Research in Developmental Disabilities 36 (2015) 150–161

Contents lists available at ScienceDirect

Research in Developmental Disabilities

Learning of writing letter-like sequences in children with physical and multiple disabilities Marjolein Jongbloed-Pereboom a,*, Ange`le Peeters a, Anneloes Overvelde b,c, Maria W.G. Nijhuis-van der Sanden c, Bert Steenbergen a,d a

Radboud University Nijmegen, Behavioural Science Institute, Montessorilaan 3, 6525 HR Nijmegen, The Netherlands Radboud University Nijmegen, Donders Centre for Cognition, Montessorilaan 3, 6525 HR Nijmegen, The Netherlands Radboud University Nijmegen Medical Centre, Scientific Institute for Quality of Healthcare, Department of Rehabilitation, Pediatric Physiotherapy, Geert Grooteplein 21, 6525 EZ Nijmegen, The Netherlands d Australian Catholic University, School of Psychology, Melbourne, Australia b c

A R T I C L E I N F O

A B S T R A C T

Article history: Received 16 July 2014 Received in revised form 26 September 2014 Accepted 2 October 2014 Available online

This study compared implicit and explicit learning instructions in hand writing. Implicit learning is the ability to acquire a new skill without a corresponding increase in knowledge about the skill. In contrast, explicit learning uses declarative knowledge to build up a set of performance rules that guide motor performance or skills. Explicit learning is dependent on working memory, implicit learning is not. Therefore, implicit learning was expected to be easier than explicit learning in children in special education, given their expected compromised working memory. Two groups of children (5–12 years) participated, children in special education with physical or multiple disabilities (study group, n = 22), and typically developing controls (n = 32). Children learned to write letter-like patterns on a digitizer by tracking a moving target (implicitly) and verbal instruction (explicitly). We further tested visual working memory, visual-motor integration, and gross manual dexterity. Learning curves were similar for both groups in both conditions; children in the study group did learn both implicitly and explicitly. Motor performance was related to the writing task. In contrast to our hypothesis, visual working memory was not an important factor in the explicit condition. These results shed new light on the conceptual difference between implicit and explicit learning, and the role of working memory therein. ß 2014 Elsevier Ltd. All rights reserved.

Keywords: Implicit Explicit Hand writing Working memory Education, Special Motor skills disorders

1. Introduction Motor skills can be learned in an explicit or an implicit manner (Masters, 1992). Explicit learning requires cognitive processes that generate declarative knowledge. Declarative knowledge is knowledge which we can describe, and which we consciously remember (Maxwell, Masters, & Eves, 2003). Implicit learning, on the other hand, is the process by which we do not show any awareness of learning the rules underlying the learning process (Berry & Dienes, 1993). Implicit learning builds up procedural knowledge, which is difficult or even impossible to access consciously and/or report verbally. Importantly, in

* Corresponding author at: Radboud University Nijmegen, Behavioural Science Institute, Montessorilaan 3, 6500 HE, P.O. Box 9104, Nijmegen, The Netherlands. Tel.: +31 243616284. E-mail address: [email protected] (M. Jongbloed-Pereboom). http://dx.doi.org/10.1016/j.ridd.2014.10.005 0891-4222/ß 2014 Elsevier Ltd. All rights reserved.

M. Jongbloed-Pereboom et al. / Research in Developmental Disabilities 36 (2015) 150–161

151

contrast to explicit learning, research has found minimal associations between implicit learning and intelligence or (mental) age (Meulemans, van der Linden & Perruchet, 1998; Reber, Walkenfield & Hernstadt, 1991; Vinter & Detable, 2003; Vinter & Perruchet, 2000), indicating that it could be easier for children with intellectual disabilities (ID) to learn in an implicit manner than in an explicit manner. Furthermore, studies on motor skill learning in healthy adults have shown that intact working memory is a necessary prerequisite for the explicit learning of motor skills, while this is no prerequisite for implicit learning (Maxwell et al., 2003). Working memory is an active processing system that keeps different types of information available for current use, for example for activities such as problem solving, reasoning, and comprehension (Baddeley, 2001). Poor memory function has consequences for different aspects of learning and cognitive ability (Cowan & Alloway, 2009). Several studies have shown that working memory deficits are related to learning disabilities and the severity of intellectual disabilities (Henry, 2001; Henry & MacLean, 2002; Maehler & Schuchardt, 2009). Working memory deficits are also related to disorders of the development of movement, posture and coordination, such as cerebral palsy (CP) (Bax, Goldstein, Rosenbaum, Leviton & Paneth, 2005; Jenks, Moor de & Lieshout van, 2009; Straub & Obrzut, 2009) and developmental coordination disorder (DCD) (Wilson, Ruddock, Smits-Engelsman, Polatajko & Blank, 2013). It is therefore likely that working memory capacities have important consequences for a child’s ability to acquire knowledge and to learn new complex skills in an educational context. Students frequently have to rely on working memory to perform a range of activities. Students with working memory impairments may struggle in classroom because they are unable to hold in mind sufficient information to allow them to complete a task (Engle, Carullo & Collins, 1991). This could lead to failures in simple task performance such as remembering classroom instructions (Engle et al., 1991), but also to problems in more complex activities involving storage and processing of information and keeping track of progress in difficult tasks (Gathercole & Alloway, 2008). Up to date, in (special) education, acquisition of cognitive skills is predominantly guided by explicit instructions, either from the teacher or from textbooks (Graham et al., 2008). Obviously, such explicit instructions place a high demand on working memory functioning and may potentially hinder proper learning of cognitive skills, given the impaired nature of working memory (Steenbergen, van der Kamp, Verneau, Jongbloed-Pereboom, & Masters, 2010). Since implicit learning is less dependent on working memory functioning (Maxwell et al., 2003), this type of learning may be useful for applications in educational contexts for children with working memory impairments. Handwriting is a complex perceptual-motor skill that requires a multitude of abilities and skills, such as visual-motor integration, motor planning, cognitive and perceptual skills, kinesthetic and tactile sensitivities (Feder & Majnemer, 2007), and linguistic awareness (Berninger, Abbott, Nagy, & Carlisle, 2010). The prevalence of handwriting problems in school-age children varies between 12% and 33% (Karlsdottir & Stefansson, 2002; Rubin & Henderson, 1982; Smits-Engelsman, Niemeijer, & Van Galen, 2001). Moreover, most children with learning disabilities experience fine motor difficulties or handwriting problems (Clements, 1966; Rourke, Ahmad, Collins, Hayman-Abello, & Warriner, 2002; Tamopol & Tamopol, 1977). Handwriting problems are among the most common reasons for referring school-age children to physiotherapy or occupational therapy services (Bosga-Stork et al., 2009; Hammerschmidt & Sudsawad, 2004). Furthermore, handwriting difficulties do not only influence a child’s success in school performance, but can also affect his/her self-esteem (Dunford, Missiuna, Street, & Sibert, 2005). At present, no consensus exists as to the most effective method for teaching handwriting in classroom. The majority of teachers use a variety of instructional practices for teaching handwriting, which are predominantly explicit procedures, such as copying, tracing, verbal description and modeling (Graham et al., 2008). In physiotherapy, especially the amount of handwriting practice is important for improving handwriting skills. It is not yet clear in what way these handwriting skills should be provided, in a more cognitive or sensorimotor focused training (Hoy, Egan, & Feder, 2011). This study attempted to gain more insight into the relation between children’s disabilities and two methods of learning abstract letter-like patterns, implicit and explicit learning. Children with physical or multiple disabilities (special education) and typically developing controls (mainstream education) learned to write unfamiliar, abstract patterns on a digitizing tablet. The patterns included all aspects of letter-like patterns. In the implicit condition (moving target) children used a stylus to track a target that moved along an invisible trajectory. The explicit condition focused on simple verbal instructions. We hypothesized that children in the control group learned new handwriting skills better than the children with physical and multiple disabilities in both conditions. Furthermore, given their compromised working memory, the study group would benefit from the implicit learning procedure compared to the explicit condition. This beneficial effect was not expected for the control group. Next to these main research questions and hypotheses, we also examined the possible effects of gross manual dexterity and visual motor integration in learning these letter-like patterns in both conditions.

