In a company blog post, Facebook says it will rely more on time spent reading an article as a metric for deciding if similar content should be more prominent in people's news feeds.
In the post by Moshe Blank, a Facebook software engineer, and Jie Xu, a research scientist, the pair says, "We've found that there are stories people don't like or comment on that they still want to see, such as articles about a serious current event." In those cases, the amount of time you spend on the article page after you've clicked to it from the news feed will be factored in. But time spent reading can be tracked only for articles published through the Facebook Instant Article tool or the Facebook mobile app. The data will be weighted so that long-form articles don't get preferential treatment.
Facebook did not immediately respond to a request for comment, but you can read the company blog post here.