Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Jul 1.
Published in final edited form as: Dev Psychol. 2015 Jun 1;51(7):879–888. doi: 10.1037/a0039424

Development of Early Handwriting: Visual-Motor Control During Letter Copying

Jennifer E Maldarelli 1, Björn A Kahrs 2, Sarah C Hunt 3, Jeffrey J Lockman 4
PMCID: PMC4478098  NIHMSID: NIHMS693484  PMID: 26029821

Abstract

Despite the importance of handwriting for school readiness and early academic progress, prior research on the development of handwriting has focused primarily on the product rather than the process by which young children write letters. In contrast, in the present work, early handwriting is viewed as involving a suite of perceptual, motor and cognitive abilities, which must work in unison if children are to write letters efficiently. To study such coordination, head-mounted eye-tracking technology was used to investigate the process of visual-motor coordination while kindergarten children (N=23) and adults (N=11) copied individual letters and strings of letters that differed in terms of their phonemic properties. Results indicated that kindergarten children were able to copy single letters efficiently, as did adults. When the cognitive demands of the task increased and children were presented with strings of letters, however, their ability to copy letters efficiently was compromised: children frequently interrupted their writing mid-letter, whereas they did not do so on single letter trials. Yet, with increasing age, children became more efficient in copying letter strings, in part by using vision more prospectively when writing. Taken together, the results illustrate how the coordination of perceptual, motor and cognitive processes contributes to advances in the development of letter writing skill.

Keywords: eye-tracking, visual-motor integration, early handwriting development

Despite the increased use of technology in society today, handwriting remains one of the most direct and efficient forms of graphic communication (Feder & Majnemer, 2007; Stevenson & Just, 2014). Acquisition of handwriting skill early in the educational process is fundamental to future academic success across many domains (Cahill, 2009; Graham, Harris, & Fink, 2000; Harvey & Henderson, 1997; Simner, 1982, 1983). For instance, the frequency with which kindergarten children make printing errors involving a change in letter shape is correlated with reading, language, and mathematics achievement in kindergarten and first grade (Simner, 1982). Furthermore, continued production of these errors in first grade is correlated with academic achievement as late as sixth grade (Moore & Rust, 1989).

Given the importance of handwriting for academic achievement, what abilities contribute to the early acquisition of this fundamental skill? Handwriting is typically taught through letter and word copying tasks (Huber & Headrick, 1999) and like reading, depends upon the integrated functioning of a suite of motor, perceptual, cognitive and linguistic abilities. When copying letters and words, children need not only control the fine movements and force of the fingers and hand, but attend to the relevant stimulus, keep in working memory the target to be copied, coordinate visual and manual movements, and integrate this perceptual and motor information with linguistic rules about orthographic structure (e.g., in English, letters are written successively from left to right). Coordinating all of these elements when copying letters and words will help ensure that handwriting is fluent, automatic and legible-- key markers of mastery of this skill (Stevenson & Just, 2014).

In the present paper, we consider how young children, who are in the process of learning to write, begin to coordinate these elements. Specifically, we focus on how young handwriters couple eye and hand movements while copying letters and words, and how they incorporate information about linguistic and orthographic structure when doing so. For comparison purposes, we also examine a group of adults on similar handwriting tasks.

Handwriting and Visual-Motor Integration

Visual-motor integration is known to play a crucial role in mastering the mechanical aspects of handwriting (Daly, Kelly, & Kraus, 2003; Kaiser, Albaret, Doudin, 2009; Tseng & Murray, 1994; Volman, van Schendel, & Jongmans, 2006). Kindergarten children with higher scores on standardized tests of visual-motor integration write faster (Tseng & Chow, 2000) and produce more legible handwriting (Cornhill & Case-Smith, 1996; Weil & Amundson, 1994) than children with lower scores. Furthermore, research using fMRI has demonstrated a close link between the visual and motor systems involved in handwriting, with handwriting practice leading to subsequent enhanced activation of the visual association cortex during letter perception in preschool children (James, 2010).

Despite this and other research suggesting that visual-motor integration is critical for handwriting fluency, little is known about how the visual and motor systems actually work together to guide handwriting. Handwriting research has traditionally taken a product-oriented approach, evaluating the accuracy and legibility of completed written products (Alston, 1985; De Ajuriaguerra & Auzias, 1975; Hamstra-Bletz & Blote, 1990). Yet to understand the normative development of handwriting skill and design more targeted strategies for instruction and remediation, a process-oriented approach must also be taken, in which investigators consider how children coordinate vision and manual action when writing (e.g., see Bouwien, Smits-Engelsman, & Van Galen, 1997).

While little research currently exists regarding how vision is used during handwriting, there has been extensive research on how vision is used during the closely related task of reading (Norton & Wolf, 2012; Rayner, 1978; 1986) and other everyday tasks that involve manual action (Hayhoe, 2000; Land, 2009). Consideration of how a process-oriented approach is used in these literatures can inform study of early handwriting development. To illustrate, in the reading literature, the focus has extended beyond whether children are able to read a word or sentence to questions about the patterning of eye movements that characterize the process of efficient reading. It is known, for instance, that younger readers evidence shorter perceptual spans, shorter saccades, and longer and more frequent fixations than more skilled readers (Starr & Rayner, 2001). As children become more proficient readers, they make fewer regressions--that is, re-fixations of letters that have already been read (Taylor, 1965; Rayner, 1978). Collectively, this work suggests that developmental changes in visual attention during the period when children are learning to write letters and words may promote fluency and automaticity in handwriting.

