How to Conduct an Interview While Filming a Video

As I’ve talked about before, conducting an interview is one of the biggest challenges the scientist videographer may ever face—especially at noisy venues such as a scientific conference. In a previous post, I described how I had conducted and filmed a series of interviews at a scientific meeting and pointed out what I had learned from the experience.

Now, I’ve created a short video that covers ten tips for conducting interviews while filming a video:

Use Video to Promote the Mission of Your Science Society

This week, the Society of Wetland Scientists (SWS) rolled out their new media initiative and YouTube Channel. Their website explains how video can be used by SWS members to share their work and why video can be beneficial to the SWS mission:

Exposure: Video can raise awareness of wetland issues, new research, and society activities.

Communication: Video augments other forms of communication, such as technical articles, but is a more accessible and modern way to share information that appeals to a broad audience.

Education: Video can enhance the public’s understanding of the importance of wetlands, can inspire current and future wetland scientists, and help in recruiting students to the study of wetland science.

The SWS New Media Team is currently soliciting videos from members and non-members with an interest in wetlands. If you are a wetland researcher or student studying wetlands…or just a wetland enthusiast, consider submitting a video (see the video preparation and submission instructions). If you’ve never made a video before, the following tutorial provides some basic guidelines for making a video with a smartphone.

How to Create a Videographic

I recently came across a Tweet from Climate Central that was illustrated with a striking videographic, which is a combination of graph and video. In this case, the graph showed where Earth’s accumulated energy (heat) ends up (land, sea, air), and it was superimposed on a video of ice floes floating on the ocean.

The idea behind such videographics is to create an attractive and memorable information product that catches people’s eye. The moving image draws your attention as you scroll through Tweets or surf through a website. My attention was definitely captured, and I took a closer look at the graph and the data it presented.

In addition to making your Tweets more visible and informative, videographics can be used on a webpage, as supplemental online material for a journal article, or for a scientific presentation. On a webpage, it can create an eye-catching visual that highlights a recent publication. More journals now accept videos and interactive graphics to accompany articles; a videographic can enhance an online article or be offered as a downloadable supplementary file. Judicious use of a videographic in a conference presentation or seminar can emphasize a key finding and make the point more memorable.

So, how do you create a videographic? It’s relatively easy if you know how to use Photoshop and a movie-editing program. Here are the steps:

1. Prepare your graph in any graphing program and save it as an image (jpg, png).

2. Open the image in Photoshop.

3. Use the “magic wand” tool to highlight the graph’s background and delete it.

4. Now save the graph with its transparent background as a .png file, which will preserve the transparency.

5. Import the new graph into iMovie (or other movie-editing program).

6. Import a video clip that illustrates what the graph depicts (clouds streaming across the sky, waves lapping on the shore, people walking).

7. In the timeline, add a ten-second segment (or whatever duration you choose) of the video. Add the graph to the timeline as a picture-in-picture image and resize/re-position as needed.

8. Export the video file and post it on your website or in a Tweet.

I made a tutorial showing exactly how to prepare your graph and then superimpose it on a video clip (see embedded video below or go to this link).

How to Improve Your On-Camera Delivery in Science Videos

Picture this scenario:

A middle-aged scientist in a white lab coat is speaking on film about his research on cancer. He’s sitting in a well-equiped laboratory and looks very authoritative. The camera gradually pans from a broad view of the room to focus in on the scientist. He begins by saying, “I’m really passionate about my work and want to share my findings with you in this video.” The only problem is that this cancer researcher does not look or sound passionate! Far from it. Instead, he sounds like a robot. He speaks in a monotone, does not smile or show any other facial expression, uses no hand gestures, sits stiffly and does not make eye contact with the viewer (his eyes are looking down or off camera). Things don’t get any better as he continues to explain the details of his research. 

Now, I can sympathize with this guy because this is how my early attempts at making videos about my research looked and sounded. I’ve improved since then, but still find it really difficult not to come across on camera like Mr. Spock (played by Leonard Nimoy in the original Star Trek series). Spock had difficulty showing emotion due to his Vulcan ancestry.

So what’s our excuse?