2. Method 2.1. Participants Children in the study group were recruited from and tested in two schools for special education for physical and/or multiple disabilities. Children in the study group had the following inclusion criteria: age between 5 and 11 years and IQ score 55 (based on personal file). Furthermore, to fulfill the test they had to be able to respond to the instructions, to be able to grasp a pencil, to respond verbally, and to discriminate colors. Twenty-nine children were eligible for participation in the

M. Jongbloed-Pereboom et al. / Research in Developmental Disabilities 36 (2015) 150–161

152

study group. After testing seven children were excluded, of which six children had insufficient attention/concentration to fulfill the tasks, and one child was absent during testing. The study group consisted of six girls and 16 boys and the mean age of the children at the onset of the study was 9;0 years (SD = 1.40, range = 6;4–10;11). The mean total intelligence score of 20 children was 78.05 (SD = 14.99, range = 58–107). Two children had no recent total intelligence scores, but the subtest scores indicated that they had a normal IQ. A detailed description of the children in the study group is presented in Table 1. Children in the control group were recruited from and tested in one school for mainstream education. Children in the control group were also between 5 and 11 years old and were included if they had no cognitive, physical, or behavioral problems. Two children were excluded because they were diagnosed with a behavioral problem (attention deficit (hyperactivity) disorder), two children received incorrect instruction from the experimenter during the experimental task and one child was absent during testing. After exclusion of these five children, the control group consisted of 20 girls and 12 boys and the mean age of the children at the onset of the study was 8;9 years (SD = 1.67 years, range = 5;7–11;3). Twentyfive children were right-handed and seven children were left-handed. For both groups, parents of the children provided written informed consent for participation of their children in the study. This study was approved by the local ethics committee. 2.2. Measures 2.2.1. Working memory The Automated Working Memory Assessment (AWMA) is a computerized tool for assessing short-term and working memory in persons aged 4–22 years (Alloway, 2007a). The three subtests to assess visuospatial working memory were used: Odd-One-Out, Mr. X, and Spatial Recall. The raw scores were converted into standardized scores, and a composite score for visual working memory (M = 100, SD = 15). Test–retest reliability of the three subtests is .88, .84, and .79, respectively, for children between 4.10 years to 22.5 years (Alloway, 2007a). 2.2.2. Visual motor integration Integration of visual and motor abilities was tested using the Developmental Test of Visual-Motor Integration (Beery VMI; 5th ed., Beery & Beery, 2004). The children had to copy, by paper and pencil, a sequence of geometric figures. The test consisted of 30 items. 1 Point was awarded to a correctly copied form, with a maximum raw score of 30. Raw scores were converted into standardized scores (M = 100, SD = 15). Reliability of the VMI is good (Beery & Beery, 2004; inter-rater: r = .91; internal consistency: .96; test–retest: .89). 2.2.3. Gross manual dexterity Gross manual dexterity was tested with the Box and Block Test (Mathiowetz, Volland, Kashman & Weber, 1985). A box, which was divided into two square compartments by a partition in the center, was placed lengthwise along the edge of the Table 1 Descriptives of the children in the study group. Participant

Gender

Age (in years; months)

Dominant hand

Diagnosis

ID

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Boy Boy Girl Boy Girl Boy Girl Girl Girl Boy Boy Boy Boy Boy Girl Boy Boy Boy Boy Boy Boy Boy

7;10 10;5 9;6 10;11 7;4 9;2 7;5 9;3 6;4 9;5 10;0 8;4 10;11 10;1 6;6 9;0 10;5 10;0 9;8 6;11 10;2 8;9

Left Left Right Right Right Right Right Right Right Left Right Right Right Right Right Left Right Right Left Right Right Right

Paroesophageal hernia Abnormality of chromosome 11 Syndrome of Dandy Walker Status after surgery for epilepsy Cardiac problems DCD Hypotonia, myopathy Severe schisis CP, PDD-NOS Intestinal disorder Intestinal disorder CP DCD, growth hormone deficiency DCD, ADHD Cardiac problems Syndrome of Stuve–Wiedeman DCD Syndrome of Leigh DCD Mitochondrial dysfunction DCD Mitochondrial dysfunction

Mild Mild Mild Borderline Borderline Borderline No Mild No Borderline Borderline Mild No No Borderline Borderline No Borderline Mild No No Borderline

DCD, developmental coordination disorder; CP, cerebral palsy; PDD-NOS, pervasive developmental disorder–not otherwise specified; ADHD, attention deficit hyperactivity disorder; ID, intellectual disability (based on IQ scores); mild ID, IQ score between 55 and 69; borderline ID, IQ score between 70 and 84, no ID, IQ score of 85 and higher.

