Using Support Vector Machines for Facet Partitioning in Multidimensional Scaling.

In this article we focus on interpreting multidimensional scaling (MDS) configurations using facet theory. The facet theory approach is attempting to partition a representational space, facet by facet, into regions with certain simplifying constraints on the regions' boundaries (e.g., concentric circular sub-spaces). A long-standing problem has been the lack of computational methods for optimal facet-based partitioning. We propose using support vector machines (SVM) to perform this task. SVM is highly attractive for this purpose as they allow for linear as well as nonlinear classification boundaries in any dimensionality. Using various classical examples from the facet theory literature we elaborate on the combined use of MDS and SVM for facet-based partitioning. Different types of MDS are discussed, and options for SVM kernel specification, tuning, and performance evaluation are illustrated.

Using the force: STEM knowledge and experience construct shared neural representations of engineering concepts.

How does STEM knowledge learned in school change students' brains? Using fMRI, we presented photographs of real-world structures to engineering students with classroom-based knowledge and hands-on lab experience, examining how their brain activity differentiated them from their "novice" peers not pursuing engineering degrees. A data-driven MVPA and machine-learning approach revealed that neural response patterns of engineering students were convergent with each other and distinct from novices' when considering physical forces acting on the structures. Furthermore, informational network analysis demonstrated that the distinct neural response patterns of engineering students reflected relevant concept knowledge: learned categories of mechanical structures. Information about mechanical categories was predominantly represented in bilateral anterior ventral occipitotemporal regions. Importantly, mechanical categories were not explicitly referenced in the experiment, nor does visual similarity between stimuli account for mechanical category distinctions. The results demonstrate how learning abstract STEM concepts in the classroom influences neural representations of objects in the world.

Individual differences in white and grey matter structure associated with verbal habits of thought.

Modality specific encoding habits account for a significant portion of individual differences reflected in functional activation during cognitive processing. Yet, little is known about how these habits of thought influence long-term structural changes in the brain. Traditionally, habits of thought have been assessed using self-report questionnaires such as the visualizer-verbalizer questionnaire. Here, rather than relying on subjective reports, we measured habits of thought using a novel behavioral task assessing attentional biases toward picture and word stimuli. Hypothesizing that verbal habits of thought are reflected in the structural integrity of white matter tracts and cortical regions of interest, we used diffusion tensor imaging and volumetric analyses to assess this prediction. Using a whole-brain approach, we show that word bias is associated with increased volume in several bilateral language regions, in both white and grey matter parcels. Additionally, connectivity within white matter tracts within an a priori speech production network increased as a function of word bias. These results demonstrate long-term structural and morphological differences associated with verbal habits of thought.

Mental models use common neural spatial structure for spatial and abstract content.

Mental models provide a cognitive framework allowing for spatially organizing information while reasoning about the world. However, transitive reasoning studies often rely on perception of stimuli that contain visible spatial features, allowing the possibility that associated neural representations are specific to inherently spatial content. Here, we test the hypothesis that neural representations of mental models generated through transitive reasoning rely on a frontoparietal network irrespective of the spatial nature of the stimulus content. Content within three models ranges from expressly visuospatial to abstract. All mental models participants generated were based on inferred relationships never directly observed. Here, using multivariate representational similarity analysis, we show that patterns representative of mental models were revealed in both superior parietal lobule and anterior prefrontal cortex and converged across stimulus types. These results support the conclusion that, independent of content, transitive reasoning using mental models relies on neural mechanisms associated with spatial cognition.

Individual differences in encoded neural representations within cortical speech production network.

When two individuals view the same item, they do not necessarily perceive an item in the same way. If an individual is presented with a stimulus to be recalled later, the information that is encoded is dependent on the features of the stimulus to which one attends. Past studies have shown that, on the group level, verbal and visual information (e.g., words and pictures) are encoded in disparate regions of the brain. However, this account conflates external and internal representational formats, and it also neglects individual differences in attention. In this study, we examined neural and behavioral patterns associated with individual differences in attention to verbal representations-both external and internal. We found that the encoded neural representation of semantic content (meaningful words and pictures) varied as a function of individual differences in verbal attention, independent of the stimulus presentation format. Individuals who demonstrated an attentional bias toward words showed similar multivariate BOLD activity patterns within an a priori speech production network when encoding object names as when encoding pictures of objects. This result indicates that these individuals encode both words and pictures verbally. These effects were not found for non-semantic stimuli (pronounceable non-words and nonsense pictures). Importantly, as expected, no individual differences in neural representation were found in a separate network of regions known to process semantic content independent of format. These results highlight inter-individual divergence and convergence in internal representations of encoded semantic content. SIGNIFICANCE STATEMENT: This study shows how tendencies to attend to word representations is associated with individual differences in encoded neural representations. Individuals who selectively attend to words instead of pictures process semantically meaningful information in language regions of the brain, regardless of whether the information was originally presented as a word or a picture. Though all participants encoded words and pictures similarly in regions that are known to represent domain-general semantic information, only the individuals who were biased towards word representations additionally processed both words and pictures in modality-specific verbal regions. These results demonstrate both the convergence and divergence between individuals that occurs during encoding of meaningful information.

Decoding individual differences in STEM learning from functional MRI data.

Traditional tests of concept knowledge generate scores to assess how well a learner understands a concept. Here, we investigated whether patterns of brain activity collected during a concept knowledge task could be used to compute a neural 'score' to complement traditional scores of an individual's conceptual understanding. Using a novel data-driven multivariate neuroimaging approach-informational network analysis-we successfully derived a neural score from patterns of activity across the brain that predicted individual differences in multiple concept knowledge tasks in the physics and engineering domain. These tasks include an fMRI paradigm, as well as two other previously validated concept inventories. The informational network score outperformed alternative neural scores computed using data-driven neuroimaging methods, including multivariate representational similarity analysis. This technique could be applied to quantify concept knowledge in a wide range of domains, including classroom-based education research, machine learning, and other areas of cognitive science.