A study of storytelling strategies
By Annie Lang & Deborah Potter

In this study, four stories submitted to NewsLab by local news stations were modified to conform with a set of seven rules for improving news comprehensibility. The rules were developed by Potter & Lang (2000) from an extensive review of the news processing literature. The seven rules are:

Let the emotions talk. Emotions compel attention, but they require more effort to process. Emotion may increase memory for a story but not for its details. Use emotion sparingly, and when the emotion is strong, keep graphics, language, and relationships simple.

Slow it down. Keep the pace of production moderate. If the story is complex, keep the pace slow to moderate and structural features to a minimum.

Dare to be quiet. When the video is complex or compelling, be quiet for one or two seconds, this allows the viewer to process the video followed by the audio contact. This will significantly improve memory for the verbal material. If both things happen at once, viewers will remember the pictures and forget the words.

Match the audio and video. When audio and video information match, understanding and memory are better.

Know how to deal with negative images. Negative pictures compel attention. Strong negative images tend to result in strong memory for the negative image, but weak memory for concurrent audio information. In addition, memory is particularly bad for audio information which precedes the negative images appearing on screen. On the other hand, memory for audio information which follows the negative images (i.e. audio which occurs after the negative images are no longer on screen) is increased.

Take a literal approach. Use concrete words and pictures when possible. Using high imagery words which build mental pictures may be better than using bad or semi-related video. Graphics which show relationships are better than graphics which present facts.

Use strong chronological narratives. Stories told as narrative, in chronological order are easier to process and are better remembered than stories with weak narratives or told in non-chronological styles.

To test the ability of these rules to improve processing and memory for news stories, two versions of each of four news stories were produced. Version 1 conformed as closely as possible to the story submitted to NewsLab by the local station. NewsLab, however, reproduced these stories both to remove identification with the original station, and so that both versions of the story would use the same reporter. The stories were re-voiced, and the stand ups reproduced. However, the script remained unchanged and as much of the original video track as possible was used. Version 2 of each story was designed to implement the seven rules, as applicable. In most cases, the original video footage was the raw material for the reproduced version. The pacing, order, and choice of video were changed to implement the seven rules. If appropriate, new graphics were produced. The scripts were rewritten, using the facts from the original story, to conform with the seven rules.

In order to test whether the new versions of the stories were indeed more comprehensible and more memorable (without becoming less compelling, interesting, or arousing) an experiment was conducted. In this experiment, 40 residents of a small Midwestern city between the ages of 25 and 65 were invited to participate in the experiment. A small payment was made to a local organization in return for their participation. Participants viewed all four news stories. Each participant saw two of the original stories, and two of the re-done stories. Different participants saw different combinations and different presentation orders of the four stories. While the participants were viewing the stories, their heart rate and skin conductance were measured. Following each story participants rated the stories on seven scales. The scales were informative, believable, interesting, understandable, enjoyable, engaging, and important. After all four stories had been viewed participants performed an unrelated task for about ten minutes. Following that task, participants were asked to write down what they could remember about each of the four stories. After participants finished completing this questionnaire, they were thanked and dismissed. Forty eight hours later subjects were telephoned and asked to list as many of the stories they had seen as possible.

The following predictions were made about the effect of Version on our dependent variables:

Hypothesis 1: Participants will rate the re-done versions as more interesting, enjoyable, engaging, believable, important, informative, and comprehensible than the original versions.

Hypothesis 2: The reproduced versions of the stories will not elicit less attention (measured by heart rate) or arousal (measured by skin conductance) in viewers compared to the original versions.

Hypothesis 3: The reproduced versions of the news stories will be better remembered overall than the original stories, both immediately after viewing the stories, and forty eight hours later.

Hypothesis 4: Participants' written descriptions of the reproduced news stories will contain more correct information and less incorrect information than their written descriptions of the original stories.

In other words, we are predicting that altering the news stories to conform with the seven rules will make the news stories easier to understand and easier to remember without making them less arousing or reducing their ability to elicit attention.

Results

Hypothesis 1

This hypothesis predicted that participants would rate the reproduced news stories as more informative, comprehensible, interesting, engaging, believable, and important than the original stories. Indeed, as expected, the reproduced stories were rated significantly higher on all measures except importance. These results are shown in Table 1. The ratings are on a 10 point scale.

Table 1: Effects of Version on Participants Ratings of News Stories

Hypothesis 2

This hypothesis predicted that implementing the seven rules would not reduce the level of attention or arousal elicited by the new stories. Heart rate can be used as a covert measure of television viewers' attention during viewing. The more attention viewers pay to the television, the lower their heart rate becomes, even when the content is arousing or excited. We tested for a Version X Time interaction on the heart rate data. That effect approached significance (F (9,297) = 1.68, p <.094) and is shown in Figure 1. As can be seen , heart rate is, if anything, slower (suggesting greater attention) during reproduced versions of the stories than it is during original versions.

This hypothesis also predicted that the reproduced versions would be at least as arousing as the original versions. Analysis of participants' skin conductance during viewing yielded no significant differences between the two versions of the stories (F < 1). Similarly, subjects self-report ratings of how aroused they felt during viewing did not differ significantly for the two versions (F < 1).

Hypothesis 3

This hypothesis predicted that viewers would have greater immediate cued recall and delayed free recall for reproduced versions of the news stories compared to original versions. The cued recall protocols were coded by two coders (inter-coder correlation r =.95) for correct facts (one point each), correct relationships (two points each), partially correct relationships (one point each), incorrect facts/relationships (minus one point). Total cued recall per story was calculated by summing the categories. Results show that participants had significantly higher (F (1,41) = 4.32, p <.044, epsilon squared =.065) recall for reproduced versions of stories (mean = 3.56) compared to original versions of stories (mean = 2.79).

During delayed free recall, participants listed the stories they could remember from the experiment. Thus, this is a measure of ability to remember the story, not the amount of information in the story that participants remember. There were no significant effects of Version on the free recall data.

Hypothesis 4

This hypothesis predicted that in addition to remembering more information from reproduced versions of stories, participants would remember more correct information and less incorrect information. As predicted there was a significant effect of Version on the amount of correct information recalled (F (1,41) = 6.83, p <.012, epsilon squared =.101). Participants recalled more correct information (mean = 1.24) from reproduced stories than they did from original stories (mean =.78). There was no significant effect of Version on the amount of incorrect information recalled.

Conclusion

The results of this experiment suggest that applying the seven rules of good storytelling to naturally occurring local news stories can significantly increase viewers' ability to understand and remember information from news stories without reducing levels of attention or excitement. It is important to note that the changes made in the stories are almost all post production or writing changes. This means, that these changes can be made without investing more resources in the reporting of the stories. As a result, local news stories can be made more user-friendly, more engaging, more interesting, and more comprehensible simply by applying these seven rules to constructing the final product.

1. Annie Lang is professor of Telecommunication at Indiana University and director of the Institute for Communication Research.

2. The stories were all day-of-air reporter packages, and covered the following topics: a grand jury hearing into possible influence peddling by members of a county commission; a public commission meeting to change its ethics code because of a lawsuit alleging sweetheard deals with former members; a preview of an economic summit conference; and developments in a court case involving airport landing rights.