The 9/11 attacks inside and outside the US: Testing four models of flashbulb memory formation across groups and the specific effects of social identity

Abstract

Flashbulb memories (FBMs) are consistent recollections of specific details of the reception context of events. Four theoretical models accounting for FBM formation (Brown & Kulik, 1977; Conway et al., 1994; Er, 2003; Finkenauer et al., 1998) were tested on average 21 and 524 days after the 9/11 attacks in the US in 2001 in two groups of participants (US, N=112; non-US, N=727). Structural equation modelling revealed that (1) a revision of Finkenauer et al.'s (1998) model provided the best fit indices in both the US and non-US groups, (2) several relations among latent variables that were predicted by the three other models (Brown & Kulik, 1977; Conway et al., 1994; Er, 2003) were not significant, (3) with respect to Finkenauer et al's model, the “direct path” (from novelty/surprise to FBMs) was found to be significant only for the US group, while the “indirect path” (from emotional states, rehearsal and event-memory to FBMs) was only significant for the non-US group. It is suggested that the specific activation of social identity in the US group can explain these differences.

Keywords:

• Flashbulb memories
• Consistency of memory
• Emotion and cognition
• Social identity
• 9/11 attacks

Acknowledgements

The present study was supported by grants 1.5.078.06 and 1.5.158.08 from the Belgian National Fund for Scientific Research. The authors gratefully acknowledge the help of Elizabeth J. Marsh (now at Duke University, USA), Ineke Wessel (now at the University of Groningen, The Netherlands), Ticu Constantin (University of Iasi, Romania), Masao Yogo (Doshisha University, Kyoto, Japan), and Albina Shayevich (previously at University of Cambridge, UK) for their enormous help in collecting and encoding the data at test and retest, and Tiziana Lanciano (University of Bari) for the data cleaning.

The authors would also like to thank members of the FNRS-research group on collective memory (Olivier Klein, Laurent Licata, Valérie Rosoux, Laurence van Ypersele, and Susann Wolff) for their useful comments at various stage of the work progress.

The authors would also thank the following people for their contribution to various aspects of the study: Guglielmo Bellelli, Liesbeth van Breda, Giovanna D'Onchia, Dario Galati, Jenny van den Heuvel, Giannangela Mastrorilli, Prisca Niel, Annarita Pugliese, Mario Refice, Maud Slaats, Catherine Demaret, Jurgen Engerisser, Sylvie Grumiau, Véronique Pellegrin, Sophie Wautier, James Pennebaker, Allen Omoto, Martin Conway, Catrin Finkenauer, Mark Snyder, Carolin Showers, Martial van der Linden, Marie-Claire Gay, Emmanuelle Zech, Anna Tcherkassof, Kate Niederhoffer, Ambroise Nève, Paula Niedenthal, Gwenola Herbette, Debra Golden, Gerard Duveen, and Annette Lafontaine. Many thanks also to Mike Friedman for his careful check of the manuscript.

Notes

¹Structural equation modelling (SEM) adopts a confirmatory approach to test hypotheses of relationships among observed indicators of latent variables. It encompasses measurement and structural modelling. The former is accomplished primarily through confirmatory factor analysis, the latter through path analysis with latent variables (Bollen, 1989).

²One could argue that attitudes should also have been assessed within the US sample. However, there were several reasons for not including them. First, we made the assumption that we would obtain ceiling effects on these variables within this group. People from one country usually exhibit highly favourable attitudes towards themselves, especially at times of high threat for their nation, which was the case after the 9/11 attacks. Second, we think that even US residents who hold less positive attitudes were not likely to express them spontaneously through explicit measures immediately after the 9/11 attacks. Only implicit measures would have been appropriate, thus making their assessment through a web questionnaire prohibitively long and difficult.

³Mahmood et al. (2004) used a single measure of memory for the reception event. They thus adopted the vividness perspective of FBMs rather than the consistency one. One would therefore need to be careful about the implications for the present study based on the consistency of FBMs.