In our daily lives, we continuously evaluate feedback information, update our knowledge, and adapt our behavior in order to reach desired goals. This ability to learn from feedback information, however, declines with age. Previous research has indicated that certain higher-level learning processes, such as feedback evaluation, integration of feedback information, and updating of knowledge, seem to be affected by age, and recent studies have shown how the adaption of choice behavior following feedback can differ with age. The neural (...) mechanisms underlying this age-related change in choice behavior during learning, however, remain unclear. The aim of this study is therefore to investigate the relation between learning-related neural processes and choice behavior during feedback learning in two age groups. Behavioral and fMRI data were collected, while a group of young and older adults performed a probabilistic learning task consisting of 10 blocks of 20 trials each. On each trial, the participants chose between a house and a face, after which they received visual feedback. In each block, either the house or the face image had a higher probability of yielding a reward. Participants were instructed to try to maximize their gains. Our results showed that less successful learning in older adults, as indicated by a lower learning rate, corresponded with a higher tendency to switch to the other stimulus option, and with a reduced adaptation of this switch choice behavior following positive feedback. At the neural level, activation following positive and negative feedback was found to be less distinctive in the older adults, due to a smaller feedback-evaluation response to positive feedback in this group. Furthermore, whereas young adults displayed increased levels of knowledge updating prior to adapting choice behavior, we did not find this effect in older adults. Together, our results suggest that diminished learning performance with age corresponds with diminished evaluation of positive feedback and reduced knowledge updating related to changes in choice behavior, indicating how such differences in feedback processing at the trial level in older adults might lead to reduced learning performance across trials. (shrink)

Tetris provides a difficult, dynamic task environment within which some people are novices and others, after years of work and practice, become extreme experts. Here we study two core skills; namely, choosing the goal or objective function that will maximize performance and a feature-based analysis of the current game board to determine where to place the currently falling zoid so as to maximize the goal. In Study 1, we build cross-entropy reinforcement learning models to determine whether different goals (...) result in different feature weights. Two of these optimization strategies quickly rise to performance plateaus, whereas two others continue toward higher but more jagged heights. In Study 2, we compare the zoid placement decisions made by our best CERL models with those made by 67 human players. Across 370,131 human game episodes, two CERL models picked the same zoid placements as our lowest scoring human for 43% of the placements and as our three best scoring experts for 65% of the placements. Our findings suggest that people focus on maximizing points, not number of lines cleared or number of levels reached. They also show that goal choice influences the choice of zoid placements for CERLs and suggest that the same is true of humans. Tetris has a repetitive task structure that makes Tetris more tractable and more like a traditional experimental psychology paradigm than many more complex games or tasks. Hence, although complex, Tetris is not overwhelmingly complex and presents a right-sized challenge to cognitive theories, especially those of integrated cognitive systems. (shrink)

From a theoretical perspective, most discussions of statistical learning have focused on the possible “statistical” properties that are the object of learning. Much less attention has been given to defining what “learning” is in the context of “statistical learning.” One major difficulty is that SL research has been monitoring participants’ performance in laboratory settings with a strikingly narrow set of tasks, where learning is typically assessed offline, through a set of two-alternative-forced-choice questions, which follow (...) a brief visual or auditory familiarization stream. Is that all there is to characterizing SL abilities? Here we adopt a novel perspective for investigating the processing of regularities in the visual modality. By tracking online performance in a self-paced SL paradigm, we focus on the trajectory of learning. In a set of three experiments we show that this paradigm provides a reliable and valid signature of SL performance, and it offers important insights for understanding how statistical regularities are perceived and assimilated in the visual modality. This demonstrates the promise of integrating different operational measures to our theory of SL. (shrink)

This paper examines whether adults can adapt to novel accents of their native language that contain unfamiliar context-dependent phonological alternations. In two experiments, French participants listen to short stories read in accented speech. Their knowledge of the accents is then tested in a forced-choice identification task. In Experiment 1, two groups of listeners are exposed to newly created French accents in which certain vowels harmonize or disharmonize, respectively, to the rounding of the preceding vowel. Despite the cross-linguistic predominance (...) of vowel harmony over disharmony, the two groups adapt equally well to both accents, suggesting that this typological difference is not reflected in perceptual learning. Experiment 2 further explores the mechanism underlying this type of phonological learning. Participants are exposed to an accent in which some vowels harmonize and others disharmonize, yielding an increased featural complexity. They adapt less well to this regularity, showing that adaptation to novel accents involves feature-based inferences. (shrink)

