Individual Research Projects

The role of sleep in early cognitive development

Early Stage Researcher: Louisa Gossé

Supervisors: Emily Jones (BBK), Frank Wiesemann (P&G), Clare Elwell (UCL)

During her PhD Louisa investigated the relationship between sleep and early neurocognitive development using a multi-method approach, and conducted two large-scale studies to ask specific questions about individual differences in sleep, social interactions with parents, parental characteristics, and their potential impact on individual infants' learning. Louisa spent September 2017 - February 2019 at the Procter & Gamble German Innovation Center (Germany) then  moved to the CBCD (Birkbeck) March 2019 until finishing the PhD in January 2021. The first study was a longitudinal study using non-invasive electroencephalography EEG, eye-tracking, parent-report measures, sleep diaries and actigraphy to study habitual sleep and its association to various markers of cognitive development in typically developing infants. Louisa tested 76 infants (ages 4-14 months) with 166 visits over the course of a year at the Procter & Gamble. After completion of the study she used linear mixed modelling (using SPSS), cluster analysis (using R) and EEG power analysis (using Matlab/EEGlab) to analyse the longitudinal data. Her first major finding is  that objective and subjective sleep measures differ significantly from each other and do not provide the same results. Moreover this lack of concordance is influenced by infant age and maternal stress. Secondly, she found that typical infant sleep patterns vary widely but generally seem to have relatively little impact on behavioral development. Finally, Louisa found that sleep quality, particularly the amount of time spent awake at night, may be related to aspects of brain development. 

The second study (conducted at the CBCD, Birkbeck, London, UK) was a cross-sectional study where Louisa focused in on how infant brain development might be influenced by quality of sleep. Here, she added another tool to amplify the methods used in the first study, namely functional near-infrared spectroscopy (fNIRS), here using the Artinis NIRS system created by NTS-Gowerlabs, a spin-out company created in part by INTERLEARN member Clare Elwell.   For this purpose Louisa designed the first for-sleep-customized combined fNIRS-EEG headgear; she also set up the first in-lab sleep study at the Birkbeck Babylab. She tested N = 25 infants (ages 5 - 8 months) of which N =11 slept in the lab with the headgear on. To analyse the data Louisa developed a novel MATLAB pre-processing pipeline,  and conducted connectivity analysis on fNIRS data. She found that individual differences in infant functional connectivity patterns during sleep, which potentially represent sleep quality differences. Because of the COVID-19 pandemic, full data collection could not be completed, so results are still preliminary. Thanks to a Wellcome grant, she has been able to continue this work in London.  The technical innovations in conducting simultaneous fNIRS/EEG that were made possible in part by her work in INTERLEARN  also have shown great promise for use in understanding sleep mechanisms in infancy and their relationship to learning and behaviour. 

Current publications resulting from Louisa's project:

Gossé, L. K., Wiesemann, F., Elwell, C. E., & Jones, E. J. H. (2022). Concordance between subjective and objective measures of infant sleep varies by age and maternal mood: Implications for studies of sleep and cognitive development. Infant Behavior and Development66, 101663. 


How touch mediates infant happiness and learning

Early Stage Researcher: Alicja Brzozowska

Supervisors: Frank Wiesemann (P&G), Denis Mareschal (BBK), Teea Gliga (BBK), Matthew Longo (BBK)

The aim of Alicja's project was to investigate if and how touch can affect learning and exploratory behaviour in human infants born at term. Discovering if the findings from animal research could be extended to this population would be of significant importance, as it could potentially lead to designing new, low-cost and powerful ways of supporting healthy infant development. The project addressed three key research questions:  1) Is touch received by infants from caregivers related to their exploratory behaviours and learning? 2) Does touch affect infant learning and exploration through modulation of arousal? If so, could this be explained by stress-buffering effects of touch? Are there different mechanisms mediating the effects of touch on learning social vs. non-social information? 

The first experiment was a broad, cross-sectional study, the primary aim of which was to identify if the amount and types of touch provided by caregivers are related to infants’ exploratory behaviours and learning, and if that relation exists, whether it is occurs indirectly, by modulating arousal. A secondary aim was to develop and validate ways of measuring parental touch and infant exploratory behaviours, which could potentially be used in later studies. Two groups of infants aged 6 – 8 months (N=39) and 11 – 13 months (N=32) were tested. The purpose of having these two age groups was to examine whether the hypothesized relations are continuous across development, or possibly more pronounced in the younger group. By the nature of early interactions and motor constraints, younger infants are more likely to receive more tactile stimulation from the caregivers, which could potentially translate into greater impact of it. The experiment used a wide number of physiological, behavioural, and observational measures to assess these relationships. Alicja found that context affected both the quantity and types of touch used in interaction. Parent-reported touch was moderately associated with touch observed in parent–child interactions and more strongly with touch used during play. 

