A few years ago, an exciting study made waves by showing that narrative spoilers could improve enjoyment… a really counterintuitive finding.
My colleague Judith Rosenbaum and I tried to replicate and extend this finding with more nuanced measures of enjoyment and appreciation from the communication literature. This project is now in press at Communication Research. We find that spoilers generally harm enjoyment, in keeping with conventional wisdom.
To make sense of these competing findings, we did another study, where we looked at individual differences. I turns out that people low on need for cognition (those who are not motivated to think hard) prefer to choose spoiled stories… probably because it makes the narrative easier to understand. But, people who had a high need for affect (who want emotion in their life), or who read a lot, experienced less enjoyment of spoiled stories compared to unspoiled stories. This experiment appears in an article over at Psychology of Popular Media Culture.
Next, I am working with students here at the VU University Amsterdam to test these (and other) aspects of spoilers with different television and film materials. People get really excited (i.e., distressed) about spoilers, and program producers and marketers are always walking a fine line in terms of how much to reveal in advance. The effects of spoilers are coming into relief through this research, which can help viewers and readers understand if they really should or should not avoid spoilers, and help program creators understand what to reveal when they are promoting their stories to potential audiences.
Interest in selective exposure to political communication, especially in online contexts, continues to grow. For example, see the recent article in Science on filter bubbles versus self-selection, and ensuing debate over its sampling technique.
One of the most important parts of my graduate training at OSU was to work with my advisor, Silvia Knobloch-Westerwick, in using behavioral measures of selective exposure in experimental settings. By unobtrusively observing media use behavior, researchers can gather measures of exposure and selectivity that are much more valid than self-report or recall measures. The selective exposure paradigm developed by Knobloch-Westerwick also allows for the control of an experimental design.
In the two recently published articles, we tested political selective exposure during the 2012 U.S. general election and the 2013 German federal election. A confirmation bias (toward attitude-consistent news articles) was evident in both elections, but the effect was stronger in the U.S.
We also looked at moderating factors such as attitude importance and source credibility, and further show that both attitude-consistent and attitude-discrepant information affect political attitudes after exposure. In short, people are biased in what they choose to view, and this reinforces their existing beliefs. But, when they do choose to view information they disagree with, it often weakens their views.