Arbeitsgedächtnisfunktionen von Kindern mit Minderleistungen in der Schriftsprache : Zur Dissoziation von Lese- und Rechtschreibfertigkeiten und zur Relevanz des IQ-Diskrepanzkriteriums

Abstract

Epidemiologische Studien aus dem deutschen Sprachraum zeigen, dass Lernstörungen im Lesen und im Rechtschreiben häufig nicht nur in Kombination, sondern auch isoliert voneinander vorkommen. Während bereits viele Befunde über die kognitiven Besonderheiten des kombinierten Lese- und Rechtschreibdefizits vorliegen, gilt dies nicht für isoliert auftretende Schriftsprachstörungen. Unklar ist etwa, inwieweit das Vorliegen einer IQ-Leistungs-Diskrepanz, wie sie von der Weltgesundheitsorganisation in der ICD-10 ( WHO, 2011) zur Diagnose einer Lernstörung gefordert wird, tatsächlich mit kognitiven Unterschieden im Vergleich zu schriftsprachbeeinträchtigten Kindern ohne Lernstörungsdiagnose einhergeht. Daher wurden in der vorliegenden Studie die Arbeitsgedächtnisleistungen von 142 Drittklässlern mit isoliertem Lesedefizit, isoliertem Rechtschreibdefizit oder einem kombinierten Lese-und Rechtschreibdefizit jeweils mit den Leistungen einer lernunauffälligen Kontrollgruppe verglichen. Über einen Vergleich von schriftsprachbeeinträchtigten Kindern mit und ohne Lernstörungsdiagnose wurde in einem weiteren Analyseschritt die Relevanz des IQ-Diskrepanzkriteriums überprüft. Die Befunde der Studie zeigen, dass Minderleistungen im Lesen und/oder im Rechtschreiben jeweils mit unterschiedlichen Dysfunktionen im Arbeitsgedächtnis einhergehen. Demgegenüber konnten keine umfassenden Unterschiede in den Arbeitsgedächtnisprofilen von schriftsprachbeeinträchtigten Kindern mit und ohne Lernstörungsdiagnose nachgewiesen werden. Die Befunde liefern somit keine Argumente für die Angemessenheit des IQ-Diskrepanzkriteriums, wohl aber für die Notwendigkeit, künftig stärker zwischen isolierten und kombinierten Minderleistungen im Lesen und Rechtschreiben zu unterscheiden. Diskutiert werden resultierende Implikationen für die Diagnostik von Lernstörungen der Schriftsprache.

Working Memory in Children with Reading and/or Spelling Disorder: on the Relevance of the IQ-Achievement-Discrepancy Criterion

Background & Aims: In transparent orthographies like German, isolated learning disorders in either reading or spelling are common and occur as often as a combined reading and spelling disorder ( Landerl & Moll, 2010). However, due to the widespread belief that reading and spelling are interchangeable constructs, most studies so far have not sufficiently differentiated between these two literacy skills. Thus, most issues surrounding the emergence of isolated or combined literacy disorder are yet unresolved: First, it is not clear to which extent learning disorders in reading and/or spelling are accompanied by different cognitive profiles. Although there is first evidence to suggest that they do differ in their cognitive characteristics, research so far has only focused on phonological information processing (e. g., Wimmer & Mayringer, 2002; Wimmer & Schurz, 2010). Second, another unresolved issue regards the validity of the IQ-achievement-discrepancy criterion. According to the diagnostic manuals ICD-10 (WHO, 2011) and DSM-IV (APA, 2000), a learning disorder in reading and/or spelling is only diagnosed when children exhibit a large discrepancy between their poor literacy skill(s) and their intellectual ability. However, this discrepancy criterion is under debate, because IQ-consistent and IQ-discrepant poor learners do not seem to differ in a range of cognitive factors (e. g., Stuebing et al., 2002). But since most of the research on the IQ-achievement-discrepancy criterion has been conducted in English speaking countries, it is questionable whether these findings also transfer to German orthography. This caution is warranted, because it has been argued that the manifestation of literacy disorders is not universal, but varies between orthographies (e. g., Vellutino et al., 2004).

