Background

People's performance on a variety of tasks is influenced by circadian rhythms and people tend to have optimal times of day during which their performance peaks (optimality of testing effect).

Young adults and older adults tend to differ in terms of what time of day is optimal, with older adults doing better in the mornings and younger adults doing better in the afternoons and evenings.

Although this is generalizing, the pattern of results tends to be that older people do as well as younger people in the mornings, but considerably less well later in the day.

Theoretical Accounts:

• Inhibitory Deficiency Model: With aging people have difficulty inhibiting tempting but incorrect responses.
• Constructive Memory Model: With aging, people rely more on schemas and gist.
• Differential Memory Search Cue (Craik): With aging people are not as adept at generating useful search cues.

• Both young adults (18-25) and older adults (60-90) completed a survey designed to determine what their optimal time of day is (MEQ--Morningness-Eveningness Questionnaire).
• All Ss were tested either at their optimal time of day or non-optimal time of day
• Included a recall/math variable.  After each study list was presented subjects either tried to recall that list or worked on math problems for 2.6 minutes.
• Ss studied 8 DRM type lists and 4 lists were used as distractors.  As is typical in these kinds of experiments there were 4 types of items on the recognition test:
•  Targets: Words that really were on the lists
• Critical Lures: Words that were not on the lists but that are related to words that were on the lists
• Distractors: Words from lists that were not presented to that subject (but that were presented to other subjects)
• Distractor Lures: The critical lures of the distractor lists
• Subjects also made remember/know judgments

Recall.

• Older subjects recalled fewer targets (39%) than younger subjects (56%) but there was no effect of the time of day at which they were tested.
• Age didn't influence recall of critical lures, but time of testing did.  People who were tested at their optimal time of day falsely recalled fewer critical lures (48%) than did subjects tested at their non-optimal time of day (60%) and people recalled more critical lures than they did items from the middle of the lists.
• The older group generatated more intrusions other than the critical lure (11% of all words produced) than did the younger group (4% of all words produced) -- Is it misleading to look at it as a proportion of all words produced?  Older group is producing fewer words overall, so of course their intrusions will be higher when taken as a percentage of all words produced even if they produce the exact same number of intrusions!

Recognition
 

• They did not discuss in detail the Study + Recall condition and focused their analysis on the Study + Math condition.
• No difference in target recognition between older (0.66) and younger group (0.67) but the older group recognized more critical lures (0.78) than the younger group (0.71)
• The time of testing didn't have much of an effect on the younger group but for the older group nonoptimal testing decreased recognition of targets and increased recognition of critical lures.
• But this effect doesn't seem to be limited to critical lures.  Importantly distractor lures also increased substantially in the nonoptimally tested older adults.

Remember/Know Judgments

• Remember judgmetns were as common for critical lures as they were for targets.
• Older individuals were somewhat more likely to claim that they "remembered" critical lures than targets whereas the opposite was true for younger individuals
• Its noteworthy that the "remember" judgments for unrelated distractors was pretty high in this experiment (14% for older adults tested during nonoptimal times)

• Younger group recalled more items than older group but recognized the same number which the authors argue is consistent most consistent with Craik's retrieval cue account.
• However, they argue that the greater false alarm rate in recognition is consistent with the inhibition account (and they don't consider the possibility that because the nonoptimally tested older adults are stumped, that they just set a lower response criterion!)

• This is just like Experiment 1 except that optimal time of day was not manipulated as a variable.  Rather people were tested at around their preferred time of day.

• As before, when it came to recall the younger group recalled more targets and the same number of critical lures as the older group.  The older group had more non-critical intrustions as a percentage of their total recall.
• Again, in recognition they focused on the Math condition.  The older group recognized fewer targets than did the younger group but there was no difference in recognition of critical lures.
• Notice in Table 3 that the Older group recognized a very large proportion of the distractor lures (31%) and a huge proportion of these were remember judgments!  If one were to use the kind of corrected recognition scores that Schacter and others have used its arguable that the older group had fewer false memories than the younger group.

 Experiment 3

Also used imagery instructions to see if this might increase false memories.

Method

• Used pictures of objects based on category norms.  Critical lure in this design is the most protypical category member.  Note that this is a slightly different approach for coming up with picture false memories than that adopted by Israel & Schacter (1997) and by Schacter, Israel & Racine (1999).
• Final design was a Age X Optimal Time of Day X Imagery Instructions design
• Note that the imagery instructions emphasized relational processing rather than individual item processing.  (Any effect of imagery will not JUST be an effect of imagery.)
• They were tested using a picture recognition test (Type / Token problem?)
• Made remember / know judgments AFTER making all of their OLD/NEW judgments (Is there a problem with doing things this way?  Also why did they do this only in Experiment 3?)
• After the recognition test there was a free recall test.

Recognition Results

• Greater recognition of targets than critical lures.  Greater recognition of critical lures than distractors.

• False memory effect seems smaller than in previous two experiments including a lower rate of false remember judgments (.15) than in the previous two experiments, but note that this experiment isn't directly comparable to those.
• Targets were recognized equally often by the older and younger groups.
• People were more likely to recognize targets at their optimal time of day but this difference didn't quite reach conventional levels of significance (p = .052).
• Imagery instructions hurt target recognition.

• Critical lures were more likely to be false recognized by the older group than the younger group, but this difference was not significant at conventional levels (p = .074).  (Note that in this case if the results were to hold up it would not be due to criterion shifts)
• Imagery (relational processing?) condition led to greater recognition of critical lures.

 
Recall
• Primacy effect but no recency effect (those of you who've had Jim's Cognition class, why would that be?)
• Target recall was better for the younger group (p = .07), for the optimally tested group, and for the no imagery group.
• Older group recalled more critical lures than did the younger group
• Time of testing didn't effect false recall in the younger group but had a substantial effect in the younger group.

• Target recall was greater for the younger group than the older group, but target recognition was similar for the two groups
• Older group was more likely to recognize critical lures (and other distractors!) than was younger group.
• Optimal time of day mattered for older group but not for the younger group.  For older group false recognition was greater during non-optimal times of testing.
• Imagery instructions decreased accurate recall
• Greater false memories for words than for pictures (Note that there really is no comparison of pictures and words, the authors are relying on a comparison across their three experiments but their third picture memory experiment uses materials and procedures that are quite different from their two word false memory experiments)
• Results differed somewhat from Norman and Schacter (1997).  They found that young adults recalled fewer CLs than older adults.  This may be a function of the greater number of items Norman and Schacter included.

 

• Differential Memory Search Cue Model correctly predicts differences in target recall without differences in target recognition.
• Both the Inhibitory Deficiency Model and Constructive Memory Frameworks predict false memory findings.

 
 

University of Arkansas

Department of Psychology

Lampinen Lab

False Memory Reading Group

False Memory Reading Group Fall 1999