An overview of Sacchi et al's cognitive study into doctored photos and their effects on memory. Designed for Edexcel International A level students as a revision resource.

1. CHANGING HISTORY:
DOCTORED
PHOTOGRAPHS AFFECT
MEMORY FOR PAST
PUBLIC EVENTS
Sacchi et al (2007)

2. Previous Research on False
Memory Construction

3. People who witness car accidents and later receive misleading information about what
they saw, will recall details incorrectly (Loftus, 1977)

4. People will remember false events from childhood- Participants in this study
were presented with a narrative describing a false event from their childhood.
25% of participants ‘remember’ being lost in a shopping mall even though this

5. Participants believed they had experienced unusual or traumatic events such as
nearly drowning and being saved by a lifeguard (Heaps and Nash, 2001)

6. Particpants were led to believe they had witnessed a demonic possession
(Mazzoni, Loftus and Kirsch, 2001)

7. Aims of the Study (Sacchi et al,
2007
 To investigate whether doctored photographs
of two well know events could change a
person’s memories of an event.
 To find out if viewing doctored images would
change the attitudes a person has towards a
past event.
 To investigate if viewing doctored images of a
past event could change behavioural
intentions in the future.

8. What do we mean by a doctored
photo?

9. To demonstrate power

10. To promote ideology

11. To fit the narrative

12. Sample
 Sample: 187 participants
 (31 male and 156 female) who were
undergraduates (92% Psychology, 8% other)
enrolled at the University of Padua or at the
University of Udine, in Italy.
 The age range was 19–39 (mean age 22.3
years).
 Participants did not receive any compensation
for involvement.

13. Event 1: Beijing
 To represent the Beijing event, a well-known
image of a student standing in front of tanks in
Tiananmen Square was used.
 For the Beijing event, a conspicuous crowd
was added on both sides of the line of tanks.

15. Event 2: Rome
 For the Rome event, a photograph depicting
peaceful demonstrators marching in front of the
Coliseum was used.
 In the photograph for the Rome event, police
officers and aggressive-looking demonstrators
were placed among the peaceful crowds
 To ensure the doctored Rome photo conveyed
violence, two versions were presented to eight
independent judges, who rated each image on a
peaceful–violent scale. The version rated more
violent was selected for the experiment.

17. Experimental Design
 Participants viewed one combination of the
photographs for the Beijing event and the
Rome event, either the original or doctored
version.
 Two original photos (N=48)
 Two doctored photos (N=44)
 The doctored Beijing photo and original Rome
photo (N=43)
 The original Beijing photo and doctored Rome
photo (N=52)

18. Experimental Design
 Three sets of multiple-choice questions were
used: manipulation check questions, critical
questions and attitude questions.
 Photographs and questions were presented in
a printed questionnaire that participants
completed in large groups in classroom
settings.
 No information about the experiment was
given.

19. Experimental Design
 Critical questions addressed aspects of
participant memories that could be biased by
the content of the doctored photographs.
 Attitude questions tested whether the doctored
material could affect attitudes towards the
events, for example rating violence.
 Finally, a blank page was left for participants to
add their comments or to point out aspects of
the event that they had found particularly
striking

20. Manipulation Check Questions
 Manipulation check questions were meant to
ascertain whether our digitally doctored material
was believable and to assess the participants’
familiarity with the events. First, participants
indicated whether they had already seen the
photograph. In this case, three options were
available: ‘Yes’, ‘No’, ‘I’m not sure’. If the false
images were plausible, then we expected the
participants who viewed the doctored version of a
photo to answer similarly to those who viewed its
original version. A second question asked
participants to rate how familiar they were with the
event on a 7-point scale in which 1 =
completely unfamiliar and 7 = completely
familiar.

21. Critical Questions
 Critical questions addressed specific aspects
of participants’ memories for each event that
we anticipated would be biased by the content
of the doctored photographs. For the Beijing
event, the focus of the critical questions was
the number of demonstrators; for the Rome
event, the focus was violent action.

22. Critical Questions

23. Attitude Questions
 Attitude questions were developed to test the
hypothesis that the misleading doctored
material could affect attitudes towards the
events. Participants rated the importance of
each event on a 7-point scale (1 ¼
insignificant, 7 ¼ important). They also rated
how violent (1 ¼ peaceful, 7 ¼ violent), and
how positive or negative (1 ¼ positive, 7 ¼
negative) the events were, according to their
memory.

