Journal of Experimental Psychology-Learning Memory and Cognition最新文献

In many real-life settings, feedback is only available for cases that decision makers accept and so may be biased toward positive events. How do people learn to distinguish good from bad alternatives from such selective feedback, and can they correct for this bias? We describe the computational problems of classification learning from biased samples and examine how exemplar and model-based methods can deal with this challenge: Model-based methods can adjust their representation of the task based on what information is available while exemplar models can impute fictive negative outcomes in missing cases to avoid positivistic biases. Importantly, these methods imply distinct assumptions about the task and reactions to missing feedback, which can be assessed empirically. In three experiments, we test whether participants rely on imputation or use a Bayesian model of the task to correct for selection bias. We find that many participants were best described by an exemplar model, most with imputation, but an almost equal proportion was best described by a Bayesian model. People best described by different models reacted somewhat differently to missing feedback. We also observe substantial stability in whether individuals were best described by model-based or exemplar models across tasks, though participants were more likely to use exemplar models when there was greater uncertainty about the task structure. Overall, our findings show that people deal with missing feedback in an adaptive manner by adopting diverse approaches that are partially stable and partially reflect assumptions made about the experimental context. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Fluent reading comprehension demands the rapid access and integration of word meanings. This can be challenging when lexically ambiguous words have less frequent meanings (e.g., the dog meaning of boxer). Indeed, readers fixate on lexically ambiguous words that are disambiguated toward their subordinate meaning for longer than matched control words embedded within identical sentence contexts. Word-meaning priming studies have shown that participants flexibly use recent experiences with ambiguous words to guide their interpretation when these words are presented in isolation, even after substantial delays. However, word-meaning priming paradigms have almost always used artificial tasks to measure word-meaning availability and we do not therefore know how priming would support lexical processing when reading for comprehension. Thus, we conducted two eye-movement experiments to examine word-meaning priming during sentence reading. Both experiments employed a 2 (ambiguity: low-ambiguity control vs. high-ambiguity) × 2 (priming: unprimed vs. primed) within-participants design, with either a 1-min delay (Experiment 1; N = 28) or a 30-min delay (Experiment 2; N = 60) between prime and test sentences. Both experiments showed greater reductions in go-past times and total reading times following priming for high-ambiguity target words than matched low-ambiguity control words, indicating that recent encounters support the processing of word meanings during sentence reading and that this effect extends beyond the simple repetition effect observed for low-ambiguity control words. This illustrates the remarkable flexibility of the human language system in using diverse input to refine stored lexical knowledge even in skilled readers. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

When performing successive recall tests without restudy, subjects' recalls exhibit intriguing variability across tests, including gaining or losing items across tests. To examine the cognitive mechanisms underlying this variability, research has focused primarily on hypermnesia, the finding that recall performance increases across tests (Erdelyi & Becker, 1974). Hypermnesia studies commonly consider conditions that impact recall levels of items gained across tests versus items maintained across tests. By contrast, analyses of recall clustering in hypermnesia studies typically collapse across maintained items and item gains. Here, I examine associative processes separately for item gains and maintained items. Experiment 1 examines these effects in final free recall, a paradigm also used to examine changes in recall across tests but less commonly linked with hypermnesia, whereas Experiment 2 uses a classic hypermnesia design. In both experiments, subjects exhibited significant temporal and semantic clustering for maintained items, but there was less evidence of these associations supporting item gains. In Experiment 1, transitions to maintained items boasted a greater proportion of same-list transitions than item gains, and in Experiment 2, there were no significant clustering effects to item gains on a test producing hypermnesia. Further, in Experiment 1, subjects exhibiting greater list-level temporal clustering of maintained items also maintained more items across tests. The results highlight the importance of episodic and semantic associations to changes in recall across tests and have implications for current theories of hypermnesia. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

It has been demonstrated that surprising information often leads to better recall. Yet, this might not apply to information that is considered to be implausible. The present study examines how surprise and plausibility judgments relate to participants' memory for numerical statements. Participants performed an estimation task in which they were presented with an incomplete numerical fact (e.g., X out of 10 bus drivers are women) for which they were asked to provide an estimation. After being presented with an answer, they indicated how surprised they were about the answer and whether they found the answer plausible. Next, participants performed a memory test to examine the effects of surprise and plausibility on recall of the presented answers. Finally, 24-48 hr later, participants provided new estimations for the numerical statements to examine whether participants had integrated the presented answer into their knowledge representation. A U-shaped relation between surprise and memory recall was found for recall on Day 1, with unsurprising and highly surprising items being remembered better than moderately surprising items. Importantly, the relationship between surprise and recall was only found for plausible items. Next, new estimations on Day 2 indicated that unsurprising and plausible items were incorporated into participants' knowledge representation more often than surprising and implausible items. Taken together, our findings support the notion that surprise enhances memory but also show that metacognitive judgments influence this effect. Moreover, our findings revealed that enhanced recall does not necessarily mean the information is fully incorporated into participants' knowledge representation. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