The computations involved in statistical learning have long been debated. Here, we build on work suggesting that a basic memory process, chunking, may account for the processing of statistical regularities into larger units. Drawing on methods from the memory literature, we developed a novel paradigm to test statistical learning by leveraging a robust phenomenon observed in serial recall tasks: that short‐term memory is fundamentally shaped by long‐term distributional learning. In the statistically induced chunking recall (SICR) task, (...) participants are exposed to an artificial language, using a standard statistical learning exposure phase. Afterward, they recall strings of syllables that either follow the statistics of the artificial language or comprise the same syllables presented in a random order. We hypothesized that if individuals had chunked the artificial language into word‐like units, then the statistically structured items would be more accurately recalled relative to the random controls. Our results demonstrate that SICR effectively captures learning in both the auditory and visual modalities, with participants displaying significantly improved recall of the statistically structured items, and even recall specific trigram chunks from the input. SICR also exhibits greater test–retest reliability in the auditory modality and sensitivity to individual differences in both modalities than the standard two‐alternative forced‐choice task. These results thereby provide key empirical support to the chunking account of statistical learning and contribute a valuable new tool to the literature. (shrink)

There are two main approaches to how statistical patterns are extracted from sequences: The transitional probability approach proposes that statistical learning occurs through the computation of probabilities between items in a sequence. The chunking approach, including models such as PARSER and TRACX, proposes that units are extracted as chunks. Importantly, the chunking approach suggests that the extraction of full units weakens the processing of subunits while the transitional probability approach suggests that both units and subunits should strengthen. Previous findings (...) using sequentially organized, auditory stimuli or spatially organized, visual stimuli support the chunking approach. However, one limitation of prior studies is that most assessed learning with the two-alternative forced-choice task. In contrast, this pre-registered experiment examined the two theoretical approaches in sequentially organized, visual stimuli using an online self-paced task—arguably providing a more sensitive index of learning as it occurs—and a secondary offline familiarity judgment task. During the self-paced task, abstract shapes were covertly organized into eight triplets (ABC) where one in every eight was altered (BCA) from the canonical structure in a way that disrupted the full unit while preserving a subunit (BC). Results from the offline familiarity judgment task revealed that the altered triplets were perceived as highly familiar, suggesting the learned representations were relatively flexible. More importantly, results from the online self-paced task demonstrated that processing for subunits, but not unit-initial stimuli, was impeded in the altered triplet. The pattern of results is in line with the chunking approach to statistical learning and, more specifically, the TRACX model. (shrink)

Decision making often takes place in social environments where other actors influence individuals' decisions. The present article examines how advice affects individual learning. Five social learning models combining advice and individual learning‐four based on reinforcement learning and one on Bayesian learning‐and one individual learning model are tested against each other. In two experiments, some participants received good or bad advice prior to a repeated multioption choice task. Receivers of advice adhered to the (...) advice, so that good advice improved performance. The social learning models described the observed learning processes better than the individual learning model. Of the models tested, the best social learning model assumes that outcomes from recommended options are more positively evaluated than outcomes from nonrecommended options. This model correctly predicted that receivers first adhere to advice, then explore other options, and finally return to the recommended option. The model also predicted accurately that good advice has a stronger impact on learning than bad advice. One‐time advice can have a long‐lasting influence on learning by changing the subjective evaluation of outcomes of recommended options. (shrink)

The aim of this study was to test the hypothesis that relevance modulates subsequent non-emotional behaviour, using a personalised mental-imagery-cued relevance manipulation paradigm. Participants had to build positive, negative and neutral mental images based on personalised scenarios that had been selected during an earlier picture-cued imagery phase. Participants imagined situations that were highly relevant for them and situations that were moderately relevant for them, depending on the effects the situations could exert on them. After each mental image, the effect of (...) the relevance manipulation was tested in a two-choice detection task. The interaction of relevance and valence was found to be predictive of the response times. Moreover, in the high-relevance condition, longer response times were observed for positive scenarios compared to negative ones, whereas in the low-relevance condition, shorter response times were observed for positive scenarios compared to negative ones. Results are consistent with the motivational theory of emotions, which posits that low-relevance stimuli trigger valence-specific attention modulations, whereas high-relevance stimuli trigger valence-specific action tendencies. (shrink)