M. Jongbloed-Pereboom et al. / Research in Developmental Disabilities 36 (2015) 150–161

153

table and the examiner sat facing the participant. The blocks were 150 colored wooden cubes of 2.5 cm2. First, the cubes were on the side of the box of the participant’s dominant hand. The participants had to transfer, the blocks from one side of the box to the other in one minute with their dominant hand. A 15-s practice period preceded testing. The procedure was then repeated with the non-dominant hand. The outcome score was the number of blocks transferred in one minute for each hand separately. Only the score of the dominant hand was included, since this was the writing hand. 2.2.4. Handwriting task The handwriting task was adapted from Overvelde and Hulstijn (2011). The writing task is a computerized task, in which handwriting is measured by using a graphics tablet (Wacom XY-tablet, Intuos 3). The pen position was recorded at a rate of 200 Hz. Nine differently colored circles (11 mm in diameter) were always visible on the computer screen, and acted as landmarks for the movement sequences. Two different trajectories (abstract patterns) each consisting of a 7-element sequence, comparable to the sequence of movements needed for writing the cursive capital ‘H’, with a total length of approximately 37 cm were used as learning sequences (see Fig. 1). During the experiment the participant was told to learn to write new figures like letterforms. Participants were instructed to guide a yellow-colored cursor (5 mm diameter) on the computer screen by moving a non-inking pen on the graphics tablet. In order to reproduce the stimulus patterns the cursor had to pass the nine landmarks in one of five possible ways, i.e. at the inside (between the circle and the center circle), at the outside, by encircling it clockwise or anticlockwise, or by stopping in the circle. The start and end positions of the sequence were marked by a white and a black circle, that produced a high and low beep, respectively, when the cursor reached them. There were two learning conditions, a moving target condition (implicit) and a verbal instruction condition (explicit). In the moving target condition a circular gray target (10 mm diameter) moved with a natural speed along the invisible prerecorded trajectory and the child was asked to keep the cursor within the moving target by moving the non-inking pen. The invisible pre-recorded trajectory was recorded with the timing characteristics of a well trained teacher and these timing characteristics incorporated natural changes and stops in velocity at the proper positions. During the experiment the moving speed was reduced to 50%, so that children were able to keep up with the target. During the verbal instruction condition no trail was presented, but a verbal instruction was given to explain the to-be-produced sequence. The participants were instructed about the order and how to pass the colored landmarks. If the participants needed more clarification of the instruction, physical prompts on the screen were given for the first three trials. The training phase consisted of 10 trials in both learning conditions. Prior to the training phases children had to practice the two conditions with shorter, simpler figures. Each learning condition was directly followed by a test phase of another 10 trials. In the test phase, children had to reproduce the learned sequences without the help of the moving target or verbal instruction. The start and end position were still marked. A set of simple loops around the outer landmarks served as an intermission between the two learning conditions. In the design, the conditions and the different trajectories (Fig. 1) were counterbalanced among participants. The recorded data were analyzed in Matlab (Matlab R2012a, Mathworks). There were four outcome measures: movement duration, number of errors, distance and fluency. These four measures were chosen because they reflect performance and deviation thereof at different levels. Movement duration (in s) was determined by measuring the required time to perform the letter-like pattern from the start position to the end position. The number of errors was determined by counting the number of circles that were incorrectly passed, with a maximum of seven errors per trial. Distance (in mm) was determined by measuring the distance between the X and Y points for all samples in the trial, and adding these distances. The absolute

[(Fig._1)TD$IG]

Fig. 1. Illustration of the abstract letter-like patterns. Note that the gray trace was not shown on screen, but the start and end positions were marked. In the moving target condition (implicit) a circular gray target moved along the invisible pre-recorded trajectory. In the verbal instructed condition (explicit), a verbal instruction was given to explain the sequence.

154

M. Jongbloed-Pereboom et al. / Research in Developmental Disabilities 36 (2015) 150–161

difference between the programmed patterns (also used for the verbal instruction condition) and the actual written patterns was the outcome variable. Fluency was determined by assessing the number of crossings through 0 in the acceleration plot, thus the number of transitions from acceleration to deceleration and vice versa. The lower this number of zero crossings was, the better the fluency of the movement was. 2.3. Procedure All children were individually tested in a quiet room in their school by the second author or by one of two trained test assistants. The children were seated comfortably on height adjusted furniture. The background tests (VMI, Box and Block Test, AWMA) and the experimental task were divided into two sessions. In the first session working memory, visual motor integration and gross manual dexterity were assessed. In the second session the handwriting task was assessed. Both sessions lasted about 45 min. 2.4. Statistical analysis Data analysis was performed with the Statistical Package for the Social Sciences (version 20.0 for Windows; SPSS Inc. Chicago, IL). Group differences on background characteristics and the background tests were evaluated by independent samples t-tests and x2-tests. For the handwriting task, the learning curves of all four outcome variables were assessed by a General Linear Model multivariate repeated measures analysis with trials (10) and condition (2: implicit/explicit), and group (2: special education/controls). The analyses were conducted for the training and test phase separately. Next, the repeated measures analysis were repeated for the separate conditions to test the influence of the covariates. Five covariates were tested: group (special education/control), gender, Box and Block Test score for the dominant hand, VMI standard score, and AWMA visual working memory standard score. In this analysis group was added as a covariate instead of a between subjects factor, since all children performed on both conditions. This way, we were able to test all covariates with an ANOVA analyses. All ANOVA analyses were conducted with Greenhouse–Geisser’s epsilon adjusted probabilities to control for violation of the sphericity assumption. An alpha level of .05 was considered statistically significant. 3. Results 3.1. Description of groups There was no significant difference between the mean ages of the two groups. The proportion boys and girls was not equal between the two groups, x2(1) = 6.48, p < 05. In the study group 27% were girls, and in the control group 63% were girls. Table 2 provides an overview of the descriptive statistics of all measures with respect to background information for the study and control group. As expected children in the study group scored significantly lower on the VMI, t(34.635) = 4.849, p < .001, r = .64, and the Box and Block Test, t(52) = 3.410, p = .001, r = .43 compared to the control group. The scores of the study group were also significantly lower on all subtests for visuospatial working memory of the AWMA: Odd-One-Out, t(52) = 4.832, p < .001, r = .56, Mr. X, t(52) = 4.673, p < .001, r = .54, Spatial Recall, t(29,376) = 4.677, p < .001, r = .65, and the composite score of visuospatial working memory, t(30,907) = 5.599, p < .001, r = .71. Additionally, we tested if these differences in scores were due to gender. Within the study group as well as within the control group there were no significant differences between boys and girls on all background measures, except for the score on the Box and Block Test. Boys and girls in both groups scored significantly different on the Box and Block Test, study group t(20) = 3.386, p = .003, r = .60, and control group t(30) = 2.486, p = .019, r = .41. In both groups the score was higher for boys than girls.

