Search

Recent studies in psychology have documented how analytic flexibility can result in different results from the same data set. Here, we demonstrate a package in the R programming language, DeclareDesign, that uses simulated data to diagnose the ways in which different analytic designs can give different outcomes. To illustrate features of the package, we contrast two analyses of a randomized controlled trial (RCT) of GraphoGame, an intervention to help children learn to read. The initial analysis found no evidence that the intervention was effective, but a subsequent reanalysis concluded that GraphoGame significantly improved children’s reading. With DeclareDesign, we can simulate data in which the truth is known and thus can identify which analysis is optimal for estimating the intervention effect using “diagnosands,” including bias, precision, and power. The simulations showed that the original analysis accurately estimated intervention effects, whereas selection of a subset of data in the reanalysis introduced substantial bias, overestimating the effect sizes. This problem was exacerbated by inclusion of multiple outcome measures in the reanalysis. Much has been written about the dangers of performing reanalyses of data from RCTs that violate the random assignment of participants to conditions; simulated data make this message clear and quantify the extent to which such practices introduce bias. The simulations confirm the original conclusion that the intervention has no benefit over “business as usual.” In this tutorial, we demonstrate several features of DeclareDesign, which can simulate observational and experimental research designs, allowing researchers to make principled decisions about which analysis to prefer.

The article discusses a study which examined whether working memory training programs can improve a range of different learning disorders which include attention deficit hyperactivity disorder (ADHD), reading and language and mathematics disorders. The study showed that the mean effect size from working memory training on visuo-spatial working memory was moderate. It also reveals the effectivity of the conventional forms of treatment involving the direct training of reading and language skills.

Abstract: In free recall tasks, when low- and high-frequency items are mixed within the to-be-remembered lists, the usual recall advantage found for high-frequency words is eliminated or reversed. Recently, this mixedlist paradox has also been demonstrated for short-term serial recall (Hulme, Stuart, Brown, & Morin, 2003). Although a number of theoretical interpretations of this mixed-list paradox have been proposed, researchers have also suggested that it could simply be a result of participant-controlled strategies (M. J. Watkins, LeCompte, & Kim, 2000). The present study was designed to assess whether this explanation could be applied to immediate and delayed serial recall. The results showed that high-frequency words were recalled better than low-frequency words in pure lists, but that this effect was eliminated in mixed lists, whether they were given under intentional or incidental learning conditions. This pattern suggests that the mixed-list paradox cannot be explained by participant-controlled strategies.

The present experiments investigated the effects of articulatory suppression on reading numbers written alphabetically (one, two, etc.) or as Arabic numerals (1, 2, etc.). For alphabetic numbers, but not for numerals, a phonological representation can be generated and used for lexical access. If such phonological encoding occurs, and if it is harmed by irrelevant articulatory activity, then articulatory suppression should have a greater effect on the reading of alphabetic than on numeral representations of numbers. In contrast to a finding reported by Baron (1977), involving alphabetic and Roman numerals, in two experiments suppression was found to have essentially equivalent effects on the reading of both types of numbers. A third experiment explored a possible explanation for these discrepant findings. It is concluded that although alphabetic and numeral representations of numbers may be processed differently there is no evidence to support the idea that reading alphabetic numbers involves phonological encoding prior to lexical access.

The present experiments address two interrelated problems; the causes of reading retardation and the possible mechanisms underlying multi-sensory teaching procedures, which involve manually tracing around words, and which reputedly help children retarded in reading. Two experiments explored the effects of manual tracing on memory for letters and non-verbal forms in normal and retarded readers. The retarded readers remembered fewer letters and gained selective benefit from tracing them. In the case of non-verbal forms the two groups performed equally and tracing was equally beneficial to memory in both groups. These findings were explained in terms of the retarded readers' limited reliance on a phonological memory code. A further experiment showed that the differential effect of tracing on the retarded readers' memory for letters was not simply a consequence of their limited reading ability. It was concluded that reading retardation is characterised by deficits of verbal, but not of visual, memory. The tracing activity involved in multi-sensory teaching may help retarded readers by providing a mnemonic aid, which compensates for their verbal memory difficulties.

