Reading Research Group
The reading research group develops theoretical and computational models of reading and spelling. It uses these models as the basis for investigating how children learn to read and spell, why some children have so much difficulty (developmental dyslexia and dysgraphia), and how reading and spelling can break down after brain damage (acquired dyslexia and dysgraphia). The neural basis of both normal and impaired reading can be investigated using brain-imaging techniques such as event related potentials (ERP) and magnetoencephalography (MEG). Our findings inform educational and clinical practice and the development of targeted treatment programs for reading and spelling disorders. Our researchers have developed a variety of testing programs which are provided on the MOTIf website. Tests on this site are available for clinicians and teachers, and free registration is required for access.
The Reading Research Group also has links with the ARC Centre of Excellence in Cognition and its Disorders (CCD), and in particular the CCD's Reading Program.
Current Projects - Part 1
A randomised control trial of the effect of phonics on grapheme-phoneme correspondence knowledge
In this randomised controlled trial we compared the effectiveness of pure phonics training and sight word training on grapheme-phoneme correspondence (GPC) knowledge in a group of 76 children with poor reading. We tested children’s knowledge for single- and multi-letter GPCs, as children need to learn both of these. The results indicate that phonics and sight word training was equally effective in improving total GPC knowledge (i.e., across single- and multi-letter GPCs). However, this stemmed from different patterns of effects of phonics and sight word training on the single- and multi-letter GPCs. This study adds to the very limited evidence available on the effect of phonics training on GPC knowledge and further, it illustrates the importance of including separate measures of GPC knowledge for single- and multi-letter GPCs.
A specific association between poor word reading and low academic self-esteem
The aim of this study was to investigate - for the first time - if there are specific associations between different types of poor reading (phonological decoding, sight word reading, reading comprehension) and different types of poor self-esteem (academic, general, social, parent-home) when potentially confounding concomitant deficits (spoken language, attention, intelligence) are taken into account. We found that poor phonological decoding and poor sight word reading were specifically associated with poor academic self-esteem, and that a large minority (one-quarter to one-third) of children with poor phonological decoding or poor sight word reading had poor academic self-esteem. We suggest that this minority might be protected from, or treated for, poor academic self-esteem using resilience training.
A Test of Everyday Reading Comprehension (TERC)
The aim of this study was to develop a standardised test of everyday reading comprehension for children in Australian schools. To this end, we developed two forms for the Test of Everyday Reading Comprehension (TERC) and developed norms for children aged 6 to 12 and in grades 1 to 6. Measures of parallel form reliability indicated that the two TERC forms were equal in difficulty and measured a common concept. Measures of inter-rater reliability indicated that scoring the two TERC forms was simple enough to minimise differences between testers. Ideally, the TERC should be used: (1) as a screening test for poor reading comprehension in primary-school children, and (2) as a tool to communicate to parents the impact that a child’s reading difficulty can have on their everyday life.
Attentional deficits in dyslexia
We have been examining temporal attention in dyslexia as measured using the “foreperiod” paradigm. In 2013, we extended the foreperiod effect to visual stimuli over shorter intervals to examine speed of processing. Our findings suggest that the preparation of visual attention is slower in some children with dyslexia but that auditory deficits may be confounded with dual target confusions. We have also been conducting a meta-analysis of research on the “attentional blink” paradigm and dyslexia. This has uncovered a number of variables related to the general difficulties associated with dyslexia in this paradigm. The variables strongly correlated with between-group differences were fixation duration, temporal variability of targets, and the exposure duration of stimuli. We are in the process of planning follow-up experiments to directly manipulate these variables. Finally, we have studied the phenomenon of attentional dyslexia. People with attentional dyslexia swap letters between words. For example, dark part may be read as “park dart” or “dart park”.
