Supporting College Choice Among International Students through Collaborative Filtering.

In this paper we use historic score-reporting records and test-taker metadata to inform data-driven recommendations that support international students in their choice of undergraduate institutions for study in the United States. We investigate the use of Structural Topic Modeling (STM) as a context-aware, probabilistic recommendation method that uses test-takers' selections and metadata to model the latent space of college preferences. We present the model results from two perspectives: 1) to understand the impact of TOEFL score and test year on test-takers' preferences and choices and 2) to recommend to the test-taker additional undergraduate institutions for application consideration. We find that TOEFL scores can explain variance in the probability that test-takers belong to certain preference-groups and, by accounting for this, our system adjusts recommendations based on student score. We also find that the inclusion of year, while not significantly altering recommendations, does enable us to capture minor changes in the relative popularity of similar institutions. The performance of this model demonstrates the utility of this approach for providing students with personalized college recommendations and offers a useful baseline approach that can be extended with additional data sources.

The Common Time Course of Memory Processes Revealed.

Magnetoencephalography (MEG) was used to compare memory processes in two experiments, one involving recognition of word pairs and the other involving recall of newly learned arithmetic facts. A combination of hidden semi-Markov models and multivariate pattern analysis was used to locate brief "bumps" in the sensor data that marked the onset of different stages of cognitive processing. These bumps identified a separation between a retrieval stage that identified relevant information in memory and a decision stage that determined what response was implied by that information. The encoding, retrieval, decision, and response stages displayed striking similarities across the two experiments in their duration and brain activation patterns. Retrieval and decision processes involve distinct brain activation patterns. We conclude that memory processes for two different tasks, associative recognition versus arithmetic retrieval, follow a common spatiotemporal neural pattern and that both tasks have distinct retrieval and decision stages.

Practices and pitfalls in inferring neural representations.

A key challenge for cognitive neuroscience is deciphering the representational schemes of the brain. Stimulus-feature-based encoding models are becoming increasingly popular for inferring the dimensions of neural representational spaces from stimulus-feature spaces. We argue that such inferences are not always valid because successful prediction can occur even if the two representational spaces use different, but correlated, representational schemes. We support this claim with three simulations in which we achieved high prediction accuracy despite systematic differences in the geometries and dimensions of the underlying representations. Detailed analysis of the encoding models' predictions showed systematic deviations from ground-truth, indicating that high prediction accuracy is insufficient for making representational inferences. This fallacy applies to the prediction of actual neural patterns from stimulus-feature spaces and we urge caution in inferring the nature of the neural code from such methods. We discuss ways to overcome these inferential limitations, including model comparison, absolute model performance, visualization techniques and attentional modulation.

Reconfiguration of parietal circuits with cognitive tutoring in elementary school children.

Cognitive development is shaped by brain plasticity during childhood, yet little is known about changes in large-scale functional circuits associated with learning in academically relevant cognitive domains such as mathematics. Here, we investigate plasticity of intrinsic brain circuits associated with one-on-one math tutoring and its relation to individual differences in children's learning. We focused on functional circuits associated with the intraparietal sulcus (IPS) and angular gyrus (AG), cytoarchitectonically distinct subdivisions of the human parietal cortex with different roles in numerical cognition. Tutoring improved performance and strengthened IPS connectivity with the lateral prefrontal cortex, ventral temporal-occipital cortex, and hippocampus. Crucially, increased IPS connectivity was associated with individual performance gains, highlighting the behavioral significance of plasticity in IPS circuits. Tutoring-related changes in IPS connectivity were distinct from those of the adjacent AG, which did not predict performance gains. Our findings provide new insights into plasticity of functional brain circuits associated with the development of specialized cognitive skills in children.

Phases of learning: How skill acquisition impacts cognitive processing.

This fMRI study examines the changes in participants' information processing as they repeatedly solve the same mathematical problem. We show that the majority of practice-related speedup is produced by discrete changes in cognitive processing. Because the points at which these changes take place vary from problem to problem, and the underlying information processing steps vary in duration, the existence of such discrete changes can be hard to detect. Using two converging approaches, we establish the existence of three learning phases. When solving a problem in one of these learning phases, participants can go through three cognitive stages: Encoding, Solving, and Responding. Each cognitive stage is associated with a unique brain signature. Using a bottom-up approach combining multi-voxel pattern analysis and hidden semi-Markov modeling, we identify the duration of that stage on any particular trial from participants brain activation patterns. For our top-down approach we developed an ACT-R model of these cognitive stages and simulated how they change over the course of learning. The Solving stage of the first learning phase is long and involves a sequence of arithmetic computations. Participants transition to the second learning phase when they can retrieve the answer, thereby drastically reducing the duration of the Solving stage. With continued practice, participants then transition to the third learning phase when they recognize the problem as a single unit and produce the answer as an automatic response. The duration of this third learning phase is dominated by the Responding stage.

