lesson 8: the crucial role of domain-specific skills in early childhood education

preschool/kindergarten

• Starts at 2.5 years, 3 grades

• Non-mandatory (98% attends), fully subsidized

• Playfull activities

• Community curricular guidelines (incl math)

primary school

• Starts at 6 years, 6 grades

• Mandatory, fully subsidized

• Formal approach

• Community curricular standards (incl math)

domain-specific skills

-        The strongest predictors of later achievement are school-entry math and reading

-        Measures of socioemotional behaviors were generally insignificant

-       Patterns of associations were similar for boys and girls, and high and low SES

number sense

understanding the meaning of numbers (quantity) → key factor in math development

o   The idea that a number represents a quantity

o  Evolutionary relevant (seen in more animals)

o   We learn to use symbolic values

SYMP test

a screening measure

Task: cross out the larger number as fast as you can

o   Trade-off between speed and accuracy

mathematical learning difficulties (MLD)

Score below a certain point = difficulties (25^th percentile) Differences across the whole school

Arithmetic -> symbolic magnitude processing

-        Is this relation uni- or bidirectional?

-        symbolic comparison → calculations at the end of primary school

-        How about the reverse relation? (left) → Association was slightly stronger

dyscalculia

o   Risk for doing poorly in math → poorly in foundational skills

o   They don’t get the same input from their calculation abilities to this domain-specific processing as the typically developing children

dyslexia

phonological awareness is important for later reading, and vice versa

network analysis (early numerical skills)

Different tasks that are interrelated

o   If a task has multiple connections, it means that this task is more relevant → central in the network

o   The width of the lines gives an indication of the strength of the relation

o  This structure is very stable (2^nd vs 3^rd year of preschool)

sex differences in STEM abilities

If we see those differences in secondary school and also in university level

• If you see this at a later age, maybe they’re already there at a young age?

• Or maybe not and they only emerge later due to some very complex social interaction types of factors?

Boys and girls are equally equipped to learn math

• There are probably other reasons why we see these differences occur later on

heterogenity

Can we already identify different subgroups of kids at this age?

• The differences between those profiles are largely quantitative

• There’s heterogeneity, but this is sort of dimensional

preschool -> grade 1 and 3 (heterogenity)

These profiles are highly predictive of later match achievements

• When we add the covariates, there’s a slightly better prediction → they do matter to a certain extent, but the difference isn’t that big

• Those early domain-specific predictors are the most critical

excellence

Looking at the very good performances at math

They also added tasks that are quite complex for this age

• patterning

• proportional reasoning

• probabilistic reasoning

probabilistic reasoning

statistics

“this bird likes to eat white berries and gets sick when he eats black berries → the bird has to choose 1 bow, which one should he pick in order to avoid the black berries?”

proportional reasoning

fair sharing tasks work well for young kids

high math achievers

To what extend are these high achievers different from the others?

• General cognitive skills: they also do very well on a cognitive level

• Domain-specific skills are the most critical to predict mathematical excellence

early math and reading

Domain-specific skills of these skills are intertwined

• Comorbidity between dyslexia and dyscalculia

• We’re looking here at predicting early calculation that’s uniquely predicted by domain-specific skills (recognition of numerals), but also by early predictors of reading (phonological awareness)

• The same hold for reading: if we predict reading, it’s predicted by phonological awareness (domain-specific), but also by the early numerical abilities

• There’s a string relation between the 2, already in kindergarten

impact of schooling

Going to school also had a big impact on the precursors

• Numerical abilities: clear schooling effect

• Early learning abilities: the same

Powerful evidence that schooling had an effect, but also specifically on those predictors

school cut-off design

There’s an arbitrary cut-off date starting school (1^st of January)

So you can identify 2 groups: before and after the cut-off

·       You can follow kids who are similar in age (differ in a few months: oct-dec <> jan-mar)

·       But differ a lot in the amount of schooling (different grade)

schooling and brain level

• Significant schooling effects were observed in bilateral, but predominantly left: temporal, parietal and inferior frontal regions associated with emergent reading and mathematics

• More specifically, reading gains were associated with increased volume and thickness in right inferior frontal gyrus, with increased thickness in left superior parietal and increased volume in right inferior temporal cortices, while mathematical gains were correlated with increased thickness in left inferior parietal cortex

• Children who exhibited larger increases in volume and thickness also showed stronger improvement in their academic achievement

School changes brain function (and structures) related to math and reading

reading gains

associated with increased volume and thickness in right inferior frontal gyrus, with increased thickness in left superior parietal and increased volume in right inferior temporal cortices

mathematical gains

correlated with increased thickness in left inferior parietal cortex

early interventions at school

study until 2016

·       Numeracy interventions are effective for all children

·       Short interventions lead already to pos effects

·       Interventions focus on a variety of early numerical abilities

study until 2025

o   Effective for all children (incl at risk, MLD)

o   Variety in early numerical abilities

o   Digital interventions have smaller effects → don’t know the reason

immediate effect

right after the intervention

early interventions at school: moderate to large effect: g = .52

long-term effect

early intervention at school:

• There’s a follow-up effect, but smaller than the immediate effect

• g = .35

play-based intervention at home

• Pos effects on early numerical skills and activities and expectations parents

• This also had an impact on the activities that parents did with their kids + higher expectations for these kids

• These interventions MIGHT work in a preventive way

storybook/dialogical reading

intervention at home

• The parent reads the story out loud

• There are a set of question which they can ask to their kid and this stimulates interaction

early childhood education

focuses on rich, well-thought-out content, high-quality interactions and a caring, play-based learning environment

positive reinforcing spiral

Parents see that their kid can do more → they give them more and more difficult things → kids learn more and want more …

domain-general interventions

are they useful for working memory (WM) and executive functions (EF)?

• honest answer is NO

perils of domain-general interventions

-        Working memory/EF interventions don’t transfer to specific skills

-        Learning is a function of some cognitive capacities, but also specific knowledge and different types of strategies

-        Learning academic skills improves WM/EF

Domain-specific mathematical skills

crucial in early childhood education

(pre)school

has a critical impact on these domain-specific math skills

domain-specific interventions

the most effective