Cognitive preparation in dyslexia
Reading impairment (or dyslexia) has been associated with difficulties in a myriad of psychophysical tasks. Dual-target tasks such as the attentional blink paradigm (temporal deficit in dual-target reporting), consistently returns lower performance in dyslexia. Our research is exploring why this might be the case and how this relates to reading abilities and subtypes of dyslexia. There are many separable cognitive capacities that contribute to dual-target performance including speed of processing, temporal information processing, task-switching and cognitive preparation. We are currently focussing on cognitive preparation in line with potential difficulties with the endogenous control of attention.
Computational modelling of reading
This project has been devoted, in part, to programming the second version of the Dual Route Cascaded (DRC) computational model of reading, DRC 2.0. Programming has been completed and testing of the DRC 2.0 model involving new experimental studies of reading aloud and visual word recognition began in 2012 and continues. This empirical work is being carried out in three overseas laboratories. Work is also underway on a second project, aimed at developing a DRC-style computational model of reading aloud and visual word recognition in Japanese. Japanese represents a unique challenge for modelling because it uses three different writing systems, rather than the single, alphabetic, writing system used by languages that were previously modelled using the DRC computational model of reading. A DRC-style computational model of reading aloud and visual word recognition in Greek (Greek alphabet) and Russian (Cyrillic alphabet) has also been constructed. Finally, we have worked on developing a version of the DRC that learns. Until now, DRC has not explicitly modelled the process of acquiring reading skill, despite its success in offering an account of skilled reading. A new DRC model, titled "learning-DRC", or L-DRC, has been created that models reading acquisition within the DRC framework. L-DRC will be used to investigate theories of reading acquisition, and offer a demonstration of how learning of various word types such as regular words, irregular words, and homographs can be acquired by a beginning reader without constant instruction from a teacher on every item.
Cross-linguistic studies of reading
Theories of reading and learning to read have until recently largely focussed on reading of just one language, English. English is written alphabetically, but that is not the only writing system that exists in the world, and a comprehensive understanding of how people will read requires that reading in other forms of writing system be studied. Hence we are carrying out basic research on reading and spelling with syllabic writing systems (the Hindi script Devanagari; Japanese kana) and logographic writing systems (Japanese Kanji). Even for languages written alphabetically, there are three types of alphabetic writing system, so as well as studying reading and spelling in English (Roman alphabet), we are also doing work on Russian (Cyrillic alphabet) and Greek (Greek alphabet). And within languages written in the Roman alphabet, there are cross-linguistic differences: we are comparing reading in German to reading in English.
Developmental graphemic buffer dysgraphia
Graphemic Buffer Dysgraphia is a spelling difficulty attributed to dysfunction in the working-memory component of the spelling system. This spelling difficulty has been described in many acquired cases over the past two decades. However, to date, only one description of its developmental form exists: Yachini and Friedmann (2010) provide a brief report of 10 Hebrew-speaking children. We are currently working with a case that we believe to be the first case of developmental graphemic buffer dysgraphia in English. We conducted extensive testing with LS and compared his performance to a small control group. LS is a 10 year old English-speaking boy who presented with spelling problems that could not be accounted for by difficulties with lexical or nonlexical processing. Rather, the majority of his errors consisted of letter substitutions (enough as emough), deletions (enough as enugh), additions (ennough), transpositions (eonugh), or a combination of these errors (eomugh). Moreover, LS showed similar errors in different output modalities (written, oral and typing), as well as in different input modalities (spelling to dictation and written naming). He also demonstrated greater difficulty when spelling longer compared to shorter words. LS's error pattern is very similar to that seen in acquired cases of Graphemic Buffer Dysgraphia.
Does increased letter spacing help all subtypes of poor readers?
There are several research studies that suggest that increasing the spacing between l e t t e r s can help poor readers increase reading accuracy and speed. In this research study, we want to find out if these spacing manipulations improve reading performance both in single words and texts. We are also investigated if children with certain subtypes of dyslexia tend to benefit more from increased spacing than others. As part of the Sydney Kids (SKIDS) Holiday Research Project (run by Dr Nicholas Badcock), we collected data from 117 normal and dyslexic readers.