In a similar vein, eye-tracking research conducted with adults as they perform common tasks suggests that they employ eye movements to plan as well as guide ongoing manual actions (Hayhoe, 2000; Hayhoe, Shrivastava, Mruczek, & Pelz, 2003; Land, 2009; Land & Hayhoe, 2001; Land, Mennie, & Rusted, 1999). For instance, when sketching from a model, adult artists plan their actions by fixating the exact feature of the model they intend to copy prior to beginning to draw as well as use vision online, following their hand as it completes the drawing action (Miall & Tchalenko, 2001). In a block copying task, Ballard and colleagues (Ballard, Hayhoe, & Pelz, 1995) found that adults also used vision to reduce the memory demands of the task by making two fixations, rather than a single fixation, before beginning to build the model: the first fixation to encode the color of the block and the second to encode the location of the block. Applied to handwriting, these findings suggest that the prospective use of vision to plan and guide action may promote efficiency and fluency when copying letters and words. Even very young children are capable of using vision prospectively when engaging in actions such as reaching and locomotion (Franchak, Kretch, Soska, & Adolph, 2011; Lockman, Ashmead & Bushnell, 1984; von Hofsten & Fazel-Zandy, 1984), but little is known about the development of young children’s ability to use vision prospectively during handwriting.

The Current Study

In the current study, we contrast how novice and experienced handwriters (i.e., children and adults) use vision to inform manual action during the completion of letter and word copying tasks. We used a copying task because such tasks are one of the most commonly employed methods of handwriting instruction in schools (Huber & Headrick, 1999) and have been shown to be more effective than tracing tasks in facilitating handwriting development (Askov & Greff, 1975; Hirsch & Neidermeyer, 1973). The two main goals of this study were to (1) explore how variation in the linguistic/orthographic properties of the stimulus to be copied influences how vision is deployed during copying and (2) examine how children and adults might differ in their use of vision when gathering the information necessary to copy stimuli of varying length (e.g., single letters and letter triads).

To study the degree to which novice and experienced writers integrate linguistic/orthographic information with visuomotor actions during handwriting, we systematically varied the phonological properties of the stimuli to be copied. Based on reading research that suggests children learn to phonologically chunk strings of individual letters to read them as one unit, (Cutler, Dahan, & Van Donselaar, 1997; Jorm & Share, 1983; Taft, 1991), we suspected that variation in the phonological properties of the stimulus to be copied might influence how vision is used to acquire information from the stimulus and subsequently affect handwriting actions. To explore this issue, we presented participants with real words, pronounceable nonsense strings, and non-pronounceable nonsense strings, all three letters in length (letter triads). James, James, Jobard, Wong, and Gauthier (2005) suggest that at the neurological level, adults process pronounceable nonsense strings in a similar manner to that of real words; non-pronounceable nonsense strings, however, are likely processed differently. Therefore, in the present study we were interested in examining how visuomotor coupling during letter copying may be enhanced (e.g., greater efficiency in terms of fewer initial fixations and re-fixations of the stimulus) as the letter triads became more word-like.

Prior research has also shown that increased cognitive demands can disrupt motor planning and the execution of manual actions in young children (Boudreau & Bushnell, 2000). To study the effects of cognitive demand on visual-motor coupling during handwriting, we presented participants with both single letters and letter triads to copy. Due to the increased cognitive demands associated with copying from a set of three letters as opposed to a single letter alone, we expected that children would take longer to copy a single letter and exhibit more re-fixations to the stimulus when the letter was embedded in the context of three letters than when it was presented alone.

To address these issues about the development of visual-motor coupling during handwriting, we used a recently developed wireless and lightweight head-mounted eye tracker that can be easily worn by young children as well as adults (Franchak et al., 2011). Using eye-tracking technology allowed us to examine the patterning of participants’ eye movements as they planned and executed handwriting actions. To our knowledge, few, if any, studies have used head-mounted eye-tracking technology to explore the development of handwriting in kindergarten children. The current study thus represents one of the first empirical descriptions of how young children use vision to guide their handwriting actions.

Method

Participants

The final sample consisted of 23 five-year-old children (mean age = 64.9 months, range 60.1 - 70.9 months; 11 females) currently enrolled in kindergarten and 11 adults (mean age = 23.4 years, range = 19.6 – 32.7 years; seven females) recruited from a university psychology department. Given that the kindergarten curricula utilized by participating schools focused on teaching letter and word writing, we chose to recruit kindergarten-age children. The child sample was 66% Caucasian, 14% African-American, 7% Hispanic, 10% Multiracial, and 3% Asian; the adult sample was 83% Caucasian and 17% Asian. Seven additional children who completed the study were excluded from analyses due to either an inability to hold their head still such that the eye-tracker could be calibrated (N = 3), an insufficient number of trials with a satisfactory track of the participant’s eye movements (N = 2), or technical / equipment malfunctions (N=2). Child participants were recruited from schools in an area surrounding a university in the Southeastern United States serving mainly middle-class families. All participants were tested at a university lab in sessions lasting approximately 30-45 minutes. For child participants, at least one parent was present during testing. All children received a toy for participating.

Materials and Equipment

Participants were asked to copy a series of visual stimuli consisting of three single letters (C,B,T), three real words (CAT, BED, TOY), three pronounceable nonsense strings (CAG, BEF, TOJ), and three non-pronounceable nonsense strings (CGA, BFE, TJO). The three simple real words included in this study were selected from the early emergent reader stage of the Reading A – Z program, a standardized program for reading and vocabulary instruction, such that they would likely be familiar to kindergarten children (Holl & Morgan, 2002). Pronounceable and non-pronounceable nonsense words were then created by modifying the three real words. The 12 visual stimuli were printed in black ink on a piece of laminated white cardstock (27.9 cm × 8.9 cm; 26.18° × 8.48° visual angle). Letters were spaced 3.8 cm (3.63° visual angle); if a box were drawn around each letter, the approximate size of each letter would be 6.7 cm × 8.9 cm, or 6.4° × 8.5° visual angle. An upright easel was used rather than a traditional desk and chair in which participants look down as they write to facilitate tracking of eye movements (see Figure 1).

Figure 1.

Figure 1

Open in a new tab

Coded areas of interest.

Eye-tracking apparatus

Participants wore a lightweight, wireless, head-mounted eye-tracker (Positive Science), which gave them the space and freedom of movement necessary to complete the handwriting task (see Figure 2). The basic structure of the eye-tracking headgear resembled a pair of eyeglasses and rested comfortably on the nose and ears. Mounted on the headgear were two miniature cameras. A scene camera (54.4° horizontal × 42.2° vertical field view) mounted at eyebrow level above the participant’s right eye recorded the participant’s view of the world and an infrared eye camera mounted on a flexible wire just below the participant’s right eye recorded eye movements. The eye-tracker had a sampling frequency of 30 Hz (Franchak et al., 2011). Information from both cameras was transmitted to a computer in real time where Yarbus software integrated the information to create a Quicktime video, which consisted of the scene video with a superimposed crosshair indicating gaze location.

