Examining the Association Between the Computer-Aided Scoring and Manual Scoring for the Beery Buktenica Developmental Test of Visual Motor Integration (Beery VMI) and Children’s Handwriting Skills: A Convergent Validity Study

ABSTRACT

Introduction

Visual motor integration (VMI) is an important underlying mechanism for children’s handwriting. The Beery-Buktenica Developmental Test of Visual-Motor Integration (Beery VMI) is commonly used to assess children’s VMI skills, using a manual scoring method which can be prone to individual subjectivity and impact its sensitivity. This study investigated the use of the Computer-Aided Scoring System for scoring the Berry VMI (CASS-Beery VMI) to assess school-aged children’s VMI and handwriting skills. Specifically, the convergent validity between the CASS-Beery VMI and three standardized handwriting assessments was investigated.

Methods

A quantitative cross-sectional design was used. Thirty-five neurotypical Australian students aged 6–9 years completed the Beery-VMI–6th edition, Minnesota Handwriting Assessment (MHA), Evaluation Tool of Children’s Handwriting – Manuscript (ETCH-M) and the Here’s How I Write Assessment – Child version (HHIW-C). The Berry-VMI was scored using its original scoring method and the CASS-Beery VMI. Thirty parents of the child participants completed the Here’s How I Write Assessment – Parent version (HHIW-P). Descriptive statistics, Spearman rho correlations and regression analyses using bootstrapping were completed.

Results

The CASS-Beery VMI error scores were significantly correlated with the Beery-VMI score, and subscales of the MHA, ETCH and HHIW-P scores (r = 0.341– −0.555; p = .045–0.001). The CASS-Beery VMI error scores and Beery VMI raw scores were found to be significant predictors of the ETCH-M number legibility scores (adjusted R² = 0.209; p = 0.017) and MHA form scores (adjusted R² = 0.220; p = 0.007). Individually, the CASS-Beery VMI score was a significant predictor of the ETCH-M word legibility scores (Adjusted R² = 0.305; p = 0.001).

Conclusion

Findings indicate that the CASS-Beery VMI can be used to assess children’s VMI skills. The significant correlations between the CASS-Beery VMI and other standardized handwriting assessments (MHA, ETCH-M and HHIW) adds to the body of convergent validity evidence that the CASS-Beery VMI can be a sensitive scoring method for Beery VMI when assessing children’s handwriting performance.

KEYWORDS:

• Handwriting
• computerized evaluation
• visual-motor integration
• occupational therapy
• children

Introduction

Pediatric occupational therapists work with children to improve their engagement in activities that are appropriate for their age and developmental milestones (Pavlos et al., 2020). For occupational therapists working in educational settings to engage students in academic and non-academic activities, up to 98% of referrals are due to children presenting with handwriting difficulties (Brown & Link, 2016; Case-Smith et al., 2014). As visual-motor integration (VMI) is one of the key foundational skills that underpin functional handwriting skills, investigation of the valid and reliable assessment of VMI skills is warranted.

The Occupation of Handwriting

Handwriting is the written form of communication that is an essential occupation and is a crucial skill for all school-aged children to master (Klein et al., 2021; Piller & Torrez, 2019). Handwriting has an imperative role in accomplishing tasks like writing, spelling, and completing mathematical equations in a timely manner (Brown & Link, 2016; Hunter & Potvin, 2020). Proficient handwriting requires the coordination of sensory, motor, perceptual, cognitive and language abilities, whilst executing the performance components of grip, posture, in-hand manipulation, letter recognition and formation, and sustained attention (Case-Smith et al., 2014; Maldarelli et al., 2014).

Despite advancements in technology and the increased use of computers for producing work, written output remains a significant part of school curriculum (Klein et al., 2021; McCarney et al., 2013; Pavlos et al., 2020). It has been identified that students in Australia spend 85% of their time handwriting (McMaster & Roberts, 2016). This illustrates the importance of mastering the skill of handwriting as it remains a predominant occupation of school students.

The link between handwriting proficiency and a child’s emotional, cognitive, and academic functioning has been well documented throughout the literature (Kadar et al., 2020; Piller & Torrez, 2019). To promote a high level of fluency, spelling, grammar and academic accomplishment, a level of automaticity in handwriting must be achieved (Klein et al., 2021). Subsequently, children who are unable to achieve proficient handwriting are likely to experience academic difficulties leading to challenges with behavioral problems and low self-esteem (Klein et al., 2021).

A qualitative study completed in the United States that explored teachers’ perspectives on children’s handwriting found that deficits in handwriting can significantly impact a child’s academic performance as they are unable to meet the required school demands, potentially resulting in academic failure (Nye & Sood, 2018). These findings are supported by other research which found that students’ perceptions of poor handwriting and academic failure can result in suboptimal self-esteem levels and contribute to a negative mind-set due to inaccurate perspectives of their capabilities (Kadar et al., 2020; Piller & Torrez, 2019). Similarly, a student may become frustrated, less motivated to perform school activities and may begin to avoid completing school-based tasks (McCarney et al., 2013).

Skilled handwriting entails the integration and development of fine-motor coordination, visual-motor, cognitive and perceptual skills including vestibular, proprioceptive, and tactile registration (Pavlos et al., 2020). These skills in addition to motor planning and temporal, spatial and force elements are inherent in the task of handwriting (McCarney et al., 2013). Fine-motor skills including in-hand manipulation, opposition of thumb to fingers to grasp the pencil and asymmetrical bilateral integration (using one hand for paper stabilization and the other hand for holding the pencil) are all required for achieving proficiency in handwriting (Axford et al., 2018).