Fractionating non-lexical reading processes: An investigation of grapheme parsing and grapheme-phoneme knowledge
Non-lexical reading is the ability to read words via grapheme parsing, grapheme-phoneme correspondence (GPC) knowledge, and phoneme blending (e.g., Coltheart, 1985, 2006). The aim of this research was to disentangle the relationship between GPC knowledge (measured by the ability to read graphemes in isolation) and grapheme parsing (measured by the ability to identify grapheme boundaries in nonwords). We used a multiple single-case study approach with children who presented with a specific problem with non-lexical reading (i.e., phonological dyslexia). We found a dissociation between GPC knowledge and grapheme parsing: Unknown GPCs could be correctly parsed, and known GPCs could be incorrectly parsed. These findings suggest that grapheme parsing and GPC knowledge may be acquired independently and that there is an editing mechanism - a re-parse mechanism - involved in non-lexical reading.
Generalisation effects in treatment of acquired language disorders Trudy Krajenbrink, Saskia Kohnen And Lyndsey Nickels
Brain damage can cause people to have difficulty spelling. Treatment for these acquired spelling impairments usually improves the words that have been practiced in treatment, but sometimes untreated words also improve. Surprisingly little is known about this generalisation process: Do treated and untreated words need to be related? Does the individual’s underlying impairment play a role? Can we predict when generalisation will occur? My PhD aims to answer these questions by conducting treatment studies testing different theories of generalisation. A better understanding of generalisation will improve our comprehension of the spelling process and will help make treatments more effective.
Getting to grips with the heterogeneity of developmental dyslexia
The aim of this study was to better understand the heterogeneity of developmental dyslexia by identifying the most common deficits in the reading systems of children with dyslexia with different poor word reading profiles. We classified the word reading profiles of 138 children with developmental dyslexia using nonword and irregular-word reading tests, and then used independent experimental tests to explore the cognitive deficits within their word reading systems. The most common deficit associated with primary sublexical impairment (i.e., poor nonword reading) was poor GPC knowledge. The most common deficits associated with primary lexical impairment (i.e., poor irregular word reading) were an impaired orthographic lexicon plus impaired links between this lexicon and the phonological lexicon and semantic knowledge. Finally, the most common deficits associated with mixed reading impairment (i.e., poor nonword reading and poor irregular word reading) were poor GPC knowledge, an impaired orthographic lexicon, poor links between this lexicon and the phonological lexicon and semantic knowledge, and poor phonological output.
Investigating the relationship between developmental dyslexia, verbal short-term memory and verbal working memory
Children with developmental dyslexia may suffer from concurrent verbal short-term memory and verbal working memory deficits. That is, these children have limited ability to store verbal information over a short period (short-term memory) or to store and manipulate information, or store information in the face of interference (working memory). Deficits in these skills have been proposed as a causal factor in poor reading, and thus the aim of this research project is to investigate the nature of these deficits in developmental dyslexia. The project attempts to achieve this by conducting: (1) a group study investigating the direct links between reading and reading-realted skills with verbal memory, (2) a randomised control trial of the effect of reading training on verbal memory, and (3) a case series investigating the effect of verbal memory training on reading in poor readers with verbal memory deficits.
Learning to spell the sounds of English
Proficient spellers make use of multiple sources of knowledge. They know how to translate sounds into letters, have detailed long-term memories of word spellings, they know spelling rules, and they can rely on their knowledge of meanings and origins of words. Learning the translations between sounds and letters is one of the most fundamental spelling skills to acquire. However, this is not a trivial skill set! There are many sounds in English and quite a few of these sounds can be spelled in more than one way. Just think of the different spellings of /ee/ in green, stream, and Pete, or /ai/ in aim, lame, and ray. In this project we asked two main questions: When do children master these different sound-letter mappings? And what are the most common errors they make? To answer these questions, we have analysed over 55,000 spelling responses from 750 children in Grades 1-7. We are expecting the results of this study to be finalised mid-2015. The results will be helpful for clinicians and special educators in deciding when to teach which sounds and the order in which they should be targeted in a training program.
