University students need projects. Something that's real and they can learn how to do it - but isn't critical because it's a learning process rather than something you want to make policy on. So early projects are often things that seem more than a bit mad! Then some papers go all out on a let's-all-dogpile on the universities and what was a light hearted publicity piece than made lots of people smile and was a boost to the people doing the work... backfires
I wonder if this is one of these projects?
The research, not the dogpile!
Yup, statistical analysis exercises can be fun like this.
For example, students might plunder the National Crime Survey for variables such as
dog ownership, gender, being in regular employment, owning no motor vehicles and
living in an urban area, along with levels of fear about walking in the street at night.
A student can then see whether a dog-owner with no car who would walk their dog in the streets to the park at odd hours because of their job has more or less fear of street crime. They might find that Fido's owner has a low level of fear and can then explore the reasons for this confidence.
If you swap the variables around, like looking at the gender of our dog-owner, you might find a difference in levels of fear.
A larger variation might arise if you substitute
cat for
dog. The owner has no need to walk a cat so won't be nipping out twice a day.
Does this affect how safe they feel? One might predict that it will. The statistics will settle it.
We might suggest that this is an advantage of keeping dogs; having to walk them every day encourages owners to explore their environment, create loose ties with other dog-owners, develop stable routines and so on, all of which support their sense of personal security.
The National Crime Survey: endlessly fascinating.