms. We should not ask does X mechanism explain this phenomenon better than Y mechanism? Instead we should ask, when do X and Y matter most when consider phenomenon Z?
A final consideration: Does all this humble limitation focus actually improve your presentation and make you look better?
So this post is centered around the idea that we need to stop focusing on looking good and we to need to start focusing on doing what’s best for science. Arguably, doing these three things can make you look better.
As a young scholar, I am always so saddened when I attend a talk by famous scholars or rising stars in academia only to see that they are biased as hell toward the prominence of their theories. Their self-aggrandizing style can hurt them.
In the end, scientists probably appreciate when other scientists act like scientists. Accordingly, following this advice may not only help produce good science; it may also help you look good. And, oh no… it seems I’ve just written another blog about “looking good” in a conference presentation.