After moving to London, Alijca started data collection for an experiment investigating the effects of stroking on neural response to faces. Aljica spent the first half of her  PhD at P&G in Frankfurt (since September 2017) and then was based in  London for the completion of the project, with frequent interchanges between beneficiaries and ESR throughout.  She submitted her PhD thesis in March 2021, had her viva in May 2021, and passed with no corrections.  Alicja has also continued her research career as a postdoctoral researcher in another European capital (Vienna), following her interests in the development of attention. 

Current publications resulting from Alicja's project:

Brzozowska, A., Longo, M. R., Mareschal, D., Wiesemann, F., & Gliga, T. (2021). Capturing touch in parent–infant interaction: A comparison of methods. Infancy26(3), 494-514.


The neural dynamics of motivated learning

Early Stage Researcher: Cécile Gal

Supervisors: Teodora Gliga (UEA), Marie Smith (BBK), Raul Muresan (Neurodynamics)

Cécile’s PhD focused on investigating individual differences in learning strategies. In the first year of her PhD, she stayed in her partner institute in Romania, where she looked at how adults select relevant information in a task as basic as visual recognition. They explored altered, ambiguous images of daily-life objects and Cécile precisely quantified the information content at each location explored. She showed that although increasing early information access increased performance — viewing the less ambiguous stimuli first before moving on to more ambiguous ones enabled participants to explore more informative areas and recognise the objects more efficiently than those who started with the most ambiguous stimuli — it was not sufficient. Indeed, introducing randomness in the task undermined the process, and led to less optimal exploration and worse performance. She concluded that a homogenous context was needed for people to choose an optimal visual sampling strategy in order to form priors and learn.  She has prepared  this into a draft of an article to be submitted.

In the second year of her PhD, Cécile moved to her academic partner in London where she designed and recorded data for two experiments with infants. These results were then analysed in Romania, when she moved back to her industrial partner in her third year. In infant study 1, Cécile recorded 10-month-olds’ (N=31) brain activity with electroencephalography (EEG) as infants observed and built knowledge on a repeated stimulus. Using state-of-the-art analysis techniques, she uncovered a key role of high-frequency (high-gamma) EEG activity for infants’ learning. Gamma rhythms represent a crucial brain function because they reflect local processing in the brain, gated by more global and widely studied low-frequency oscillations. However, they remain mostly undescribed and overlooked in infants as their identification is methodologically challenging. For the first time in infants, Cécile was able to show that 10-month-olds’ high-gamma activity is 1) modulated by learning and 2) by low-frequency oscillations (ERPs). This opens exciting and novel avenues for investigating how the human brain learns. In the second infant study (N=70), Cécile explored how 15-month-olds’ memory, inhibition and attention control abilities influence how they sample information during free play. This dataset is unique because it employed a range of complementary methods – eye-tracking and exploratory behaviour, in a large sample. Analysis is still ongoing to show how individual differences in executive functions selectively influence curiosity-driven exploration.

Cécile is just finishing her thesis, and  has secured a six-month postdoctoral position at Birkbeck funded by the Wellcome Trust, during which she will be able work further on publishing and disseminating her work after submission of her thesis. 


Tailoring the individual learning experience using online educational games

Early Stage Researcher: Susanne de Mooij

Supervisors: Han van der Maas (Prowise), Maartje Raijmakers (Oefenweb), Natasha Kirkham (BBK), Iroise Dumontheil (BBK) 

In her first year at Birkbeck (October 2017-October 2018) Susanne performed a study with 43 primary school children (8-11 years old) where they performed online maths tasks and executive functioning tasks. While doing this, she collected behavioural, eye & mouse tracking and video recordings data. Susanne's aim was to look at (1) the impact of time pressure, (2) how the level of executive functioning affect learning experience and performance, (3) what information eye and mouse tracking conveys regarding strategy use and misconceptions/hesitations, and (4) how eye and mouse tracking map on to each other. She found that time pressure can have a negative effect on maths performance depending on the level of executive functioning. Furthermore, eye and mouse tracking gives insight into individual differences in strategy use and misconceptions/hesitations. 