Recently, working memory has been shown a promising construct to investigate literacy disorders as well as the validity of the IQ-achievement-discrepancy criterion. According to Baddeley (1986) working memory comprises a central executive system responsible for the monitoring of complex cognitive processes. Furthermore, it consists of two subsidiary systems, the phonological loop and the visual-spatial sketchpad, responsible for the passive storage and rehearsal of phonological-based and visual-spatial information, respectively.

The aims of the present study were twofold: First, it was investigated whether learning disorders in reading and/or spelling are associated with different working memory profiles when compared to normally achieving children. Second, it was explored whether working memory functioning differs between IQ-consistent and IQ-discrepant poor readers and/or spellers.

Methods: Based on their performance in standardized school achievement measures, children were allocated to one of the following groups: Children were identified as having an isolated reading disorder (R−S+), when they exhibited poor reading skills of T < 40, but showed at least average spelling skills of T ≥ 40. In contrast, children were identified as having an isolated spelling disorder (R+S−), when they had at least average reading skills of T ≥ 40, but poor spelling skills of T < 40. Children with combined reading and spelling disorder (R−S−) had to perform below average in both reading and spelling ( T < 40). To investigate working memory functioning related to IQ-discrepancy, half the children of each literacy group met the IQ-achievement-discrepancy criterion, whereas the other half did not due to lower intelligence. To be included in the control group children had to score at least average in both reading and spelling with T ≥ 45. Further, all children were normally achieving in mathematics ( T ≥ 40) and showed at least average nonverbal intelligence ( IQ ≥ 85).

Children completed the computerized Working Memory Test Battery for Children Aged Five to Twelve Years (AGTB 5 – 12; Hasselhorn et al., 2012) to assess working memory subcomponents: The phonological loop was assessed with five tasks, the visual-spatial sketchpad with two tasks, and the central executive with four tasks.

Results: In all statistical analyses, the three literacy groups were examined separately from each other. In a first step, working memory functioning in each group was compared to a control group of children with unimpaired literacy development. For each working memory subcomponent, separate multivariate analyses of variance (MANOVA) were conducted. Results revealed different patterns of working memory dysfunction in the three literacy groups: Isolated reading disorder was associated with little impairments in the phonological loop, but marked impairments in the central executive. In contrast, children with isolated spelling disorder only showed severe deficits in the phonological loop, but not in the other working memory subcomponents. Children with combined reading and spelling disorder exhibited severe impairments in both the phonological loop and the central executive. The latter result provides first evidence to suggest that the working memory profile associated with combined reading and spelling disorder is best described as an additive combination of the isolated deficits.

In a second step, it was investigated whether working memory functioning differs as a function of IQ-achievement-discrepancy. Since it has been hypothesized that the IQ-discrepancy is not associated with differences in working memory performance, the null hypothesis was assumed to be preserved instead of being rejected as in conventional hypothesis testing. Therefore, the alpha level was set on .10 instead of using the conventional level of .05 to increase statistical power and thereby reducing the probability that the null hypothesis might erroneously been accepted ( Type II error). With respect to children with isolated reading disorder and those with combined reading and spelling disorder, no differences in working memory performance between IQ-consistent and IQ-discrepant children emerged. However, as regards children with isolated spelling disorder, results were slightly different: Although there were no differences between IQ-consistent and IQ-discrepant poor spellers in the visual-spatial sketchpad and the central executive, this was not the case for phonological loop functioning. Here, IQ-consistent children showed lower performance than IQ-discrepant children in two measures.

Discussion: Results suggest that learning disorders in reading and/or spelling are each associated with different working memory profiles when compared to normally-achieving children. Reading and spelling should therefore not be considered as interchangeable constructs as done in most German research so far. In contrast, differentiating these two literacy skills seems necessary when investigating their causal cognitive factors.

Overall, there were no differences in working memory performance between IQ-consistent and IQ-discrepant poor learners – except for children with isolated spelling disorder. Here, IQ-consistent poor spellers showed in two out of eleven working memory measures lower performance than IQ-discrepant poor spellers. However, it is questionable, whether this little difference is sufficient to assume distinct cognitive profiles for these two groups: Given their marked difference in overall intelligence, it is surprising that their working memory functioning was that similar. To sum, these findings cast doubt on the validity of the IQ-achievement-discrepancy criterion for diagnosing learning disorders in reading and/or spelling.