Cognitive Predictors of Timed and Untimed Early Arithmetic Performance
Olivia Wassing,Marcie Penner-Wilger
Abstract:Children perform worse on timed than untimed arithmetic tasks (e.g., Tsui & Mazzoco, 2006), yet children are evaluated on both types of tasks early in their education. The purpose of present study was to investigate whether the set of cognitive predictors differs for performance on timed versus untimed arithmetic tasks. The cognitive predictors of interest in the current study, visuospatial short-term memory (STM), phonological working memory (WM), and phonological awareness, were chosen as they have been previously shown to be important in early arithmetic development. Participants in the study were drawn from the Count Me In project and consisted of 116 children tested in Grade One (62 male, M = 82.98 months), 88 of whom also participated in Grade Two (41 male). Results demonstrated that in Grade One there was no difference in the set of cognitive predictors between performance on the timed and untimed arithmetic tasks (i.e., phonological awareness and visuospatial STM; R = .238 and .198 respectively). In Grade Two, however, the predictor sets varied for performance on the timed (i.e., phonological awareness and phonological WM; R = .251) and untimed (i.e., phonological awareness and visuospatial STM; R = .306) arithmetic tasks. Overall, phonological awareness was the dominant predictor for both tasks in Grade One and Grade Two, accounting for 8.1-12.7 % of unique variance in arithmetic performance. Therefore, consideration should be given to including phonological awareness in diagnostic tests to identify children at risk of mathematical difficulties. PREDICTORS OF EARLY ARITHMETIC 3 Cognitive Predictors of Timed and Untimed Early Arithmetic Performance Arithmetic is one of the primary academic skills children learn in their early education. Next to learning to read and write, arithmetic skills are a priority of childhood education. Furthermore, the basic arithmetic skills that children are initially taught have been demonstrated to be the building blocks for further math achievement (Geary & Brown, 1991; Geary, Hoard, Nugent, Bailey, 2013; Watts, Duncan, Siegler, Davis-Kean, 2014). For example, proficiency in simple addition is predictive of the development of more complex addition, and the early development of multiplication skills (Geary & Brown, 1991), and arithmetic ability in Senior Kindergarten is strongly related to arithmetic performance at age 15 (Watts et al., 2014). During these early educational years (i.e., Senior Kindergarten to Grade 3), children are introduced to arithmetic in an array of forms. Children need to master the four arithmetic operations (i.e., addition, subtraction, multiplication and division), which can be presented in written or verbal format, and may or may not involve time pressure (e.g., mad minutes). However, until recently, much of the previous research regarding arithmetic achievement typically investigated the predictors of general arithmetic achievement (e.g., Alloway et al., 2005; Berg, 2008; Bull, Espy, &Wiebe, 2008; De Smedt et al., 2009; Leather & Henry, 1994; Navarro et al., 2011; Swanson, 2004). A limited number of studies have focused on the complex relations between different forms of arithmetic presentation and underlying cognitive abilities (e.g., Cowan & Powell, 2014; De Smedt, Taylor, Archibald, & Ansari, 2010; Fuchs et al., 2006; Jordan, Whylie, & Mulhern, 2010; LeFevre et al., 2010; Schwenck, Dummert, Endlich, & Schneider, 2015; Sowinski et al., 2015). This research, concerning the predictors of arithmetic achievement, is necessary for the understanding of different arithmetic deficits that children can potentially encounter. Furthermore, it is necessary to help understand why some children perform better on different PREDICTORS OF EARLY ARITHMETIC 4 types of arithmetic tasks compared to others, which can help in the creation of diagnostic tests to identify arithmetic deficits and hopefully lead to effective interventions. For the purpose of the present study, we were specifically interested in whether the relations between cognitive predictors and arithmetic performance differ depending on the presence or absence of time pressures incorporated into the arithmetic tasks. The effect of time pressure on arithmetic performance, and more general academic performance, has been investigated in previous studies (e.g., Dreyden & Gallagher, 1989; Kellogg, Hopko, & Ashcraft, 1999; Schwartz, Evans, & Agur, 2015; Tsui & Mazzocco, 2006). In general, individuals tend to perform worse on timed tasks versus untimed tasks, perhaps due to a speed-accuracy trade-off, the effort to strike a balance between the demands of speed and accuracy (Bogacz, Wagenmakers, Forstmann, & Nieuwenhuis, 2010). Another explanation that has been considered is anxiety, regarding arithmetic (Tsui & Mazzocco, 2006) and other tests (Schwartz et al., 2015). Ashcraft and Kirk (2001) have suggested that anxiety can act like a secondary