We watched the Netflix series called, "You Are What You Eat: A Twin Experiment". It was interesting and they did a decent job in showing how the industrial food supply is causing health issues for everyone (i.e. food getting less healthy in general) but where they did a less than stellar job is in showing/interpreting the graph with the blood testing results at the end.
The graph had a green dot for the median value for people on the omnivore diet and a yellow one for those on the vegan diet for different columns like LDL, HDL, Glucose, etc. The problem is that they never explained the margin-of-error bar that these dots where on and seemed to gloss over the "Statistical significance" for each of the columns.
They were pointing out what appeared to be dramatic results where some of them had a statitistical significance value as high as .98. To say that there was statitistical significance that number needed to be 0.05 or less. It's like they kind of left it up to the viewer to interpret results without understanding what they were looking at.
It was an interesting randomized study with the "identical twins" component, but it only lasted 8 weeks and an ever better study would likely have been a randomized cross over study. This is one where they start with an A and B group, do the intervention on one and the placebo on the other, then switch so that BOTH groups get to do and be evaluated for the intervention/placebo.
All in all it was entertaining if you like the subject matter but just be aware that there wasn't as much difference in the results as it might seem.