A current issue in psycholinguistic research is whether the language difficulties exhibited by children with developmental language disorder [DLD, previously labeled specific language impairment ] are due to deficits in their abilities to pick up patterns in the sensory environment, an ability known as statistical learning, and the extent to which explicit learning mechanisms can be used to compensate for those deficits. Studies designed to test the compensatory role of explicit learning mechanisms in children with DLD are, (...) however, scarce, and the few conducted so far have led to inconsistent results. This work aimed to provide new insights into the role that explicit learning mechanisms might play on implicit learning deficits in children with DLD by resorting to a new approach. This approach involved not only the collection of event-related potentials, while preschool children with DLD [relative to typical language developmental controls] were exposed to a continuous auditory stream made of the repetition of three-syllable nonsense words but, importantly, the collection of ERPs when the same children performed analogous versions of the same auditory SL task first under incidental and afterward under intentional conditions. In each of these tasks, the level of predictability of the three-syllable nonsense words embedded in the speech streams was also manipulated to mimic natural languages closely. At the end of both tasks’ exposure phase, children performed a two-alternative forced-choice task from which behavioral evidence of SL was obtained. Results from the 2-AFC tasks failed to show reliable signs of SL in both groups of children. The ERPs data showed, however, significant modulations in the N100 and N400 components, taken as neural signatures of word segmentation in the brain, even though a detailed analysis of the neural responses revealed that only children from the TLD group seem to have taken advantage of the previous knowledge to enhance SL functioning. These results suggest that children with DLD showed deficits both in implicit and explicit learning mechanisms, casting doubts on the efficiency of the interventions relying on explicit instructions to help children with DLD to overcome their language difficulties. (shrink)

The main goal of this study was to identify factors motivating pragmatic transfer in advanced learners of English. Based on a cross-cultural comparison of requesting behavior between Koreans and Americans, this study determined the impact of individual subjective motives on pragmatic language choice. Two different groups of subjects participated in this study: 30 Korean participants and 30 American college students. Data were collected by using a Discourse Completion Task. Korean participants provided the data for Korean and English versions (...) of DCT. Semi-structured interviews were also conducted with 13 Korean ESL learners who showed the most and least amount of pragmatic transfer. Findings showed evidence of pragmatic transfer in the request responses given by Korean ESL learners in their requestive strategies, level of directness, and perspectives of head acts. The interview data revealed that Korean students were conscious of differing rules for making requests. Learners’ judgment of L2 pragmatic norms, perception of their own language, and their attitudes of the target language influence language use. Furthermore, findings showed that purpose of learning English, different types of motivation, and the length of intended residence contribute to the extent of pragmatic transfer. (shrink)

Stress induction reduces people's ability to modify their instrumental choices following changes in the value of outcomes, but the mechanisms underpinning this effect have not been specified because previous studies have lacked crucial control conditions. To address this, the current study had participants learn two instrumental responses for food and water, respectively, before water was devalued by specific satiety. Choice between these two responses was then measured in extinction, reacquisition and Pavlovian to instrumental transfer tests. Concurrently during these tests, (...) a negative emotional appraisal group evaluated aversive images, whereas a control group evaluated neutral images, at the same time as choosing between the two instrumental responses. Negative emotional appraisal abolished the impact of water devaluation on instrumental choice in the extinction test, but did not affect instrumental choice in the reacquisition or PIT tests. These findings suggest that negative emotional appraisal selectively impaired participants’ ability to retrieve the expected value of outcomes required to make goal-directed instrumental choices in the extinction test, and that this effect was not due to task disengagement, nullification of the devaluation treatment or impaired knowledge of response-outcome relationships. (shrink)

We demonstrated unconscious learning of task-irrelevant perceptual regularities in a Serial Reaction Time task in both visual and auditory domains. Participants were required to respond to different letters or syllables which occurred in random order. Unbeknownst to participants, the color of the two letters or the tone of the syllables varied according to certain rules. Reaction times indicated that people indeed learnt both the color and tonal regularities indicating that task-irrelevant sequence structure can be learned perceptually. (...) In a subsequent prediction test of knowledge of the color or tonal cues using subjective measures, we showed that people could acquire task irrelevant knowledge unconsciously. (shrink)