Cognitive processes involving working memory, planning and ideas generation are required to produce skilled handwriting (McCarney et al., 2013). It is important that the cognitive process of forming letters and words reaches a level of automaticity for children in elementary school, to facilitate the quality of their handwriting and academic work (McCarney et al., 2013). When producing letters, children not only require the use of fine-motor movements, but they also need to attend to relevant stimuli, use their working memory, coordinate visual movements, and simultaneously integrate perceptual and motor information (Maldarelli et al., 2014). The coordination of these skills when writing will ensure that a child’s handwriting is automatic, fluent, and legible (Maldarelli et al., 2014).

Visual-Motor Integration

Visual-motor integration (VMI) is defined as the process of integrating visual perceptual and motor coordination skills simultaneously, and visual perception refers to the brain’s capacity to receive and interpret what is seen (Beery & Beery, 2010; Klein et al., 2011). This involves the ability of an individual’s eyes and hands to work together in an efficient and smooth pattern to master the visual perceptual skills that assist with learning the visual representation of letters and guiding letter construction (Abou-El-Saad et al., 2017; Klein et al., 2011). Likewise, motor coordination involves the body movements required to perform motor output which enables the necessary motor patterns needed to form letters and numbers (Brown & Link, 2016). VMI skills allow individuals to organize, monitor and execute motor tasks; for example, tying shoelaces or catching a ball (Abou-El-Saad et al., 2017). As a child’s ability to copy and recognize letters improves, so too do their VMI skills resulting in increased capability to print and write letters accurately and efficiently (Abou-El-Saad et al., 2017).

Assessment tools are often used by occupational therapists to develop therapeutic goals, inform intervention processes, and measure the success of therapy (Pavlos et al., 2020). Measuring the effectiveness of therapeutic interventions frequently depends on the therapist’s access to quality information and assessment. Consequently, occupational therapists often rely on sound assessment tools and measures to implement appropriate interventions to address the client’s needs (Pavlos et al., 2020). With advances in technology, the use of computerized scoring has enhanced accuracy and reliability and minimized subjectivity pertaining to manual scoring when grading participants’ performance on standardized tests. The psychometric body of evidence of any standardized assessment is dynamic and is always growing. The same applies to specific VMI assessments.

Computer-Aided Scoring System for the Beery-Buktenica Developmental Test of Visual-Motor Integration (CASS-Beery VMI)

The Beery-VMI is an assessment tool commonly used in occupational therapy practice to inform children’s handwriting skills. However, the sensitivity of Berry-MVI in detecting handwriting difficulties has been an ongoing debate (Pfeiffer et al., 2015; Prunty et al., 2016). The Berry-VMI requires the therapist to manually score a child’s attempt at copying a series increasingly more complex geometric shapes and designs using specific scoring criteria detailed in its manual and the therapists’ own knowledge of the assessment and VMI skills. Hence, the scoring can be time consuming, and the results can be influenced by assessors’ subjectivity and bias. This potentially impacts the Berry-VMI’s sensitivity and possibly contributes to the weak evidence about the relationship between VMI and handwriting skills in children. Finding an efficient, reliable, and objective way to score children’s VMI skills is therefore needed. Technology advancements provide an opportunity to reduce scoring time and address the need to minimize the potential subjectivity and bias of manual scoring that can impact the sensitivity and scoring accuracy of the Berry-VMI.

The CASS-Beery VMI is a computerized scoring system to assess performance of Beery VMI by analyzing picture copying images (Liu, 2019). The CASS-Beery VMI was based on the image registration technique, which was originally created to align an object with its ideal shape (Zitova & Flusser, 2003). In the current study, participants completed the Beery VMI test on paper sheets, and the products of picture copying were scanned using ratio 1:1 to create enhanced digital images (Beery & Beery, 2010). Using CASS-Beery VMI, the scanned images were put through a transforming process that optimally matched the scanned images to the corresponding templates by resizing, rotating, and repositioning.

Following the transforming process of image registration, two scores – an error score and an effort score – were calculated to evaluate the amount of discrepancy based on the pixel-by-pixel comparison (Liu, 2019). Error score represented the discrepancy between the transformed image and the template, and effort score represented the discrepancy between the transformed image and the original drawn picture to indicate the adjustment involved in resizing, rotating, and repositioning. Higher error and effort scores indicate that participants demonstrate poor VMI ability. Within the Beery-VMI test, each picture copying item received both an error and effort score. Participants’ average scores were then calculated to provide them with separate overall error scores and effort scores. However, there is limited publishing evidence available about the convergent validity of the CASS-Beery VMI, and hence warrant the current study.

This study aims to examine the convergent validity between the CASS-Beery VMI and three standardized handwriting assessments by examining the relationship between school-aged children’s performance on the CASS-Beery VMI and their English language handwriting performance on three standardized handwriting assessments (e.g., the Evaluation Tool of Children’s Handwriting – Manuscript [ETCH-M], Minnesota Handwriting Assessment [MHA] and Here’s How I Write Assessment [HHIW] scale). It will also investigate the relationship between the children’s performance on the Beery-VMI– 6th edition, using its original scoring method, and the CASS-Beery VMI. The research questions posed are:

1. What are the relationships between children’s VMI skills assessed using the CASS-Beery VMI and the original Beery-VMI scoring method with Beery-VMI?

