They had 16 different psych experiments (recruit subjects, tell them a story about a "victim", ask them about the victim's moral virtue). For 15 of the 16 experiments, the subjects were recruited online via Amazon Turk. I'd be concerned that Amazon Turk-recruited psych subjects are not even reliably representative of WEIRD populations (the point of crateless's comment), let alone non-WEIRD populations.

From their summary table, all 16 of their experiments (with various stories being told to the subjects) sound like "the victim was the chosen target of lone bad person". So - no victims of unsafe consumer products, nor of cancer, nor natural disaster, nor famine, nor systematic racism/sexism/etc., nor malicious or failed government, nor of war, etc. Extrapolating this study's results, to those types of victims, would be very difficult to justify. But (my impression) most of the current hot topics in the news, in public policy debate, and on HN involve those types of victims.

SO: The current title - "Victims are perceived as more morally virtuous than non-victims" - seems, in the general context here on HN, extremely overstated.