How statistics shape reading and spelling

The central aim of the project is to unravel how learning and knowledge of distributional (statistical) information supports reading and its acquisition. In the first subproject we create standardized and reliable measures of reading and spelling. These methods are not only essential for the current project, but will be available for researchers and clinicians working with reading and spelling impairments. The second subproject identifies the statistical learning profiles of primary school children with or without isolated or combined reading and spelling deficits. We aim to explore whether and how domain-general statistical learning skills can provide a causal background to reading and spelling impairments. The aim of the third subproject is to identify how individual differences in statistical learning explains variance in reading and spelling skills in unselected samples of beginning readers (1st graders), intermediate readers (3rd graders) and proficient readers (adults). Subproject 3 adds to the existing literature by providing a more fine-grained analysis of the association with numerous different measures of statistical learning. Subproject 4 involves EEG experiments in which we explore how orthographic patterns with varying frequency (from non-existent to high frequency) are processed by children and adults, and whether there is a difference in the time course of processing. Overall, the project uses a wide range of methods and research perspective to examine how statistical learning contributes to reading, spelling and their development.

The central presumption of the project is that the ability to extract statistical regularities – either from letter-based stimuli, or in general – is associated to reading and spelling, as well as their development. We are aimed to confirm such an association, and examine how statistical learning and distributional knowledge contributes to reading and spelling skills. In Subproject 1 we will standardize measures of reading and spelling. The central research questions are how the development of reading and spelling can reliable be measured in group- or individual assessment sessions. Subproject 2 will unravel whether and how deficits in statistical learning explain reading and/or spelling deficits. We will recruit children with 1) isolated reading deficit, 2) isolated spelling deficit 3) combined reading and spelling deficits, and compare their performance to typically developing peers on domain-general and reading-related statistical learning tasks. The central questions are whether and how the three deficit groups differ from each other; and whether and how domain-general and domain-specific statistical learning mechanisms differ in explanatory power in predicting reading and spelling deficits. The central research questions of Subproject 3 are whether and how domain-general and domain-specific statistical learning abilities explain reading and spelling skills in groups with different reading experience. Subproject 4 involves EEG experiments to clarify how distributional differences in orthographic patterns are represented in the human mind; how different patterns are processed; and whether and how the time course of processing differs between children and adults.

The current project explores how extracting and learning statistical regularities in general and from letter-based stimuli contribute to reading and spelling as well as their development. The speciality of the project is that it uses a diverse methodology, incorporating neuropsychological research on developmental disorders, developmental research, an experimental approach as well as neuroscience methodology. While the role of statistical learning in reading and reading disorders is already a hot topic, the current project advances research from two perspectives: on the one hand, statistical learning is not assessed with a single measurement, but with a wide variety of methods including tasks focusing specifically on letter-based stimuli. This approach is in accordance with the current views of statistical learning, and gives the opportunity to have a more fine-grained analysis of the association between learning and reading and spelling. On the other hand, literacy is also approached in a more detailed manner. We will focus on both reading and spelling skills, both in the case of unselected participants, as well as in the case of children with developmental deficits. Although this view is not new, it has not yet been applied to the current research domain.
Literacy is among the most important academic skills. It is essential that all members of western societies develop expertise in reading. This development is supported by compulsory primary education. Identifying how statistical learning contributes to reading and spelling skills can open way to further applied research in how reading instructions can be advanced with the inclusion of statistical learning perspectives. Furthermore, the current project can help developing individualized remediation techniques for individuals with reading and spelling deficits. Such a remediation is essential to support individuals with developmental deficits in becoming fully integrated in the literate society. The new methods standardized during the project will also prove helpful for researchers, teachers and clinicians.

The environment surrounding us is full of repeating patterns. Many of these regularities are simply picked up from the environment incidentally. Such regularities are not exploited in institutional reading instruction. In reading instructions, children are generally taught letters and their corresponding speech sounds, as well as to read whole words. The appearance of word is not random though. Some words are more frequent than others. Some words appear in certain environments more frequently. This also applies to letter patterns. Certain letter patterns are more frequent than others – even though they have no independent meaning. The current project is aimed at examining how knowledge on distributional regularities, and the ability to learn these statistics can contribute to reading and spelling. To this end, we will develop the most important research methods, examine typical readers as well as children who struggle with reading and spelling. Finally, we will use neuroscience methods to unravel the biological foundations of the link between distributional information and literacy skills. The successful completion of the project is expected to facilitate the development of reading instructions, as well as the development of remediation methods for children with reading and spelling difficulties.