This paper presents a follow-up study of JM, a developmental dyslexic first described by Snowling, Stackhouse, and Rack (1986b). Tests of phonological processing revealed that JM's deficit was primarily located at the level of output phonology (speech production). Further studies explored the mechanisms by which JM has managed to learn to read in the face of these deficits. Tests of nonword reading supported the view that JM has learnt to read by gradually expanding his sight vocabulary, supported by his intact visual and semantic processing skills.

We present a four-year follow-up study of JM, a developmental phonological dyslexic previously described by Snowling, Stackhouse, and Rack (1986). JM has made some progress in reading and spelling although these skills remain severely impaired. His reading and spelling skills are consistent with arrest in the logographic phase of literacy development as described by Frith (1985); reading errors are primarily visual, he is severely impaired in reading non-words, and his spelling errors are predominantly dysphonetic. In contrast to these severe decoding problems, his semantic and syntactic skills are excellent, and reading comprehension is consistently better than accuracy. He shows a range of deficits in phonological skills. He has severe impairment of output phonology, as evidenced by difficulties in repeating words and nonwords and systematic mispronunciations in his spontaneous speech. Naming skills are impaired and he has severely impaired verbal short-term memory skills. It is argued that JM's reading and spelling problems may plausibly be linked to his underlying phonological problems, particularly at the level of output phonology.

Lovegrove, Martin, and Slaghuis (1986) have argued that low-level visual deficits may commonly be a cause of reading retardation in children. It is argued here that these deficits do not provide a plausible explanation for the sorts of reading problems seen in children retarded in reading. In the minority of cases of reading retardation that do seem to reflect an underlying visual deficit, the problem would appear to depend upon higher-level visual processes.

The development of digit span was studied in relation to mental age in two large groups of severely subnormal (SSN) individuals, one diagnosed as suffering from Down's syndrome and the other of mixed aetiology, and a group of normal subjects matched for mental age. Digit span in the Down's syndrome and SSN mixed aetiology groups was generally lower than that expected on the basis of mental age. A cross-sectional and longitudinal study showed a failure of development of digit span in both SSN groups which resulted in an increasing lag between digit span and mental age as mental age increased. Some possible explanations for the lack of short-term memory development in Down's syndrome and other SSN subjects are discussed, together with the possible implications of this for other aspects of cognitive development.

We present a study of the accuracy, consistency, and speed of word naming in a dyslexic boy, JM, who has severe impairments in the ability to use sub-lexical, phonological reading strategies. For words that he can recognise, JM's naming latencies do not differ from those of control subjects matched for reading age, and he is generally consistent from one occasion to the next. He can also match printed homophones with their definitions–-a skill that requires access to well-specified orthographic representations. The data are interpreted as evidence for the creation of efficient recognition devices for words within JM's sight vocabulary. However, he shows a continuing inability to use phonological decoding strategies to deal with words that he cannot recognize by sight. Overall we argue our results pose problems for stage models of reading development, and that they may best be interpreted within a connectionist framework of the development of word recognition skills.

This paper reports two studies investigating the nature of comprehension deficits in a group of 7–8 year old children whose decoding skills are normal, but whose reading comprehension skills are poor. The performance of these poor comprehenders was compared to two control groups, Chronological-Age controls and Comprehension-Age controls. The first study examined whether these comprehension difficulties are specific to reading. On two measures of listening comprehension the poor comprehenders were found to perform at a significantly lower level than Chronological-Age controls. However, they did not differ from a group of younger children matched for reading comprehension skills. This indicates that the observed comprehension difficulties are not restricted to reading, but rather represent a general comprehension limitation. The second study investigated whether these comprehension difficulties can be explained in terms of a memory deficit. The short-term and working memory skills of these three groups were examined. The poor comprehenders did not differ from their Chronological-Age controls on either of these tasks. In conclusion, it is argued that working memory processes are not a major causal factor in the creation of the comprehension difficulties identified in the present group of poor comprehenders.