Susanne continued the use of eye and mouse tracking further in Oefenweb/Prowise, where she continued her work from October 2018. She focused on collecting eye and mouse tracking in their database to investigate individual differences in learning at massive scale and to validate the use of both techniques. She set up the tracking of mouse data in the company's software and investigated the emergence of misconceptions in the form of A/B testing.   Impressively, Susanne used these techniques with 150,000 users of the Oefenweb ages 5-13 years to understand how individual children learn to adapt their behaviour during a series of online math challenges.  Because of the scale of data collection and the sensitivity of her new measures, Susanne was able to pull out subtle timing effects and non-linearities in learning over development. 

Susanne has continued her research career in Amsterdam in a postdoctoral position, and is well-placed for an exciting career combining theoretical research to applied educational problems. 

Current publications resulting from Susanne's project:

de Mooij, S. M., Kirkham, N. Z., Raijmakers, M. E., van der Maas, H. L., & Dumontheil, I. (2020). Should online math learning environments be tailored to individuals’ cognitive profiles?. Journal of Experimental Child Psychology191, 104730.

de Mooij, S. M., Raijmakers, M. E., Dumontheil, I., Kirkham, N. Z., & van der Maas, H. L. (2021). Error detection through mouse movement in an online adaptive learning environment. Journal of Computer Assisted Learning37(1), 242-252.

de Mooij, S. M., Dumontheil, I., Kirkham, N. Z., Raijmakers, M. E., & van der Maas, H. L. (2022). Post‐error slowing: Large scale study in an online learning environment for practising mathematics and language. Developmental Science25(2), e13174.


The role of attentional processes in reading skills

Early Stage Researcher: Giada Guerra

Supervisors: Fred Dick (CBCD), Adam Tierney (CBCD), Jurgen Tijms (IWAL), Anniek Vaessen (RID), Milene Bonte (IWAL/RID)

Giada's project aimed at characterising the auditory attentional mechanisms related to reading skills in normally developing and dyslexic children. Furthermore, she aimed to gain understanding on whether and how individual differences in this fundamental cognitive function interact with or support the reading intervention implemented in RID and IWAL. After an initial six months in London, Giada's project took place primarily in the Netherlands; she returned again to London for the final six months during the first COVID lockdown period).    

Towards the goal of identifying the individual differences in auditory attention processes in dyslexic children, Giada first tested a sample of healthy adults (N=37) with non-linguistic and linguistic sustained auditory selective attention paradigms. The results showed significant individual differences in sustained auditory selective attention even in normally reading adults. In addition, she examined whether non-linguistic selective attention ability related to the same ability in the speech domain, finding a strong correlation between the two measures. The relation persisted after including in the model the mental rotation task score to control for overall performance differences. Finally, she found that two 1-hour sessions of training with feedback on the non-linguistic paradigm was able to significantly improve participants performance in the non-linguistic task. 

In her first developmental study, Giada aimed to identify the auditory attention components that predict reading fluency skills in the normally-developing population. With respect to the main questions addressed in this project, the results of this study also served the purpose of constraining predictions for her further studies which used a longitudinal design. Here participants were third- and fourth-grade children (N = 63; 33 boys and 30 girls), recruited from an elementary school in Amsterdam area.  

In a second developmental study, Giada investigated behavioural and neural correlates of auditory selective attention using EEG (electroencephalography). From autumn 2018, participants with dyslexia (63 dyslexic readers and 50 typical readers) were recruited at the RID institutes in Amsterdam. The study tested the hypothesis that children with dyslexia have impaired selective auditory attention, which predict their difficulties in perceiving speech in challenging acoustic conditions and in audio-visual association learning. Audio-visual association learning abilities were measured by employing a dynamic assessment task asking participants to learn novel symbol-speech sound associations.  

Finally, in a last study, Giada investigated attentional and reading-specific predictors of RID treatment outcome. 

Giada defended her thesis in a Birkbeck viva in summer 2021; as an example of a new and successful cross-European initiative started by the EID, Giada will be the first PhD student to receive a joint PhD at Maastricht University in the Netherlands, where RID-based Milene Bonte is also a faculty member.  In keeping with her training at RID and IWAL in the healthcare sector, Giada was been hired as assistant psychologist at a specialized developmental clinical unit in London, where she has combined her interests in learning and improving children's welfare.  

Current publications resulting from Giada's project:

Holt, L. L., Tierney, A. T., Guerra, G., Laffere, A., & Dick, F. (2018). Dimension-selective attention as a possible driver of dynamic, context-dependent re-weighting in speech processing. Hearing Research366, 50-64.

Guerra, G., Tijms, J., Vaessen, A., Tierney, A., Dick, F., & Bonte, M. (2021). Loudness and intelligibility of irrelevant background speech differentially hinder children's short story reading. Mind, Brain, and Education15(1), 77-87.