2. Is the performance of children on the CASS-Beery VMI related to their English language handwriting performance as measured by the ETCH-M, MHA, and HHIW (parent and child versions)?

Methods

Research Design

A quantitative cross-sectional research design was chosen to address the research questions.

Participants

A non-convenience sample of 35 neurotypical children (female = 17; male = 18) aged six to nine years studying in grades prep to grade three at a state primary school located in metropolitan Melbourne, Australia, were recruited (180 invitations sent; response rate 19%). The inclusion criteria for children were: (1) no known developmental or learning disability; (2) enrolled in grade prep to three at a state primary school; and (3) having sufficient English language skills to participate in data collection. The parent/guardians of the child participants were also invited to participate in the study. Parent participant inclusion criteria were: (1) having awareness of their child’s handwriting skills; and (2) having sufficient English language skills for data collection. Thirty of the 35 parents (85%) participated. Only the children who met the inclusion criteria and had a parent/guardian agreed to participate were included as participants of this study.

Instrumentation

In this study, children completed the Beery-VMI (Beery & Beery, 2010), the ETCH-M (Amundson, 1995), the MHA (Reisman, 1999) and HHIW (Goldstand et al., 2013) self-assessments. Parents completed a demographic questionnaire and the HHIW assessment for their child. The demographic questionnaire provided information about their child’s age, grade level, gender, languages spoken at home and any developmental delays or learning difficulties that may be present.

Beery-Buktenica Development Test of Visual-Motor Integration

The Beery-VMI is a standardized, norm-referenced assessment which integrates both visual and motor abilities (Beery & Beery, 2010). The tool is designed for use with participants aged two to 100 and comprises 30 items and three subsections which include visual-motor integration, visual perception and motor coordination (Axford et al., 2018; Hunter & Potvin, 2020). For the purposes of this study, the VMI subtest which takes approximately 10–15 minutes to administer was used. The VMI component of the Beery-VMI involves participants copying both simple line drawings and more complex figures, and manually scores children from zero (no shape resemblance) to one (shape resemblance) based on their performance (Duiser et al., 2014). The tool has exhibited excellent test-retest reliability (0.84–0.88), internal consistency (0.81–0.82) and inter-rater reliability (0.90) (Axford et al., 2018). Additionally, the tool has strong construct validity based on the construction of test items and strong predictive validity regarding academic outcomes (Pavlos et al., 2020). Existing research that explores the strong psychometric properties of the Beery-VMI indicates that it is a “gold standard” assessment. The CASS-Beery VMI was also applied for scoring the Berry-VMI picture copying images completed by the participants (referred to as CASS-Beery VMI score/s hereafter). Participants received an error and effort score for each of the 24 VMI test design items, and an average was then calculated for each participant to give them an overall error and effort score.

Evaluation Tool of Children’s Handwriting–Manuscript

The ETCH assessment is a standardized, criterion-referenced tool that characterizes handwriting difficulties by judging a child’s handwriting speed and legibility (Amundson, 1995). The tool is useful for evaluating manuscript and cursive handwriting; for this study, however, only the manuscript handwriting test was used. The tool is intended for use with children aged six to 12 years of age, taking approximately 30 minutes to complete and involving six tasks that examine alphabetical and numerical writing, dictation, sentence composition and near-point and far-point copying (Hunter & Potvin, 2020). Students are scored on three aspects which include word, letter, and numeral legibility (Brossard-Racine et al., 2012). Numerous studies have examined the psychometric properties of the ETCH and have found it to be a reliable and valid measure (Hunter & Potvin, 2020). Test-retest reliability ranged from 0.63 to 0.77, it has intra-rater reliability of 0.80 and inter-rater reliability of 0.84 when completed with experienced raters (Hunter & Potvin, 2020).

Minnesota Handwriting Assessment

The MHA is a norm-referenced assessment tool that assesses a child’s manuscript handwriting through rate and legibility while completing near-point copying tasks (Reisman, 1999). The tool is based on six criteria which include legibility, rate, form, size, spacing and alignment (Pfeiffer et al., 2015). The tool is designed to assess students in grades one and two and takes approximately two and a half minutes to complete and ten minutes to score (Saleem & Gillen, 2018).

The MHA has well documented psychometric properties with inter-rater reliability ranging from 0.87 to 0.99 and intra-rater reliability from 0.97 to 1.00 (Pfeiffer et al., 2015). Additionally, the MHA has test-retest reliability of 0.62 to 0.89 and moderate-to-high criterion validity (Saleem & Gillen, 2018).

Here’s How I Write Assessment

The HHIW is a standardized, criterion-referenced self-assessment that examines children’s perceptions of their own handwriting capabilities using a picture card interview format (referred as HHIW-C hereafter) (Cermak & Bissell, 2014). The assessment is used with students in grades two to five to self-assess their handwriting performance (Cermak & Bissell, 2014). Participants are presented with 24 cards simultaneously that explore various aspects of handwriting, including staying on the lines and letter formation (Cermak & Bissell, 2014). The picture-card interview shows images of a child who has mastered the skill and a child having difficulty with completing the skill. Children are then instructed to select the image that best represents their abilities. A parent version of the HHIW (HHIW-P) has been adapted from the teacher version for use in the current study. Items, meaning and the overall content in the assessment remained the same. The parent version asks a series of statements in parallel with the 24 cards used with the children. For each of the statements, parents are instructed to tick the “always” or “usually” box based on which best represents their child’s handwriting abilities. The tool takes approximately 10–15 minutes to administer (Cermak & Bissell, 2014). The HHIW is reported to have acceptable content validity (0.80 to 0.90) and test-retest reliability (0.65 to 1.00) and good responsiveness (Goldstand et al., 2018). Lastly, the HHIW is designed to discriminate between children with or without handwriting difficulties (Goldstand et al., 2018).