Figure 2.

Figure 2

Open in a new tab

Child wearing head-mounted eye-tracker (Positive Science, LLC). The authors received signed consent for this child’s likeness to be published in this article.

Eye-tracker calibration

Upon arrival to the lab, a ten-minute familiarization period was conducted during which the eye-tracking equipment was presented and the participants and/or parents were allowed to ask questions. Then, the experimenter placed the headgear on the participant and calibration began. The calibration procedure was modeled after previous research that employed head-mounted eye-tracking systems with infants and young children (Evans, Saint-Aubin, & Landry, 2009; Franchak et al., 2011). Children were seated in a wooden chair (30.5 cm in height) 60 cm in front of the previously described easel. Adults sat on the floor 60 cm in front of the easel, such that adult participants maintained the same eye level with calibration stimuli as the children. Next, calibration stimuli (a red heart, yellow sun, blue butterfly, purple star, and green tree; 3.8 cm in diameter; 3.6° visual angle) were fastened with magnetic tape and placed in the four corners and center of the easel. Participants were asked to hold their head still and look at each shape one at a time, as indicated by a research assistant. Between three and five points were registered for calibration (a minimum of three are required).

After calibration was obtained, the five shapes were removed and three calibration check stimuli were randomly placed on the easel surface to assess the quality of the calibration. These stimuli consisted of a blue circle (3.8 cm in diameter; 3.6° visual angle) with a small black dot in the center. Participants were asked to look at each black dot one at a time, again as indicated by the research assistant. If the crosshair indicating the participant’s gaze location fell inside the blue circle, the calibration was determined to be spatially accurate within 1.8° visual angle (minimum radius of error in any direction). If the crosshair fell outside the edge of the blue circle, the calibration process was repeated until an accurate calibration (i.e., within 1.8° visual angle) was obtained.

The eye-tracker was unable to attain a reliable track during 0.76% of adults trials (1/132) and 18% of child trials (50/276), as children touched or bumped the eye-tracking glasses and tended to look down as they wrote, thus making it more difficult for the eye-tracker to track their pupils in comparison to those of adults.

Testing Procedure

After calibration, the writing stimuli were presented to each participant in a random order. Stimuli were affixed at the top of the easel via magnetic tape and covered with a piece of laminated black card stock of identical size. In order to provide each participant with a specific area in which to copy the stimulus (32 cm × 13 cm; 29.9° × 12.4° visual angle), two pieces of laminated black cardstock were affixed to the easel with magnetic tape (see Figure 1). At the start of each trial, each participant was given a black dry erase marker, the black cover over the stimulus was removed, and the participant was asked to copy the stimuli in the blank white space below. Trials lasted until the participant finished writing. At the end of the testing session, the calibration check stimuli were again randomly presented for all participants to test whether the eye-tracker had shifted or otherwise lost accuracy during the testing. An accurate calibration was maintained for all trials.

Dependent Measures

Trial duration was initially scored and defined as the period of time from when the black cover over the stimulus was removed until the participant finished writing. Next, gaze videos were scored for eye-movements via OpenSHAPA, a computerized video coding system. A fixation was defined as when the gaze crosshair, indicating the participant’s gaze location, rested stable for three or more consecutive frames (i.e., greater than 99.9 ms) (Franchak et al., 2011). Fixations to three areas of interest were recorded: stimulus, current writing area, and future writing area (see Figure 1) across two time periods, before writing and during writing. Fixations to each letter of the stimulus were coded separately. For example, if a participant fixated first on ‘C’ and then shifted their gaze to fixate on ‘A’ in ‘CAT’, this was coded as two fixations to the stimulus. When the gaze crosshair rested on a letter, the crosshair had to intersect the contour of the letter or be inside the letter to be coded.

‘Before writing’ was defined as the time period after the black cover was removed and the stimulus was revealed, until the participant’s marker touched the whiteboard surface. ‘During writing’ was defined as the time period from when the participant’s marker contacted the whiteboard the first time, until the marker left the whiteboard after completing the last letter. In the before writing phase, we recorded the total number of fixations made to letters of the stimulus and the duration of each fixation. In the during writing phase, the total number and duration of fixations made to letters of the stimulus, current writing area (e.g., the area where the participant is actively writing) and future writing area (e.g., the area to the right of where the participant is currently writing, presumably where future letter(s) would be written) were coded (see Figure 1).

A primary coder coded all gaze videos, with a second reliability coder independently scoring 20% of all gaze videos. Correlation coefficients ranged from r = .91 - .93, with respect to the number of fixations made to different areas of interest, and r = .90 - .95, with respect to mean fixation duration.

Results

Overview

Data analyses were conducted to investigate the underlying patterning of visuomotor coordination when young children and adults copied individual letters and letter triads. We initially examined overall differences in trial duration, as function of age, stimuli characteristics, and writing phase (before writing, during writing). Subsequently, we conducted analyses of the eye-tracking data to understand how children and adults used vision to gather information from the stimulus and guide writing actions. For all analyses, single letter trials were analyzed individually as they were quantitatively different from letter triad trials.

GEE Modeling

In order to account for the possible correlation among data from trial to trial and to accommodate missing data from trials during which the eye-tracker could not attain a reliable track, Generalized Estimating Equation (GEE) models were used (Liang & Zeger, 1986). For every analysis, we used an iterative stepwise elimination procedure (Agresti & Finlay, 2009). The statistical modeling process began with every predictor (i.e., age, phase, writing condition, and trial order) and associated interactions included in the model. At each iteration, the least significant predictor at the highest level of interaction was removed. This eventually resulted in a model composed entirely of those factors and interactions that significantly affected the dependent variable. Thus, in the following section we highlight only statistically significant findings. Alpha was set to .01.