Table 2 Descriptive statistics of the study and control group. Study group (n = 22)

Control group (n = 32)

M

SD

Age IQ score VMI standard score* Box and Block score (dominant hand)*

9.0 78.1 83.9 40.1

1.4 15.0 11.5 9.1

8.8 – 97.7 47.6

AWMA standard score Odd-one-out* Mr. X* Spatial Recall* Composite Score visual working memory*

97.7 93.1 94.4 93.6

19.1 16.6 21.6 20.4

118.5 115.2 118.0 120.7

* Significantly different, p < 0.001.

M

SD 1.7 – 8.00 6.9

12.6 17.3 11.6 12.0

M. Jongbloed-Pereboom et al. / Research in Developmental Disabilities 36 (2015) 150–161

155

3.2. Handwriting task For each outcome measure of the handwriting task, first the main effects of trial, condition and group are presented followed by the interactions. Next, significant effects of the covariates are presented for the separate conditions. The means and standard deviations of the different outcome variables for the training and test phase are summarized in Table 3 for the study and control group, the F values of the repeated measures ANOVA’s are listed in Table 4. The results of the handwriting task are also depicted in Fig. 2. 3.2.1. Movement duration 3.2.1.1. Training phase. As expected, during the training phase movement duration decreased. Second, movement duration was larger in the verbal instruction condition than in the moving target condition. Third, children in the control group reproduced the stimulus patterns faster than children in the study group. There was a significant interaction of trials and condition, movement duration decreased more in the verbal instruction condition than in the moving target condition across the trials. As expected, movement duration of the moving target condition was almost stable whereas movement duration in the verbal instruction condition decreased during the training phase. There was a significant interaction between condition and group, differences in movement duration between the conditions were larger for the study group than for the control group. Finally, we found an interaction between trials and group, movement duration decreased more in the study group than in the control group. The study group started slower than the control group, but during training differences between groups decreased. We further investigated the effects of the covariates on the training phase trials of the moving target condition. There was only an effect of gender, girls performed the trials in the training phase slower than boys. For the verbal instruction condition, the score on the Box and Block Test had a significant effect on the training phase trials. A higher score on the Box and Block Test was associated with a smaller movement duration. Furthermore, an effect of group, indicated that children in the control group reproduced the verbal instruction trials faster than children in the study group, as was also shown above for the interaction effect of condition by group. 3.2.1.2. Test phase. There was a main effect of trials, movement duration decreased during the test phase. A significant main effect of condition, shows that movement duration was higher in the verbal instruction (explicit) condition than in the moving target (implicit) condition. We found no main effect of group for the test phase on movement duration, nor did we find any interactions. For both the moving target condition and the verbal instruction condition, only the score on the Box and Block Test had a significant effect on the test phase trials. The higher the score on the Box and Block Test, the lower the movement duration. 3.2.2. Number of errors 3.2.2.1. Training phase. Again, a main effect of trials was found, the number of errors decreased during the training phase. As expected, there was a main effect of condition, children made more errors in the explicit condition. The significant main effect of group exemplified that the number of errors was lower for the control group than the study group. Table 3 Results of the writing task. Training phase

Movement duration Implicit Explicit Number of errors Implicit Explicit

Test phase

Study group

Control group

Study group

Control group

M (SD)

M (SD)

M (SD)

M (SD)

9.49 (1.75) 21.72 (7.86)

9.14 (.61) 15.31 (3.80)

10.68 (4.49) 12.19 (4.82)

9.38 (1.84) 10.09 (2.76)

1.21 (.97) 1.81 (1.33)

.52 (.33) .29 (.43)

1.47 (1.35) 1.66 (1.59)

.35 (.41) .21 (.39)

Distance Implicit Explicit

134.83 (65.72) 272.10 (133.29)

69.69 (30.43) 107.11 (34.43)

130.17 (68.96) 159.08 (118.96)

76.02 (34.87) 95.51 (38.69)

Fluency Implicit Explicit

253.47 (66.25) 642.06 (261.45)

268.14 (36.12) 496.45 (155.36)

305.03 (149.92) 345.60 (160.22)

291.78 (83.78) 322.65 (126.68)

Mean results per 10 trials of the training and test phases. Movement duration per trial was measured in seconds, number of errors per trial was assessed by counting the number of circles that were incorrectly passed, with a maximum of seven errors per trial, the distance per trial was measured in millimeters deviation from the pre-recorded trial, fluency was measured as the number of zero-crossings in the acceleration plot.

156

Table 4 Results of the repeated measures analysis. Condition

Group

Trials  condition

Trials  group

21.75 (4.6, 239.0)*** 2.47 (6.4, 332.1)*** 6.84 (5.4, 282.6)*** 15.93 (4.4, 229.6)***

55.41(1, 52)*** 9.29 (1, 52)** 35.82 (1, 52)*** 43.56 (1, 52)***

14.29 (1, 52)*** 42.18 (1.52)*** 53.18 (1, 52)*** 3.95 (1, 52)

13.61 (4.9, 255.2)*** 1.47 (6.4, 333.5) 7.54 (5.3, 282.8)*** 10.60 (4.8, 249.6)***

3.30 1.12 3.52 2.07

(4.6, (6.4, (5.4, (4.4,

239.0)** 332.1) 282.6)** 229.6)

Test phase Movement duration Number of errors Distance Fluency

3.43 1.43 2.04 2.20

3.60 (1.52) 29.01 (1, 52)*** 13.58 (1, 52)*** 0.30 (1.52)

0.84 1.79 0.45 1.05

1.53 1.11 1.72 0.72

(5.7, (7.1, (6.0, (5.9,

297.7) 367.3) 311.3) 305.1)

(5.7, (7.1, (6.0, (5.9,

297.7)** 367.3) 311.3) 305.1)*

4.15 1.43 1.33 1.40

Effect of covariates on conditiony Training phase Movement duration Number of errors Distance Fluency

Test phase Movement duration Number of errors Distance Fluency Values are F values (df). y df (1.48). * p < .05. ** p < .01. *** p < .001.