24. Procedure
 On the first page participants saw both
photographs then answered the question ‘Can
you tell what major public event of the past 15
years is depicted in each of the following
photos?’ in a blank space next to each image.
 On the next page, one of the two photographs
appeared again, this time accompanied by a
caption indicating the event and when it took
place. On this page, participants also found the
manipulation check questions and two short filler
exercises.

25. Procedure
 Manipulation check questions assessed if
photographs were believable and participant’s
familiarity with the events. They indicated
whether they had already seen the photograph
and rated how familiar they were with the
event.

26. Procedure
 On the next page, participants responded
based on their memories of the event (being
asked not to look back at the photograph).
They were then presented with the critical
questions specific for that event and the
attitude questions

27. Procedure
 Critical questions addressed aspects of
participant memories that could be biased by
the content of the doctored photographs.
 Attitude questions tested whether the doctored
material could affect attitudes towards the
events, for example rating violence.
 Finally, a blank page was left for participants to
add their comments or to point out aspects of
the event that they had found particularly
striking

28. Procedure
 During the debriefing, participants saw both
the original and the doctored version of each
photograph, and the real purpose of the study
was revealed.

29. Results

30. Results

31. Results

32. Results

33. Evaluation
 Validity
How familiar are participants with these original
events? 35% of participants were ‘completely
unfamiliar’ with the 1989 Tiananmen Square
protest. Participants were instructed to answer
the critical and attitude questions according to
their memory, but how could they comply with
this instruction if they did not know or remember
the event at all? Perhaps they simply gave
fabricated answers.

34.  Non-Invasive Procedure- We do not know
whether the participants responded on the
basis of a modified memory or simply based
their answers on the photograph they were
shown
 Age of participants (M=22.3) is young. They
would not have witnessed these events
directly and would have fragmented memories
of reading/hearing about them if at all

35.  If the manipulation led participants to rate an
event more violent and more negative, would
they also be less likely to say they would
participate in a similar event in the future?
 This inspired a second study

36. Study 2

37. Study 2- Aims
 To test the hypothesis that the exposure to a
doctored photograph of a past public event
could affect people’s behavioural intentions a
second study was conducted.

38. Study 2- Sample
 A total of 112 participants (35 male, 73
females, 4 did not specify gender) enrolled at
University in Italy. The age range was 50–84
(mean age 64.9). About 56% of the
participants were retired, 20% were still
working and the remaining 24% did not
indicate their occupation. Participants did not
receive any kind of compensation for
involvement.

39. Study 2- Procedure
 The same photographs from Experiment 1
were used as stimulus material and
participants viewed only 1of 4 possible
combinations (as in the first experiment). The
questions were the same, however one
question was added for the Rome event to
rate how likely they would be to take part in
a similar demonstration.

40. Study 2- Results
 When asked if they would take part in a similar
demonstration those who saw the doctored
photograph gave significantly lower ratings
compared to participants in the original
condition.

44. Conclusions
 Viewing modified images affects the way people remember past public
events and also their attitudes and behavioural intentions.
 The effect was similar for younger and older adults, regardless of
whether the specific events were recent enough to be remembered
first hand.
 The authenticity of doctored images may have led participants to engage in
the
reconstructive process of remembering and to retrieve bits of information
that were
consistent with the misleading suggestion.
 According to these findings, anybody intending to deceive people and
affect their opinion by circulating such material would have a good
chance of being successful.
 If viewing false pictures during the retrieval stage
affects recollection of well-known events, what happens when we are
exposed to

45. General Evaluation
 Applications to Real life- ‘Anybody intending
to deceive people and affect their opinion by
circulating such material would have a good
chance of being successful’.
 Powerful and potentially significant effect of
doctored photographs through the media

46. General Evaluation
Ethics
 Participants had the right to withdraw and had
informed consent.
 They were deceived as to the true nature of
the study but were informed during their
debriefing
 Photographs were not necessarily distressing
in themselves, although consideration can be
given to context (e.g. are participants from
those countries or knew people affected by
those events for example)

47. General Evaluation
Reliability
 Study can be reasonably well replicated with
the source material
 Various replication studies have been done
since then, with similar findings
e.g. Slate Magazine

48. Full Article Link
 https://webfiles.uci.edu/eloftus/Sacchi_Agnoli_
Loftus_ACP07.pdf