Misinformation Effect

Memory and the Misinformation Effect
Contents

Why does the Misinformation Effect occur?
There is a general acceptance, supported by research, for the misinformation effect: The introduction of misleading postevent information will impair the memory of an original event (Toland, Hoffman & Loftus, 1991). However, although this phenomenon is well-established its interpretation is not. This debate about why the misinformation effect occurs relates to a fundamental issue about human memory - whether or not memory traces are permanently stored.

Two of the mechanisms which have been offered to explain the misinformation effect are: the alteration hypothesis and the coexistence hypothesis (Toland et al., 1991). Both hypotheses share the assumption
They claimed that the misinformation effect was not because of memory impairment but the result of a bias that existed in the standard testing procedure. Their argument was that this problem with the testing procedure creates a response bias and the possibility that subjects are being influenced by demand characteristics.

The repsonse bias is explained in terms of misinformation acceptance (subjects report the misinformation when the original information was never stored) which is not the result of a memory impairment.

To counteract these problems McCloskey & Zaragoza (1985) proposed a modified test in which a novel item was introduced. The modified test procedure followed the classical misinformation test procedure until the recognition test stage:
An item was observed (e.g. a hammer) and a postevent item was introduced (e.g. a screwdriver), but during the forced-choice recognition test a novel item (e.g. a wrench) was introduced. This novel item was neither presented originally nor presented during the postevent information phase. The novel item was, therefore, tested against the original event
If the misinformation impaired the subjects' ability to remember the original information, then misled subjects could not perform as well as the controls. However, if the misinformation has no effect on the original memory there would be no difference between control and misled conditions. The results of McCloskey and Zaragoza's experiments supported the latter hypothesis. They concluded that misinformation has no effect on memory.

After McCloskey and Zaragoza - Where to now?
This research of McCloskey & Zaragoza's (1985) has sparked much criticism especially because their results did not rule out the possibility that more subtle forms of impairment may be occurring. However, there has been a noticeable trend in current misinformation research towards designing experiments which avoided the pitfalls of the standard testing procedure they pointed out. These new studies also show an increased sensitivity towards possible memory impairment.

Tversky & Tuchin (1989) introduced still another modification to their experimental procedure. Using a "Yes/No" recognition test instead of a forced choice one they found evidence that misleading postevent information does interfere with original event memory. It was concluded that the differences between results from the standard versus modified testing procedures were because