(1, 52)* (1, 52) (1.52) (1, 52)

(6.0,310.5) (7.1, 367.9) (5.9, 305.2) (5.9, 304.6)

Condition  group 16.87 8.68 17.63 10.32

1.38 1.59 0.22 0.14

Trials  condition  group

(1.52)*** (1.52)** (1, 52)*** (1, 52)**

1.59 1.37 4.05 1.18

(4.9, (6.4, (5.3, (4.8,

255.2) 333.5) 272.8)*** 249.6)

(1.52) (1.52) (1.52) (1, 52)

0.80 1.41 0.45 1.05

(6.0, (7.1, (5.9, (5.9,

310.5) 367.3) 305.2) 304.6)

Group

Gender

Box and Block

VMI

AWMA

Moving target Verbal instruction Moving target Verbal instruction Moving target Verbal instruction Moving target Verbal instruction

0.16 5.02* 0.53 4.70* 1.99 25.48*** 1.29 0.82

7.93** 1.84 2.86 1.75 0.07 0.30 6.85* 2.07

0.36 7.77** 4.16* 10.41** 8.46** 6.43* 0.01 9.76**

1.38 3.71 0.11 6.43* 8.32** 1.94 2.50 0.64

0.21 0.10 0.35 0.73 0.30 9.12*** 0.08 0.02

Moving target Verbal instruction Moving target Verbal instruction Moving target Verbal instruction Moving target Verbal instruction

1.49 0.32 2.58 3.40 0.17 0.05 2.45 0.04

0.62 2.03 0.03 0.00 2.26 3.90 1.05 2.60

10.05** 6.29* 2.41 4.67* 12.16** 8.19** 6.46* 5.40*

3.21 0.44 1.60 1.97 3.46 4.32* 2.85 0.29

0.42 0.01 0.08 0.03 0.23 0.60 0.16 0.01

M. Jongbloed-Pereboom et al. / Research in Developmental Disabilities 36 (2015) 150–161

Trials Training phase Movement duration Number of errors Distance Fluency

M. Jongbloed-Pereboom et al. / Research in Developmental Disabilities 36 (2015) 150–161

157

There was a significant interaction between condition and group. Differences in the number of errors were larger for the verbal instruction than for the moving target, and these differences were larger for the study group than for the control group. In the moving target condition, only the score on the Box and Block Test had a significant effect on the training phase trials. The higher the score on the Box and Block Test, the lower the number of errors. The effect of group was not present for these trials. For the verbal instruction condition, again the score on the Box and Block Test had a similar significant effect on the test trials. Besides the Box and Block Test, the score on the VMI also had a significant effect. The higher the standard score on the VMI, the lower the number of errors. The effect of group was present for these trials, the number of errors was lower for the control group than the study group. 3.2.2.2. Test phase. There was a significant main effect of group on number of errors. The number of errors was significantly higher for children in the study group than for children in the control group. We found no main effects on the number of errors for trials or condition for the test phase. There were no significant effects of the covariates on the trials of the moving target condition. For the verbal instruction condition, the score on the Box and Block Test had again a significant effect on the test phase trials. The higher the score on the Box and Block Test, the lower the number of errors. The effect of group was no longer present. 3.2.3. Distance 3.2.3.1. Training phase. All three main effects were significant. The difference in absolute distance decreased during the training phase, distance was larger in the verbal instruction condition than in the moving target condition. Furthermore, the trials of the children in the control group were performed with less difference in distance to the programmed patterns than the children in the study group. For the variable distance, there were four interactions. First, there was an interaction of trials and condition. Distance in the moving target condition was almost stable during the training phase whereas distance in the verbal instruction condition decreased during training. At the end of the training, the distance was almost similar for both conditions. Second, differences in distance between the moving target condition and the verbal instruction condition were larger for the study group than the control group, condition by group. Third, we found an interaction between trials and group. Distance decreased more in the study group than in the control group during the training phase. Finally, a third order interaction of trial by condition by group was found. Distance decreased more in the verbal instruction condition than in the moving target condition but the degree of the decrease was dependent on the group; distance decreased more for the study group than the control group (see Fig. 2). We further investigated the effects of the covariates on the training phase trials of the moving target condition. The effect of group was not present for these trials, but there was again an effect of the Box and Block Test score and there was also an effect of the VMI. The higher the score on the Box and Block Test or the VMI, the lower the difference in distance was with the programmed pattern. For the verbal instruction condition, again the score on the Box and Block Test had a similar significant effect on the verbally instructed test trials. Furthermore we found an effect of group, the distance from the ideal pattern was larger for the study group than the control group. We also found an effect of working memory. If the score on the AWMA was high, the difference in distance was small. 3.2.3.2. Test phase. There was only a significant main effect of group in the test phase, the distance from the ideal pattern was larger for the study group than the control group. We found no main effect of trials or condition on distance in the test phase. For the moving target condition, only the score on the Box and Block Test had a significant effect on the test phase trials. The higher the score on the Box and Block Test, the smaller the distance from the ideal pattern. No group effect was present for the moving target trials. For the verbal instruction condition, again the score on the Box and Block Test had a similar significant effect on the test trials. The score on the VMI also had a significant effect. The higher the score on the VMI or on the Box and Block Test, the smaller the difference in distance was the programmed pattern. The effect of group was no longer present. 3.2.4. Fluency 3.2.4.1. Training phase. Fluency increased during the training phase, and fluency was better in the moving target condition than in the verbal instruction condition. There was no main group effect on fluency. Besides the main effects there were two interactions, trials by condition and condition by group. Fluency increased more in the verbal instruction condition than in the moving target condition. Furthermore, differences in fluency between the moving target condition and the verbal instruction condition were larger for the study group than the control group. For the moving target condition, we found an effect for gender. Girls had higher, thus worse, fluency scores compared to boys. We did not find this effect for the verbal instruction condition. We did find an effect of the Box and Block Test for the verbally instructed trials. Fluency was better when Box and Block Test scores were higher.