For each variable of interest, data from the child sample were first compared to data from the adult sample (we refer to differences between children and adults as effects of “age group” throughout the results). The data from the child sample were then analyzed separately, with age treated as a continuous predictor (referred to as “children’s age”) to explore developmental changes within the child group. Findings from the child sample alone are reported only if the results revealed additional information relative to the combined adult and child analysis. Trial order was included as a predictor in analyses performed on the child sample alone only, as we did not hypothesize that in-session practice would influence the behavior of adults who were are already proficient writers.

Trial Duration

Single letters

A GEE model was used to examine the effects of age group (children, adults) and phase (before writing, during writing), on trial duration for the single letter trials. A significant main effect of age group was found χ2(1) = 32.1, p < .001, such that it took children a greater amount of time to copy a single letter than adults (see Figure 3). A follow-up GEE model was used to examine the effects of children’s age (treated continuously), phase, and trial order on trial duration for the single letter trials in the child sample alone. A significant main effect of trial order was found, χ2(1) = 6.62, p = .01, such that children copied single letters more efficiently as they progressed throughout the task and acquired practice (MTrial 1 = 2647 ms; MTrial 2 = 2327 ms; MTrial 3 = 1975 ms).

Figure 3.

Figure 3

Open in a new tab

Mean duration of total trial by age group, phase, and writing condition. Error bars represent standard error.

Letter triads

A GEE model was used to examine the effects of age group (children, adults), phase (before writing, during writing), and writing condition (real words, pronounceable nonsense strings, and non-pronounceable nonsense strings) on trial duration for the letter triad trials. Significant main effects of age group, χ2(1) = 66, p < .001, and phase, χ2(1) = 333, p < .001, were qualified by an interaction of age group and phase, χ2(1) = 13, p < .001. The interaction indicates that although children took significantly longer to complete both the before and during writing phases of the task in comparison to adults, this difference was significantly more pronounced during writing (see Figure 3).

Not surprisingly, the above results demonstrate that children take longer than adults to copy both single letters and letter triads. But in addition to this difference in overall trial duration, children and adults showed markedly different patterns in regard to the duration of each phase when completing letter triad trials as compared to single letter trials (see Figure 3). Adults efficiently processed letter triads, spending only an additional 200 ms before writing as compared to single letter trials, and requiring approximately three times as long during writing relative to single letter trials. This was not the case for the children. Children spent nearly twice as long examining the stimulus before beginning to write letter triads relative to single letters. Children also spent six times as long actually copying letter triads relative to single letters during writing. Given the relatively lengthy durations of each phase of the task for the children as compared to the adults, we next examined the source(s) of these differences and addressed the apparent additional challenges children encounter when copying letter triads.

Visual Information Gathering and Visual Guidance during Handwriting

The following analyses address where children and adults looked as they performed the task. Visual attention to the target stimulus was considered to reflect visual information gathering and occurred both before writing and during writing (e.g., when individuals interrupted their writing to look back at the target stimulus). Visual attention to the current and future writing areas was considered to reflect visual guidance and planning during the act of handwriting. Variation in frequency, duration and sequencing of children’s and adults’ fixations to these areas of interest in the different task phases was examined.

Information Gathering: Analysis of Fixations to the Stimulus

Single letters

Participants’ fixations to the stimulus letter on single letter trials lasted on average 463 ms. Notably, neither the number of fixations nor the average fixation duration differed between children and adults. A significant main effect of phase revealed that both children and adults made significantly more fixations to the stimulus letter before beginning to write (M = .93) than during writing (M = .07), χ2(1) = 34.28, p < .001. The average number of fixations made to the stimulus letter before writing is just less than one because both children and adults sometimes made sub-fixation level saccades (i.e., a participant’s eye gaze rested on the stimulus letter for less than 100 ms). This occurred on four adult (N=3) and eight (N=5) single letter trials. Collectively, these analyses indicate that both children and adults efficiently acquired the visual information necessary to copy single letters, making approximately one fixation to the stimulus letter before writing and rarely looking back up at the stimulus letter during writing.

Letter triads

The mean duration of a fixation to triad stimulus letter was 322 ms and did not vary by writing condition or phase, but approached significance for age group such that children made fixations to triad stimuli of slightly greater duration than those of adults (MAdults = 284 ms, MChildren = 336 ms, p = .011). A follow up analyses of the child sample alone revealed a significant interaction of children’s age and phase (χ2(1)=17.95, p <.001). Post-hoc testing indicated that with increasing age, the mean duration of fixations made to letters of triad stimuli by children increased in the during writing phase (p=.006), but did not significantly change in the before writing phase (p = .37) (see Figure 4). One possible reason for this difference is that during writing younger children often interrupted their writing mid-letter, while older children did so less frequently (see following section -- Information Gathering: Patterns of Visual Fixations).

Figure 4.

Figure 4

Open in a new tab

Interaction of age and phase on the mean duration of fixations made to letters of the stimulus by children copying letter triads. Individual data points represent the mean fixation duration for each child and the lines are the predicted mean durations as a function of age derived from the GEE model. Individual trials are shown by the letters.

Next, we examined variation in the number of fixations participants made to letter triads as a function of age group, writing condition, and phase. The GEE analysis revealed a significant main effect of writing condition, χ2(2) = 17.6, p < .001, suggesting that both children and adults were sensitive to the phonological properties of the stimuli. All participants made significantly fewer fixations to the stimulus when copying real words (M = 1.66), as compared to pronounceable nonsense strings (M = 1.92, p = .002), and non-pronounceable nonsense strings (M = 2.33, p < .001). The number of fixations participants made to the pronounceable and non-pronounceable nonsense strings did not significantly differ from one another.

This GEE model also revealed significant main effects of age group, χ2(1) = 36.7, p < .001, and phase, χ2(1) = 14.2, p < .001, which were qualified by a significant Age Group × Phase interaction, χ2(1) = 8.6, p < .005 (see Figure 5). Pairwise comparisons used to test the effect of age group within phase revealed that in both phases, children made significantly more fixations to letters of the stimulus than adults did (ps < .002); however, this difference was significantly more pronounced in the during writing phase. This pattern suggests that when copying letter triads, children and adults differed in how they acquired visual information from the target stimulus during both task phases. We explore this possibility next by considering how participants acquired visual information about letter triads both before and during writing.