Modeling the distinct phases of skill acquisition.

A focus of early mathematics education is to build fluency through practice. Several models of skill acquisition have sought to explain the increase in fluency because of practice by modeling both the learning mechanisms driving this speedup and the changes in cognitive processes involved in executing the skill (such as transitioning from calculation to retrieval). In the current study, we use hidden Markov modeling to identify transitions in the learning process. This method accounts for the gradual speedup in problem solving and also uncovers abrupt changes in reaction time, which reflect changes in the cognitive processes that participants are using to solve math problems. We find that as participants practice solving math problems they transition through 3 distinct learning states. Each learning state shows some speedup with practice, but the major speedups are produced by transitions between learning states. In examining and comparing the behavioral and neurological profiles of each of these states, we find parallels with the 3 phases of skill acquisition proposed by Fitts and Posner (1967): a cognitive, an associative, and an autonomous phase. (PsycINFO Database Record

Plasticity of left perisylvian white-matter tracts is associated with individual differences in math learning.

Plasticity of white matter tracts is thought to be essential for cognitive development and academic skill acquisition in children. However, a dearth of high-quality diffusion tensor imaging (DTI) data measuring longitudinal changes with learning, as well as methodological difficulties in multi-time point tract identification have limited our ability to investigate plasticity of specific white matter tracts. Here, we examine learning-related changes of white matter tracts innervating inferior parietal, prefrontal and temporal regions following an intense 2-month math tutoring program. DTI data were acquired from 18 third grade children, both before and after tutoring. A novel fiber tracking algorithm based on a White Matter Query Language (WMQL) was used to identify three sections of the superior longitudinal fasciculus (SLF) linking frontal and parietal (SLF-FP), parietal and temporal (SLF-PT) and frontal and temporal (SLF-FT) cortices, from which we created child-specific probabilistic maps. The SLF-FP, SLF-FT, and SLF-PT tracts identified with the WMQL method were highly reliable across the two time points and showed close correspondence to tracts previously described in adults. Notably, individual differences in behavioral gains after 2 months of tutoring were specifically correlated with plasticity in the left SLF-FT tract. Our results extend previous findings of individual differences in white matter integrity, and provide important new insights into white matter plasticity related to math learning in childhood. More generally, our quantitative approach will be useful for future studies examining longitudinal changes in white matter integrity associated with cognitive skill development.

Cognitive tutoring induces widespread neuroplasticity and remediates brain function in children with mathematical learning disabilities.

Competency with numbers is essential in today's society; yet, up to 20% of children exhibit moderate to severe mathematical learning disabilities (MLD). Behavioural intervention can be effective, but the neurobiological mechanisms underlying successful intervention are unknown. Here we demonstrate that eight weeks of 1:1 cognitive tutoring not only remediates poor performance in children with MLD, but also induces widespread changes in brain activity. Neuroplasticity manifests as normalization of aberrant functional responses in a distributed network of parietal, prefrontal and ventral temporal-occipital areas that support successful numerical problem solving, and is correlated with performance gains. Remarkably, machine learning algorithms show that brain activity patterns in children with MLD are significantly discriminable from neurotypical peers before, but not after, tutoring, suggesting that behavioural gains are not due to compensatory mechanisms. Our study identifies functional brain mechanisms underlying effective intervention in children with MLD and provides novel metrics for assessing response to intervention.

Detecting math problem solving strategies: an investigation into the use of retrospective self-reports, latency and fMRI data.

This research explores how to determine when mathematical problems are solved by retrieval versus computation strategies. Past research has indicated that verbal reports, solution latencies, and neural imaging all provide imperfect indicators of this distinction. Participants in the current study solved mathematical problems involving two distinct problem types, called 'Pyramid' and 'Formula' problems. Participants were given extensive training solving 3 select Pyramid and 3 select Formula problems. Trained problems were highly practiced, whereas untrained problems were not. The distinction between untrained and trained problems was observed in the data. Untrained problems took longer to solve, more often used procedural strategies and showed a greater activation in the horizontal intraparietal sulcus (HIPS) when compared to trained problems. A classifier fit to the neural distinction between trained-untrained problems successfully predicted training within and between the two problem types. We employed this classifier to generate a prediction of strategy use. By combining evidence from the classifier, problem solving latencies, and retrospective reports, we predicted the strategy used to solve each problem in the scanner and gained unexpected insight into the distinction between different strategies.

