Oral Defense Reminder Sheet

Oral Defense Reminder Sheet

Cognitive processing of news as a function of story structure: A comparison between inverted pyramid and chronology.

Miglena Sternadori
Aug. 15, 2008

Purpose

This study explored the cognitive processing and subjective evaluation of written news when presented either as an inverted pyramid or a chronological presentation, and how those structures affect men and women differently in terms of cognition and evaluation.

Research Questions

  • How does news story structure affect cognitive resource allocation, memory and comprehension?
  • Do people enjoy reading one story structure more than another?
  • How do sex differences interact with story structure to affect these dependent concepts?

Theoretical Framework

  • Limited capacity model of motivated mediated message processing (LC4MP; 2000, 2006).
  • “Construction-integration” model of text processing (Kintsch, 1988).

Method

  • Design: 2 (Structure) x 2 (Story) within-subjects.
  • Participants: 51 undergraduate students and 7 working adults.
  • IVs: Story structure, operationalized in two levels: chronological and inverted pyramid model.
  • Stimuli: Four different news stories, manipulated so that there is an inverted pyramid and chronological version of each.

Results

H1: Chronologically structured stories will lead to slower STRTs than will inverted pyramid stories.

  • Supported: F(1,44)=3.92, p=.05, η2part=.08
  • Chronological stories (M=412.7, SD=122.89) elicited larger (slower) STRTs than the inverted pyramid stories (M=396.4, SD=121.13).

H2a: Recognition memory of content in chronologically structured news stories will be more accurate than recognition memory of content from inverted pyramid stories.

  • Not supported.

H2b: Cued recall of items in chronologically structured news stories will be greater than cued recall of items in inverted pyramid news stories.

  • Not supported.

H3: Chronologically structured stories will lead to better Sentence Verification Technique scores than will inverted pyramid stories.

  • Not supported
  • Effect was opposite the hypothesized direction such that inverted pyramid stories elicited higher SVT scores (M=.66, SD =.13) than chronological stories did (M=.64, SD=.11).
  • The main effect of Story Structure on sensitivity approached significance, (F(1,49)=2.91, p=.09, η2part=.06), such that sensitivity was higher for inverted pyramid stories.

H4: Chronologically structured stories will elicit more self-reported reading enjoyment than will inverted pyramid stories.

  • Not supported.

H5a: For all stories, women on average will show slower STRTs than men.

  • The difference approached significance (t51=-1.4, one-tailed p = .08, mean difference= -43.39), with the mean STRT for men (M=384.3, SD=109.8) across stories lower (faster) than for women (M=427.7, SD=114.29).

H5b: For all stories, women on average will show better recognition memory than men.

  • Not supported.

H5c: For all stories, women on average will show better-cued recall than men.

  • Not supported.

H5d: For all stories, women on average will score better on the Sentence Verification Technique task than will men.

  • The results approached significance (t55=-1.24, p=.11, mean difference=-.03), with women’s score (M=.67, SD=.07) higher than men’s score (M=.64, SD=.08).
  • For the sensitivity measure A’, the results also approached significance (t55=-1.43, one-tailed p=.08, mean difference=-.04), with women demonstrating higher sensitivity (M=.74, SD=.09) than men (M=.69, SD=.13).
  • For the criterion bias B’’, there was a significant difference between the sexes (t55=1.89, p=.03, mean difference=.16), such that women showed a lower (more liberal) criterion bias (M=-.18, SD=.34) than men (M=-.02, SD=.29).

H6a: There will be an interaction between sex and story structure, such that the mean STRT for women during reading of chronological stories will be greater (slower) than the mean STRT for women during reading of inverted pyramid stories and the mean STRTs for all men.

  • The Story Structure x Sex interaction was significant, (F(1,45)= 5.23, p=.03, η2part=.10) but opposite of the hypothesized direction.
  • Women’s STRT means for the two structures (M=427.67, SD=161.13 for inverted pyramid stories, and M=425.20, SD=163.47 for chronological stories) were both significantly greater (slower) that men’s STRT mean for inverted pyramid stories (M=365.14, SD=180.92) but not significantly different from the men’s mean for chronological stories (M=400.29, SD=183.53).

H6b: There will be an interaction between sex and story structure, such that women’s mean recognition for chronological stories will be greater than women’s mean recognition for inverted pyramid stories and all of the recognition means for men.

  • Not supported.

H6c: There will be an interaction between sex and story structure, such that women’s mean cued recall for chronological stories will be greater than women’s mean cued recall for inverted pyramid stories and all of the cued recall means for men.

  • Not supported.

H6d: There will be an interaction between sex and story structure, such that women’s mean Sentence Verification Technique (SVT) score for chronological stories will be greater than women’s mean SVT score for inverted pyramid stories and all of the SVT means for men.

  • Not supported

Theoretical Implications

From the text comprehension literature, Britton has argued that slower STRTs indicate heavier use of the cognitive capacity “by the meaning that is produced in the reader’s cognitive system” (Britton et al., 1983, p. 41). Simple messages produce slower STRTs, possibly by allowing more room for construction of meaning. By contrast, the media effects literature contains what Lang et al. (2006) call “standard STRT theory” (p. 370).

This study replicated Britton’s findings (including the lack of significant differences in recall between the structures), and so one simple way to explain the findings is in light of Britton’s comprehensibility hypothesis, which suggests that difficult texts elicit faster reaction times due to the “frequent breakdowns” that leave available capacity for the secondary task (Britton et al., 1978, p. 589).

The analysis was based on research that has used audiovisual messages. But since reading and viewing are fundamentally different, they likely involve different mechanisms of cognitive processing. Written messages take more cognitive effort, are not temporally demanding and allow readers to control their own pace. Future research should explore the possible differences in the cognitive processing of broadcast and written messages.

In the realm of sex differences in reading, the finding that women’s reading STRTs are slower than men’s can mean different things. Women’s STRTs might be slower because they are more engaged than men in constructing meaning from text. Or, women’s STRTs might be slower because reading is relatively effortless to them, so they allocate few resources to the process, and thus have few left to react to the secondary task.

The finding that women had lower (more liberal) criterion bias (B’’) than men, indicating that women were more likely to try maximize hits at the risk of some false alarms (Shapiro, 1994) is somewhat surprising, given the literature suggesting that women are more averse to risk than men (Byrnes, Miller, & Schafer, 1999).

Practical Implications

The findings of this study inform the debate about how to make print and online news more attractive and readable. Somewhat surprisingly, they imply that the inverted pyramid structure is perhaps not as “perverted” (Fry, 1999, p. 24). Although much evidence has accumulated against the inverted pyramid, it is worth considering that it may turn out to be a superior form of communicating written messages, after all.

Another practical implication is that the frequent use of the inverted pyramid structure in print and online news does not seem to present a problem for female readers. This suggests that other explanations need to be explored for the so-called “gender gap” in news readership.