158

M. Jongbloed-Pereboom et al. / Research in Developmental Disabilities 36 (2015) 150–161

3.2.4.2. Test phase. There was a main effect of trials, fluency increased during the test phase. There was no significant main effect of condition or group. For the moving target condition and the verbal instruction condition, we only found effects of the Box and Block Test. For both conditions a higher Box and Block Test score was associated with better fluency (lower fluency scores).

[(Figure_2)TD$IG]

Figure 2. Mean movement time, mean number of errors, mean distance and mean fluency per trial for the training and test phases of the two learning conditions.

M. Jongbloed-Pereboom et al. / Research in Developmental Disabilities 36 (2015) 150–161

159

4. Discussion 4.1. Discussion The aim of the present study was to investigate the efficacy of implicit and explicit learning conditions of handwriting sequences in children with physical and multiple disabilities (study group) and typically developing children (control group). We expected that children with compromised working memory would benefit from implicit learning. Our results showed that typically developing children outperformed children with disabilities in both learning conditions of the task, although learning effects were most pronounced for the study group. Differences between the two conditions were bigger for children with physical and multiple disabilities than for typically developing children. In contrast to our expectation, children in the study group did show learning effects for the verbally instructed, explicit condition. That is, children in the study group learned the patterns in both the implicit and explicit condition, but the results indicated an advantage of implicit learning for these children. For the typically developing children the implicit and explicit test phase were similar, for these children there was not an advantage of implicit learning. Finally, motor performance (measured with the VMI and Box and Block Test) was related to the results on the writing task in both conditions, which was not the case for visual working memory. Thus, our a priori expectation that children with physical and multiple disabilities would reproduce new handwriting sequences better when these sequences were learned implicitly compared to explicitly, was only partly confirmed by the results. Overall, the implicit condition seemed easier than the explicit condition. This was evidenced by the findings of stable performance during training (also due to how implicit training was presented to the children), and the lack of relapse during the first trials of the implicit test phase. During the explicit training trials, a large learning effect was observed (see also Figure 2). In the test phase, only reproduction of the patterns was faster in the implicit condition than the explicit condition for both groups of participants. Although we observed clear differences between conditions and groups in the training phase, these differences largely disappeared during the test phase. So although learning curves were different for the two conditions, the study group did also show learning effects in the explicit condition. The finding that implicit learning is intact in children with physical and multiple disabilities is in line with other investigations on comparable clinical populations. These include children with Developmental Coordination Disorder (DCD; Wilson, Maruff & Lum, 2003; Candler & Meeuwsen, 2002); children with an intellectual disability (Vinter & Detable, 2003; Vinter & Perruchet, 2000); and children with autism spectrum disorders (Brown, Aczel, Jime´nez, Kaufman & Grant, 2010). We expected explicit learning to be impaired in the children with physical and multiple disabilities. Our results did not confirm this. This is contrary to the theory on implicit and explicit learning (Steenbergen et al., 2010) and to the results of a recent study of Capio, Poolton, Sit, Eguia, and Masters (2013) in which implicit and explicit learning of overhand throwing in children with intellectual disability was examined. They found clear beneficial effects for implicit learning, but not for explicit learning (Capio et al., 2013). In contrast, a study of Watanabe, Ikeda, and Miyao (2010) found that children with attention deficit hyperactivity disorder (ADHD), who are at risk for working memory problems (Doyle, 2006), learned a serial reaction time task in an explicit manner. Although learning curves were similar to typically developing children, overall performance was lower. In studies on patients with dementia results on implicit and explicit learning are also ambiguous. These patients have severe problems with explicit memory, but not implicit memory. There is ample evidence that implicit learning is intact in patients with Alzheimer’s Disease (Van Halteren-van Tilborg, Scherder & Hulstijn, 2007), but clear evidence that explicit learning is impaired is lacking. It is important to note here that the strict theoretical distinction between explicit and implicit learning may be difficult to apply empirically. That is, both forms of learning most likely exist on a continuum. For the present handwriting study it is well possible that some children did indeed generate explicit knowledge in the implicit condition. A similar argument was also mentioned by Willingham (1998) in the case of serial reaction time tasks. This facet of implicit and explicit learning warrants further study to advance the conceptual distinction between both forms of learning. Children in our study group had significantly lower scores on visual working memory and motor performance than controls. We hypothesized that working memory deficits would affect explicit learning (Maxwell et al., 2003; Steenbergen et al., 2010), but our results did not confirm this. Visual working memory was only related to distance in the training phase (with higher, better scores on the AWMA, the difference in distance was small) but there were no effects on the other measures. For our handwriting task, we did find a consistent effect for motor performance (measured with the Box and Block Test and VMI). In general, if motor performance was better, participants were faster, made less errors, were more accurate and more fluent. In our study, only visual working memory was assessed because of maintenance and manipulation of visual and spatial representations. However, in the explicit verbal instruction condition, verbal working memory might be involved as well. Children with DCD have more pronounced visual working memory deficits than verbal working memory deficits (Alloway, 2007b). But it would have been interesting to see if verbal working memory might affect explicit learning in the study group, also since this group was quite diverse. Most participants in our study were diagnosed with DCD. Other studies on handwriting in children with DCD found that these children were slower, less accurate, and less fluent in handwriting (Blank, Smit-Engelsman, Polatajko, & Wilson, 2012; Chang & Yu, 2010; Rosenblum & Livneh-Zirinski, 2008; Prunty, Barnett, Wilmut, & Plumb, 2013). Our study also found that children in special education were less accurate and made more errors on the writing task than typically developing children