This paper explores the relationship between speech and spelling in a single-case study of developmental dyslexia. JM, a developmental dyslexic with a well-documented history of speech, reading and spelling difficulties, was examined when he was 13–14 years old. He still had subtle articulation difficulties causing some disfluency and his use of phonetic voicing was atypical. We argue that these difficulties were recapitulated in his spelling where he was more sensitive to the prosodic aspects of words than normal spellers, exhibiting a strong tendency to spell accurately words which are stressed on the first, rather than the second syllable. He also had more difficulty with phonetic voicing and spelling errors reflected this uncertainty. Thus, when word-specific (orthographic) spelling information is unavailable, JM, like all spellers, must make use of phonological spelling strategies. In his case, these are compromised because of underlying phonological speech problems. It is argued that, while young children make use of a phonological ‘frame’ on which to organize orthographic information, dyslexics, like JM, who have inadequate phonological representations, are unable to do so. This has a detrimental effect on their acquisition of spelling.

Children (ranging in age from 5 ½ to 9 ½ years) were given four tests of phonological skill. The relationships between these measures and their predictive relationship with reading and spelling ability were investigated. Performance at phonemic segmentation, rhyme sound categorisation, and alliteration sound categorisation improved with age, but all age groups performed onset‐rime segmentation at a similar level. Although phonemic segmentation was an excellent predictor of reading and spelling ability, onset‐rime segmentation was not. Rhyme and alliteration sound categorisation scores did account for statistically significant variance associated with reading and spelling ability, but they were poorer predictors than was phonemic segmentation. It is concluded that phonemic awareness is an important predictor of reading and spelling ability, even in the early stages of development. Our findings question the emphasis that has sometimes been placed on rhyming skills as predictors of reading and spelling ability. [See also letters to the editors regarding this article: http:dx.doi.org10.1598RRQ.33.3.1and http:dx.doi.org10.1598RRQ.34.3.1] SE ADMINISTRARON cuatro pruebas de habilidad fonológica a grupos de niños (de 5½ a 9½ años de edad). Se investigaron las relaciones entre estas medidas y su poder predictivo sobre las habilidades de lectura y escritura. El desempeño en segmentación fonémica, categorización de rimas y categorización de aliteraciones mejoró con la edad, pero todos los grupos etarios se desempeñaron en un nivel similar en segmentación ataque‐rima. Si bien la segmentación fonémica fue un predictor excelente de las habilidades de lectura y escritura, la segmentación ataque‐rima no lo fue. Los puntajes obtenidos en categorización de rimas y aliteraciones no explicaron la variancia estadísticamente significativa asociada con las habilidades de lectura y escritura; fueron predictores más pobres que la segmentación fonémica. Se concluye que la conciencia fonémica es un predictor importante de las habilidades de lectura y escritura, aún en las fases iniciales del desarrollo. Nuestros hallazgos cuestionan el énfasis que se ha puesto, a veces, en las habilidades de rima como predictoras de las habilidades de lectura y escritura. KINDER (im Alter von 5½ bis 9½ Jahren) bekamen vier Tests zur Feststellung der phonologischen Fähigkeiten. Die Zusammenhänge zwischen diesen Messungen und ihrer Aussagekraft hinsichtlich der Lese‐ und Buchstabierfähigkeit wurden untersucht. Die Leistungen im phonemischen Bereich, die Kategorien von Klang‐Rhythmus und Alliteration, zeigten eine zunehmende Verbesserung mit der Altersstufe, aber alle Altersgruppen zeigten sich auf ähnlichem Leistungsniveau im Bereich der Inlautung. Obgleich die phonemische Segmentation ein ausgezeichnetes Indiz zur Vorhersage von Leseund Buchstabierfähigkeiten war, so wenig war es der Bereich nach dem Anlaut. Die Werte in den Kategorien von Reim und Alliteration erklärten die statistisch signifikanten Abweichungen in Zusammenhang mit der Lese‐ und Buchstabierfähigkeit, aber sie waren schlechtere Vorhersageinstrumente als die phonemische Gliederung. Die Schlußfolgerung ist, daß phonemische Aufmerksamkeit ein wichtiger Gradmesser der Lese‐ und Buchstabierfähigkeit ist, sogar auf einer frühen Entwicklungsstufe. Unsere Ergebnisse stellen die Bedeutung in Frage, die manchmal den Fähigkeiten zur Reimbildung und deren Vorhersagekraft hinsichtlich der Lese‐ und Buchstabierfähigkeit zugesprochen wird. ON A fait passer quatre tests d'habileté phonologique à des enfants ayant de 5.6 à 9.6. On a examiné les relations entre ces mesures et leur valeur prédictive pour la lecture et l'écriture. La segmentation phonémique, la catégorisation des rimes et la catégorisation des allitérations se développent avec l'âge, mais tous les groupes d'âge ont à peu près le même niveau pour la segmentation attaque/rime. Quoique la segmentation phonémique soit un excellent prédicteur de la lecture et de l'écriture, ce n'est pas le cas de la segmentation at taque/rime. Les scores de catégorisation des rimes et des allitérations sont bien statistiquement associés avec le savoir‐lire et le savoirécrire mais sont de moins bons prédicteurs que la segmentation phonémique. On en conclut que la conscience phonémique est un important prédicteur du savoir‐lire et du savoir‐écrire, même en tout début d'apprentissage. Ces résultats mettent en question l'accent que l'on met parfois sur la compétence relative aux rimes par rapport à la lecture et à l'écriture.