Letter position dyslexia and the acquisition of letter position coding skills in developing readers
While much research has investigated how skilled adult readers code letter position within words (to distinguish between anagrams such as slime and smile), less is known about how these skills develop in children learning to read. Furthermore, even less is known about the deficit underlying children with Letter Position Dyslexia (LPD), who make excessive migration errors (reading slime as ‘smile’) in the absence of any other obvious reading or spoken language difficulty. Using single-case and group study designs, this research project investigates how letter position coding skills typically develop in children learning to read, and attempts to identify the specific impairment underlying children with letter position dyslexia. The findings are considered in light of current theories and models of letter position coding and reading development.
Letter processing in typical reading and dyslexia
This project is focussed on early letter and character string processing in dyslexia, with an eye towards developing and implementing a computational model of different dyslexias. While skilled readers can effortlessly identify letters and order them, letter-position dyslexics find the task very difficult. For example, they might read ‘wrap’ and ‘pirates’ as ‘warp’ and ‘parties’. Letter-by-letter reading is another form of dyslexia that is characterised by a strong sensitivity to letter-confusability, which results in very slow "one letter at a time" reading of words (which skilled readers do not do). Results like these show that letter identification and letter-position coding are not bullet-proof. In the first part of this project, we conducted experiments designed to test whether or not visual similarity between two adjacent letters influences processing. For example, is LCOAL easier to identify as local than LOACL (where the OC pair is visually similar, but the CA pair is not)? In another project, we have explored the question of whether increasing letter spacing improves reading in children with dyslexia. There are several research studies suggesting that increasing the spacing between l e t t e r s can help poor readers increase reading accuracy and speed. In this research study, we want to find out if these spacing manipulations improve reading performance both in single words and texts.
Masked priming is the facilitation in responding to a target word that is preceded by a related prime, where the prime is presented so briefly that they are not consciously available to the reader. For example, one type of masked priming is repetition priming: you are faster in responding to the target 'WINDOW' if it is preceded by the prime 'window' even though you are not consciously aware of the prime. Contrary to the commonly held view that masked priming effects reflect automatic activation of the target by the prime, we now know that masked priming effects are highly task-dependent. This task-dependent nature is explained by a theoretical framework which regards perception as a Bayesian decision making process and that when a prime is masked and veiled from conscious awareness, people treat the prime and target as a single perceptual object.
Neurophysiological indicators of dyslexia
This research program focuses on neurophysiological indicators of reading impairment. There seem to be subtle differences in the way the brains of people with reading difficulties respond to sounds. We use electroencephalography (EEG) to examine the electrical response the brain emits when it detects a sound. This response, the 'event-related potential' or ERP, shows a slightly different pattern in reading impairment. This has the potential to contribute to the early detection of reading difficulties. One component of this research has been to validate a wireless gaming EEG system for research. Our research has been promising and using this technology we will be able to examine EEG and ERP activity in a much broader population to better characterise the brain responses in question.