I think there are three basic reasons why some scientists come across on camera as being stiff and robotic: personality, training, and fear of the camera. People who are naturally gregarious or funny come across well on camera, but someone who is introverted may seem stiff or robotic. It’s possible to go against your natural demeanor, but you will likely find it difficult. I’m a naturally reserved, quiet person and feel terribly awkward when I try to be more extroverted. Also, I have to fight the years of training and experience talking to an audience of scientists, during which I cultivated a demeanor of calm confidence and authority. My talks at conferences and in seminars have been successful because those audiences expected a serious, academic delivery. But what works for an audience of scientists can be a detriment on camera. My serious, authoritative demeanor could be misinterpreted as arrogance or just a nerdy attitude. In addition, the camera not only adds ten pounds to your apparent body weight, it drains your energy. Consequently, it’s necessary to be more personable and to raise your energy level when being filmed above that normally used with a live audience. If you are like me and have a more reserved demeanor, you will have to work much harder than your colleague who is naturally gregarious and likeable.

Also, many people—even experienced speakers—freeze up when the camera is turned on them. They get that “rabbit in the headlights” look on their faces, and their bodies seem to turn to stone. Whenever a camera was turned on, I found it difficult to gather my thoughts and speak coherently. This reaction is a bit like stage fright and can make you look like someone with “Stuck in Their Heads” syndrome. Extreme self-consciousness is the culprit here.

After watching many, many videos made by science professionals (or videos in which a scientist appears), I realized that there were quite a few people out there with the “Stuck in Their Heads” problem. I’ve wanted to make a video tutorial about how to improve on-camera delivery, but put it off because I did not think I was the best person to tackle this topic. I thought it was better to hear tips about on-camera delivery from someone who does it well. However, it finally occurred to me that people might want to hear how a scientist with this problem has faced the problem and eventually improved.

In the video below, I briefly explain what I think are the main problems someone faces when trying to speak on camera and a few ideas of how to overcome them (direct link to video).

As you saw, there are several ways to improve your on-camera delivery if you are having problems. I focused on the most common issues and how to overcome them. My take-home message to you is not to give up if your delivery is poor at first. Keep practicing and you will improve. Even though I’m not as engaging or likable or convincing as, say, Neil deGrasse Tyson, and never will be, I have improved. More importantly, I feel less self conscious and thus more comfortable speaking on camera.

One bonus to learning to speak with more energy and confidence on camera is that it can help you in other stressful, speaking situations such as a job interview seminar or a TED talk. If you have an upcoming presentation, film yourself practicing your talk and try to apply some of the tips I cover in the video. I think you’ll find it’s well worth the effort.

Can Artificial Intelligence Help Scientists Be Better Communicators?

This post is part of a series about Artificial Intelligence (AI). In this concluding post, I explore the possibilities of AI to help scientists be better communicators.

As I’ve talked about before, many scientists have difficulty communicating their work in a way that is interesting and compelling, both intellectually and emotionally. This situation is improving, as more people recognize the importance of addressing the growing anti-science movement in the U.S. and the need for credible and articulate scientists to state the case for science. Once upon a time, scientists could safely remain in their ivory towers and talk among themselves about science. But no longer. Scientists are increasingly called upon to talk to the media (the AAAS has even published media interview tips for scientists), to testify before Congressional committees, to give public lectures, and to explain the “broader impacts” of their research on society. Consequently, efforts are underway to train the next generation of scientists to be better communicators (e.g., through academic programs focused on science communication). Science students today also seem to have a greater interest in developing better communication skills than when I was a student (just my personal impression).

Despite the advances in communication technology and emphasis on the new media to communicate information, though, I find that students still struggle with many of the same issues that plagued earlier generations when it comes to explaining science. And as the volume of science information grows exponentially, staying abreast of the literature and communication technology will be increasingly difficult for these future scientists.

The following are a few ways in which AI may help.

Designing a More Effective Science Message

One of the difficulties faced by a science communicator is how to design a message that resonates with a particular target audience. Few scientists are trained in communication theory and often rely on their default mode—explaining their work as they would for a technical audience. But what if the intended audience is not trained in science? How would you know if your message is appropriate in content and tone? AI might help, for example with the tone. The IBM Watson Tone Analyzer “uses linguistic analysis to detect three types of tones from text: emotion, social tendencies, and language style. Emotions identified include things like anger, fear, joy, sadness, and disgust. Identified social tendencies include things from the Big Five personality traits used by some psychologists. These include openness, conscientiousness, extroversion, agreeableness, and emotional range. Identified language styles include confident, analytical, and tentative.” One of the intended uses is to optimize a message intended for a particular audience. A message that shows strong emotions and is less analytical in style may be perceived more favorably by the general public, for example. You can try it out by inserting a piece of text into a dialog box and get an analysis of the overall tone of the message as well as a sentence by sentence breakdown. There are links to additional information about what a particular tone conveys and how to improve the tone of a message.