Memory for series of abstract graphic forms was tested using a recognition/ reconstruction procedure in two experiments with 8–9-year-old children. The children either looked at the forms and traced around each of them in turn (VT condition), or looked at the forms for an equivalent length of time and pointed to them (V condition). Memory for the forms, but not their order was improved by tracing. The mechanism responsible for this was investigated by comparing the effects of motor and visual interference interpolated between stimulus presentation and recognition testing. Motor interference was most disruptive in the VT condition while visual interference tended to be most disruptive in the V condition. These results support the proposal that tracing produces a distinct motor memory trace and that this additional source of information aids visual recognition.

On the basis of a study of the reading skills of 56 developmental dyslexics, Castles & Coltheart (1993) have proposed that there are at least two varieties of dyslexia in children. In the first, there is an impairment of lexical reading and in the second, the operation of the sublexical procedure is impaired. We take issue with their, proposal on theoretical grounds and argue that methodologlcal weaknesses in their study seriously weaken their claims.

We report an experiment investigating the mechanisms responsible for short-term memory span and its development by examining the relationship between memory span and speech rate for words and nonwords of differing spoken lengths. Linear functions related memory span to speech rate for both words and nonwords in children of different ages. The functions for nonwords had equivalent slopes (interpreted as reflecting a contribution from a speech-based process) but lower intercepts (interpreted as reflecting a contribution from a long-term memory component) than the functions for words. Children in both age groups studied showed evidence of a relationship between speech rate and memory span and part of the difference in memory span between age groups appears to reflect a difference in speech rate. However, there is also evidence that the long-term memory component of memory span shows greater efficiency in older children.

Recent convolution-based models of human memory (e.g. Lewandowsky & Murdock, 1989), have accounted for a wide range of data. However such models require the relevant mathematical operations to be provided to the network. Connectionist models, in contrast, have generally addressed different data, and not all architectures are appropriate for modelling single-trial learning. Furthermore, they tend to exhibit catastrophic interference in multiple list learning.In this paper we compare the ability of convolution-based models and DARNET (Developmental Associative Recall NET work), to account for human memory data. DARNET is a connectionist approach to human memory in which the system gradually learns to associate vectors, in one trial, into a memory trace vector. Either of the vectors can then be retrieved. It is shown that the new associative mechanism can be used to account for a wide range of relevant experimental data as successfully as can convolution-based models with the same higher-level architectures. Limitations of the models are also addressed.

Parents and Children Together (PACT), a new parent-delivered teaching programme, has led to gains in children's early language and reading skills. Kelly Burgoyne, Rachel Gardner, Helen Whiteleyand Charles Hulmereport on the project