AI Analysis on Megan Pozzi's Speaking Performance

This is Megan Pozzi, who was the winner of Queensland University of Technology's 2013 The Three Minute Thesis (3MT) competition and the people's choice winner. We conducted an analysis using Rehearso's AI on her performance and here is the summary of what we got:

The AI says that her pitch and intonations perform the best. If you watched her speech, I think you would agree with that. Good use of intonations would grab the audience's attention and it would not sound boring. Together with her expressions, you'd just keep listening to her. Our AI also shows that she uses very minimal fillers, the most she used is only 'so' and 'like', but that is also only 5 times all added together. Interestingly, the AI points out that her hand gestures look stiff with quite some evidence points. When we checked on the timepoints, they are all pointing to moments when she clenched her hands. This hands clenching gesture should not be a problem, but it may look nervous and insecure when you start holding them too hard. During a speech, we are always encouraged to leave our hands naturally by the side when we are not using them to express something. Imagine if she could do that, it would look more confident and natural. One thing she did very well is she looks happy, which is an important emotion to have when you want to influence and look confident:

This is the AI score for her happiness emotion flow. She got a pretty smooth flow by starting it low and all the way to peak and slowing down again. Generally, a score of above 50 would mean the emotion is obvious and significant. If you do not wish to be perceived otherwise, your emotion flow for other irrelevant emotion (depending on your speech context) should be as follow:

All the while, the score is below 50, and I would say it is actually not significant at all even though you see an increase starting from 01:07 but it is only at 20–far from 50.

No wonder she is the People's Choice. Her score for her authenticity perception is high. It means she was genuinely interested in whatever she is sharing with her audience and was doing it sincerely. Similarly, a score of 50 and above means it is significant. As you can see, it is 60 and above from the beginning to the end.

This is an example of how we can use AI to help us in fine-tuning our speaking performance. In the software, the scores are being broken down in detail on your vocal, body language, and audience perception goal–from worst to best. You would be shown where you need improvement together with the evidence point of interest for you to decide if you need to do something about it. You also can know if you achieve your perception goal outcome. In Megan Pozzi's case, she achieved her goals. She is being perceived as authentic, persuasive, confident, and a good storyteller–the four top areas where she did best. This is why she is the winner. What do you think?