To explain your research to the public in a video, turn the paper's order upside down. Start with the everyday question your work speaks to, give the one finding a non-specialist should remember in plain words, explain how you found it in a sentence or two, and end with what it does not mean, because that is the part the public most often gets wrong. Keep it to two to three minutes, which is 300 to 450 spoken words. Replace every technical term with an everyday one or define it once, draw a picture for each step instead of showing your figures, and have a non-scientist watch it before you post. Then check it with your press office and co-authors.

A public video is not a short paper

A paper is ordered for peers: background, methods, results, discussion. A public audience will not wait two minutes for the result. They want to know, in the first fifteen seconds, why this is about them.

The guide to turning a research paper into a video summary is for a seminar, a journal club or your own reading, where the viewer already knows the field and wants the effect size with its confidence interval. This post is for the opposite viewer: someone curious, intelligent, and with no background at all, who found the video because a friend shared it or because they searched a question.

A structure that works: and, but, therefore

The simplest reliable structure for public science is one the biologist and filmmaker Randy Olson calls ABT: and, but, therefore. It fits almost any study.

• And: here is something we know, and here is something else.
• But: here is the gap, the puzzle or the surprise.
• Therefore: here is what we did, and what we found.

It works because it gives the viewer a reason to keep watching, the "but", before you ask them to follow a method. Write your ABT in three sentences before you write anything else. If you cannot, you do not yet know what the video is about.

Swap the jargon

Every field has words that feel ordinary to the people in it and opaque to everyone else. Some of the worst offenders are words the public also uses, with a different meaning.

Go through your script with a highlighter and mark every word a sixteen-year-old would not use. Replace it, or define it once in passing and then use it the same way every time.

A worked example: sound and plant growth

To show the method, here is a script for an imaginary study. It is made up for illustration, and the numbers are not real results.

Suppose your group grew 200 tomato seedlings in two identical rooms. In one room, a speaker played a recording of wind through leaves for six hours a day; the other room was silent. After four weeks, the seedlings in the sound room had stems that were, on average, a little thicker, and no taller. Your paper calls this a modest effect on stem diameter with no effect on height, and spends three pages on acoustic calibration.

The public version, about 330 words, a little over two minutes:

You may have heard that playing music to your plants helps them grow. It is a nice idea. Gardeners have talked about it for decades.Plants have no ears, but they do respond to being touched and shaken. A tomato plant grown in a windy spot grows a thicker, sturdier stem than one grown indoors. So we wondered: is it the movement, or could the sound of the wind alone make a difference?We grew two hundred tomato seedlings in two identical rooms. Same light, same water, same soil, same temperature. The only difference was sound. In one room, a speaker played the sound of wind through leaves for six hours a day. The other room was silent.After four weeks, we measured every plant. The plants in both rooms were the same height. But the plants that heard the wind had slightly thicker stems. Not hugely. About the thickness of a sheet of paper more, on average. But it was a steady difference, and unlikely to be down to chance.So what does that mean? It does not mean music makes plants grow. Our plants did not grow taller, and we played wind, not music. It also does not mean the sound is doing what real wind does. We do not yet know how a plant could detect sound at all.What it does suggest is that plants may be sensitive to more of their surroundings than we assumed, and that sound might be one more signal a plant uses to prepare for rough weather.The next question is how. That is what we are working on now.

What it does: the opening meets a belief the viewer already has. The "but" is the question about sound versus movement. The method is one paragraph with no equipment named. The finding has one comparison a person can picture (a sheet of paper) instead of a millimetre figure. And the longest paragraph is about what the result does not mean, because "Scientists prove music makes plants grow" is the headline this study would attract.

Scenes

Notice there is no bar chart. A chart is the right way to show this result to a reviewer. For the public, two stem cross-sections side by side say the same thing in a second. More on finding pictures for findings with no obvious image is in how to illustrate abstract ideas.

Honesty is the hard part

Public research videos fail in one direction far more often than the other: they overstate. Under the pressure to be interesting, "associated with" becomes "causes", "in mice" disappears, and "may" becomes "will".

Before you publish, read your script against your paper and check:

• Every causal word is justified. If your study was observational, the script must not say "causes".
• The population is named. In mice, in cells, in 200 tomato seedlings, in adults in one country.
• The size is honest. If the effect was small, say small, and give a comparison people can picture.
• The "does not mean" paragraph exists. Write the misleading headline your study could attract, then write the sentence that prevents it.
• Co-authors have seen it. They are named on the work, and they should approve how it is described.
• Your institution's press office knows. Many have guidance on public communication, and if the paper is under embargo, the video must wait until the embargo lifts.

If the work is not yet peer-reviewed, say so on screen and in the description.

Make it

1. 1

   Write your and, but, therefore

   Three sentences: what is known, the gap or surprise, what you did and found. This is the spine of the script.

2. 2

   Write 300 to 450 words for a non-specialist

   Open on an everyday question, one paragraph of method, one plain finding with one comparison, a paragraph on what it does not mean, and the next question.

3. 3

   Run the jargon pass

   Highlight every term a sixteen-year-old would not use. Replace it or define it once.

4. 4

   Paste it in and keep your wording

   In openCanviz, paste the script, choose Keep my wording so your careful phrasing survives, and pick whiteboard for a process or painted for an outdoor or field study.

5. 5

   Replace charts with pictures and check every scene

   Redraw any scene that shows a different result from the narration, or labels that are wrong. Fix each in its own scene.

6. 6

   Record it in your own voice and test it

   Viewers trust a researcher speaking about their own work. Record over each scene, then show it to someone outside science and ask them to tell you what you found.

The test in the last step is the one that matters. If your non-scientist friend says "music makes tomatoes grow", the video has failed, however good it looks. Rewrite the "does not mean" paragraph and try again.

Where to put it

A two to three minute horizontal video works on your lab website, YouTube and in a conference talk. For social platforms, cut the strongest thirty to sixty seconds (usually the "but" and the finding) into a vertical version that links to the full one. Always add captions; many people watch without sound. Put the paper's full reference and a link to it in the description, so the curious can go further.

Common questions

Should I appear on camera? You do not have to. A drawn video narrated in your own voice carries most of the trust of a face, and is much quicker to update. If your funder or institution wants a talking-head introduction, film ten seconds of that and use the drawn video for the explanation.

How much of the method should I include? One paragraph, describing what you compared and how you measured it in everyday terms. The public trusts a result more when they can picture how it was obtained, but they do not need the equipment list.

What if my research has no obvious picture? Draw the comparison, not the apparatus. Two groups side by side, a before and after, a quantity growing or shrinking. Almost every finding is one of those.

Can I make it in other languages? Yes. Make it once, duplicate the project, and switch the narration language. Have a native speaker who knows the field check the translated script for technical terms before you post it.