ObjectiveThe Spatial Learning Task of Lhermitte and Signoret is an object-location arbitrary associative learning task. The task was originally developed to evaluate adults with severe amnesia. It is currently used in populations where the memory system either is not yet fully developed or where it has been compromised (e.g. epilepsy, traumatic brain injury, electroconvulsive therapy, cerebrovascular disease and dementia). Normative data have been published for paediatric cohorts and for older adults, however no data exist for (...) the intervening adult years.MethodHere, we address this gap, collecting normative data from 101 adults aged 18–45.ResultsOur data indicate that performance on the Spatial Learning Task is not influenced by age, gender, level of education or overall IQ. Less than 10% of the variance in learning scores is associated with variability in verbal memory. Ninety percent of participants achieved perfect scores on two successive trials (T2Cr) within five or fewer trials on the Spatial Learning Task. A T2Cr score of 6 is suggestive of impairment and a T2Cr score of 7 or more is statistically abnormal.ConclusionThese data expand the clinical utility of the Spatial Learning Task in the adult population. Future work should examine performance in lower IQ cohorts, including intellectual disability, and explore sensitivity to disease factors such as laterality of mesial temporal lobe damage. (shrink)

The Iowa Gambling Task (IGT) and the Soochow Gambling Task (SGT) are two experience-based risky decision-making tasks for examining decision-making deficits in clinical populations. Several cognitive models, including the expectancy-valence learning model (EVL) and the prospect valence learning model (PVL), have been developed to disentangle the motivational, cognitive, and response processes underlying the explicit choices in these tasks. The purpose of the current study was to develop an improved model that can fit empirical data better than (...) the EVL and PVL models and, in addition, produce more consistent parameter estimates across the IGT and SGT. Twenty-six opiate users (mean age 34.23; SD 8.79) and 27 control participants (mean age 35; SD 10.44) completed both tasks. Eighteen cognitive models varying in evaluation, updating, and choice rules were fit to individual data and their performances were compared to that of a statistical baseline model to find a best fitting model. The results showed that the model combining the prospect utility function treating gains and losses separately, the decay-reinforcement updating rule, and the trial-independent choice rule performed the best in both tasks. Furthermore, the winning model produced more consistent individual parameter estimates across the two tasks than any of the other models. (shrink)

In prior research, signal detection theory has been widely utilized to assess metacognitive ability. However, the SDT metacognitive model requires the use of a two-alternative forced-choice task, while confidence must also be measured discretely. In our model, participants’ cognitive ability and their confidence in the cognitive task were used to estimate their metacognitive abilities. Therefore, in this study, a metacognitive model that can be applied to various cognitive tasks was developed. This model implements the item response theory (...) and Q-learning models to estimate cognitive ability; participants’ metacognitive ability is defined as the discrepancy between their confidence in their cognitive ability and their actual cognitive ability. The entire procedure was divided into two experiments. In experiment 1, two different cognitive tasks were used to estimate metacognitive ability and to examine overall discriminative and convergent validity. Notably, the parameters representing metacognitive ability did not correlate with cognitive ability but were positively correlated between the two tasks. In experiment 2, we performed a similar analysis using a different task to test the replicability of experiment 1. The results for experiment 2 were replicated for discriminative and convergent validity, albeit with weak results. Our metacognitive model exhibited high interpretability and versatility. (shrink)

This paper reports an experiment which investigates a possible cognitive antecedent of event-splitting effects (ESEs) experimentally observed by Starmer and Sugden (1993) and Humphrey (1995) – the learning of absolute frequency of event category impacting on the learning of probability of event category – and reveals some evidence that it is responsible for observed ESEs. It is also suggested and empirically substantiated that stripped-down prospect theory will accurately predict ESEs in some decision making tasks, but will not perform (...) well in others. This contention, it is argued, is indicative of fundamental descriptive shortcomings in the economic conception of choice under uncertainty and may entail implications beyond the direct concerns of this paper. (shrink)

We argue performance in the serial reaction time task is associated with gradations of awareness that provide examples of fringe consciousness [Mangan, B. . Taking phenomenology seriously: the “fringe” and its implications for cognitive research. Consciousness and Cognition, 2, 89–108, Mangan, B. . The conscious “fringe”: Bringing William James up to date. In B. J. Baars, W. P. Banks & J. B. Newman , Essential sources in the scientific study of consciousness . Cambridge, MA: The MIT Press.], and address (...) limitations of the traditional SRT procedure, including criticism of exclusion generation tasks. Two experiments are conducted with a modified SRT procedure where irrelevant stimulus attributes obscure the sequence rule. Our modified paradigm, which includes a novel exclusion task, makes it easier to demonstrate a previously controversial influence of response stimulus interval on awareness. It also allows identification of participants showing fringe consciousness rather than explicit sequence knowledge, as reflected by dissociations between different awareness measures. The NEO-PI-R variable Openness to Feelings influenced the diversity of subjective feelings reported during two awareness measures, but not the degree of learning and awareness as previously found with traditional SRT tasks [Norman, E., Price, M. C., & Duff, S. C. . Fringe consciousness in sequence learning: the influence of individual differences. Consciousness and Cognition, 15, 723–760.]. This suggests possible distinctions between two components of fringe consciousness. (shrink)

