Readers Remember Misleading Headlines

Wednesday's Example of Media Bias   —   Posted on May 27, 2015

from a post by TJ Anderson at Content Customs:

…A new study shows that headlines can have more of an effect on a reader’s interpretation of an article than the text in the article itself – even if the whole article is read.

Everybody’s familiar with tabloid-style headlines that are clearly exaggerations or fabrications. Misleading headlines in supermarket tabloids and gossip magazines are to be expected. But what happens when the line between tabloid and hard news starts to blur? …..

What happens when one of the most well-known, supposedly unbiased news outlets is just as misleading [as the tabloids]? Check out this CNN article from September 2014 titled:

“Ebola in the air? A nightmare that could happen.”

This headline is definitely going to get some clicks (“ebola” was the top search term in 2014), but the experts interviewed for the story claimed that the chances of ebola mutating to spread through the air are actually very small. The headline could just as easily have been:

“Ebola in the air? Experts say it’s unlikely.”

The Australian study, published in the Journal of Experimental Psychology: Applied, gave participants four articles to read – two factual pieces and two opinion pieces, all of which were 400 words or less.

The articles also presented different slants in their headlines. For example, one of the factual pieces concerned burglary rates, which had decreased by 10 percent in the past decade but showed a 0.2 percent rise in the last year. Readers read two articles on this topic, one titled:

Number of Burglaries Going Up” and one called:

Downward Trend in Burglary Rate”

When the study participants faced a surprise quiz after reading the articles, they were better at recalling information that was congruous [compatible; matched] with the headline. In other words, readers could remember more details about the declining trend in the article titled “Downward Trend…” while also having better retention of the 0.2 percent increase in the article titled “Number of Burglaries Going Up.”

The headlines told readers what to focus on, and those are exactly the details they retained. On the other hand, most readers were able to infer that the burglary rate would decrease next year regardless of article headlines. …

The main problem here is that publishers are posting articles with…headlines that…end up actually leaving readers with skewed versions of the truth. This happens even if the whole article is read. Thus, the study suggests that content creators are doing a serious disservice to the their readers by using headlines such as these. The question is this: if publishers and article writers know that readers retain information from the headline more than anything else in the article, don’t they have a responsibility to avoid headlines that bend the truth? …

(from a 12/29/14 post by TJ Anderson at