that's the analysts being responsible statisticians instead of making wildly overblown claims. but when I read the entire report, what's convincing is that they find that a disproportionate % of accidents with pedestrians/cyclists at intersections and in other low-speed situations involve hybrid vehicles. the differences are statistically significant at conventionally-acceptable (95% confidence level) intervals.

and those results are also consistent with pedestrians/cyclists not being able to hear the hybrid vehicle coming. so don't dismiss it as "inconclusive"