This article discusses how sequential sampling models can be integrated in a cognitive architecture. The new theory Retrieval by Accumulating Evidence in an Architecture (RACE/A) combines the level of detail typically provided by sequential sampling models with the level of task complexity typically provided by cognitive architectures. We will use RACE/A to model data from two variants of a picture–word interference task in a psychological refractory period design. These models will demonstrate how RACE/A enables interactions between sequential sampling (...) and long-term declarative learning, and between sequential sampling and task control. In a traditional sequential sampling model, the onset of the process within the task is unclear, as is the number of sampling processes. RACE/A provides a theoretical basis for estimating the onset of sequential sampling processes during task execution and allows for easy modeling of multiple sequential sampling processes within a task. (shrink)

Past research found performance differences between monolingual and bilingual children in the domain of executive functions. Furthermore, recent studies have reported advantages in processing efficiency or mental effort in bilingual adults and children. These studies mostly focused on the investigation of “cold” EF tasks. Studies including measures of “hot” EF, i.e., tasks operating in an emotionally significant setting, are limited and hence results are inconclusive. In the present study, we extend previous research by investigating performance in a task of (...) the “hot” EF domain by both behavioral data and mental effort via pupillary changes during task performance. Seventy-three monolingual and bilingual school children solved the Iowa Gambling Task in two different conditions. In the standard task, characterized by constant gains and occasional losses, children did not learn to improve their decision-making behavior. In a reversed task version, characterized by constant losses and occasional gains, both monolinguals and bilinguals learned to improve their decision-making behavior over the course of the task. In both versions of the task, children switched choices more often after losses than after gains. Bilinguals switched their choices less often than monolinguals in the reversed task, indicating a slightly more mature decision-making strategy. Mental effort did not differ between monolinguals and bilinguals. Conclusions of these findings for the bilingual advantage assumption will be discussed. (shrink)

To what extent can external incentives influence students’ effort and learning in online course contexts? While cognitive science research has found that monetary incentives can increase goal-directed cognitive effort in certain laboratory tasks, attempts to use monetary incentives to increase students’ academic performance in naturalistic settings has shown mixed results. In two experiments, we tested the influence of a monetary incentive on immediate and delayed tests of computer-based educational performance. In Experiment 1, participants were assigned to receive monetary incentives (...) for correct quiz responses, or receive no additional incentive for correct responses other than finding out their score, and we found no significant difference in total score across groups. In Experiment 2, we used a within-subjects design to test whether participants performed better when they were provided monetary incentives for correct responses on quiz questions. Here, participants performed significantly better on incentivized quiz questions. Thus, monetary incentives may increase performance in online learning tasks when participants can anchor the “stakes” of an incentive compared to no external incentive. These findings highlight potential benefits of external incentives for promoting effort and learning in online contexts, although further research is needed to determine the most useful educationally-relevant extrinsic incentives, as well as potential negative effects of incentives on long-term intrinsic motivation. (shrink)

It is a common belief that males have superior visuospatial abilities and that differences in this and other cognitive domains (e.g., math) contribute to the reduced interest and low representation of girls and women in STEM education and professions. However, previous studies show that gender-related implicit associations and explicit beliefs, as well as situational variables, might affect cognitive performance in those gender-stereotyped domains and produce between-gender spurious differences. Therefore, the present study aimed to provide information on when, how and who (...) might be affected by the situational reactivation of stereotypic gender-science beliefs/associations while performing a 3D mental rotation chronometric task (3DMRT). More specifically, we assessed the explicit beliefs and implicit associations (by the Implicit Association Test) held by female and male students of humanities and STEM majors and compared their performance in a 3DMRT after receiving stereotype- congruent, incongruent and nullifying experimental instructions. Our results show that implicit stereotypic gender-science associations correlate with 3DMRT performance in both females and males, but that inter-gender differences emerge only under stereotype-reactivating conditions. We also found that changes in self-confidence mediate these instructions’ effects and that academic specialization moderates them, hence promoting 3DMRT performance differences between male and female humanities, but not STEM, students. Taken together, these observations suggest that the common statement “males have superior mental rotation abilities” simplifies a much more complex reality and might promote stereotypes which, in turn, might induce artefactual performance differences between females and males in such tasks. (shrink)

