Schreiber, T.A. & Sergent, S.D. (1998). The role of commitment in producing misinformation effects in eyewitness memory. Psychonomic Bulletin & Review, 5, 443-448.

Purpose of the Experiments

• Examined the effects of misleading information on recognition memory
• Original Loftus test

1. See A
2. Hear B
3. Test A or B?

• McCloskey & Zaragoza modified test

1. See A
2. Hear B
3. Test A or C?

• Findings: Typically much more difficult to find misinformation effects with modified test than with original test.
• Possible reasons for this:

1. Non-impairment: Misleading information does not impair memory. People do worse on original test when misled because they are responding to social demand.
2. Blocking hypothesis: If misinformation is retrieved first it will block access to original information. On the modified test a there is nothing to cue memory for the misinformation

• Effects of interpolated tests.

1. If an interpolated test causes someone to retrieve misinformation, it should make that misinformation more accessible increasing the chances that it will be retrieved on the final test and, if the blocking hypothesis is correct, lead to impairment on the modified test.
2. Some folks have found this while others have not

• Possible reasons for the discrepant findings

1. Differences in procedure: People who have found evidence supporting blocking hypothesis have included the misinformation as an option on an interpolated recognition test in which the correct answer was not an option. People who have not found this (Belli) allowed people to respond on the interpolated test with the correct answer
2. Differences in materials: People who have not found this used M&Zs original materials.

• Purpose of the experiments reported here is to see how you can explain these discrepant findings.

EXPERIMENT 1 & 2

• Used M & Z's slides and compared a control condition to a misled + Commit condition.
• Procedure

1. View slides (e.g. screwdriver)
2. Read narrative (e.g. Control: Tool; Misled: Hammer)
3. Test 1 (e.g. Wrench or Hammer?)
4. Test 2 (e.g. Wrench or Screwdriver?)

• Results: Misled subjects were much less accurate than control subjects in both experiments

EXPERIMENT 3

1. Type of Test: Original, Modified
2. PEI: misled, misled + commitment, control
3. Number of slide presentations: one or two

Discussion

• Compatible with the blocking hypothesis