[dataflow]

Huh? The point is that your attention (and vision) isn't nearly as divided when driving with a stop sign than with a roundabout. You don't have to multitask nearly as much; you do one thing at a time. Less division of attention = less car accidents.

[uoaei]

It doesn't sound like you're familiar with any of the studies concerning the safety of traffic flow if you think that that is the only factor that determines intersection safety.

[dataflow]

> It doesn't sound like you're familiar with any of the studies concerning the safety of traffic flow if you think that that is the only factor that determines intersection safety.

I'm pretty sure "the only factor that determines intersection safety is division of attention" was not a stance I was taking (when is an event ever a function of just 1 variable in the real world?), but if you'd prefer to take a swipe at me regardless, it would be nice if you could make your response constructive and actually link to some studies that show I said something contrary to reality, if you're well-studied in traffic flow.

Hopefully my reply clarified for you what my earlier point was, even if you think the point was wrong.

[uoaei]

The implied point of the above comment was that, while specific linear relationships can be hypothesized, and shown in certain limited controlled experiments, the nonlinear interactions with every single other factor makes such an assertion not only meaningless but also misleading.

"Well-studied", no, I would not claim that. I follow the zeitgeist of urban planners as they discuss topics like this on the fora in which they congregate. I read some of the studies they post and discuss, and I have picked up on some of the memes present in that community. Something that comes up again and again and again and again (and which is immediately apparent upon reading the research) is the primacy of roundabouts for intersections, because they are the safest of the popular options and less prone to congestion than stop signs, stop lights, etc. Even when they do get congested, the outcomes are better for everyone on average, since everyone waits a similar amount of time compared to the asymmetry of e.g. intersections with stoplights. One common pitfall in thinking about these things is only thinking about individual wait times in a subjective sense, and not aggregate wait times in a systemic sense. The latter perspective provides much clarity.

Here, I did your Googling for you. Where do I send the invoice?

https://www.iihs.org/topics/bibliography/ref/1248