Publications

1. *Prem, Boadu, Saju, Nisbet, & Cummine. (Submitted, 2024). Functional Brain Activation during Visual Statistical Learning is Related to Rapid Automatised Naming of Digits and Objects: an fMRI Study. Scientific Studies of Reading (in Review)

Statistical learning refers to the brain’s ability to detect patterns in the environment and is important for language acquisition in both infants and adults. Notably, SL has been shown to activate brain regions that are implicated in language/reading; however, the extent to which this brain activity is related to reading-related tasks is unknown. Here, we used fMRI to explore the brain-behavior relationships between SL and reading processes. Adult participants (N = 44) completed a Visual SL (VSL) task inside the MRI that included two phases: 1) learning, wherein participants were presented with novel triplets of shapes varying in transitional probabilities, and 2) testing, which involved selecting best-fitting triplets that assessed their learning of these probabilities. Outside the scanner, participants completed 3 Rapid Automatized Naming (RAN) Tasks consisting of quickly repeating strings of digits, letters, and objects. Using a whole-brain correlational approach, we found that brain activity during both phases of the VSL task was significantly correlated to performance on RAN digits and objects (p < 0.001, uncorrected), but not letters. These findings may indicate that VSL (non-linguistic, as used in the current study) is sensitive to symbols that carry semantic information in and of themselves (i.e., digits and objects).

2. *Prem, P., Eze, P., Chew, R., Saju, S., Nisbet, K., & Cummine, J. "The Crossing Path(ways) of Reading Processes and Statistical Learning: Evidence from Diffusion Tensor Imaging Study."

There is a well established relationship between basic reading processes and statistical learning abilities. Further, the white matter pathways that support basic reading processes have been well defined. The extent to which statistical learning abilities are also related to these same underlying white matter pathways has yet to be explored. This study investigated the relationship between microstructural properties of four key reading-related white matter tracts: the Inferior Fronto-Occipital Fasciculus (IFOF), Left Uncinate Fasciculus (UF), Arcuate Fasciculus (AF), and Corticospinal Tract (CST); and performance on reading (i.e., Word Identification, and Word Attack) and statistical learning (i.e., auditory, ASL & visual, VSL) tasks. Diffusion Tensor Imaging (DTI) data were collected from 61 participants. Diffusion metrics of Fractional Anisotropy (FA) and Mean Diffusivity (MD) values were extracted from each tract. Results revealed significant associations between diffusion metrics of each of the tracts and reading performance. Interestingly, only the CST was related to performance on the ASL and VSL task (p < 0.05), suggesting that structural connectivity within motor pathways may contribute to statistical learning. These findings provide unique information about the potential connections between reading and statistical learning processes, further advancing models of reading and implicit learning.

3. *Prem, P., Saggu, S. K., Boadu, A., Saju, S., Nisbet, K., & Cummine, J. (Submitted, 2024). “Neuroanatomy of Visual and Auditory Statistical Learning: A Volumetric Account of Cortical and Subcortical Regions.” Human Brain Mapping (under Review) 

A human’s ability to implicitly extract patterns from the environment, known as statistical learning (SL), is ubiquitous across the lifespan. We explored the extent to which varying forms of SL performance, namely visual vs. auditory, are differentially related to cortical vs. subcortical brain volume. Methods: High resolution T1-weighted Magnetic Resonance Imaging (MRI) scans were taken of 61 participants then volumetric measures were extracted from cortical (planum temporale, fusiform gyrus, middle temporal gyrus, inferior frontal gyrus) and subcortical (caudate, nucleus accumbens, putamen, globus pallidus) regions of interests. SL involved two phases: training and testing. In the familiarization phase, participants were passively and sequentially presented with structured triplets of unfamiliar and non-linguistic shapes (Visual-SL), and environmental sounds (Auditory-SL). In the testing phases, participants had to identify trained triplets or pairs or complete the triplet/pair of the missing object using a series of alternative-forced-choice tasks for both VSL and ASL. Results: Hierarchical regression analyses revealed that ASL performance was significantly related to the volume of the IFG, whereas VSL performance was significantly associated with all examined subcortical structures (p < 0.05). Together these findings advance our understanding of the domain-general and domain-specific aspects of SL, and provide evidence for a complex and interconnected network of brain regions involved in learning.