160

M. Jongbloed-Pereboom et al. / Research in Developmental Disabilities 36 (2015) 150–161

during the entire writing task for both conditions. However, our results did not confirm differences in movement duration. Initially, during training, children in special education were slower than typically developing children only in the explicit condition. In the implicit condition we did not expect to find differences in movement duration between group due to how the implicit training trials were presented. When children were familiar with the letter-like patterns (in the test phase), this difference disappeared for the explicit condition and there were also no differences between groups in movement duration for the implicit test trials. Prunty et al. (2013) discussed that children were not slower on the actual writing task, but took more time to pause in between writing words. This could explain why we did not find differences in movement duration. In our study, we only measured time on task, not time in between trials. A strength of the study is that we tested a heterogeneous group of children, representative of the classroom situation in special education, in particular for schools that focus on children with physical and/or multiple disabilities. Although differences between implicit and explicit learning in handwriting are not yet clear, it was evident that implicit training was effective and feasible for both the study and control group. Furthermore, for children who experience difficulties in learning it could be more rewarding and motivational to learn in an implicit way. Because errors and feedback are reduced in implicit learning, children do not have to experience task failures (Capio, Sit, Abernethy & Masters, 2012), which could have a positive effect on children’s self esteem. Moreover, implicit learning is more robust under stress than explicit learning (Mullen, Hardy, & Oldham, 2007). Therefore, we would like to recommend teachers and specialists to implement more different, and implicit teaching methods in classroom. In future studies, it would be interesting to extent the amount and time of practice during learning. Most investigations on learning have multiple training sessions, e.g. different blocks during a serial reaction time task (e.g. Wilson et al., 2003) or a training period of several days with a pre- and post-test (e.g. Capio et al., 2013). Our study investigated motor learning within a single training session. A learning effect was found within that one session, but retention of the writing patterns was not assessed. Extending the training period and adding a retention test could give more insight into which ways implicit and explicit learning differ, and the role of working memory. Furthermore, IQ and age effects should also be assessed in explicit learning since explicit learning is not only dependent on working memory (Maxwell et al., 2003), but also on age and IQ (Meulemans et al., 1998; Reber et al., 1991). 5. Conclusion To conclude, the control group performed better than the study group both in implicit and explicit learning. Interestingly, children in the study group were able to learn both implicitly and explicitly, although children in the study group showed larger differences in learning in the explicit than the implicit condition compared to the control group. This was also partly due to how training trials in the implicit condition were presented. The effects of motor performance and visual working memory were not as expected, motor performance was to a great extent related to the writing task, visual working memory was not. These results are not in line with the current theory on implicit and explicit learning. More research is warranted to further investigate differences in implicit and explicit learning and the role of working memory therein, especially in populations of atypically developing children. Acknowledgements This study was funded by the Faculty of Social Sciences of the Radboud University Nijmegen. We would like to thank all the children and their parents and teachers for participating in the study, Wouter Hulstijn for his help in setting up the experiment and the Matlab analysis, and Mandy Bentvelsen and Kim Ruijs for helping collect data. References Alloway, T. P. (2007a). Automated working memory assessment. London: Pearson Assessment. Alloway, T. P. (2007b). Working memory, reading and mathematical skills in children with developmental coordination disorder. Journal of Experimental Child Psychology, 96, 20–36. Baddeley, A. D. (2001). Is working memory still working? American Psychologist, 56, 851–864. Bax, M., Goldstein, M., Rosenbaum, P., Leviton, A., & Paneth, N. (2005). Proposed definition and classification of cerebral palsy. Developmental Medicine & Child Neurology, 47, 571–576. Beery, K. E., & Beery, N. A. (2004). Beery VMI: The Beery-Buktenica development test of visual-motor integration (5th ed.). Minneapolis: Pearson. Berninger, V. W., Abbott, R. D., Nagy, W., & Carlisle, J. (2010). Growth in phonological, orthographic, and morphological awareness in grades 1 to 6. Journal of Psycholinguistic Research, 39, 141–163. Berry, D. C., & Dienes, Z. (1993). Towards a working characterisation of implicit learning. In D. C. Berry & Z. Dienes (Eds.), Implicit learning: Theoretical and empirical issues. Hove: Lawrence Erlbaum Associates. Blank, R., Smits-Engelsman, B., Polatajko, H., & Wilson, P. (2012). European Academy for Childhood Disability (EACD): Recommendations on the definition, diagnosis and intervention of developmental coordination disorder. Developmental Medicine and Child Neurology, 54, 54–93. Bosga-Stork, I., Overvelde, A., Van Bommel-Rutgers, I., Van Cauteren, M., Halfwerk, B., Nijhuis-van der Sanden, R., et al. (2009). Inventarisatie van verwijzingspatroon, onderzoek en behandeling van kinderen met schrijfproblemen. Een digitale enqueˆte Nederlands Tijdschrift voor Kinderfysiotherapie, 2, 14–18. Brown, J., Aczel, B., Jime´nez, L., Kaufman, S. B., & Grant, K. P. (2010). Intact implicit learning in autism spectrum conditions. Quarterly Journal of Experimental Psychology, 63, 1789–1812. Candler, C., & Meeuwsen, H. (2002). Implicit learning in children with and without developmental coordination disorder. American Journal of Occupational Therapy, 56, 429–435.