Figure 5.

Figure 5

Open in a new tab

Mean number of fixations made to letters of the stimulus triad in each phase, across writing condition. Error bars represent standard error.

Information Gathering: Patterns of Visual Fixations

Before writing

To understand how adults and children gathered information from the triad stimulus before beginning to write, we investigated the order in which they fixated the individual letters of the stimulus. Data were collapsed across writing condition due to the large number of possible different combinations of fixations for a given trial—for example, for a series of n fixations there are 3n different possible combinations in the task.

Adults most commonly made just one (36 trials, 39.13%) or two (31 trials, 33.70%) fixations to letter(s) of the triad stimulus before writing. Their first fixation was generally to the second letter in the triad stimulus (78 trials, 84.78%). After the initial fixation, adults often began to write (30 trials, 32.61%) or fixated the first letter and then proceeded to write (18 trials, 19.56%). In contrast, children typically made several fixations to letters of the triad stimulus before writing (2 fixations on 53 trials (34.87%), 3 fixations on 41 trials (26.97%), 4 or more fixations on 44 trials (28.95%), while rarely looking at a single letter and immediately proceeding to write (14 trials, 9.12%). Like adults, however, children often started by fixating the middle letter (77 trials, 50.66%) and from there typically fixated the first letter (63 of the 77 trials; 81.81%). While children typically made several more fixations after these initial two, it is interesting to note that children’s last fixation was most frequently to the first, or leftmost letter in the triad, which they were just about to begin writing (91 trials, 60%). This suggests that the kindergarten children in this sample have already internalized the orthographic rule that English proceeds from left to right.

During writing

During writing, adults and children differed dramatically in their visual fixation patterns. Once adults had begun to write, they rarely looked back up to re-fixate the stimulus. In contrast, children did so often (see Figure 5). In fact, children re-fixated at least one letter of the stimulus during writing on nearly all codeable letter triad trials (148/164, 89.02%). We next explored children’s fixation patterns during writing in greater detail.

First, we examined at what point in the writing process children stopped to re-fixate letters of the triad stimulus. Specifically, we asked if children were more likely to make re-fixations between writing letters or while writing a letter. (Note that when doing so while in the middle of writing a letter, children paused the writing action rather than continuing to write). A GEE model was used to analyze the likelihood that a child would re-fixate a triad stimulus letter as a function of age, writing condition, and time period (between writing letters, while writing a letter). This analysis revealed a significant main effect of time period (χ2(1) = 9.34, p =.002), which was qualified by a significant interaction of age and time period, χ2(1) = 13.52, p <.001. As indicated in Figure 6, although children across age were more likely to re-fixate the stimulus between writing letters rather than while writing a letter, this difference became more pronounced with age.

Figure 6.

Figure 6

Open in a new tab

The probability that a child will look back up at a letter triad stimulus during writing and re-fixate a letter in both time periods, across writing condition. The lines indicate the predicted likelihood derived from the GEE model; data points are also included for each child (represented by letters) for each trial (n=9) for both the between and while writing a letter time periods.

Next, we explored which letter(s) children actually re-fixated when they looked back up at the stimulus in each time period. To address this issue, we coded each fixation as being to a past, current, or future letter and analyzed each time period (between writing letters, while writing a letter) separately using GEE models. When children stopped while in the middle of writing a letter to re-fixate a triad stimulus letter, they were equally likely to fixate the letter they were currently copying (P = .60) or a future letter (P = .70). In contrast, when children stopped in between letters, they were more likely to re-fixate a future letter (P = .91) than a letter they had just completed (P = .30), (main effect of letter location, χ2(1) = 48.26, p < .001). (The probabilities in these analyses sum to more than 1.00 because the categories are not mutually exclusive; e.g., it is possible for a child to fixate both a current and future letter during a single trial).

Taken together, our analyses of the differences between how children and adults acquire visual information from triad stimuli indicate that adults used vision more efficiently (e.g. made fewer fixations) both before and during. However, children did show improvements in efficiency and planning with age. Older children were more likely to finish writing a single letter without looking back up at the triad stimulus than were younger children. Also, when older children did look back up after finishing the letter, they predominantly used vision prospectively to fixate the next letter(s) to be copied. These findings mirror the well documented finding in the reading literature indicating that as children become more proficient readers they make fewer regressions, or re-fixations of previously read words (Rayner, 1986; Taylor, 1965). Our results also indicate that children’s eye-gaze behaviors were strategic, reflecting awareness of the English orthographic rule that words are written left to right.

Visual Guidance of Action Execution: Analysis of Fixations to the Writing Areas

The final step of analysis addressed how vision was deployed online and prospectively to the writing areas when creating the written product (see Figure 1). The current writing area was defined as the area where the participant was actively writing; the future writing area was defined as the area to the right of where the participant was actively writing, where the next letter(s) would presumably be written (see Figure 1). Data yielded from single letter trials and from the last letter of letter triad trials were not included in these analyses since they did not require participants to plan for writing future letters.

GEE models revealed that all participants made significantly more fixations to the current writing area (M = 3.13) than to the future writing area (M = 1.18), χ2(1) = 106, p < .001 as well as fixations of greater duration to the current writing area (M = 1378 ms) than to the future writing area (M = 298 ms), χ2(1) =790, p < .001. Not surprisingly, however, children (M = 1116 ms) made longer fixations to both writing areas than adults (M = 608 ms), χ2(1) = 123, p < .001, likely reflecting the greater amount of time that children spent planning and executing handwriting actions as compared to adults.

