TV or not TV? Does the immediacy of viewing images of a momentous news event affect the quality and stability of flashbulb memories?

Abstract

The flashbulb accounts of 38 participants concerning the September 11th 2001 terrorist attack reported at both 28 hours and 6 months following the event were examined for quantity, quality, and consistency as a function of the time lapse between first learning of the event and initial viewing of media images. The flashbulb accounts of those who reported seeing images at least an hour after learning of the event differed qualitatively, but not quantitatively, from accounts of participants who reported seeing images at the same time as or within minutes of learning of the event. Delayed viewing of images resulted in less elaborate and generally less consistent accounts across the 6-month interval. The results are discussed in terms of factors affecting flashbulb memory formation and individual differences in connectedness to the event.

Keywords:

• Flashbulb memory
• Consistency
• Media images
• 9/11 terrorist attack
• Image-viewing delay

Acknowledgements

A portion of the data set was used in an Honours thesis conducted by the third author. We thank our research assistants, Janice Edwards and Alexis Manson, for coding the data.

Notes

¹With only three means, Fisher's protected ts hold the family-wise error rate at the level of α for the overall F test (see Howell, 2002, p. 391 for a discussion).

²The degrees of freedom for this analysis reflect the loss of data for one participant who failed to respond to this question.

³We thank an anonymous reviewer for raising the possibility that differences in the initial images viewed may play a role in FBM.

⁴A reviewer has suggested that participants in the hours group might have had a trait preference for a non-visual processing mode. However, we found that the narratives of only two participants in the hours group suggested that they could possibly be conceived of as individuals who failed to seek out images when there appears to have been an opportunity to do so. Our participants’ estimates of amount of image viewing of the event over the subsequent 6-month interval could be taken to reflect participant preference for visual versus non-visual processing. When we repeated the analysis of participants’ estimated number of post-event image viewings, but excluded the data from these two participants, the estimates were still significantly reduced in the hours group compared to those for participants in the other two groups, whose mean ratings were identical, (F = 3.270, p = .051, ?² = .17). Thus we believe that a trait preference for non-visual processing was playing little, if any, role as a self-selection criterion in our study.

⁵We are grateful to anonymous reviewers for raising issues concerning environmental location influences.

⁶Given unequal group ns in the set of analyses, the Satterthwaite correction to the degrees of freedom was applied.

⁷As a result of differences in connectedness, the overall magnitude of estimates of visual media exposure found for our Canadian sample might possibly be less than that found with a United States sample and exceed that for a non-North American sample.