EXCITED to have my second science communication video out today!
This was a collaboration with ASU’s Risk Innovation Lab, as I co-wrote the script with Dr. Maynard. In addition, I used the great video making setup in the lab’s facilities (instead of suffering in my own home with a small unstable whiteboard and terrible lighting).
The Process. To produce such videos, you first need a good to-the-point succinct script. This is the toughest part for me personally. Once you have that, you need to create the drawings to go along (I enjoy this part the most, though that’s not the case for everyone). Then you’re ready to film!
If you have professional lighting equipment, great camera, and a sturdy whiteboard, you can do it in < 2 hrs, which is how long it took me (my first video took much much longer- in fact I had to re-record sections on the next day). This part is probably the most tedious and frustrating– for one, try writing in a straight line and with good enough handwriting!! Then you have to record the voiceover- so, read the script you wrote. This can take many tries, but it seems like the simplest part to me!
Finally, you need to edit the video- so, take your recordings and synchronize them so that the images go with the script perfectly. This is not as horrible as you might imagine (iMovie makes it straightforward), but it does take some time. Overall, this video took me about 7 hours to make. My previous one (HERE) took about 16!!
I am very happy with this work (especially the epi detective with a sizable mustache), but I wonder about one element. Originally, I wrote the p-value explanation a bit longer. We then shortened it, but I am curious which version does a better job explaining the concept. Here’s the first writeup:
“One standard practice in analyzing data is to look at the P-VLUE (or probability value) to determine if the findings are true or are simply due to chance.
For this, a p-value cut off is set at 0.05: this means that the probability of findings being caused by random chance is 5% or less. P-values above this 0.05 threshold, meaning the probability of chance findings is more than 5%, are considered NOT statistically significant.
In other words: researchers across various scientific fields have arbitrarily decided that out of 100 findings, they are comfortable accepting that 5 of those will actually not be true but will be caused by random variations. And this amount of error is the most they are willing to accept (thus the 0.05 cutoff value).”
So, which explanation leaves a non-expert with a better understanding (in the video, it starts at 4:13)? Let me know if you have an opinion, because I honestly can’t tell which is more effective.