Data Collection and Management

The student researcher administered the Beery-VMI, ETCH, MHA and HHIW. Prior to administering these assessments, the student researcher had successfully completed two pediatric placements demonstrating appropriate skills to engage children. The student researcher also received trainings through reviewing the assessment manuals and practicing administering and scoring the assessments under supervision and guidance provided by two experienced pediatric occupational therapists. All assessments were administered face-to-face with child participants over two 30–minute sessions, conducted in a quiet space in the school at a time that was mutually convenient. The assessment manuals were used to ensure the correct administration specifications were followed and the standardized assessments were randomly ordered across the participants to decrease the risk of test order bias effect arising (Amundson, 1995; Beery & Beery, 2010; Goldstand et al., 2013; Reisman, 1999). The HHIW-P was distributed to child participants to pass onto their parents who received the HHIW-P form, instructions on how to complete the form and a return envelope. Parents were asked to return the completed forms to the provided drop-box located at the school administration office. The student researcher scored the ETCH, MHA, HHIW and manual scoring of the Berry-VMI, while the CASS-Berry VMI was completed by a member of the research team developed the CASS-Berry VMI.

Data Analysis

The CASS-Beery VMI, Beery-VMI, MHA, ETCH-M, HHIW scores, along with the demographic data, were entered onto the Statistical Package for Social Sciences Program (SPSS) version 27 (International Business Machines Corporation [IBM], 2020). Descriptive statistics including the mean, median, standard deviation, and interquartile ranges (IQR) were calculated. Bootstrapping, a statistical procedure that involves resampling a single dataset to create a larger sample size, was applied in this study due to the small sample size (Weisberg, 2010). Spearman’s rho correlations with bootstrapping of up to 1000 participants examined the degree of association between the CASS-Beery VMI scores and the Beery-VMI and between the CASS-Beery VMI scores and the three handwriting assessments (Pierson, 2017). For each handwriting subscale that was significantly correlated with more than one score among the CASS-Beery VMI error and effort scores and the Beery-VMI, a multivariate regression with bootstrapping was conducted. The regression analyses only included the VMI scores that were found to be significantly correlated as independent variables.

Procedures

Ethics was approved by Monash University Human Research Ethics Committee (approval number 26,536) on the 16th December 2020 and the Victorian Department of Education and Training (approval number 2021_004348) on the 15th February 2021. The study requested written consent from parents/guardians and verbal assent from children to participate. Information packs (comprising parent and child explanatory statements, parent and child consent forms and a demographic screening questionnaire) were distributed to 180 students in grade prep to grade three to take home for parents’ consideration of participation.

Results

Table 1 provides the participants’ demographic information and descriptive statistics of the handwriting and VMI assessments and their subscales are reported in Table 2.

Table 1. Participant demographic information (n = 35).

Demographic variable	Frequency	(%)
Child participant’s age (years) 6 7 8 9	7 18 8 2	20 51.5 23 6
Child participant’s gender Male Female	18 17	51.5 48.5
Child participant’s grade level Prep Grade 1 Grade 2 Grade 3	3 17 10 5	8.5 48.5 28.5 14
Child participant’s hand preference in handwriting Right Left	32 3	91.5 8.5

Table 2. Descriptive statistics of participant sample CASS-Beery VMI, beery-VMI, MHA, ETCH-M and HHIW (n = 35, unless otherwise stated).

Instrument subscale and total scores	Mean	Standard deviation	Median	Range (Min-Max)	Interquartile range (Q1, Q3)
CASS-Beery VMI Error total score Effort total score	30.59 123.05	4.62 24.22	29.66 120.87	18.53 (22.05–40.58) 110.98 (83.16–194.14)	26.64, 33.50 107.85, 138.35
Beery VMI Raw score Standard score	15.74 85.40	4.36 20.81	15 89	17 (6–23) 112 (0–112)	12, 19 76, 98
MHA Rate total score Legibility total score Form total score Alignment total score Size total score Spacing total score	32.46 30.74 27.86 26.80 23.40 28.26	3.41 3.38 4.24 5.43 7.15 4.57	34 32 29 28 23 30	12 (22–34) 12 (22–34) 15 (19–34) 20 (14–34) 29 (5–34) 17 (17–34)	34, 34 30, 33 24, 31 24, 31 18, 29 26, 31
ETCH-M Total letter legibility Total word legibility Total number legibility	91.80 90.70 88.30	10.76 9.35 16.52	95.60 94 95.50	55.3 (44.7–100) 33.4 (66.6–100) 54.5 (45.5–100)	88.60, 99.10 85, 100 81.80, 100
HHIW HHIW-C total score HHIW-P total score (n = 30) HHIW level of agreement total score (n = 30)	85.69 76.47 19.53	8.16 9.43 3.76	88 75.5 20	28 (68–96) 44 (59–103) 15 (9–24)	79, 92 69.75, 81.50 17, 23

CASS-Beery VMI=Computer-Aided Scoring System for Beery Buktenica Developmental Test of Visual Motor Integration; Beery-VMI=Beery Buktenica Developmental Test of Visual-Motor Integration; MHA=Minnesota Handwriting Assessment; ETCH-M=Evaluation Tool of Children’s Handwriting – Manuscript; HHIW=Here’s How I Write.