Discussion

Handwriting is an assembled skill, dependent upon the integrated functioning of perceptual, motor, cognitive and linguistic abilities. Handwriting that is mechanically fluent and efficient enables individuals to devote attentional resources to the content that they wish to express through their manual actions on a page. The mastery of handwriting to the point of fluency and efficiency, however, takes young children many months, if not years (Stevenson & Just, 2014). By focusing on how visual-motor coupling during handwriting is influenced by linguistic and orthographic information as well as cognitive demands, the present study offers new insights into the challenges that young children encounter and the process by which they master this motor skill. The findings also highlight the adaptive strategies that young children evidence during this learning process. To our knowledge, the present study is the first to employ head-mounted eye-tracking technology to investigate the process by which young children integrate visual, motor and cognitive-linguistic information during handwriting.

Overall, children took longer to copy both individual letters and letter strings than adults did. When the copying task was broken down on the basis of where individuals were looking (to the stimulus or writing area(s)), our results indicate that children encountered different challenges depending on where they were looking. When looking at the stimulus, children showed a far greater number of fixations to the stimulus, but only slightly longer fixations than adults. This finding is consistent with data on children’s visual patterns during word reading, where differences between children and adults are most pronounced when considering the number of fixations that they make (Starr & Rayner, 2001). This is most likely because young children’s perceptual span is smaller than that of adults (Haikio, Bertram, Hyona, & Niemi, 2009; Rayner, 1986) and thus children devote most of their attention to foveal information. In contrast, when looking to the writing area, children evidenced the same number of fixations as adults, but children’s fixations lasted far longer, suggesting that children’s slower manual actions may have influenced how long they were attending to the area in which they were writing. Taken together, the results identify some reasons why young children’s use of vision to guide copying was notably less efficient than that of adults.

Children’s abilities were taxed even further when they were faced with the task of copying letter strings, as opposed to single letters. Children took six times as long to copy letter triads as compared to a single letter (see Figure 3). Furthermore, when copying the letter triads, they frequently interrupted their writing mid-letter to re-fixate the stimulus, while they almost never did so during single letter trials. This pattern suggests that the increased cognitive demands associated with attending to and copying three letters as compared to one adversely affected children’s performance. Additionally, older children were less impacted than younger children by this increased cognitive demand. When copying letter triads, all children frequently re-fixated the stimulus during writing; however, the youngest children were most likely to interrupt their writing mid-letter to do so. In contrast, older children were more likely to finish copying a letter before doing so. And when children did look back up after finishing the letter, they predominantly used vision prospectively to look at the next letter(s) to be copied. The latter finding suggests that during handwriting, young children are beginning to show the ability to engage in prospective control, a hallmark of skilled manual behavior in other task domains (Land, 2009).

Furthermore, children were able to derive benefits from the linguistic and orthographic properties of English when copying letter strings. In the before writing phase, children, just like the adults in the study, exhibited fewer fixations to real words than to non-word letter strings, suggesting they benefited from differentiating the linguistic structure of the target stimulus. Children also manifested knowledge of English orthography in their eye-gaze patterns. They most frequently fixated the leftmost letter of a letter triad immediately before beginning to write, suggesting that they recognized that English is written from left to right. Additionally, the sequence of children’s first two fixations to the stimulus before writing was similar to that of adults: Children began by fixating the middle of the letter string and subsequently moved to the leftmost letter. In sum, although relatively shorter perceptual spans and developing manual skills may have resulted in young children coupling vision and manual actions somewhat inefficiently when they copy letter strings, our results suggest that they have already begun to master some basic English language orthographic rules that help guide visual information gathering when copying words.

Conclusions & Implications

To our knowledge, this study provides the first empirical description of how typically developing young children use vision to plan and guide their handwriting actions and introduces a new method for examining handwriting in children. Much like research on early reading (Norton & Wolf, 2012; Rayner, 1986), the present study suggests that early handwriting is dependent on the coordination of perceptual, motor and cognitive processes. The coordination of these processes is critical to efficient and fluid handwriting. We further suggest that the eye-tracking methods detailed in this study can illuminate how this coordination process unfolds in young children as they are learning to write letters and words.

Additionally, these methods might be especially useful for studying children who experience difficulty when learning to write, such as children with dysgraphia who have underlying visual-spatial difficulties (Berninger, 2004). It should be noted that in the present study, all of the children were in high quality kindergarten programs and likely had a great deal of experience writing letters in pre-kindergarten, in the home, or in both. In future research, it would also be informative to examine the development of visual-motor control in handwriting in populations of children from more diverse socio-economic environments, where exposure to early literacy in the home and access to high quality early education may vary (Magnuson, Meyers, Ruhm, & Waldfogel, 2004).

In conclusion, despite the increased reliance on keyboards for typing by school-age children, handwriting still remains a fundamental skill to be fostered in early childhood. Handwriting proficiency is critically linked to academic achievement (Cahill, 2009; Moore & Rust, 1989; Simner 1982, 1983), and reading has been shown neurologically to be more strongly related to handwriting than typing (James & Engelhardt, 2012). By incorporating eye-tracking technology into the study of early handwriting, investigators can gain insights into how children begin to coordinate the perceptual, motor and cognitive abilities that underlie this critical skill and develop research-informed strategies to promote handwriting in young children.

Acknowledgments

This research was supported in part by two National Institutes of Health grants (5RO1HD043842 & 5RO1HD067581) and by the Tulane University Flowerree Summer Research award. We would like to thank the parents and children who participated in this study.