task, consuming one’s attentional resources potentially used for the test at hand, such as arithmetic. Interestingly, some research has demonstrated that some individuals perform better on timed tasks, compared to others (Jordan & Montani, 1997). Unfortunately, there is little research on this topic that is specific to children in their early education (e.g., Jordan & Montani, 1997; Tsui & Mazzocco, 2006). Further research is needed to address to the questions of why there is a difference between timed and untimed achievement, specifically regarding early education. One possible place to look for answers to these questions is at the cognitive level, which has not been explicitly investigated in the previous research. Various cognitive abilities have been revealed in previous research to be related to general arithmetic performance (e.g., Berg, 2008; Cowan & Powell, 2008; De Smedt, Taylor, PREDICTORS OF EARLY ARITHMETIC 5 Archibald, & Ansari, 2010; Geary, 2011; LeFevre et al., 2010). Furthermore, it has been revealed that there are potentially different subtypes of mathematic learning disorders, including a semantic/long-term memory subtype, which often co-occurs with language-based disabilities; a procedural/working memory subtype, with an impaired strategy use and problem solving; and a visual-spatial-motor-subtype (Geary, 1993). All three of the subtypes reflect an arithmetical deficit, accompanied by other deficits, due to neuronal damage or a cognitive deficit. Therefore, more research is needed that is focused on cognitive level and what roles it may play in different aspects of arithmetic to further understand the deficits faced by these individuals, as well as how to identify and intervene to improve their learning experience. For the interests of this present study, three cognitive abilities were utilized in our investigation and include visuo-spatial short-term memory, phonological working memory (i.e., phonological loop) one component of Baddeley and Hitch’s (1974) multicomponent working memory model, and phonological awareness. All of the aforementioned cognitive abilities have consistently been demonstrated to play an important role in early arithmetic performance (e.g., Alloway et al., 2005; Geary, 2011; Jordan et al., 2010). In the present study, the relation of visuo-spatial short-term memory to timed and untimed arithmetic tasks was investigated. However, it is important to note the methodological confusion between visuo-spatial working memory (i.e., visuo-spatial sketchpad) and visuospatial short-term memory. In the present study, the forward Corsi Blocks task, was used as a measure of visuo-spatial short-term memory. For the Corsi Blocks task, children witness sequences of increasing length and are then asked to replicate the sequence in the order presented. However, it is important to note, previous research has utilized this task, and similar ones (e.g., Rabbits), to assess visuo-spatial working memory (e.g., Cirino, 2011; LeFevre et al., PREDICTORS OF EARLY ARITHMETIC 6 2010; Simmons, Singleton, & Horne, 2008; Sowinski et al., 2015). The issue is that visuo-spatial working memory encompasses the capabilities of temporarily maintaining and manipulating visuo-spatial information (Baddeley, 2001). Yet, using the forward version of the Corsi Blocks task merely involves the ability of temporarily maintaining, and no manipulation, of visuo-spatial information. To assess visuo-spatial working memory with the Corsi Blocks task one would have to utilize the reverse version, since it involves maintaining and manipulating the information. Therefore, for the purpose of the present study, previous research that utilized the forward Corsi Blocks task as a measure for visuo-spatial working memory will be discussed as visuo-spatial short-term memory. Previous research indicates visuo-spatial short-term memory offers a unique contribution to early arithmetic performance when problems are presented visually (e.g., Berg, 2008; Cirino, 2011; De Smedt et al., 2009; Geary, 2011; LeFevre et al., 2010; Simmons, Singleton, & Horne, 2008; Simmons, Willis, & Adams, 2012; Soto-Calvo, Simmons, Willis, & Adams, 2015; Szucs, Devine, Soltesz, Nobes, & Gabriel, 2013, 2014). However, visuo-spatial short-term memory is reported to not be related to, or a significant predictor of, arithmetic tasks in some studies (e.g., Passolunghi & Lanfranchi, 2012; Passolunghi, Lanfranchi, Altoe, & Sollazzo, 2015). In contrast, Lefevre et al., (2010) reported that visuo-spatial short-term memory was related to, and predicted unique variability for, six out of seven arithmetic outcomes under investigation, including both timed and untimed arithmetic tasks. Other studies have reported that visuo-spatial short-term memory is a unique predictor of arithmetic fluency tasks for adding small sums (timed tasks; Cirino, 2011; Sowinski et al., 2015). Furthermore, visuo-spatial short-term memory at the beginning of Grade One has been reported as a significant predictor of standardized arithmetic tests adm