Finding Appropriate Material for Your Science Message

AI may be particularly useful in reducing the time involved in finding material to include in a message as well as to locate media that can be freely reused (such as in the public domain) or purchased for a fee. I know I spend a lot of time searching for footage, images, animations and music that I can freely use in a science video. Because I can’t review everything available, I probably miss a lot of really good material. Search engines can locate photos or videos posted on the Internet based on provided keywords and criteria (e.g., size, resolution, format). However, I may still end up with thousands of candidate media, not necessarily ranked according to what I might need. Artificial intelligence systems may improve such searches. Google, which used algorithms (rules set by humans) in the past to respond to search queries, is transitioning to deep neural networks, which can learn to respond to new search queries and other tasks such as figure out where a photo was taken. Improvements in search tools could make finding the right media for a video or other information product much easier.

Creating Media for Your Science Message

Another way AI might help is in creating new media such as art or music that can be used in an information product. If you need a painting or jingle but are not artistic or musically inclined, you may one day be able to generate what you need using an AI system trained to do this. An example of artwork created by an AI system and a 3-D printer is a new painting by Rembrandt…or rather one created by a computer based on information from 346 of Rembrandt’s paintings. The video below shows the amazing process by which this 3-D painting was created:

There are also efforts to develop AI systems that can compose music. Google is apparently working on such a system, although not everyone is impressed with the result. If you want to play around with a music-composing system (based on language algorithms), check out Wolfram Tones. You can select a music style and change up the instruments and other aspects to create a unique tune.

Teaching Science Professionals to Communicate Like Normal People

Scientists are traditionally taught to maintain a serious demeanor when speaking to an audience of our peers so that we are judged to be credible sources of information. But this approach doesn’t work so well with the average person. By hiding our emotions, we can come across on camera as “robotic”, for lack of a better term. So it’s rather ironic to consider if an AI can help scientists be better communicators.

The computer, Watson, was trained to “recognize” different emotions displayed in a film and to assess and rank the ones that would work best in a movie trailer about that film. In the same way, an AI system could be trained to evaluate video footage showing scientists or students conducting their research or discussing the challenges they faced and select the best clips in terms of conveying emotion (enthusiasm, humor, curiosity, tension). But we don’t really need a computer to tell us which footage shows a particular emotion—we are much better at this than any machine or program currently available.

However, an in-depth analysis of a person’s on-camera delivery of information might be used to train science professionals to be better communicators. A video clip could be fed into a computer like Watson to be assessed on the basis of both content and tone. The speaker would be evaluated and scored according to various criteria. They could then try to alter some aspect of their performance and see how it affects their scores. This immediate feedback from a machine might be a faster, more efficient, and less painful way for someone to improve their communication skills. A problem could be identified early and  eliminated before it becomes a habit.

I’m not aware of any system that can analyze a video of a person speaking, but there are AI-based personality tests that use what someone has written (an essay, a letter). One example is the Watson-based service, Personality Insights. You can see the outcome for famous people (Gandhi, Barack Obama) or you can insert your own text. I gave it a try by inserting the text from one of my blog posts. Here’s what it said about me:

You are unconventional, somewhat indirect and skeptical. You are authority-challenging: you prefer to challenge authority and traditional values to help bring about positive changes. You are philosophical: you are open to and intrigued by new ideas and love to explore them. And you are unstructured: you do not make a lot of time for organization in your daily life. Your choices are driven by a desire for discovery. You are relatively unconcerned with both tradition and taking pleasure in life. You care more about making your own path than following what others have done. And you prefer activities with a purpose greater than just personal enjoyment.

This analysis is eerily correct about some aspects of my personality. However, I got different results when I tried different text. For example, a second blog post generated the statement that my “choices are driven by a desire for organization”–opposite to the preceding analysis. Other aspects remained the same: authority-challenging, love of discovery, going my own way rather than following others. It’s necessary to provide sufficient text for a strong analysis, and the service warns you if you’ve given too little text. It also provides a more in-depth breakdown of the various traits that were analyzed.

As I said above, I don’t think there are any AI systems that can analyze a person’s performance in a video. However, it seems that this might be possible using a combination of existing AI systems such as the movie trailer and the personality test described above.

Conclusion

In this post, I’ve mentioned only a few ways AI might be used to improve science communication. Some of these systems, such as better search engines for locating media, are already available to us. Others will need more work and testing. I started this series with a hypothetical, futuristic scenario about a scientist using an AI system to create a video proposal. I don’t know if such a thing would ever exist or even be of widespread use in the scientific community. But it was fun thinking about it and learning more about AI systems.

This post is final part of a series about Artificial Intelligence (AI) and its potential role in science communication. You can find the first post in the series here.