Current Projects - Part 2
Order mttaers: A computational model of letter position coding
In order to recognize a printed word, the reading system first has to identify the word's letters and arrange them in the correct order. The last ten years have seen an explosion of research into how the reading system coordinates the positioning of letters. The goal of this project is to understand why we can recognize mttaers as a distortion of matters and where the limits of this flexibility are. Understanding how these processes operate in skilled readers is critical to better understanding how they fail in dyslexias such as letter-position dyslexia. While skilled readers can effortlessly identify letters and order them, letter-position coding is not a trivial problem. Friedmann and Gvion (2001) reported results from several dyslexic adults who had no trouble with letter identification, but struggled to order letters correctly (Friedmann and Rahamim, 2007, also reported similar problems in a group of dyslexic children). These letter-position dyslexics find anagram words especially difficult. For example, letter-position dyslexics might read wrap and pirates as warp and parties. Letter-by-letter reading (Fiset et al., 2005) is another form of dyslexia that is characterized by a strong sensitivity to letter-confusability, which results in very slow "one letter at a time" reading of words (which skilled readers do not do). Results like these show that letter identification and letter-position coding is not bullet-proof. In the last ten years, several computational models have attempted to explain letter-position-coding, in particular the flexibility of the coding system (Davis, 1999; 2010; Whitney, 2001; Kinoshita & Norris, in press). Though they vary in their approach, the models rely on abstract mathematical descriptions of "information" and "variance" to explain how letter-position is coded (and where the flexibility arises). The explicit nature of this information (what parts of the visual input are used) is underspecified, as are the processes that result in ambiguity about the information (i.e., what makes the information gathering noisy). This research seeks to develop a computational model of visual word recognition that incorporates early visual processing. Such a model will make explicit the nature of the information used by readers to encode letter identities and positions. The long-term goal is to have a model that can account for dyslexic performance.
Order of acquisition of grapheme-phoneme correspondences in children
Previous research has shown that the acquisition of grapheme-phoneme correspondence (GPC) knowledge for single-letter graphemes is influenced by children’s knowledge of letter names, which often comprise the relevant phoneme correspondence (e.g., T contains the phoneme “t”; McBride-Chang, 1999; Treiman, Tincoff, Rodriguez, Mouzaki, & Francis, 1998). Beyond letter name knowledge, very little is known about the factors that influence children’s acquisition of single-letter or multi-letter GPCs. In this study, we assess children in Kindergarten to Grade 3 for their knowledge of single- and multi-letter GPCs in order to address two aims. First, we investigated the order of acquisition of single- and multi-letter GPCs in typical early readers. And second, we explored the influence of four factors - grapheme status, phoneme status, GPC frequency, and GPC consistency - on the order in which children acquire GPC knowledge. The results of this study suggest that while Australian children with less than one year of reading instruction have learned many single-letter GPCs, they have not yet learned many multi-letter GPCs by Grade 3. Further, grapheme status, phoneme status, and GPC frequency – but not GPC consistency - appear to influence the acquisition of GPC knowledge in early readers.
Orthographic learning in children
When encountering a written word for the first time, such as VOLDEMORT, a child needs to first map each letter on to its respective sound (i.e., phonological decoding), which subsequently establishes an association between the spoken word /voldemort/ and its written form VOLDEMORT. Within a few exposures, the child can then automatically recognise the word without having to decode it again. This transition from laboriously sounding out a novel word to automatically recognising it is referred to as the process of learning to read by sight. This is an important skill that a child must develop to become a proficient reader. This project seeks to better understand this process of learning to read by sight. We have systematically examined this learning process in both typically-developing and impaired readers, and explored what language and cognitive factors modulate success in this process. More specifically, this project has examined the influence of phonological skills, contextual information, vocabulary knowledge, and paired-associate learning ability on sight word acquisition in both typically developing children and children with dyslexia. More recently, we have been working on projects including the investigation of sight word learning in children with cochlear implants and children with dyslexia, sight word learning and its consolidation over sleep, rapid and automatic word recognition in typically-developing children and children with dyslexia.
Orthographic learning in dyslexia and the hearing-impaired
This ongoing research project explores how children acquire representations of individual written words, and how this may be impaired in children with different developmental disorders. Although it is known that learning alphabetic decoding skills is a key foundation of learning to read, children must ultimately move beyond this to recognising individual words fluently and automatically. Much less is known about how this advanced stage of word reading is acquired. In a new project on this theme, we have begun to explore orthographic learning in two groups of children who often experience phonological problems: namely, profoundly deaf children with Cochlear Implants and children with phonological dyslexia. A special focus is to find differences and similarities between the two groups in: 1) orthographic learning and orthographic representations, 2) the cognitive skills that affect orthographic learning and reading in general, and 3) the possible effects that semantic context may have on orthographic learning in these children. The performance of children with dyslexia and children with Cochlear Implants (CI) will also be compared with that of peers with normal hearing and typical reading development. This project is in collaboration with the National Acoustic Laboratories (NAL).
