I'd go one step further. The worst case is they made up data enough to seem plausible, knowing full well replication is impossible, but their institution wanted

I'd go one step further. The worst case is they made up data enough to seem plausible, knowing full well replication is impossible, but their institution wanted their name in a major publication so they were told to make shit up and get published.