This new result isn’t in contradiction to the original test; it affirms it. It shows why poverty is passed down, despite government programs to increase social motility.
The similarity of the kids’ socioeconomic background in the first test was actually useful; it reduced the socioeconomic of the parenting without presuming causality.
What drives the marshmallow test? The emotional security of the pay-off. Poor kids lack that security, and act in ways that perpetuate poverty.
Has no-one ever noted the similarity between this test and the forgiven sin of Charlie Bucket in the chocolate factory?