4. *Simpson, D., Prem, P., & Nash, K. (Submitted, 2023). “Increased attention decreases the convincingness of belief-confirming evidence.” Journal of Experimental and Social Psychology (under Review). Preprint available at SSRN: doi.org/10.2139/ssrn.4671269

We tested whether increased attentiveness would increase or decrease confirmation bias. We defined confirmation bias (in the sense of biased information interpretation) as the difference between how convincing people found belief-confirming arguments to be, and how convincing they found belief-disconfirming arguments to be. The arguments in question were arguments for and against the economic benefits of raising the minimum wage, all of which were modified from real research articles. People in the experimental condition were told up front that they would be regularly tested on their comprehension of the arguments, which would presumably provide an incentive to read the arguments more attentively. They were then given a comprehension question after every argument that they read. In Study 1, we found that this manipulation caused participants to find belief-confirming evidence marginally less convincing, while having no effect on their evaluation of the convincingness of belief-disconfirming evidence. This marginally decreased their overall biased information interpretation score. When we replicated this pattern with a high-powered preregistered study (Study 2), the results were significant. Possible implications are discussed.

5. ^Ellis, Prem, Kecinski, Nash, Messer, & Lusk. Feelings Vs. Behavior - A Neuroeconomic Perspective on Food and Water Stigma.

Dealing with large-scale societal problems, such as water scarcity, often requires changes in behavior that consumers resist. Some sustainable, cost-effective, and safe solutions are even rejected because of a psychological response of disgust, such as food produced with recycled water to supplement traditional water supplies and crickets as a replacement for water-intensive proteins like beef. Adding the “right” positive elements to a stigmatized item has been shown to mitigate disgust. However, this can be difficult and expensive to do as it requires the stigmatized object to go through a process that restores its “natural” and “pure” state. This study combines fMRI and a choice experiment to test if a behavioral intervention that emphasizes the existing, positive dimensions of a stigmatized object can lessen the weight of disgust in the decision process, mitigating its effects. Results suggest that it can. A video that promotes recycled water was shown to ameliorate consumers’ aversion to food produced with it. However, consumers’ greater acceptance of recycled irrigation water came not from overcoming whatever psychological reactions of disgust they experienced but from reweighting its importance in their decisions. These findings are relevant to many real-world scenarios where stigma-driven behavior produces inefficiencies, such as farm-level adoption of recycled irrigation – a safe, cost-effective and environmentally sound solution to water shortages.

6. ^Prem, P., & Nash, K. Experimentally Induced Anxiety Reduces Behavioural Dishonesty in Low-Trait Anxious Individuals: Neural Evidence from EEG.

Prior research has found that neural markers of trait anxiety predict reduced behavioral dishonesty in the Broken Promise (BP) paradigm, i.e., modified trust game with an antecedent promise stage. However, it remains unexamined if state anxiety, i.e., acute feelings of anxiousness, is also associated with dishonesty. Here, we examined the interaction between TA and SA on dishonesty along with neural markers of negative arousal, i.e., P300 amplitudes to aversive stimuli (sensitive to state anxiety). 112 adult participants completed the Behavioral Inhibition System scale to measure trait anxiety, before being randomly assigned to an established SA manipulation or control condition. They then completed an oddball passive listening task while wearing EEG headset to measure P300 amplitudes, and the BP game that measured dishonesty. We find that 1) all participants high in trait anxiety are reliably low in dishonesty, 2) while those low in trait anxiety are higher in dishonesty, but experiencing state anxiety significantly reduces their dishonesty, and 3) this effect is further moderated by mean P300 amplitude difference measure, meaning low trait anxiety individuals in state anxiety condition with high mean P300 difference (evidence of increased state anxiety) show significantly higher decrease in dishonesty. Results support state anxiety's association with dishonesty. We speculate dishonesty can be muted by transient self-sanctioning emotions.

7. ^Boadu A., Prem, P., Saju, S., Nisbet, K., & Cummine, J. "The Neural Correlates of Statistical Learning and Reading Fluency."

While SL is important for reading and language proficiency, the neural mechanisms, especially regarding domain generality or specificity, remain less well understood. Here we examined whether SL mechanisms operate through domain-general or domain-specific pathways across different stimuli. Methods: Forty-eight participants underwent functional Magnetic Resonance Imaging (fMRI) while engaging in auditory and visual SL tasks, followed by language fluency assessments. Following standard preprocessing pipelines, brain regions were analyzed for activation during exposure to auditory statistical learning (ASL) and visual statistical learning (VSL) with abstract stimuli. Participants reading fluency scores were correlated with brain activity during SL using a whole-brain approach. Results: Brain-behavior relationships were observed in the middle and superior frontal gyri and the cingulate gyrus across both sensory domains, whereby increased activity was associated with faster fluency. These findings are in line with a domain-general processing hypothesis whereby there is an integrated neural network that supports SL processing in both auditory and visual domains. The results contribute to the ongoing investigations into the neural basis of SL mechanisms and indicate an interplay between different sensory modalities in SL.

8. ^Borle, C., Prem, P., Boadu, A., Saju, S., Nisbet, K., & Cummine, J. "Functional Connectivity during Visual and Auditory Statistical Learning."