People’s knowledge about the world comes from many sources, including fictional ones such as movies and novels. In three experiments, we investigated how people learn and integrate information from fictional sources with their general world knowledge. Subjects read a series of short stories that contained information about the real world. After a short delay, all participants took a general knowledge test. Subjects did indeed use information from the stories to answer general knowledge questions. Prior reading of facts boosted participants’ abilities to produce both obscure and better-known facts, and the effect held for both correct and incorrect facts (misinformation). Repeated reading of the stories increased the effect. After a delay of one week, effects of story exposure were strongest for items that also had been tested in the first session. Subjects were aware of use of story information, but interestingly, story exposure also increased belief that the facts had been known prior to the experimental session, even for misinformation answers that were rarely produced without story reading.
In three experiments, we examined part-set cuing effects in younger and older adults. Participants listened to lists of category exemplars, and later recalled them. Across conditions, recall was uncued, or cued with a subset of studied items. In Experiment 1, younger and older adults were cued with some of the category names, and they remembered fewer never-cued categories as compared to a free recall condition. In Experiment 2, a similar effect was observed when the cues were category exemplars (as opposed to category names). There was also an age difference: by some measures, a small number of cues impaired older adults more than younger adults. In Experiment 3, we replicated this result and also found that older adults were disproportionately slow in the presence of cues. In all experiments, older adults showed robust part- set cuing effects and in some cases, they were disproportionately impaired by the cues.
Readers rely on fiction as a source of information, even when fiction contradicts relatively well-known facts about the world (Marsh, Meade, & Roediger, 2003). Of interest was whether readers could monitor fiction for errors, in order to reduce suggestibility. In Experiment 1, explicit warnings about errors in fiction did not reduce students’ reliance on stories. In Experiment 2, all subjects were warned before reading stories written at 6th or 12th grade reading levels. Even though 6th grade stories freed resources for monitoring, suggestibility was not reduced. In Experiment 3, suggestibility was reduced but not eliminated when subjects pressed a key each time they detected an error during story reading. Subjects do not appear to spontaneously monitor fiction for its veracity, but can do so if reminded on a trial-by-trial basis.
The present review addresses whether multiple-choice tests may change knowledge even as they attempt to measure it. Overall, taking a multiple-choice test boosts performance on later tests, as compared to non-tested control conditions (the testing effect). This benefit is not limited to simple definitional questions, but holds for SAT II questions and for items designed to tap concepts at a higher level in Bloom’s (1956) taxonomy of educational objectives. Students, however, can also learn false facts from multiple-choice tests; testing leads to persistence of some multiple-choice lures on later general knowledge tests. Such persistence appears due to faulty reasoning rather than an increase in the familiarity of lures. Even though students may learn false facts from multiple-choice tests, the positive effects of testing outweigh this cost.
Of interest was whether prior testing of related words primes false memories in the Deese-Roediger-McDermott (DRM) paradigm. After studying lists of related words, subjects made old-new judgments about 0, 3, or 6 related items before being tested on critical non-presented lures. When the recognition test was self-paced, prior testing of list items led to faster false recognition judgments, but did not increase the rate of false alarms to lures from studied lists. Critically, this pattern changed when decision-making at test was speeded. When forced to respond quickly, presumably precluding the use of monitoring processes, clear test-induced priming effects were observed in the rate of false memories. The results are consistent with an activation-monitoring explanation of false memories and support that retrieving veridical memories can be a source of memory error.
In contrast to laboratory free recall (which emphasizes detailed and accurate remembering), conversational retellings depend upon the speaker’s goals, audience, and the social context more generally. Because memories are frequently retrieved in social contexts, retellings of events are often incomplete or distorted, with consequences for later memory. Selective rehearsal contributes to the memory effects, as does the schema activated during retelling. Retellings can be linked to memory errors observed in domains such as eyewitness testimony and flashbulb memories; in all of these situations, people retell events rather than engage in verbatim remembering.