This study investigated (a) whether L2 semantic processing is modulated by automatic activation of L1 translations, (b) whether L1 translation activation involves both phonological and orthographic representations, and (c) whether these phonological and orthographic representations of L1 translations are accessed along a similar time course. To this end, 48 Hebrew-English bilinguals and 48 native English speakers with no Hebrew knowledge performed a semantic relatedness judgment task in English. Critical prime-target pairs (n = 96) were semantically unrelated, but their translations in Hebrew could include form overlap. Specifically, complete translation-overlap pairs shared both a phonological and an orthographic lexical form (e.g., "beak" and "source" = מקור /makor/), whereas partial translation-overlap pairs shared either a phonological form (e.g., "skin" and "light" = /or/) or an orthographic form (e.g., "book" and "barber" = ספר) in Hebrew. Stimulus onset asynchrony (SOA) of the prime-target L2-English words was further manipulated to reveal the time course of phonological and orthographic translation activation. Results showed that complete overlap in the translation lead Hebrew-English bilinguals, but not native English speakers, to judge semantically unrelated pairs as related in meaning and to do so more quickly irrespective of SOA. For partial translation overlap in phonology, the percentage of "yes" responses was affected only in the short SOA (300 ms), and under partial translation overlap in orthography, only in the long SOA (750 ms). These findings suggest that L1 translation activation during L2 word processing spreads to both phonological and orthographic representations but at different time points along processing. (PsycInfo Database Record (c) 2025 APA, all rights reserved).

Speech intonation conveys a wealth of linguistic and social information, such as the intention to ask a question versus make a statement. However, due to the considerable variability in our speaking voices, the mapping from meaning to intonation can be many-to-many and often ambiguous. Previous studies suggest that the comprehension system resolves this ambiguity, at least in part, by adapting to recent exposure. However, these studies have largely been limited to single-talker exposure, leaving open how listeners adapt to input from multiple talkers. Four experiments herein address this question. Listeners were exposed to a male and/or female talker producing statements ("It's raining.") and declarative questions ("It's raining?"). After exposure, listeners categorized the utterances of both talkers (Experiments 1-3) or a novel test talker (Experiment 4) as statements or questions. In all four experiments, intonation adaptation was found, and it was neither strictly talker-dependent nor strictly talker-independent. While listeners tracked production patterns unique to each talker, adaptation sometimes generalized across talkers. We relate these findings to research on segmental speech perception, which has found that talker-dependence is conditioned on phonetic contrast or cue distributions. (PsycInfo Database Record (c) 2025 APA, all rights reserved).

The present study uses event-related potentials (ERPs) to investigate lexicosemantic prediction in native speakers (L1) of English and advanced second language (L2) learners of English with Swedish as their L1. The main goal of the study was to examine whether learners recruit predictive mechanisms to the same extent as L1 speakers when a change in the linguistic environment renders prediction a useful strategy to pursue. The study, which uses a relatedness proportion paradigm adapted from Lau et al. (2013), focuses on the N400, an ERP component that is sensitive to the ease of lexical access/retrieval, including lexical prediction. Participants read 800 prime-target pairs, presented word by word and divided into two blocks, while they searched for animal words. Unknown to them, some of the pairs were semantically associated, which is known to reduce the amplitude of the N400 via spreading semantic activation. Most importantly, the proportion of semantically related pairs increased in the second experimental block (via fillers), thereby increasing the reliability of the primes as predictive cues and encouraging prediction. Results from 36 L1-English speakers and 53 L2 learners showed an N400 reduction for related (remain-stay ) relative to unrelated targets (silver-stay ) across blocks. Crucially, this N400 reduction for related targets was significantly larger in the block that encouraged prediction, in both L1 and L2 speakers, consistent with the possibility that both groups recruited similar predictive mechanisms when the context encouraged prediction. These results suggest that, at high levels of proficiency, L2 speakers engage similar predictive strategies to L1 speakers. (PsycInfo Database Record (c) 2025 APA, all rights reserved).