M. Jongbloed-Pereboom et al. / Research in Developmental Disabilities 36 (2015) 150–161

161

Capio, C. M., Poolton, J. M., Sit, C. H. P., Eguia, K. F., & Masters, R. S. W. (2013). Reduction of errors during practice facilitates fundamental movement skill learning in children with intellectual disabilities. Journal of Intellectual Disability Research, 57, 295–305. Capio, C. M., Sit, C. H. P., Abernethy, B., & Masters, R. S. W. (2012). The possible benefits of reduced errors in the motor skill acquisition of children. Sports Medicine, Arthroscopy, Rehabilitation, Therapy & Technology, 4, 1–4. Chang, S. H., & Yu, N. Y. (2010). Characterization of motor control in handwriting difficulties in children with or without developmental coordination disorder. Developmental Medicine and Child Neurology, 52, 244–250. Clements, S. D. (1966). Minimal brain dysfunction in children: Terminology and identification. Washington, DC: US Department of Health Education and Welfare. Cowan, N., & Alloway, T. P. (2009). The development of working memory. In M. L. Courage & N. Cowan (Eds.), Development of memory in infancy and childhood (2nd ed.). Hove, England: Psychology Press. Doyle, A. E. (2006). Executive functions in attention-deficit/hyperactivity disorder. Journal of Clinical Psychiatry, 67, 21–26. Dunford, C., Missiuna, C., Street, E., & Sibert, J. (2005). Children’s perceptions of the impact of Developmental Coordination Disorder on activities of daily living. British Journal of Occupational Therapy, 68, 207–214. Engle, R. W., Carullo, J. J., & Collins, K. W. (1991). Individual differences in working memory for comprehension and following directions. Journal of Educational Research, 84, 253–262. Feder, K. P., & Majnemer, A. (2007). Handwriting development, competency, and intervention. Developmental Medicine and Child Neurology, 49, 312–317. Gathercole, S. E., & Alloway, T. P. (2008). Working memory and learning: A practical guide. London: Sage Publications. Graham, S., Harris, K. R., Mason, L., Fink-Chorzempa, B., Moran, S., & Saddler, B. (2008). How do primary grade teachers teach handwriting? A national survey. Reading and Writing, 21, 49–69. Hammerschmidt, S. L., & Sudsawad, P. (2004). Teachers’ survey on problems with handwriting: Referral, evaluation, and outcomes. American Journal of Occupational Therapy, 58, 185–192. Henry, L. A. (2001). How does the severity of a learning disability affect working memory performance? Memory, 9, 233–247. Henry, L. A., & MacLean, M. (2002). Working memory performance in children with and without intellectual disabilities. American Journal on Mental Retardation, 107, 421–432. Hoy, M. P., Egan, M. Y., & Feder, K. P. (2011). A systematic review of interventions to improve handwriting. Canadian Journal of Occupational Therapy, 78, 13–25. Jenks, K. M., De Moor, J., & Van Lieshout, E. C. D. M. (2009). Arithmetic difficulties in children with cerebral palsy are related to executive function and working memory. Journal of Child Psychology and Psychiatry, 50, 824–833. Karlsdottir, R., & Stefansson, T. (2002). Problems in developing functional handwriting. Perceptual and Motor Skills, 94, 623–662. Maehler, C., & Schuchardt, K. (2009). Working memory functioning in children with learning disabilities: Does intelligence make a difference? Journal of Intellectual Disability Research, 53, 3–10. Masters, R. S. W. (1992). Knowledge, nerves and know-how: The role of explicit versus implicit knowledge in the breakdown of a complex motor skill under pressure. British Journal of Psychology, 83, 343–358. Mathiowetz, V., Volland, G., Kashman, N., & Weber, K. (1985). Adult norms for the Box and Block Test of manual dexterity. American Journal of Occupational Therapy, 39, 386–391. Maxwell, J. P., Masters, R. S. W., & Eves, F. (2003). The role of working memory in motor learning and performance. Consciousness and Cognition, 12, 376–402. Meulemans, T., Van der Linden, M., & Perruchet, P. (1998). Implicit Sequence Learning in Children. Journal of Experimental Child Psychology, 69, 199–221. Mullen, R., Hardy, L., & Oldham, A. (2007). Implicit and explicit control of motor actions: Revisiting some early evidence. British Journal of Psychology, 98, 141–156. http://dx.doi.org/10.1348/000712626X114336 Overvelde, A., & Hulstijn, W. (2011). Learning new movement patterns: A study on good and poor writing comparing learning conditions emphasizing spatial, timing and abstract characteristics. Human Movement Science, 30, 731–744. Prunty, M. M., Barnett, A. L., Wilmut, K., & Plumb, M. S. (2013). Handwriting speed in children with Developmental Coordination Disorder: Are they really slower? Research in Developmental Disabilities, 34, 2927–2936. Reber, A. S., Walkenfield, F. F., & Hernstadt, R. (1991). Implicit and explicit learning: Individual differences and IQ. Journal of Experimental Psychology, 17, 888–896. Rosenblum, S., & Livneh-Zirinski, M. (2008). Handwriting process and product characteristics of children diagnosed with developmental coordination disorder. Human Movement Science, 27, 200–214. Rourke, B. P., Ahmad, S. A., Collins, D. W., Hayman-Abello, B. A., & Warriner, E. M. (2002). Child clinical/pediatric neuropsychology: Some recent advances. Annual Review of Psychology, 53, 309–339. Rubin, N., & Henderson, S. E. (1982). Two sides of the same coin: Variation in teaching methods and failure to learn to write. Special Education: Forward Trends, 9, 17–24. Smits-Engelsman, B. C., Niemeijer, A. S., & Van Galen, G. P. (2001). Fine motor deficiencies in children diagnosed as DCD on poor grapho-motor ability. Human Movement Science, 20, 161–182. Steenbergen, B., van der Kamp, J., Verneau, M., Jongbloed-Pereboom, M., & Masters, R. S. W. (2010). Implicit and explicit learning: Applications from basic research to sports for individuals with impaired movement dynamics. Disability and Rehabilitation, 32, 1509–1516. Straub, K., & Obrzut, J. E. (2009). Effects of cerebral palsy on neuropsychological function. Journal of Developmental and Physical Disabilities, 21, 153–167. Tamopol, L., & Tamopol, M. (1977). Brain function and reading disabilities. Baltimore: University Park Press. Van Halteren-van Tilborg, I. A. D. A., Scherder, E. J. A., & Hulstijn, W. (2007). Motor-skill learning in Alzheimer’s disease: A review with an eye to the clinical practice. Neuropsychology Review, 17, 203–212. Vinter, A., & Detable, C. (2003). Implicit learning in children and adolescents with mental retardation. American Journal on Mental Retardation, 108, 94–107. Vinter, A., & Perruchet, P. (2000). Implicit learning in children is not related to age: Evidence from drawing behavior. Child Development, 71, 1223–1240. Watanabe, K., Ikeda, H., & Miyao, M. (2010). Learning efficacy of explicit visuomotor sequences in children with attention-deficit/hyperactivity disorder and Asperger syndrome. Experimental Brain Research, 203, 233–239. Willingham, D. B. (1998). A neuropsychological theory of motor skill learning. Psychological Review, 105, 558–584. Wilson, P. H., Maruff, P., & Lum, J. (2003). Procedural learning in children with developmental coordination disorder. Human Movement Science, 22, 515–526. Wilson, P. H., Ruddock, S., Smits-Engelsman, B., Polatajko, H., & Blank, R. (2013). Understanding performance deficits in developmental coordination disorder: A meta-analysis of recent research. Developmental Medicine and Child Neurology, 55, 217–228.