Contributor Information

Jennifer E. Maldarelli, Department of Psychology, Tulane University

Björn A. Kahrs, Department of Psychology, Tulane University

Sarah C. Hunt, Department of Psychology, Tulane University

Jeffrey J. Lockman, Department of Psychology, Tulane University

References

  1. Agresti A, Finlay B. Statistical Methods for the Social Sciences. Dellen Publishing Company; San Francisco: 1986. [Google Scholar]
  2. Alston J. The handwriting of seven to nine year olds. British Journal of Special Education. 1985;12:68–72. doi:10.1111/j.1467-8578.1985.tb00609.x. [Google Scholar]
  3. Askov EN, Greff KN. Handwriting: Copying versus tracing as the most effective type of practice. The Journal of Educational Research. 1975;69:96–98. [Google Scholar]
  4. Ballard D, Hayhoe M, Pelz J. Memory representations in natural tasks. Cognitive Neuroscience. 1995;7:66–80. doi: 10.1162/jocn.1995.7.1.66. doi:10.1162/jocn.1995.7.1.66. [DOI] [PubMed] [Google Scholar]
  5. Berninger VW. Understanding the “graphia” in developmental dysgraphia. In: Dewey D, Tupper D, editors. Developmental motor disorders: A neuropsychological perspective. The Guilford Press; New York, New York: 2004. pp. 328–350. [Google Scholar]
  6. Boudreau JP, Bushnell EW. Spilling thoughts: Configuring attentional resources in infants’ goal-directed actions. Infant Behavior & Development. 2000;23:543–566. doi:10.1016/S0163-6383(01)00059-5. [Google Scholar]
  7. Bouwien CM, Smits – Engelsman, Van Galen GP. Dysgraphia in children: Lasting psychomotor deficiency or transient developmental delay? Journal of Experimental Child Psychology. 1997;67:164–184. doi: 10.1006/jecp.1997.2400. doi:10.1006/jecp.1997.2400. [DOI] [PubMed] [Google Scholar]
  8. Cahill SM. Where does handwriting fit in? Strategies to support academic achievement. Intervention in School and Clinic. 2009;44:223–228. doi:10.1177/1053451208328826. [Google Scholar]
  9. Cornhill H, Case-Smith J. Factors that relate to good and poor handwriting. The American Journal of Occupational Therapy. 1996;50:732–739. doi: 10.5014/ajot.50.9.732. doi:10.5014=ajot.50.9.732. [DOI] [PubMed] [Google Scholar]
  10. Cutler A, Dahan D, Van Donselaar W. Prosody in the comprehension of spoken language: A literature review. Language and Speech. 1997;40:141–201. doi: 10.1177/002383099704000203. doi:10.1177/002383099704000203. [DOI] [PubMed] [Google Scholar]
  11. Daly CJ, Kelley GT, Krauss A. Relationship between visual-motor integration and handwriting skills of children in kindergarten: A modified replication study. The American Journal of Occupational Therapy. 2003;57:459–462. doi: 10.5014/ajot.57.4.459. doi:10.5014/ajot.57.4.459. [DOI] [PubMed] [Google Scholar]
  12. De Ajuriaguerra J, Auzias M. Preconditions for the development of writing in the child. Foundations of Language Development: A multidisciplinary Approach. 1975;2:311–328. [Google Scholar]
  13. Evans MA, Saint-Aubin J, Landry N. Letter names and alphabet book reading by senior kindergartners: An eye movement study. Child Development. 2009;80:1824–1841. doi: 10.1111/j.1467-8624.2009.01370.x. doi:10.1111/j.1467-8624.2009.01370.x. [DOI] [PubMed] [Google Scholar]
  14. Feder KP, Majnemer A. Handwriting development, competency, and intervention. Developmental Medicine & Child Neurology. 2007;49:312–317. doi: 10.1111/j.1469-8749.2007.00312.x. doi:10.1111/j.1469-8749.2007.00312.x. [DOI] [PubMed] [Google Scholar]
  15. Franchak JM, Kretch KS, Soska KC, Adolph KE. Head-mounted eye tracking: A new method to describe infant looking. Child Development. 2011;82:1738–1750. doi: 10.1111/j.1467-8624.2011.01670.x. doi:10.111/j.1467-8624.2011.01670.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Graham S, Harris KR, Fink B. Is handwriting causally related to learning to write? Treatment of handwriting problems in beginning writers. Journal of Educational Psychology. 2000;92:620. doi:10.1037/0022-0663.92.4.620. [Google Scholar]
  17. Hamstra-Bletz L, Blote AW. Development of handwriting in primary school: A longitudinal study. Perceptual and Motor Skills. 1990;70(3):759–770. doi: 10.2466/pms.1990.70.3.759. doi:10.2466/PMS.70.3.759-770. [DOI] [PubMed] [Google Scholar]
  18. Harvey C, Henderson S. Children’s handwriting in the first three years of school: consistency over time and its relationship to academic achievement. Handwriting Review. 1997;11:8–25. [Google Scholar]
  19. Haikio T, Bertram R, Hyona J, Niemi P. Development of the letter identity span in reading: Evidence from the eye movement moving window paradigm. Journal of Experimental Child Psychology. 2009;102:167–181. doi: 10.1016/j.jecp.2008.04.002. doi:10.1016/j.jecp.2008.04.002. [DOI] [PubMed] [Google Scholar]
  20. Hayhoe M. Vision using routines: A functional account of vision. Visual Cognition. 2000;7:43–64. doi:10.1080/135062800394676. [Google Scholar]
  21. Hayhoe M, Shrivastava A, Mruczek R, Pelz J. Visual memory and motor planning in a natural task. Journal of Vision. 2003;3:49–63. doi: 10.1167/3.1.6. doi:10:1167/3.1.6. [DOI] [PubMed] [Google Scholar]
  22. Hirsch E, Niedermeyer FC. The effects of tracing prompts and discrimination training on kindergarten handwriting performance. The Journal of Educational Research. 1973;67(2):81–86. [Google Scholar]
  23. Huber RA, Headrick AM. Handwriting Identification: Facts and Fundamentals. CRC Press; Boca Raton, FL: 1999. [Google Scholar]
  24. Holl B, Morgan F. Reading A – Z. 2002 Retrieved from http://www.readinga-z.com/