Correlation Results

Table 3 provides the Spearman’s rho correlation results for the association between the Beery-VMI and subscale scores of the ETCH-M, MHA and HHIW, as well as the association between the CASS-Beery VMI scores and the handwriting assessment subscales.

Table 3. Spearman’s rho correlation coefficients (two-tailed) between the VMI assessments and the handwriting assessments after bootstrapping, (n = 35; bootstrapped sample of 1000, unless otherwise stated).

Assessment subscales				Beery VMI raw score	CASS-Beery VMI error score	CASS-Beery VMI effort score
MHA legibility	Correlation coefficient			.299	−.408*	−.174
	Sig. (2–tailed)			.081	.015	.4318
	^aBootstrap	Bias		−.009	.008	.004
		Std. error		.146	.160	.160
		BCa 95% Confidence interval	Lower	.018	−.684	−.473
			Upper	.537	−.049	.157
MHA spacing	Correlation coefficient			.245	−.410*	−.141
	Sig. (2–tailed)			.157	.015	.419
	^aBootstrap	Bias		−.008	.010	.000
		Std. error		.158	.150	.168
		BCa 95% Confidence interval	Lower	−.051	−.677	−.460
			Upper	.513	−.083	.187
MHA alignment	Correlation coefficient			.327	−.424*	−.244
	Sig. (2–tailed)			.055	.011	.158
	^aBootstrap	Bias		−.008	.007	.008
		Std. error		.135	.157	.161
		BCa 95% Confidence interval	Lower	.065	−.709	−.536
			Upper	.548	−.090	.100
MHA size	Correlation coefficient			.200	−.335*	−.102
	Sig. (2–tailed)			.249	.049	.560
	^aBootstrap	Bias		−.008	.004	.006
		Std. error		.161	.175	.161
		BCa 95% Confidence interval	Lower	−.109	−.646	−.401
			Upper	.472	.033	.221
MHA form	Correlation coefficient			.355*	−.463**	−.147
	Sig. (2–tailed)			.037	.005	.399
	^aBootstrap	Bias		−.012	.006	.003
		Std. error		.150	.169	.158
		BCa 95% Confidence interval	Lower	.029	−.772	−.549
			Upper	.603	−.102	.173
MHA rate	Correlation coefficient			.068	−.420*	−.211
	Sig. (2–tailed)			.698	.012	.223
	^aBootstrapc	Bias		−.004^b	.014	.003
		Std. error		.134^b	.157	.179
		BCa 95% Confidence interval	Lower	−.183^b	−.666	−.541
			Upper	.307^b	−.031	.167
ETCH word legibility	Correlation coefficient			.307	−.519**	−.359*
	Sig. (2–tailed)			.073	.001	.034
	^aBootstrap	Bias		−.004	.006	.004
		Std. error		.176	.151	.147
		BCa 95% Confidence interval	Lower	−.071	−.774	−.624
			Upper	.638	−.200	−.050
ETCH letter legibility	Correlation coefficient			.270	−.555**	−.161
	Sig. (2–tailed)			.116	.001	.354
	^aBootstrap	Bias		−.001	.019	.001
		Std. error		.153	.134	.164
		BCa 95% Confidence interval	Lower	−.048	−.766	−.476
			Upper	.537	−.240	.167
ETCH number legibility	Correlation coefficient			.457**	−.341*	.009
	Sig. (2–tailed)			.006	.045	.961
	^aBootstrap	Bias		−.002	.010	−.005
		Std. error		.143	.166	.183
		BCa 95% Confidence interval	Lower	.136	−.622	−.376
			Upper	.715	0.17	.354
HHIW child & parent agreement^c	Correlation coefficient			.243	−.147	.067
	Sig. (2–tailed)			.196	.437	.724
	^aBootstrap	Bias		−.012	.012	.004
		Std. error		.173	.199	.188
		BCa 95% Confidence interval	Lower	−.133	−.514	−.312
			Upper	.556	.263	.433
HHIW child scores	Correlation coefficient Sig. (2–tailed) ^aBootstrap	Bias Std. error BCa 95% Confidence interval	Lower Upper	0.78 .738 .000 .243 −.382 .590	.417 .060 −.013 .165 .015 .682	.249 −.212 .356 .012 .244 −.676
HHIW parent scores^c	Correlation coefficient Sig. (2–tailed) ^aBootstrap	Bias Std. error BCa 95% Confidence interval	Lower Upper	.425 .055 −.020 .217 −.060 .777	−.459* .036 .020 .203 −.763 .026	.219 .340 −.008 .250 −.307 .661
Beery VMI raw score	Correlation coefficient				−.466**	−.203
	Sig. (2–tailed)				.009	.282
	^aBootstrap	Bias			.018	.007
		Std. error			.146	.190
		BCa 95% Confidence interval	Lower		−.697	−.547
			Upper		−.148	.183

^aUnless otherwise noted, bootstrap results are based on 1000 bootstrap samples. ^b Based on 999 samples. * Correlation is significant at the 0.05 level (2–tailed). ** Correlation is significant at the 0.01 level (2–tailed). All correlation coefficients rounded to three decimal places. CASS-Beery VMI=Computer-Aided Scoring System for Beery Buktenica Developmental Test of Visual Motor Integration; Beery-VMI=Beery Buktenica Developmental Test of Visual-Motor Integration; MHA=Minnesota Handwriting Assessment; ETCH-M=Evaluation Tool of Children’s Handwriting – Manuscript; HHIW=Here’s How I Write; Std. error=standard error; BCa=bias-corrected and accelerated. ^c HHIW parent scores and HHIW child & parent agreement, n = 30.

