Seven Design Principles for Presentation Skills: What 52 Studies Say Makes Speakers Improve

By Noam Shemla · The Studies

TL;DR Van Ginkel and colleagues at Wageningen University gathered 52 peer-reviewed publications, from 1990 to 2012, on how university courses build the ability to present. They mapped every study onto one framework and kept each feature of a course that more than ten of them had studied. The result is seven design principles, covering what students are asked to do, what they do and how they are assessed.

• Paper: Van Ginkel, S., Gulikers, J., Biemans, H., & Mulder, M. (2015). Towards a set of design principles for developing oral presentation competence: A synthesis of research in higher education. Educational Research Review, 14, 62–80.
• Design: A systematic review. Four databases searched in 2012 (ERIC, Scopus and three Web of Science indexes), then the reference list of every study kept.
• Scope: 52 peer-reviewed publications on higher education, 1990 to 2012: 41 empirical studies and 11 conceptual papers.
• Method: Every publication mapped onto Biggs’s 3P model of teaching and learning. A feature of the course studied in more than ten publications, about a fifth, became a candidate principle.
• Output: Seven design principles, each set out with the theory and the evidence behind it.
• Access: Subscription-only at the journal. The same text is open as Chapter 2 of the first author’s doctoral thesis, which is what this summary is read from.

Most advice on getting better at presenting is somebody’s opinion. Stan van Ginkel and colleagues at Wageningen University went looking for the version research supports. They collected two decades of studies on how university courses develop presentation skill and asked what the courses that worked had in common.1

They were answering a complaint from the field itself: that studies of presentation teaching each tested one ingredient, often without a comparison group, and reached contradictory conclusions, so nobody could say what a complete course should contain.

How the review was built

A first search on phrases such as “oral presentation skills” returned more than 300 results, fewer than 3% of them relevant: most described presentations without testing anything that improved them. So the team searched instead for the teaching (teaching, pedagogy, learning) combined with the skill, with verbs of improvement (develop, improve, enhance) and with higher education.

That narrower search found 25 publications. Two researchers screened them independently and kept 15. Reading the reference lists of those 15 added 37 more, most of which the search had missed because they named a specific ingredient, such as peer feedback or modelling, rather than “the learning environment”.

What the 52 publications were

Business was the most common field because presenting is treated there as a professional skill. Most of the studies, 36 of the 52, taught presenting inside an ordinary subject course; 16 were in courses devoted to public speaking.

One framework for every study

To compare such different studies, the authors put each one into the same frame: John Biggs’s 3P model of university teaching. What the students bring and what the course provides (presage) shape how they learn (process), which shapes what they can do at the end (product).

Two researchers first mapped three frequently cited studies independently, to agree on a method. The first author then mapped the other 49, each checked with the second author, and the individual maps were merged into one.

The learning processes box lists ten ways the studies thought students learned, from being taught and watching others to practising, handling feedback, reflecting on their own talks, giving feedback to peers and setting goals for the next one. That box is what turned a list of ingredients into principles: each principle had to come with a reason, in theory or in evidence, for why the ingredient should work.

The authors wrote each principle to a template from design research: to reach a given aim in a given setting, give the intervention certain characteristics, for stated reasons. The research team then argued each draft on four tests: whether its evidence convinced, whether it was distinct from the others, whether it could be used in a real course, and whether it could be read.

The seven principles

The principles follow Biggs’s idea of aligned course design: what students are asked to do, what they do, and how they are assessed.

The seven principles, how often each appeared, and its clearest direct test

What students are asked to do

Objectives. A talk has too many parts to improve at once, so goals should name one: the structure, the delivery, the audience. They should also be concrete, such as rehearsing the talk twice before giving it, rather than general. In the experiment the review leans on most, students who wrote their own specific goals outperformed students given a single general goal, on the content of their talks. Goals set by a teacher can work as well, the review notes, if students accept them.

The task. The review found three things about the assignment itself. Its content should matter to the students: in one experiment, students given a topic nearer their interests scored higher. Tasks should run from simple to complex across a course, which several studies linked to lower apprehension. And the audience should be real: presenting to people rather than to a camera, and to professionals or clients rather than classmates, went with more confidence and better talks.

What students do

Watching a model. The theory here is Bandura’s: people learn complex behaviour partly by watching someone perform it, and seeing someone like them succeed builds belief in their own ability. The one study to compare the two kinds of model found that watching a peer lifted students’ self-efficacy about their own talk, while watching an expert left it unchanged. Other studies found that both kinds helped performance; the authors judged the evidence for peer models the stronger.

Practice. Practice was the nearest thing to a universal ingredient: 47 of the 52 publications included it, with courses asking for anywhere from two to five talks. The most detailed study, in a business curriculum, found that students who gave two presentations did significantly better than those who gave one, while a third added no significant benefit, as if performance followed an S-shaped learning curve and the steep part came early. A later study saw the same pattern: clear gains from the first talk to the second, none from the second to the third.