The goal of the present investigation was to perform a registered replication of Jones and Macken's (1995b) study, which showed that the segregation of a sequence of sounds to distinct locations reduced the disruptive effect on serial recall. Thereby, it postulated an intriguing connection between auditory stream segregation and the cognitive mechanisms underlying the irrelevant speech effect. Specifically, it was found that a sequence of changing utterances was less disruptive in stereophonic presentation, allowing each auditory object (letters) to be allocated to a unique location (right ear, left ear, center), compared to when the same sounds were played monophonically. Due to its importance for theoretical accounts of auditory distraction and because the results were somewhat equivocal, it is important to replicate this influential study with enhanced statistical power. The present replication (N = 60) confirmed that the disruptive effect of a changing-state sequence ("V-J-X") as compared to a steady-state sequence ("J-J-J")-the changing-state effect-is reduced significantly with stereophonic presentation, suggesting that listeners perceptually grouped the presented sound into three separate steady-state streams, which produce much less interference with seriation compared to the monophonic presentation. However, in contrast to the original study, stereophonic sequences tended to be slightly more disruptive than monophonic steady-state sequences, suggesting that the change in location may also cause some interference on its own. Moreover, there was also a significant steady-state effect, with both steady-state conditions being more disruptive than silence. The results are discussed with regard to interference-by-process and attentional accounts of auditory distraction. (PsycInfo Database Record (c) 2025 APA, all rights reserved).

Listeners can use both lexical context (i.e., lexical knowledge activated by the word itself) and lexical predictions based on the content of a preceding sentence to adjust their phonetic categories to speaker idiosyncrasies. However, this phonetic retuning is difficult for listeners to achieve using lexical context when adjusting to idiosyncrasies in word onsets. In this situation, sentence context could help by boosting lexical knowledge. In a series of experiments, we tested for the interplay between lexical context and sentence context. Using the sentence-guided retuning paradigm from Jesse (2021), either a preceding sentence context or following lexical context disambiguated the perceptually ambiguous onset of short words as /s/ or /f/. At test, listeners categorized steps from an /sɑ/-/fɑ/ continuum. Evidence for phonetic retuning, in terms of more responses at test in line with prior disambiguation during exposure, was found when sentence context had disambiguated the critical sound during exposure. In contrast, lexical knowledge activated by the word itself only produced trends across a subset of steps. When sentence and lexical context disambiguated the idiosyncrasy in the same words, the change of the overall retuning effect across steps followed the pattern observed in the experiment with only lexical disambiguation. Furthermore, this modulation of the retuning effect was not observed when sentence and lexical context disambiguated the idiosyncrasy in different items. This pattern of results suggests an interplay between these two types of contexts. Sentence context therefore helps with retuning to talker idiosyncrasies in word onsets when the lexical context can fail listeners. (PsycInfo Database Record (c) 2025 APA, all rights reserved).

Is the developing lexicon phonologically detailed or are representations underspecified? Experimental results from toddlers suggest phonological specificity. By contrast, the featurally underspecified lexicon theory (Lahiri, 2018; Lahiri & Reetz, 2010), motivated by evidence such as the cross-linguistic prevalence of phenomena such as coronal assimilation (rainbow → rai[m]bow), proposes that coronal sounds are unspecified for place of articulation even in the adult lexicon. The featurally underspecified lexicon, therefore, predicts that asymmetries in mispronunciation sensitivity are also present in the developing lexicon. Recent research (Ren et al., 2019) has rejected this, reporting similar sensitivity to mispronunciation of coronals and noncoronals at 19 months. Using a more sensitive experimental paradigm, we provide new evidence demonstrating a lack of asymmetries at 18 months, but mispronunciation sensitivity for coronals disappears by 24 months. In an intermodal preferential looking study, growth curve analysis shows that 18-month-olds are sensitive to mispronunciations of words with a coronal (e.g., duck vs. *buck) and noncoronal (e.g., bird vs. *dird) onset. At 24 months, mispronunciations of coronal-onset words were treated just like the accurate pronunciations. We conclude that coronals are underspecified in the developing lexicon at 24 months. We propose a model under which initial representations are phonetic in nature and require exact acoustic input, whereas phonological coronal underspecification at the lexical level emerges gradually as a result of exposure to variation in the input such as coronal assimilations that only become detectable patterns with growing lexical and segmentation skills. (PsycInfo Database Record (c) 2025 APA, all rights reserved).