Orthographic learning in the brain: New insights from fixation related potentials
This research investigated the neural activity elicited by unfamiliar words that are repeatedly read in a naturalistic context. We used fixation-related potentials (FRPs) to measure brain responses to familiar and unfamiliar words that were presented four times in paragraphs of text. The FRPs indicated that when read for the first time, familiar words produced a stronger positive peak than unfamiliar words over the frontal-right scalp region at approximately 150 ms (P150), whereas unfamiliar words elicited a larger positive peak than familiar words over the right parietal region at approximately 300 ms (P300). The early P150 effect was only observed for the first encounter of the words in the text. In contrast, the P300 effect was observed at the first and second encounters. These patterns suggest an interplay between perceptual and attentional demands relating to the initial stages of integrating new orthographic forms into the lexicon.
Phonics training for English-speaking poor readers
Around 5% of English speakers have a significant problem with learning to read words. Poor word readers are often trained to use letter-sound rules to improve their reading skills. This training is commonly called phonics. Well over 100 studies have administered some form of phonics training to poor word readers. However, there are surprisingly few systematic reviews or meta-analyses of these studies. The primary aim of this Cochrane Review was to measure the effect that phonics training has on the literacy skills of English-speaking children, adolescents, and adults whose reading was at least one standard deviation (SD), one year, or one grade below the expected level, despite no reported problems that could explain their impaired ability to learn to read. A secondary objective was to explore the impact of various factors, such as length of training or training group size that might moderate the effect of phonics training on poor word reading skills. The systematic review and analysis revealed that phonics training appears to be effective for improving some reading skills. Specifically, statistically significant effects were found for nonword reading accuracy (large effect), word reading accuracy (moderate effect), and letter-sound knowledge (small-to-moderate effect). For several other outcomes, there were small or moderate effect sizes that did not reach statistical significance but may be meaningful: word reading fluency, spelling, phonological output, and reading comprehension. The effect for nonword reading fluency, which was measured in only one study, was in a negative direction, but this was not statistically significant. Future studies of phonics training need to improve the reporting of procedures used for random sequence generation, allocation concealment, and blinding of participants, personnel, and outcome assessment.
Phonological processing deficits in specific reading disability and specific language impairment: Same or different?
The aim of the current study was to understand if phonological processing deficits in specific reading disability (SRD) and specific language impairment (SLI) are the same or different. To this end, the current study compared more specifically defined groups of SRD and SLI children for the type and severity of their phonological processing deficits. In four separate analyses, a different combination of reading (irregular word reading and nonword reading) and spoken language measures (receptive language and receptive-expressive language) were used to divide 73 children into three subgroups: poor readers with average spoken language (SRD), poor readers with poor spoken language (SRD+SLI), and average readers with poor spoken language (SLI). In each analysis, the three groups were compared for their neural representations for phonemes, phoneme discrimination, phoneme awareness, rapid naming, and phonological short-term memory. Children with SRD showed moderate to large deficits in their neural representations of phonemes, their phoneme discrimination, their phoneme awareness, and their rapid naming, which were worse when SRD was defined as poor letter-sound reading than whole-word reading. Children with SRD+SLI showed more severe deficits than the SRD group for some phonological processing measures but not others, and showed a unique additional deficit in phonological short-term memory. Interestingly, children with SLI were free from phonological processing deficits no matter how SLI was defined. These data demonstrate that whether or not phonological processing is the same or different in SRD and SLI depends upon an interaction between how SRD and SLI is defined and how phonological processing is measured.
