Jaywalk3r

The actual study is behind a paywall, so I can't access it. So, going by the article, which is less than optimal …

One city was studied, so our n = 1. That's not very useful for drawing broad conclusions. We can't extrapolate the results to other cities. I've ridden in lots of cities, and just like with bike lanes, some implement sharrows well and some implement them poorly. Around here, sharrows tend to be better implemented than bike lanes, so I take the sharrowed streets whenever I have a choice between the two. That preference is dependent on the city in which I'm riding.

The study doesn't seem to attempt to look at how much of the increase in bicycle use is due to implementing the bike lanes versus due to sharrows. The direct safety benefit of bicycle infrastructure is usually smaller than the indirect benefits, at least in cities with room for rapid modal share growth for bikes. Safety in numbers is generally more important than infrastructure, but nothing gets more people out on bikes like infrastructure.

There was not a statistically significant difference between the decreased rates of accidents on sharrowed streets and non-changed streets (and perhaps with striped bike lane streets, the article's wording is ambiguous there). This matters, because one of the important results of (properly implemented) sharrows is showing bicyclists (and motorists) proper lane positioning for bicycles. That knowledge carries over well to non-sharrowed streets, making it very difficult to isolate the effects of sharrows simultaneously with bike lanes.

In general, I would expect bike lanes to get more people out riding and sharrows to have a more positive impact on motorist behavior. The study does not appear to isolate the impacts of these two factors, which each enhance bicycling safety.

It might be that the study was well-designed, but the article described it poorly. But if the article offers a reasonable description of the study, then there are lots of reasons to accept the results only with a giant grain of salt.