I'm using here bayesian inference to find expected probability of one inductive argument "no wedding ring -> not married":
P(not married | no ring & my data) = 62% ±14%
In other words, given the data I have, if one sees that a person has NO wedding ring, one should rationally expect that this person is NOT married with only 62% confidence level.
Prior distributions are such, that prior expectation of this probability is:
P(not married | no ring) = 96% ±11%
Hence, my analysis falsifies the hypothesis, that this inductive argument is accurate.
Here's a little article of mine, describing what data I used and some basic logic of how it's connected to the probabilities I estimate https://medium.com/@chainforced/the-absence-of-a-wedding-ring-is-an-uninformative-sign-of-singleness-of-random-people-fbf1f924088f