Profiling working memory deficits in children with developmental dyslexia
The aim of the present study is to profile verbal short-term and working memory deficits against a dual route model of dyslexia. Sixty-three children with developmental dyslexia (DD), and 11 controls, were compared on the forward and backward digit span of the WISC-IV and various components of the dual route model of reading (e.g. irregular and nonword reading accuracy and fluency, comprehension, letter-sound knowledge, blending). Children were classified as DD without memory problems, DD with a short-term memory only problem, DD with a working memory only problem, or DD with both short-term and working memory deficits. The results suggest that the part of reading that is most affected by memory difficulties is in the early letter-learning stage. Accounting for memory deficits during this stage of learning may assist beginning readers and should be investigated.
One of the most persistent difficulties for poor readers is the ability to read both accurately and quickly. This combination of speed and accuracy is called reading fluency. After years of intensive training, poor readers usually become accurate readers but still decode text slowly. Such poor fluency is thought to impair the ability to comprehend written texts and hence hinder the ultimate goal of reading: gaining new knowledge.
The aim of this line of research is twofold. First, we want to discover what underlies the development of reading fluency and second we want to develop effective treatment or strategies for children that are struggling to learn to read fluently. We are currently conducting two concrete research projects in this area:
1. A reading fluency training study that tests if word reading speed, text reading speed, and reading comprehension of children with reading fluency problems can be improved with the aid of a syllable training program;
2. A study investigating the influence of text presentation (in terms of letter font and letter spacing) on the reading fluency of normally developing children and children with dyslexia. This project is conducted in collaboration with MultiLit.
Socio-emotional health in dyslexia
Our research explores why individuals with reading difficulties have negative mental health outcomes. In collaboration with Dr Mark Boyes (Curtin University), we are focussing on coping and resilience as protective factors with a future goal of developing interventions targeting these abilities with respect to reading difficulties.
Testing spelling skills
Skilled spellers make use of multiple sources of knowledge including: (1) how to translate sounds into letters, (2) long-term memories for whole words, (3) morphological knowledge. Over recent years, we have developed and normed tests assessing these skills. We are in the process of analysing the norm data and will make the tests available on the online test interface www.motif.org.au to allow teachers and clinicians more targeted assessment and ultimately treatment of spelling difficulties.
The effect of verbal working memory training on short-term and working memory in children with dyslexia
The present study aims to assess the efficacy of verbal working memory training on memory and reading skills in children with dyslexia. Five children with dyslexia are currently participating in a within-subjects double-baseline study. Children were included if their sight-word and/or nonword reading fell one standard deviation (SD) below the mean for their age on the Castles & Coltheart 2, and if either their composite scores, or 2 or more of the subscales, fell one SD below the mean for their age on the verbal working memory scale of the Automated Working Memory Assessment (AWMA). Children receive 8 weeks of verbal working memory training, and 8 weeks of reading training, with an 8-week consolidation period in between. Participants are assessed at baseline, 8 weeks later, and after each training block, and after the consolidation period. In 2014, results will be reported for experimental (trained and untrained) verbal and visuospatial short-term memory and working memory tasks. Results will also be reported for measures of reading accuracy and fluency, attention, and classroom-based memory skills. The results of this study have the potential to increase the efficacy of current training programs for both working memory and reading deficits, and guide best practice decisions for these children.
The Letter-Sound Test (LeST) with norms for children in Kindergarten to Grade 3
A core component of decoding instruction is teaching grapheme-phoneme correspondences (GPCs; also known as letter-sound mappings). The aims of this study were to: (1) develop the first comprehensive standardised test of GPC knowledge in English, and (2) test the hypothesis that children have learned the most common GPC rules after just four years of reading instruction (i.e., by Grade 3 in Australian schools). The Letter-Sound Test (LeST) was developed and provides grade-based normative data based on an Australian sample (n=337). The data challenge the claim that children have acquired the most common GPC rules by Grade 3.
