Gallo, David A.(2004). “Using recall to reduce false recognition: diagnostic and disqualifying monitoring.” Journal of Experimental Psychology: Learning, Memory, and Cognition, vol. 30, no.1. 120-128.
In 1983, Tulving proposed what he called the recall to reject process, in which a person will reject a lure as ‘old’ if it reminds him of the associated item that was actually studied. There are two ways that this type of recall could reduce false recognition: disqualifying and diagnostic recall to reject.
Disqualifying recall to reject happens when the subject remembers information that eliminates the possibility that the event in question occurred. This strategy is rule based and depends on whether the task allows a rule to be applied.
Diagnostic recall to reject occurs when a subject recalls information that suggests that the lure did not occur. This process is heuristic based, in that the subject compares the memories for the lure to the memories of the study items, and rejects the lure if its evidence is not as strong as that of the study items. Recalled information develops expectations about what should be remembered.
This study investigates whether these strategies could reduce false recall in a converging associates task (one in which studied items relate to a central nonstudied item). Because these related lures are often falsely recognized and because the recall to reject strategies depend on correct recall, it was unclear whether a recall to reject strategy could be employed.
A preliminary study presented words from categories and then took a recognition test. The manipulation was the presentation of the category labels (at test, at study, neither, or both). The greatest level of recall occurred when labels were presented at study and test; the least level of recall occurred when they were never presented. Presenting labels at test only did not affect false recognition of related lures. There were differences in recall, but not in false recognition, which indicates no support for the recall to reject process.
Experiment 1
Gallo wanted to ensure that subjects used the strategy just in case making familiarity judgements is easier and more natural. Some subjects received instructions on how to use the recall to reject strategy, and were asked to decide if they were able to recall all items from the category after each recognition decision. He also varied the length of the lists, making it more difficult to be sure that all words from the category had been recalled. This would make the disqualifying recall to reject process more difficult but not the diagnostic process. Labels were presented at study and test and category lists were blocked. Three exemplars were used for each category.
Subjects were told that category labels would appear. They saw words and categories on a computer, then completed a distractor task, then completed a recognition test. They were told there would be only one exemplar from each category on the test. Subjects in the strategic condition were told that each category contained three words, and in the varied condition they were told that the word lists ranged from two to four. Both groups were told to use a recall to reject strategy (if you can recall all words and the test word isn’t one of them, its old). They then reported whether they had recalled all words from that category. They completed a questionnaire asking if they used recall, then took a cued recall test.
Subjects in the strategic strategy were more likely to report using a recall to reject process, and their reaction times were slower. Recall was greater for categories that had positive all-recall judgements than for negative recall-all judgments. Subjects in the strategic constant and varied condition had similar judgments, but recall was greater for those in the constant condition. Cued recall of studied items did not differ across both constant length conditions, but in the varied condition, recall was greater for standard instructions than strategic. Recall was greater when an item had been correctly recognized or correctly rejected. However, when looking at data in which all words had not been recalled, this relationship disappeared and recall associated with false alarms did not differ from that associated with correct rejections. Also, true recognition was greater for categories that were exhaustively recalled and false recognition was greater when they weren’t exhaustively recalled.
Experiment 2
This experiment was identical to the first except five items were studied in each constant condition, and the varied condition studied four or six items. This made it more difficult to exhaustively recall all items, which is what seemed to reduce false recognition in Experiment 1.
Increasing the category length was successful at minimizing full recall of the categories. In the strategic condition, subjects reported trying to use the process, but it resulted in reducing both true and false recognition relative to the standard condition. Again, recall was greater when an item had been recognized, but there was no negative relationship between recall and false recognition. These results indicate that the results found in Experiment 1 were due to exhaustive recall.
General Discussion
There is no support for a diagnostic recall to reject strategy, since recalling some of the items did not reduce false alarms. There is evidence that a disqualifying recall to reject process might be used when the subject call recall all items.
These results indicate that neither of these processes is used in DRM paradigms, because subjects would not be able to recall all words that relate to the central item, and only remembering some does not reduce false recognition. Instead Gallo suggests that subjects might figure out what types of features to look for and use this general knowledge to make recognition decisions.
The success of the disqualifying recall to reject strategy suggests that true recall is separate from false recall. This could happen in two ways. First, based on the source monitoring framework, evaluating the evidence for each event could allow people to discriminate between false and true memories (a diagnostic process). A dual process frame work suggests that the two memories are qualitatively different, using familiarity based attribution. In this process the subject only needs a good intuitive sense to make the distinction.