Significance:
Our research programme takes on the grand challenge to prepare our children well in learning to read and reason mathematically in their early school years by leveraging on the Science of Learning through the use of neuroimaging methods together with cognitive measures to inform practice. Uncovering the underpinnings of brain networks related to literacy and numeracy to allow earlier identification of difficulties in learning with existing screeners, will help differentiate whether these learning difficulties stem from weaknesses in reading, mathematical or other cognitive abilities, such as working memory. The long-term goal of this research is to value-add and enhance the existing learning support programmes with precise measures to monitor development of literacy and numeracy from the start of primary school. This enables earlier characterization of difficulties in learning, provide targeted remediation, level-up disadvantaged students and bring their learning trajectory closer to that of their peers.
Aims and Objectives:
Although there is growing evidence showing that learning difficulties in literacy and numeracy may stem from common factors, current diagnostic approaches and interventions treat learning difficulties in these two domains independently. This research will join recent global efforts in developing transdiagnostic approaches to understand developmental learning disabilities not only by providing novel evidence to validate different subtypes of learning difficulties in literacy and numeracy, but also by conceptualizing an exemplary model of how such evidence can be translated into educational practice to augment current approaches in diagnostics, monitoring and interventions. The goals of the research programme are:
1. To identify the core deficits and behavioural indicators for learning disabilities (LD) in the areas of reading (RD), math (MD) and their co-morbidity (RMD)
development within a local bilingual sample of Singaporean children.
2. To provide precise whole-brain mapping using magnetic resonance imaging (MRI) of the functional and structural networks supporting reading, math
and working memory (where working memory is hypothesized to be the underlying domain affected in RMD) from preschool to lower primary children,
as well as to elucidate how these networks are compromised by the respective difficulties.
3. To uncover the differences in functional networks of reading, math and working memory (visuospatial and verbal) between RD, MD and RMD, and
provide neural evidence from ecologically valid tasks to validate the neurocognitive tests that are sensitive in distinguishing RMD from RD and MD.
4. To examine purported subgroups of RD and MD and RMD enrolled in learning support programmes using identified sensitive neurocognitive measures
to (a) check the sensitivity and specificity of the school screeners based on the assignment of students to the LD groups (RD, MD, RMD) according to
the neurocognitive measures, and (b) to examine neurocognitive profiles of individuals who respond well to the intervention programmes (e.g., graduate
out) versus non-responders who remain in the system over the years.
5. To target the environmental factor of socioeconomic status (SES), which is important in discerning children’s abilities and elucidate the extent of
environmental disadvantages contributing to poorer learning in contrast to pre-existing dysfunction in the brain networks.
Approach:
Employing an accelerated longitudinal approach to discern the potential subtypes or mechanisms underlying reading difficulties (RD), math difficulties (MD) and their comorbidity (RMD), we will study changes in brain and cognitive/behavioural measures in the cohorts with two dimensions in mind. (1) A trajectory approach starting from basic science methodologies to translational neuroscience for education. Relevant structures and functions of the brain networks that are associated with RD, MD, and working memory difficulties will be correlated with existing cognitive/behavioural measures to identify sensitive cognitive/behavioural measures, respectively. Magnetic resonance imaging (MRI) with precise whole-brain mapping will identify the key brain networks involved. These findings will be integrated with functional near-infrared spectroscopy (fNIRS) mapping of brain functions during more ecological tasks performed in naturalistic environments. The identified sensitive cognitive/behavioural measures for RD, MD, and working memory difficulties can then be used in the classrooms by educators to discern and monitor learning progress. It will allow more agile interventions for children to move in between the tiers (i.e., mainstream, learning support, or more targeted interventions like the dyslexia programme) based on their response to intervention. (2) A developmental approach to characterize and understand the neural and performance changes from 5-6 years (K2-P1) to 7-8 years (P2-P3) and from 6-7 years (P1-P2) to 8-9 years (P3-P4). In addition to neural and cognitive factors, this research programme will target the environmental factor of socioeconomic status (SES), which is important in discerning children’s abilities. This will help decipher to what extent environmental disadvantages can contribute to poorer learning in contrast to pre-existing dysfunction in the brain networks. This approach enables the selection of more sensitive methods in identifying children in need of interventions at the right time in their development.
The research programme will have 5 integrated projects. Projects 1 to 4 will study the same two cohorts of children (N=240). Cohort 1 (n=120) will start at age 5-6 years with a follow-up time point at 7-8 years. Cohort 2 (n=120) will start at age 6-7 years with a follow-up time point at 8-9 years. Each proposed cohort consists of typically developing (TD) children, children with RD, children with MD and children with RMD as identified with screening and standardized assessments. We will derive the criteria for group classification from a larger sample of children (N=600 in each cohort) assessed at multiple time points longitudinally from the “Growth in Bilingual & Biliteracy Proficiency: Environmental, Individual & Experiential Factors (GiBBEr)” study using data-driven approaches such as cluster analysis. This sampling strategy enables the ability to address attrition to have adequate power for change analyses.
Project 1 will characterize reading, math, and executive function domains in children using existing tests conducted in the schools. Project 2 will employ MRI and task-fMRI to examine the function, structure, and connectivity of reading, math and working memory networks in the cohorts. Project 3 will employ fNIRS to understand brain networks while doing a naturalistic reading task and Project 4 will apply fNIRS while the children are doing a naturalistic math task. Project 5 is proposed to be an exploratory study starting in the second year of the research programme after some sensitive measures are identified. These will be applied to follow children identified from intake using literacy and numeracy indices at P1 in 5 neighbourhood schools (N=125). The identified students are followed through pull-out early intervention known as the “Learning Support Programmes” of reading and math, and monitored through P2 and P3. Sensitivity and specificity of the measures will be identified, and their utility as measures to monitor response to intervention will be evaluated.
Feasibility:
We are leveraging on the existing larger cohort study “Growth in Bilingual & Biliteracy Proficiency: Environmental, Individual & Experiential Factors (GiBBEr)” to synergise and minimize duplication of efforts. We will derive the criteria for group classification from a larger sample of children (N=600 in each cohort) assessed at multiple time points longitudinally from the GiBBEr study using data-driven approaches such as cluster analysis. This sampling strategy enables the ability to address attrition to have adequate power for change analyses. Our Co-PI and collaborators from the Ministry of Education developed the curriculum used for learning support programmes in the schools and has strong networks with neighbourhood schools to engage with for implementing the Learning Support in Action project that seeks to translate neuroscience findings for education.
This interdisciplinary research programme will draw expertise from cognitive neuroscience, clinical neuropsychology, education, and psychology with colleagues from National Institute of Education and Psychology at NTU and NUS, and stakeholders at Ministry of Education and the Dyslexia Association of Singapore (DAS). Support from the Science of Learning (SoL) research thrust at the Centre for Research and Development in Learning (CRADLE), and Centre for Research in Child Development (CRCD) provide the infrastructure necessary to translate empirical findings into educational and clinical practice. Existing state-of-the-art neuroimaging facilities at NTU, such as the Cognitive Neuroimaging Centre (CoNIC) and the server farms for high-computing data analytics and storage are available to the team.
Reading difficulties, Math difficulties, working memory, comorbidity, response to intervention, screeners