The review is careful not to make practice a principle on its own. Several studies concluded it works only alongside other ingredients, such as an attentive audience and feedback between attempts.

How they are assessed

Feedback. Thirty-six publications treated feedback as essential, and the review draws out four qualities. It should be explicit, so the student can reflect on it. It should be tied to the particular talk and situation, so students do not draw the wrong general rule from it. Its timing should match the skill: in one experiment, immediate feedback worked better for habits that are nearly automatic, such as eye contact, and delayed feedback for changes that need deliberate thought, such as lengthening an introduction. And its intensity matters: students who got any feedback outperformed students who got none, and students sensitive to criticism improved more when it was tactful rather than blunt.

Peers. Involving classmates in assessment was argued for as active, collaborative learning that also teaches students the criteria by making them apply those criteria. In a controlled study, students who got feedback from peers as well as a tutor improved more than those with the tutor alone. The authors point out the catch themselves: the peer group also simply received more feedback, so the source and the amount cannot be separated. Not every student liked being assessed by peers, especially when they felt unsure of the criteria, and several researchers recommend training peers before they assess.

Self-assessment. Watching a recording of your own talk, often with a written portfolio, was linked in several studies to better talks, higher confidence and lower apprehension. Two experiments found no effect, and the review explains why they may have missed one: in the first, both groups watched their own videos, and in the second, self-assessment began only after most of the improvement had already happened.

The authors’ closing argument is that the seven belong together, because a course aligns what it asks, what students do and how it assesses. None of the 52 studies had examined all seven as a set; most tested one or two ingredients alone.

Has it held up?

The authors tested the principles themselves. They built presentation courses on all seven, then varied only who gave the feedback, assigning 144 first-year students at random, within each of five courses, to feedback from a teacher, from peers, from peers guided by a tutor, or from their own review of a video.2

A second follow-up asked whether feedback needs a person at all. Thirty-six students presented either to a virtual audience, receiving feedback the system generated from their voice and movement, or face to face, with feedback from an expert. Both groups improved in knowledge, skill and attitude, with no significant difference between them.3

Both tests come from the same team, and the second is small. The review itself has become a standard reference in the field: CrossRef lists more than 130 works that cite it.

What this study cannot tell you

• It counts; it does not weigh. There is no meta-analysis and no pooled effect size. A feature studied in many publications is not thereby shown to work better than one studied in few, and the 20% rule selects what researchers chose to study.
• Much of the evidence is weak by design. Few studies tested one ingredient on its own against a comparison group, and the authors say so for the task, for modelling and for self-assessment.
• The population is narrow. University students, two thirds of the publications from the United States, business the most common field. Working adults are not covered.
• It stops in 2012, before automated feedback on talks, from software rather than a person, was widely available.
• Some comparisons are tangled, such as peer feedback against simply more feedback, which the authors point out themselves.

What a speaker can take from it

The review describes a course, but the ingredients work for one person too. Improve one thing per talk and name it in advance. Speak to real people about something you care about. Watch someone a step ahead of you rather than only the stars. Then give the talk again soon, because the studies found the largest gain between the first attempt and the second.

Ask for feedback that is specific and about that talk. Take it on the spot for habits like eye contact, and give yourself time with it for bigger changes like restructuring an opening. Learn the criteria well enough to give feedback to others, and watch your own recording with those criteria in hand, not just to wince at it.

This article summarises a published review, read from Chapter 2 of the first author’s open doctoral thesis, which reproduces it. The counts, the κ value, the principles and the studies described are the authors’ own; the framework and funnel are redrawn in our words. The follow-up results are the authors’ own from the thesis; their confidence intervals are ours.

Measured

• How many of the 52 publications studied each feature of a course
• Agreement between two researchers on which publications to keep
• In the follow-up, rubric scores for talks before and after feedback from four sources

Inferred, not measured

• That each feature helps: drawn from theory and from studies that mostly tested features one or two at a time
• That the seven work best together: the authors’ argument, not yet tested as a set when the review was written

References

1. Van Ginkel, S., Gulikers, J., Biemans, H., & Mulder, M. (2015). Towards a set of design principles for developing oral presentation competence: a synthesis of research in higher education. Educational Research Review, 14, 62–80. https://doi.org/10.1016/j.edurev.2015.02.002
2. Van Ginkel, S., Gulikers, J., Biemans, H., & Mulder, M. (2017). The impact of the feedback source on developing oral presentation competence. Studies in Higher Education, 42(9), 1671–1685. https://doi.org/10.1080/03075079.2015.1117064
3. Van Ginkel, S., Gulikers, J., Biemans, H., Noroozi, O., Roozen, M., Bos, T., Van Tilborg, R., Van Halteren, M., & Mulder, M. (2019). Fostering oral presentation competence through a virtual reality-based task for delivering feedback. Computers & Education, 134, 78–97. https://doi.org/10.1016/j.compedu.2019.02.006