Association Between the CASS-Beery VMI Scores and the Beery-VMI

The results indicated a statistically significant correlation between the CASS-Beery VMI error scores and the Beery-VMI raw scores rho = −0.466, p < 0.01). There was no statistically significant correlation between the CASS-Beery VMI effort scores and the Beery-VMI raw scores.

Association Between the Beery-VMI and the ETCH-M, MHA and HHIW Subscales

The Beery-VMI raw scores were found to have a statistically significant positive correlation with the MHA form scale (rho = 0.355, p < 0.05) and the ETCH-M number legibility (rho = 0.457, p < 0.01). No statistically significant correlations were found between the Berry-VMI raw scores and the HHIW-C and HHIW-P scores and the remaining subscales of the ETCH-M and MHA.

Association Between the CASS-Beery VMI and the ETCH-M, MHA and HHIW Subscales

Statistically significant negative correlations between the CASS-Beery VMI error score and the MHA legibility score rho = −0.408, p < 0.05) and MHA spacing score rho = −0.410, p < 0.05) were obtained, as well as the MHA alignment rho = −0.424, p < 0.05), size rho = −0.335, p < 0.05), form rho = −0.463, p < 0.01) and rate scores rho = −0.420, p < 0.05). There was a statistically significant correlation with the CASS-Beery VMI error scores and ETCH-M word legibility rho = −0.519, p < 0.01), as well as the CASS-Beery VMI error scores and the ETCH-M letter legibility rho = −0.555, p < 0.01). Statistically significant correlations were also found between the CASS-Beery VMI error scores and the ETCH-M number legibility and HHIW-P scores rho = −0.341, p < 0.05;rho = −0.459, p < 0.05, respectively). There were no statistically significant correlations between the CASS-Beery VMI effort scores with the handwriting subscales, besides the ETCH-M word legibility. Additionally, the HHIW-C scores and CASS-Beery VMI error scores were non-significant.

Regression Analysis Results

Table 4 provides the regression analysis results.

Table 4. Multivariate regression analysis between the CASS-Beery VMI error and effort scores and the Beery-VMI with the MHA form and ETCH-M word and number legibility subscales (n = 35; bootstrapped sample of 1000).

	Before bootstrapping					After bootstrapping^b
	B^a	SE B	b	t	p	Bias	SE B	p	BCa 95.0% CI [LL, UL]
Dependent variable: ETCH-M (number legibility)
Independent variables^c: (Constant) Beery VMI raw scores CASS-Beery VMI error scores	95.051 1.522 −1.081	32.602 0.777 0.862	0.383 −.245	2.915 1.960 −1.254	0.008 .063 .223	−0.926 −0.022 0.044	34.299 .712 0.986	0.016 0.056 0.299	[29.522, 159.462] [.192, 2.953] [−2.845, 0.957]
Dependent variable: ETCH-M (word legibility)
Independent variables^d: (Constant) CASS-Beery VMI error scores CASS-Beery VMI effort scores	140.805 −1.253 −0.105	14.581 0.412 0.071	−0.523 −0.255	9.657 −3.044 −1.485	0.000 0.006 −0.281	2.644 0.010 −0.026	15.750 0.409 0.092	0.001 0.006 0.193	[114.813, 179.634] [−2.045, −0.375] [−0.355, −0.004]
Dependent variable: MHA (form)
Independent variables^e: (Constant) Beery VMI raw scores CASS-Beery VMI error scores	35.326 0.225 −0.355	8.183 0.195 0.216	0.233 −0.332	4.317 1.153 −1.642	0.000 0.261 −.417	−.804 .008 .023	11.328 0.195 .295	.015 .269 .264	[11.493, 54.743] [−.137, .638] [−.864, .261]

^aBootstrapping specifications: (a) sampling method: simple; (b) number of samples: 1000; (c) CI level: 95%; and (4) CI type: BCa. ^b Unless otherwise noted, bootstrap results are based on 1000 bootstrap samples. CASS-Beery VMI=Computer-Aided Scoring System for Beery Buktenica Developmental Test of Visual Motor Integration; ETCH-M=Evaluation Tool of Children’s Handwriting–Manuscript; MHA=Minnesota Handwriting Assessment; Beery-VMI=Beery Buktenica Developmental Test of Visual-Motor Integration; B: unstandardized beta coefficient; SE B: standard error for the unstandardized beta; b: standardized beta; t: t test statistic; p: probability statistic; CI: confidence interval; LL: lower limit; UL: upper limit; BCa: bias-corrected and accelerated.

^cR=0.275; R2=0.209; p=0.017

^dR=0.346; R2=0.305; p=0.001

^eR=0.266; R2=0.220; p=0.007

ETCH-M number legibility

Two independent variables (Beery-VMI raw and CASS-Beery VMI error scores) were included in the regression analysis. The regression model was significantly predictive of the ETCH-M number legibility percentage score (Adjusted R² = 0.209, p = 0.017). However, neither of the independent variables (Beery-VMI raw and the CASS-Beery VMI error scores) made a unique significant contribution to the overall regression model’s variance.

ETCH-M word legibility

Two independent variables (CASS-Beery VMI error and effort scores) were involved in the regression equation. The regression model was significantly predictive of the ETCH-M word legibility percentage score (Adjusted R² = 0.305, p = 0.001). The CASS-Beery VMI error score was found to have a statistically significant unique contribution to the overall regression model’s variance (β =-1.25, SE = 0.41, p = 0.006).