The interest in crossmodal correspondences, including those involving sounds and involving tastes, has experienced rapid growth in recent years. However, the mechanisms underlying these correspondences are not well understood. In the present study (N = 302), we used an associative learning paradigm, based on previous literature using simple sounds with no consensual taste associations (i.e., square and triangle wave sounds at 200 Hz) and taste words (i.e., sweet and bitter), to test the influence of two potential mechanisms in establishing (...) sound–taste correspondences and investigate whether either learning mechanism could give rise to new and long‐lasting associations. Specifically, we examined an emotional mediation account (i.e., using sad and happy emoji facial expressions) and a transitive path (i.e., sound‐taste correspondence being mediated by color, using red and black colored squares). The results revealed that the associative learning paradigm mapping the triangle wave tone with a happy emoji facial expression induced a novel crossmodal correspondence between this sound and the word sweet. Importantly, we found that this novel association was still present two months after the experimental learning paradigm. None of the other mappings, emotional or transitive, gave rise to any significant associations between sound and taste. These findings provide evidence that new crossmodal correspondences between sounds and tastes can be created by leveraging the affective connection between both dimensions, helping elucidate the mechanisms underlying these associations. Moreover, these findings reveal that these associations can last for several weeks after the experimental session through which they were induced. (shrink)

Human participants and recurrent (“connectionist”) neural networks were both trained on a categorization system abstractly similar to natural language systems involving irregular (“strong”) classes and a default class. Both the humans and the networks exhibited staged learning and a generalization pattern reminiscent of the Elsewhere Condition (Kiparsky, 1973). Previous connectionist accounts of related phenomena have often been vague about the nature of the networks’ encoding systems. We analyzed our network using dynamical systems theory, revealing topological and geometric properties that (...) can be directly compared with the mechanisms of non-connectionist, rule-based accounts. The results reveal that the networks “contain” structures related to mechanisms posited by rule-based models, partly vindicating the insights of these models. On the other hand, they support the one mechanism (OM), as opposed to the more than one mechanism (MOM), view of symbolic abstraction by showing how the appearance of MOM behavior can arise emergently from one underlying set of principles. The key new contribution of this study is to show that dynamical systems theory can allow us to explicitly characterize the relationship between the two perspectives in implemented models. (shrink)

In this contribution we analyze how the actions of rational agents feed back on their beliefs. We present two agent-based computer simulations studying complex social interactions in which agents that follow utility maximizing strategies thereby deteriorate their own long-term quality of beliefs. We take these results as a starting point to discuss the complex relationship between rational action couched in terms of maximizing utility and the emergence of informational inequalities.

Environment learning is essential in everyday life. In individuals with Down syndrome (DS), this skill has begun to be examined using virtual exploration. Previous studies showed that individuals with DS can learn and remember paths in terms of sequences of turns and straight stretches, albeit with some difficulty, and this learning is supported by their cognitive abilities. This study further investigates environment learning in the DS population, newly examining their ability to learn a path from actual movements, (...) and to learn increasingly long paths, and how their performance relates to their visuo-spatial abilities and everyday spatial activities. A group of 30 individuals with DS and 30 typically-developing (TD) children matched for receptive vocabulary performed a 4 × 4 Floor Matrix task in a grid comprising 16 squares (total area 2.3 × 2.3 meters). The task involved repeating increasingly long sequences of steps by actually moving in the grid. The sequences were presented in two learning conditions, called Observation (when participants watched the experimenter’s moves), or Map (when they were shown a map reproducing the path). Several visuo-spatial measures were also administered. The results showed a clear difference between the two groups’ performance in the individual visuo-spatial measures. In the Floor Matrix task, after controlling for visuo-spatial reasoning ability, both groups benefited to the same degree from the Observation condition vis-à-vis the Map condition, and no group differences emerged. In the group with DS, visuo-spatial abilities were more predictive of performance in the Floor Matrix task in the Observation condition than in the Map condition. The same was true of the TD group, but this difference was much less clear-cut. The visuo-spatial working memory and visualization tasks were the strongest predictors of Floor Matrix task performance. Finally, the group with DS showed a significant relation between Floor Matrix task performance in the Observation condition and everyday spatial activity. These results enlarge on what we know about path learning in individuals with DS and its relation to their visuo-spatial abilities. These findings are discussed within the frame of spatial cognition and the atypical development domain. (shrink)