Effects of repetition on memory for pragmatic inferences

Abstract

Social interaction requires active inferential processing on the part of the listener. Such inferences can affect memory. For example, after hearing the karate champion hit the cinder block, one might erroneously recollect having heard the verb broke (Brewer, 1977)--a reasonable inference, but one not logically necessitated. The mechanisms behind this type of erroneous recollection have not been much explored. Experiments in the present article assessed the influence of repetition, response deadline, and age (cf. Jacoby, 1999), in an effort to demonstrate the dual contributions of familiarity and recollection underlying this phenomenon. For older adults, repetition at encoding increased the later likelihood of erroneously recognizing pragmatic inferences. For younger adults, repetition exerted the opposite effect. Both age groups, however, benefited from a second study-test trial. Experiment 2 demonstrated a similar interaction on a cued recall test for younger adults, whereby repetition exerted different influences as a function of time permitted during retrieval. Implications for theories of memory and discourse processing are considered.

The effects of frontal lobe functioning and age on veridical and false recall

“…Scoring procedures were similar to those used by McDermott and Chan (2006). Responses were classified into five categories: (1) Correct for verbatim responses, although synonyms, tense, and singular-plural form switching were acceptable, (2) false for pragmatic inferences, (3) other for responses that did not fit either the correct or false categories, (4) forget, and (5) new.…”

“…In fact, the equations provided by the Glisky lab (age-adjusted scores) classified 58 of the 60 young adults in the current experiment as low frontals. 2 To measure recall performance, we used the same pragmatic inference sentences as presented in McDermott and Chan (2006). The 48 sentences were separated into three sets of 16 sentences each for counterbalancing purposes.…”

Strength-based criterion shifts in recognition memory

“…In these investigations, participants have read texts either once or multiple times and then have performed recognition tests that intermixed sentences from the repeated texts (strong probes) and ones from the unrepeated texts (weak probes). The results have revealed the mirror effect of higher hits and lower false alarms for strong than for weak test sentences (Brainerd, Reyna, & Estrada, 2006;McDermott & Chan, 2006). 1 The false alarm differences were interpreted as reflecting criterion shifts.…”

The role of unconscious memory errors in judgments of confidence for sentence recognition

“…Thus, we hypothesized that an inferred sentence would produce a high rate of false alarms in recognition memory without participants being aware of the errors. The 24 schema-inference pairs used in this experiment were taken from Brewer (1977) and Birch and Brewer (1995), and most are given in the Appendix of McDermott and Chan (2006).…”