MHA form

Two independent variables (Beery-VMI raw and CASS-Beery VMI error scores) were included in the regression analysis. The regression model was significantly predictive of the MHA form (Adjusted R² = 0.220, p = 0.007). Neither the Beery-VMI raw or CASS-Beery VMI error scores made a unique standalone contribution to the overall regression model’s variance.

Discussion

This study adds to the body of convergent validity evidence about the CASS-Beery VMI. In addition, the study provides evidence that the computerized scoring system, CASS-Beery VMI, can be used with Beery-VMI and is a useful tool for screening and evaluating children’s visual-motor skills.

Relationship Between Children’s Performance Based on the CASS-Beery VMI Scores and the Beery-VMI

The CASS-Beery VMI error scores compared to the Beery-VMI represents fair validity rho = −0.466, p < 0.01), while there was low validity between the CASS-Beery VMI effort scores and the Beery-VMI raw scores rho = −0.203). While these findings highlight that there is not a strong association between children’s scores of the CASS-Beery VMI and the Beery-VMI, it indicates the potential that CASS-Beery VMI scores can add additional information to the Beery-VMI for assessing children’s handwriting skills.

This is the first known study to measure the relationship between children’s performance using the original Beery-VMI manual scoring and the CASS-Beery VMI, and as such there is no previously published data available for comparison purposes. While study findings indicate that the CASS-Beery VMI (effort scores) were indicative of ETCH-M word legibility, the CASS-Beery VMI (error scores) appear to be more sensitive than the Beery-VMI scores in the assessment of children’s handwriting skills. Indeed, the CASS-Beery VMI (error scores) correlated with all the subscales of the MHA and ETCH-M and the HHIW-P scores. This also provides evidence of its convergent validity with the three standardized handwriting assessments.

Conversely, the Beery-VMI raw scores only correlated with the MHA form and ETCH-M number legibility. This indicates that the CASS-Beery VMI is appropriate and beneficial for clinical use with Beery-VMI in occupational therapy practice to examine children’s English handwriting formation and number legibility. Applying the CASS-Beery VMI with the Beery-VMI enhances the capacity of the Beery-VMI to infer English handwriting assessment results comprehensively in multiple handwriting elements and ultimately handwriting performance.

A multivariate regression analysis was conducted with the Beery-VMI, CASS-Beery VMI error scores and the handwriting subscales. The results identified that the CASS-Beery VMI error scores and the Beery-VMI are jointly predictive of the ETCH-M number legibility scores as a model. However, the Beery-VMI was a stronger indicator of predicting number legibility than the CASS-Beery VMI error score (Adjusted R² = 0.124, p=0.022; Adjusted R² = 0.111, p = 0.057, respectively). Whilst the Beery-VMI and CASS-Beery VMI error scores were also jointly predictive of the MHA form scores as a model, the CASS-Beery VMI error scores were a stronger indicator of this handwriting element than the Beery-VMI (Adjusted R² = 0.138, p = 0.038; Adjusted R² = 0.071, p = 0.067, respectively). The regression analysis for ETCH-M word legibility identified the CASS-Beery VMI error score as the unique predictor when the effort score was also included in the model. This also provides evidence of the CASS-Beery VMI’s convergent validity as a significant predictor of number legibility.

Children’s CASS-Beery VMI scores are likely to correlate with their performance on handwriting assessments compared to the Beery-VMI due to the CASS-Beery VMI using an objective scoring system and the removal of common scoring disadvantages (such as human error, subjectivity, scoring bias, rater severity or rater leniency) that are present when scoring the Beery-VMI manually (Liu, 2019). These findings are supported by Chang et al. (2009), who explored the utility of computer-assisted assessment to evaluate the handwriting performance of Chinese hand writers in terms of size, spacing, alignment and resemblance to a standard model. Chang et al. (2009) also identified that computerized scoring methods accurately identified the legibility and size of letters and therefore objectively assessed handwriting performance. It was noted that traditional scoring methods relied on the subjective judgment of the scorer, whereas assessments using computerized scoring methods were more automatic and objective (Chang et al., 2009).

This supports our study’s results as the CASS-Beery VMI scores which use a computerized scoring system correlated with handwriting performance to a greater extent than the Beery-VMI scores which use a subjective scoring system. The findings also indicate that the CASS-Beery VMI error score which calculates the discrepancy between the participants’ drawings and the models is a more reliable and stronger scoring element for neurotypical children, compared to the effort score that focuses on the adjustment involved within the CASS-Beery VMI.

Relationship Between Children’s Performance on the CASS-Beery VMI Scores and Their English Language Handwriting Skills

The regression analysis results indicated that the CASS-Beery VMI scores were predictive of children’s performance on the MHA form subscales and ETCH-M word and number legibility (22% of unique variance accounted for, p = 0.007; 30.5% of unique variance accounted for, p = 0.001; 20.9% of unique variance accounted for, p = 0.017, respectively). These findings indicate that the CASS-Beery VMI error scores were predictive of children’s performance in different elements of handwriting such as legibility and formation. Therefore, participants who scored lower on the CASS-Beery VMI error scores generally performed better on the handwriting assessments. These results suggest that the CASS-Beery VMI is a sensitive and useful tool for indicating children’s handwriting skills on many aspects such as legibility, speed, formation, and size. The regression results are also evidence of the CASS-Beery VMI’s convergent validity with aspects of handwriting assessed by the MHA form subscales and ETCH-M word and number legibility.

