Memory abilities in children with mathematical difficulties: Comorbid language difficulties matter Giselle Reimann ⁎, Janine Gut, Marie-Claire Frischknecht, Alexander Grob Faculty of Psychology, University of Basel, Switzerland abstract article info Article history: Received 22 March 2012 Received in revised form 25 September 2012 Accepted 27 October 2012 Keywords: Mathematical difficulty Language difficulty Memory abilities Selective attention Children Full-scale IQ The present study investigated cognitive abilities in children with difficulties in mathematics only (n = 48, M=8 years and 5 months), combined mathematical and language difficulty (n =27, M=8 years and 1 month) and controls (n =783, M=7 years and 11 months). Cognitive abilities were measured with seven subtests, tapping visual perception, selective attention, memory, and reasoning, as well as full-scale-IQ. Children with difficulties in mathematics only differed in their cognitive abilities, not only from controls, but also from children with comorbid language difficulties. Children with mathematical difficulties only performed worse than controls in a selective attention measure, but not in any working memory measure, meanwhile children with difficulties in mathematics and language performed worse than controls in verbal working memory com- ponents, but not selective attention. Theoretical and practical implications are discussed. © 2012 Elsevier Inc. All rights reserved. 1. Introduction Mathematical deficits are usually persistent over the first school years (Andersson, 2010; Aunio & Niemivirta, 2010; Passolunghi & Siegel, 2004), and may be even associated with earlier school dropout and unemployment (Parsons & Bynner, 2005). Most literature points to working memory (Baddeley, 2000; Baddeley & Hitch, 1974) as a pos- sible explanation for mathematical learning difficulties (e.g., Alloway, Gathercole, Kirkwood, & Elliott, 2009). The phonological loop is sup- posed to be of importance when performing calculations, as the instruc- tions have to be remembered and subresults stored (Schuchardt, Kunze, Grube, & Hasselhorn, 2006; Swanson & Jerman, 2006). The visual– spatial sketchpad can serve as a mental blackboard that supports num- ber representation (Alloway, 2006) and representation of some forms of conceptual knowledge (Geary, 2004). The central executive repre- sents the controlling component of working memory and is assumed to manage the different processing steps of calculation and to enable children to use more elaborated, memory-based strategies in lieu of simpler visually-based strategies (Andersson & Lyxell, 2007). However, research evidence with respect to group differences be- tween children with and without difficulties in mathematics is mixed. There is evidence for underperformance of mathematically impaired children in every single memory subcomponent, i.e., the phonological loop (Andersson & Lyxell, 2007; D'Amico & Guarnera, 2005; Geary, Hoard, Byrd-Craven, Nugent, & Numtee, 2007; Schuchardt et al., 2006; Swanson & Beebe-Frankenberger, 2004), the visual–spatial sketchpad (D'Amico & Guarnera, 2005; Geary et al., 2007; McLean & Hitch, 1999), and the central executive (Andersson, 2010; Andersson & Lyxell, 2007; D'Amico & Guarnera, 2005; Geary et al., 2004, 2007; Kyttälä, Aunio, & Hautamäki, 2010; Mabott & Bisanz, 2008; McLean & Hitch, 1999; Passolunghi & Cornoldi, 2008; Passolunghi & Siegel, 2001; Passolunghi & Siegel, 2004; Swanson & Beebe-Frankenberger, 2004; Swanson & Jerman, 2006). On the other hand, mathematically impaired children and controls were found to be equal with respect to the phonological loop (Geary, Hoard, & Hamson, 1999; Landerl, Bevan, & Butterworth, 2004; Landerl, Fussenegger, Moll, & Willburger, 2009; McLean & Hitch, 1999; Passolunghi & Cornoldi, 2008; Passolunghi & Siegel, 2004; van der Sluis, van der Leij, & de Jong, 2005), the visual– spatial sketchpad (Geary, Hamson, & Hoard, 2000; Passolunghi & Cornoldi, 2008; Schuchardt et al., 2006; van der Sluis et al., 2005), or the central executive (Iuculano, Moro, & Butterworth, 2011; Landerl et al., 2004; Schuchardt et al., 2006; van der Sluis et al., 2005). We think that two factors may distinguish studies reporting lower scores on any memory subcomponent from studies that do not: Restriction of IQ and restriction of language abilities. Studies that re- strict IQ and language or reading abilities of children with mathematical difficulties to be in the average range often report no group differences with respect to any memory subcomponent (e.g. Geary et al., 2000; Landerl et al., 2004; Schuchardt et al., 2006; van der Sluis et al., 2005), whereas studies that only restrict IQ (e.g., Geary et al., 2004, 2007; Swanson & Beebe-Frankenberger, 2004) or only restrict language or reading abilities (e.g., Andersson, 2010; Andersson & Lyxell, 2007; D'Amico & Guarnera, 2005; Passolunghi & Cornoldi, 2008; Passolunghi & Siegel, 2001, 2004) or neither (e.g., Kyttälä et al., 2010) are more likely Learning and Individual Differences 23 (2013) 108–113 ⁎ Corresponding author at: Faculty of Psychology, University of Basel, Missionsstrasse 60/62, 4055 Basel, Switzerland. Tel.: +41 61 267 06 37; fax: +41 61 267 35 26. E-mail address: giselle.reimann@unibas.ch (G. Reimann). 1041-6080/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.lindif.2012.10.017 Contents lists available at SciVerse ScienceDirect Learning and Individual Differences journal homepage: www.elsevier.com/locate/lindif