Disordered reward processing and functional connectivity in trichotillomania: a pilot study.

BACKGROUND: The neurobiology of Trichotillomania is poorly understood, although there is increasing evidence to suggest that TTM may involve alterations of reward processing. The current study represents the first exploration of reward processing in TTM and the first resting state fMRI study in TTM. We incorporate both event-related fMRI using a monetary incentive delay (MID) task, and resting state fMRI, using two complementary resting state analysis methodologies (functional connectivity to the nucleus accumbens and dual regression within a reward network) in a pilot study to investigate differences in reward processing between TTM and healthy controls (HC). METHODS: 21 unmedicated subjects with TTM and 14 HC subjects underwent resting state fMRI scans. A subset (13 TTM and 12 HC) also performed the MID task. RESULTS: For the MID task, TTM subjects showed relatively decreased nucleus accumbens (NAcc) activation to reward anticipation, but relative over-activity of the NAcc to both gain and loss outcomes. Resting state functional connectivity analysis showed decreased connectivity of the dorsal anterior cingulate (dACC) to the NAcc in TTM. Dual regression analysis of a reward network identified through independent component analysis (ICA) also showed decreased dACC connectivity and more prominently decreased basolateral amygdala connectivity within the reward network in TTM. CONCLUSIONS: Disordered reward processing at the level of NAcc, also involving decreased modulatory input from the dACC and the basolateral amygdala may play a role in the pathophysiology of TTM.

Neural predictors of individual differences in response to math tutoring in primary-grade school children.

Now, more than ever, the ability to acquire mathematical skills efficiently is critical for academic and professional success, yet little is known about the behavioral and neural mechanisms that drive some children to acquire these skills faster than others. Here we investigate the behavioral and neural predictors of individual differences in arithmetic skill acquisition in response to 8-wk of one-to-one math tutoring. Twenty-four children in grade 3 (ages 8-9 y), a critical period for acquisition of basic mathematical skills, underwent structural and resting-state functional MRI scans pretutoring. A significant shift in arithmetic problem-solving strategies from counting to fact retrieval was observed with tutoring. Notably, the speed and accuracy of arithmetic problem solving increased with tutoring, with some children improving significantly more than others. Next, we examined whether pretutoring behavioral and brain measures could predict individual differences in arithmetic performance improvements with tutoring. No behavioral measures, including intelligence quotient, working memory, or mathematical abilities, predicted performance improvements. In contrast, pretutoring hippocampal volume predicted performance improvements. Furthermore, pretutoring intrinsic functional connectivity of the hippocampus with dorsolateral and ventrolateral prefrontal cortices and the basal ganglia also predicted performance improvements. Our findings provide evidence that individual differences in morphometry and connectivity of brain regions associated with learning and memory, and not regions typically involved in arithmetic processing, are strong predictors of responsiveness to math tutoring in children. More generally, our study suggests that quantitative measures of brain structure and intrinsic brain organization can provide a more sensitive marker of skill acquisition than behavioral measures.

Weak task-related modulation and stimulus representations during arithmetic problem solving in children with developmental dyscalculia.

Developmental dyscalculia (DD) is a disability that impacts math learning and skill acquisition in school-age children. Here we investigate arithmetic problem solving deficits in young children with DD using univariate and multivariate analysis of fMRI data. During fMRI scanning, 17 children with DD (ages 7-9, grades 2 and 3) and 17 IQ- and reading ability-matched typically developing (TD) children performed complex and simple addition problems which differed only in arithmetic complexity. While the TD group showed strong modulation of brain responses with increasing arithmetic complexity, children with DD failed to show such modulation. Children with DD showed significantly reduced activation compared to TD children in the intraparietal sulcus, superior parietal lobule, supramarginal gyrus and bilateral dorsolateral prefrontal cortex in relation to arithmetic complexity. Critically, multivariate representational similarity revealed that brain response patterns to complex and simple problems were less differentiated in the DD group in bilateral anterior IPS, independent of overall differences in signal level. Taken together, these results show that children with DD not only under-activate key brain regions implicated in mathematical cognition, but they also fail to generate distinct neural responses and representations for different arithmetic problems. Our findings provide novel insights into the neural basis of DD.