The current study’s findings on the relationship between children’s handwriting skills and their CASS-Beery VMI scores are supported by Abou-El-Saad et al. (2017) research involving a sample of 200 neurotypical children. They found that children’s performance on VMI assessments increased their ability to copy letters accurately and indicated their handwriting readiness (Abou-El-Saad et al., 2017). Their study also concluded that a child’s VMI capabilities were a significant predictor of handwriting skills in young children, which aligns with certain handwriting elements in our study (Abou-El-Saad et al., 2017). This is consistent with Volman et al. (2006) suggestion that VMI is the most significant predictor of children’s handwriting, particularly the legibility component.

Additionally, the results from our study support previous research which has explored the relationship between VMI and handwriting and reported moderate correlations (Duiser et al., 2014; Pfeiffer et al., 2015). Our findings revealed moderate to strong correlations between the CASS-Beery VMI and handwriting test scores (MHA, ETCH-M and HHIW). Whilst previous literature has supported a level of correlation between VMI results and handwriting scores, the results from the current study reinforce the need to use established standardized handwriting assessments in conjunction with computerized scoring systems like the CASS-Beery VMI when evaluating children’s handwriting performance.

The CASS-Beery VMI error scores also correlated with the HHIW-P scores (r = −0.466, p < 0.01) but did not significantly correlate with the HHIW-C scores. This may be due to children’s knowledge surrounding handwriting and perceiving their handwriting performance very differently from adults and the assessment outcome. While both children’s and parents’ perspectives regarding handwriting capabilities are important, the findings illustrate that parents may be able to provide important insight into their child’s handwriting and, therefore, they should be actively involved in the evaluation and decision-making processes during occupational therapy sessions.

Implications for Practice

This study found that the CASS-Beery VMI is a more sensitive scoring system that can be applied to the Beery-VMI to provide a reliable and comprehensive overview of children’s VMI and English handwriting performance, compared to the manual scoring. The CASS-Beery VMI should therefore be used in practice to score the Beery-MVI output from children for assess children’s English handwriting due to its higher sensitivity in indicating more elements of handwriting performance. Therefore, the CASS-Beery VMI has the potential to reduce the assessment burden that parents and children currently experience in identifying VMI and handwriting issues. Occupational therapists should also actively involve children and parents in interpreting assessment results, goal-setting and handwriting assessment choice. Integrating and discovering children’s and parents’ perspectives of their child’s VMI and handwriting skills will enable child and family-centered practice whilst providing a holistic view of a child’s handwriting performance.

Limitations

All participants were recruited from a single primary school using convenience sampling which may lead to geographical bias and volunteer bias when interpreting the results. The sample size was small (n = 35) limits the generalizability of the study’s findings to a larger population; however, bootstrapping was used during the analysis to accommodate the small sample size. Children’s and parents’ perceptions of their child’s handwriting capabilities were measured using a self-report scale which can be prone to social desirability bias. Although all the participants were neurotypical children (based on parental/caregiver reports), the findings create a benchmark for future studies involving participants with known diagnoses.

Future Research

Further studies using a randomized sampling method to recruit larger sample sizes are recommended to improve the generalizability of the findings. Future research that explores the potential of the CASS-Beery VMI with Beery-VMI in predicting children’s handwriting performance on different handwriting assessments, such as the Print Tool or the Test of Handwriting Skills, is also warranted. It is further recommended that the study is reproduced in children with a known developmental delay or diagnoses to explore the associations between the CASS-Beery VMI and the handwriting performance of this population. While this study provides promising evidence that CASS -Berry VMI is sensitive to indicate children’s VMI and handwriting skills, the requirement to scan the VMI booklet and analyzed by a specialized software reduce its feasibility. Further study of the CASS-Berry VMI’s reliability and validity more broadly viable is suggested.

Conclusion

The study findings provide preliminary evidence of the CASS-Beery VMI’s convergent validity with three standardized handwriting assessments. The findings also suggest that the CASS-Beery VMI scoring system can be used with Beery-VMI to screen and evaluate primary school-aged children’s VMI skills. The outcomes also suggest that the CASS-Beery VMI scores can predict children’s performance on handwriting assessments and their overall English handwriting capabilities. When exploring children’s handwriting and VMI skills, occupational therapists should be aware of children’s and parents’ perceptions of their child’s handwriting performance and how parents’ opinions may correlate with their child’s handwriting skills.

Key Points for Occupational Therapy

• The CASS-Beery VMI is a computerized scoring system suitable for application with the Beery-VMI to assess visual-motor integration skills.

• Preliminary convergent validity evidence between the CASS-Beery VMI and three standardized handwriting assessments was obtained

• The CASS-Beery VMI can enhance the Beery-VMI to comprehensively indicate children’s handwriting skills.

• Children’s and parents’ perceptions of handwriting performance should be considered when evaluating handwriting.

Authors’ Declaration

All authors contributed to conceptualization, methodology, analysis, review, editing and approval of the final submitted version. The first author was also responsible for data collection and writing the original draft.

Ethics Committee Approval

This study was approved by the Monash University Research Ethics Committee (26536) on 16th December 2020 and the Victorian Department of Education and Training (004348) on 15th February 2021.

Provenance and Peer Review

Not commissioned.

Acknowledgements

The authors would like to acknowledge the families who volunteered their time to participate in the study and the assistance of the primary school involved in recruiting participants.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.