  25. James KH. Sensori-motor experience leads to changes in visual processing in the developing brain. Developmental Science. 2010;13:279–288. doi: 10.1111/j.1467-7687.2009.00883.x. doi:10.1111/j.1467-7687.2009.00883.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. James KH, Engelhardt L. The effects of handwriting experience on functional brain development in pre-literate children. Trends in Neuroscience and Education. 2012;1:32–42. doi: 10.1016/j.tine.2012.08.001. doi:10.1016/j.tine.2012.08.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. James KH, James TW, Jobard G, Wong AC, Gauthier I. Letter processing in the visual system: Different activation patterns for single letters and strings. Cognitive, Affective, & Behavioral Neuroscience. 2005;5:452–466. doi: 10.3758/cabn.5.4.452. doi:10.3758/CABN.5.4.452. [DOI] [PubMed] [Google Scholar]
  28. Jorm AF, Share DS. An invited Article: Phonological recoding and reading acquisition. Applied Psycholinguistics. 1983;4:103–147. doi:10.1017/S0142716400004380. [Google Scholar]
  29. Kaiser ML, Albaret JM, Doudin PA. Relationship between visual-motor integration, eye-hand coordination, and quality of handwriting. Journal of Occupational Therapy, Schools, & Early Intervention. 2009;2:87–95. doi:10.1080/19411240903146228. [Google Scholar]
  30. Land MF. Vision, eye movements, and natural behavior. Visual Neuroscience. 2009;26:51–62. doi: 10.1017/S0952523808080899. doi:10.1017/S0952523808080899. [DOI] [PubMed] [Google Scholar]
  31. Land MF, Hayhoe M. In what ways do eye movements contribute to everyday activities? Vision Research. 2001;41:3559–3565. doi: 10.1016/s0042-6989(01)00102-x. doi:10.1016/S0042-6989(01)00102-X. [DOI] [PubMed] [Google Scholar]
  32. Land MF, Mennie N, Rusted J. The roles of vision and eye movements in the control of activities of daily living. Perception. 1999;28:1311–1328. doi: 10.1068/p2935. doi:10.1068/p2935. [DOI] [PubMed] [Google Scholar]
  33. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73:13–22. Doi:10.1093/biomet/73.1.13. [Google Scholar]
  34. Lockman J, Ashmead D, Bushnell E. The development of anticipatory hand orientation during infancy. Journal of Experimental Child Psychology. 1984;37(1):176–186. doi: 10.1016/0022-0965(84)90065-1. doi:10.1016/0022-0965(84)90065-1. [DOI] [PubMed] [Google Scholar]
  35. Magnuson KA, Meyers MK, Ruhm CJ, Waldfogel J. Inequality in preschool education and school readiness. American Educational Research Journal. 2004;41:115–157. doi:10.3102/00028312041001115. [Google Scholar]
  36. Miall RC, Tchalenko J. A painter’s eye movements: A study of eye and hand movement during portrait drawing. Leonardo. 2001;34:35–40. doi:10.1162/002409401300052488. [Google Scholar]
  37. Moore RL, Rust JO. Printing errors in the prediction of academic performance. Journal of School Psychology. 1989;27:297–300. doi:10.1016/0022-4405(89)90044-7. [Google Scholar]
  38. Norton ES, Wolf M. Rapid automatized naming (RAN) and reading fluency: implications for understanding and treatment of reading disabilities. Annual Review of Psychology. 2012;63:427–52. doi: 10.1146/annurev-psych-120710-100431. doi:10.1146/annurev-psych-120710-100431. [DOI] [PubMed] [Google Scholar]
  39. Rayner K. Eye movements in reading and information processing. Psychological Bulletin. 1978;85(3):618–660. doi:10.1037/0033-2909.124.3.372. [PubMed] [Google Scholar]
  40. Rayner K. Eye movements and the perceptual span in beginning and skilled readers. Journal of Experimental Child Psychology. 1986;41:211–236. doi: 10.1016/0022-0965(86)90037-8. doi:10.1016/0022-0965(86)90037-8. [DOI] [PubMed] [Google Scholar]
  41. Simner ML. Printing errors in kindergarten and the prediction of academic performance. Journal of Learning Disabilities. 1982;15:155–159. doi: 10.1177/002221948201500306. doi:10.1177/0022219482019482015000306. [DOI] [PubMed] [Google Scholar]
  42. Simner ML. The warning signs of school failure: An updated profile of the at-risk kindergarten child. Topics in Early Childhood Special Education. 1983;3:17–27. doi:10.1177/027112148300300305. [Google Scholar]
  43. Starr M, Rayner K. Eye movements during reading: Some current controversies. Trends in Cognitive Sciences. 2001;5(4):156–163. doi: 10.1016/s1364-6613(00)01619-3. doi:10.1016/S1364-6613(00)01619-3. [DOI] [PubMed] [Google Scholar]
  44. Stevenson N, Just C. In early education, why teach handwriting before keyboarding? Early Childhood Education Journal. 2014;42:49–56. doi:10.1007/s10643-012-0565-2. [Google Scholar]
  45. Taft M. Reading and the mental lexicon. Lawrence Erlbaum Associates; Hove, U.K.: 1991. [Google Scholar]
  46. Taylor SE. Eye movements while reading: Facts and fallacies. American Educational Research Journal. 1965;2:187–202. [Google Scholar]
  47. Tseng MH, Chow SM. Perceptual-motor function of school-age children with slow handwriting speed. The American Journal of Occupational Therapy. 2000;54:83–88. doi: 10.5014/ajot.54.1.83. doi:10.5014/ajot.54.1.83. [DOI] [PubMed] [Google Scholar]
  48. Tseng MH, Murray EA. Differences in perceptual-motor measures in children with good and poor handwriting. Occupational Therapy Journal of Research. 1994;14(1):19–36. [Google Scholar]
  49. Volman MJM, van Schendel BM, Jongmans MJ. Handwriting difficulties in primary school children: A search for underlying mechanisms. The American Journal of Occupational Therapy. 2006;60:451–460. doi: 10.5014/ajot.60.4.451. doi:10.5014/ajot.4.451. [DOI] [PubMed] [Google Scholar]
  50. von Hofsten C, Fazel-Zandy S. Development of visually guided hand orientation in reaching. Journal of Experimental Child Psychology. 1984;38:208–219. doi: 10.1016/0022-0965(84)90122-x. doi:10.1016/0022-0965(84)90122-X. [DOI] [PubMed] [Google Scholar]
  51. Weil MJ, Amundson SJC. Relationship between visiomotor and handwriting skills of children in kindergarten. The American Journal of Occupational Therapy. 1994;48:982–988. doi: 10.5014/ajot.48.11.982. doi:10.5014/ajot.48.11.98. [DOI] [PubMed] [Google Scholar]

RESOURCES