The links between oral language skills and reading comprehension difficulties
In this project, the links between oral language skills and reading comprehension were explored through a randomized controlled trial. Children with specific reading comprehension difficulties received an 8-week program of oral vocabulary training. We found that, immediately after training, gains in oral vocabulary led to gains on an experimental reading comprehension test. At follow-up (8 weeks after the completion of training), the scores of the trained group had continued to improve on the experimental reading comprehension test and had also improved on a test of listening comprehension. The results support a link between oral vocabulary skills and reading comprehension. Further investigation is underway looking at methods of assessing reading comprehension and exploring individual variation within the language skills of poor comprehenders.
The nature of the speech output system as revealed by studies of picture naming, colour naming and reading aloud
This project focuses on what we can learn about the nature of the speech output system using a variety of techniques including picture naming (especially picture-word interference effects), colour naming (especially the phonological Stroop effect), and reading aloud (especially the masked onset priming effect, coarticulation effects in nonwords reading, and word-word interference effects). We are especially interested in how phonemes are positionally encoded as part of the speech production process.
The Reading Training Study: Phase 1
The aims of this study were to: (1) compare sight word training and phonics training in children with dyslexia, and (2) determine if different orders of sight word and phonics training have different effects on the reading skills of children with dyslexia. One group of children (n=36) did 8 weeks of phonics training (reading via grapheme–phoneme correspondence rules) and then 8 weeks of sight word training (reading irregular words as a whole), one group did the reverse (n=36), and one group did phonics and sight word training simultaneously for two 8-week periods (n=32). We measured the effects of phonics and sight word training on sight word reading (trained irregular word reading accuracy, untrained irregular word reading accuracy), phonics reading (nonword reading accuracy, nonword reading fluency), and general reading (word reading fluency, reading comprehension). Sight word training led to significant gains in sight word reading measures that were larger than gains made from phonics training, phonics training led to statistically significant gains in a phonics reading measure that were larger than gains made from sight word training, and both types of training led to significant gains in general reading that were similar in size. Training phonics before sight words had a slight advantage over the reverse order.
The Reading Training Study: Phase 2
The aim of Phase 2 of the Reading Training Study (RTS) is to test the reliability of the promising outcomes of Phase 1 of RTS using a new state-of-the-art online LiteracyPlanet platform (www.literacyplanet.com). Around 120 children with dyslexia have been allocated to one of two training groups. One group has done 8 weeks of phonics training (i.e., reading with the letter-sound rules) followed by 8 weeks of sight-word training (i.e., reading whole words by sight). A second group has done the same training in the reverse order. All children have been tested for their reading accuracy, fluency and comprehension before and after each type of training. They have also been tested for each component of their word reading system as defined by the Dual Route Model of Reading. Testing was completed in November 2013. The results will be analysed in 2014.
The Spelling Training Study
In the first phase of this training study we recruited 14 primary school aged children with developmental spelling difficulties. These children were ‘phonetic spellers’ and tended to spell words the way they sound: for example, ‘said’ as ‘sed’. We tested children twice before the training and once after the training. All participants showed significant improvements in spelling their trained words. Eight of the children also showed generalisation for a small number of words that were not specifically trained. These words tended to have the following characteristics: they shared letters with the trained words (for example, tough and though), they occurred more often in written texts, and they were shorter. These results show that sight word training can improve spelling beyond just the words that are the focus of training. Unfortunately, spelling knowledge gained as a result of an intervention is often not perfectly maintained over time. We are currently investigating, in a second training study, if minimal revision can alleviate some of this loss.
Research Group Leader
- Distinguished Professor Anne Castles
- Dr Nicholas Badcock
- Distinguished Professor Anne Castles
- Emeritus Professor Max Coltheart
- Dr Saskia Kohnen
- Dr Eva Marinus
- Associate Professor Genevieve McArthur
- Professor Lyndsey Nickels
- Dr Hua-Chen Wang
- Dr Teresa Shubert
- Dr Robin Litt
- Dr Danielle Colenbrander
- Yvette Kezilas
- Trudy Krajenbrink
- Yu Li
- Signy Wegener
- Sana Haidry
Current Administrative Staff
- Dr Erin Banales
Current External Associates
- Associate Professor Sachiko Kinoshita