[Hnrobert42]

I wonder how you would collect that data.

[redtexture]

7 of the best GPS dog collars and trackers to keep tabs on your pet

https://mashable.com/roundup/best-gps-dog-collars/

[Hnrobert42]

But how would you test the inference with that data? You don’t need to know when the dog crossed the road but why.

Which reminds me of this Far Side cartoon: https://www.pinterest.com/pin/251990541629282256/

[mjevans]

I suspect a study to prove the exact means of connection is very difficult.

A retrospective study would be easier. Based on a survey of a given vets files (not quite a random sample, but for this case it might be good enough to establish a bias in facts), could trace the incidence rate of 'yard escape injuries' between 'intact' and neutered males, as well as 'intact' and spayed females.