False Memories for Converging Associate Tasks
Deese's paradigm
Thematic lists of words are presented for people to study. The critical lure is the highest semantic associate of all words presented on a thematic list. However, the critical lure is missing from the study list presented at acquisition. When participants are given a free recall test, they falsely recall the critical lure at similar rates as they correctly recall words that were presented on the study lists.
Implicit Activation and Response Hypothesis
Underwood suggested that words encoded at acquisition stimulated the activation of their semantic associates. In a converging associate task (e.g. Deese's paradigm), study words are all related to a common non-presented semantic associate and all study words are likely to implicitly activate that common associate. Therefore, words presented on the test would only be studied once, but the non-presented semantic associate of all of the words would be repeatedly activated with the presentation of each list item.
Semantic associate pops into a person's head without the word being consciously thought of but instead unconsciously processed
Consciously
Specific words pop into a persons head and they are aware that these words are popping into their head (but the words are automatically popping into the person's thought; the generation of the words are not under conscious control)
Overarching assumptions of this research:
Younger and older children should perform at comparable levels on a converging associate task if the process underlying false recognition is an automatic process, because older children and adults have better developed consciously controlled memory processes (e.g. elaborative rehearsal) than younger children.
Experiment 2A and 2B
Exposure rates were manipulated to test the unconscious quality of the activation of semantic associates at acquisition. Since the activation of semantic associates is thought to be an unconscious process decreasing the exposure rate should not influence the number of false memories generated for critical lures but should interfere with the number of correctly remembered presented words.
Participants
120 participants (40 first and 40 fifth graders, and 40 undergraduate students)
Materials
12 category lists of objects were constructed and the objects were represented as black and white line drawings. These lists were composed of 11 pictures but the first (high frequency critical lure) and last pictures (low frequency critical lure) were removed from the list. The 12 lists were divided into two groups of 6 lists (groups A and B) that were presented to the participants.
Procedure
Participants were shown the drawing for six of the 12 lists. The presentation of the lists was blocked so that all items in a list were shown before any of the items from subsequent lists were shown. After all lists were shown, half of the participants were given a recognition test for the drawings. The remaining half of the participants returned after a three-day delay to take the recognition test.
Signal Detection Analysis
A' is a measure of a person's ability to discriminate between old and new recognition items. The better a person is at making this discrimination the more sensitive they are to the differences between test items. A' scores range from 0.00 to 1.00, with higher scores indicating greater sensitivity than lower scores and .50 indicating chance performance. Chance performance indicates that sensitivity is at a point where a person is just as likely to find a new item to be old as to find an old item to be old.
BD" is a measure of response bias, a measure of how willing a person is to accept an item as old on a recognition test. Measurements of BD" are given on a scale of -1.00 to 1.00, with a score of -1.00 being extremely liberal and a score of 1.00 being extremely conservative. A liberal response bias simply means that a person is willing to call almost anything on a recognition test old. A conservative response bias means that a person must feel very certain that recognition items are old before calling them old.
Results
Immediate Recognition
Participants were very good at recognizing which pictures they were shown (.94) and accepted very few of the related critical lures (.09) on the immediate recognition test. Also, A' (.97) was very high, expressing that all participants no matter of their age were very effective at discriminating between presented items and items that were not presented but were very similar to those that were.
Delayed recognition
Participants were not as good at recognizing which pictures had been presented following a three day delay, and they accepted more of the related critical lures as well. Younger children recognized significantly fewer of the presented words than did the fifth graders and undergraduates. First and fifth graders falsely recognized significantly more of the related critical lures than the undergraduates.
After the delay, sensitivity as measured by A' decreased but response bias as measured by BD" did not increase.
High vs. Low frequency critical lures
On both the delayed and immediate recognition test, high frequency related critical lures were recognized as old significantly more often than low frequency related critical lures.
Participants
80 undergraduate students.
Materials
12 lists of 14 words were constructed. Like in experiment 1, the first and last of the list words were removed and reserved as high and low frequency related critical lures.
Procedure
The lists were presented on a computer monitor, and were separated from subsequent lists by the words "next list" presented on the monitor in red letters. Half of the subjects viewed the list words at an exposure rate of 2 seconds per word. The remaining half of the participants viewed the words at an exposure rate of 20 ms per word.
After all of the lists were displayed participants took a recognition test. The 60 item recognition test asked participants to determine if a word was old or new and then make a remember/know judgment.
Results
Exposure Rate of 2 seconds
Participants who were exposed to words at an exposure rate of 2 seconds per words correctly recognized far more study item as old (.89) than they did falsely recognize related critical lure items as old (.25). These participants were able to discriminate between studied and non-studied items very well (A' = .94), and their response bias was fairly conservative (BD" = .28).
Exposure Rate of 20 ms
Participants that were only allowed 20 ms to view each word falsely recognized related critical lures as often as they correctly recognized studied words. Also, it is important to notice that the response criterion shifted when the exposure rate was shorter. Previous results had not illustrated a criterion shift as measured by response bias but decreasing the amount of exposure time increased participants willingness to call a recognition item old.
Remember/Know judgements
More remember judgments were made for studied items than related critical lures.
High and Low Frequency words
People accepted more of the high frequency related critical lures as old on the recognition test for both 2 second and 2 ms exposure rates. More remember judgments were made for high frequency related critical lures in comparison to low frequency related critical lures.
The manipulation of age and rates of exposure suggest that an automatic process is responsible for falsely recognizing related critical lures.
The automatic nature of this process is further illustrated by the fact that more remember judgments are made for study items than for related critical lure items. This suggests that people are not capable of remembering the generation of these items but are capable of remembering the presentation of the study items.
|
|
|
|
|
|