Development of gross and fine motor skills is widely accepted as an important part of typical childhood development. Considerable information is available on the ages and stages at which children acquire various fundamental motor skills (Branta, Haubenstricker, & Seefeldt, 1984; Clark, 1994, 2005). It is also known that refined motor skills are necessary for object manipulation, daily life skills, and the ability to effectively maneuver oneself through space, as well as other important sensory and cognitive abilities including planning, problem solving, temporal–spatial awareness, and perceptual skills (Ayres, 1969; Logan, Robinson, Wilson, & Lucas, 2012; McCormack & Atance, 2011; Piek, Dawson, Smith, & Gasson, 2008).

Although acquisition of specific motor skills is relatively well understood, the ability to plan and organize motor movements has been examined less frequently. Historically, identification of deficits in motor performance and motor planning did not occur until children reached school age and functional problems in school-related tasks were noted (Gubbay, 1978). Nonetheless, early identification of problems in these areas is important. Approximately 58% of U.S. children ages 3–5 yr are enrolled in a preschool program (U.S. Department of Education, Institute of Education Sciences, 2009), and with the advent of early intervention services, it has become increasingly important to understand typical motor performance and to identify motor performance problems at younger ages, particularly in the preschool population.

Occupational therapy practitioners using a sensory integration frame of reference are particularly concerned with identifying and treating problems in motor performance, especially in the area of motor planning (Watling, Koenig, Davies, & Schaaf, 2011). Motor planning skills are viewed as a component of the larger process of praxis, which is one subtype of sensory integration and processing (Ayres, 1972a; Miller, Anzalone, Lane, Cermak, & Osten, 2007). A. Jean Ayres (1979), noted occupational therapist and neuroscientist, stated, “Praxis is the ability to conceive of, organize, and execute a motor action” (p. 183). Ayres determined that motor planning difficulties had their origin in decreased somatosensory and vestibular sensory processing. Problems in motor planning in the young child were believed to result in difficulties with the acquisition of normal developmental motor milestones, daily life skills such as eating and dressing, and development of play skills (Ayres, 1972a, 1979).

Deficits in praxis have been identified in school-age children for more than 40 years, and although many assessments address motor coordination, few assessments of praxis exist. With little consensus on the best way to examine motor planning, it is often assessed through imitation of gestures. However, Ayres developed assessment tools that also examined block design construction, visual–motor skills, ability to follow verbal commands, motor sequencing, and bilateral integration (Ayres & Mailloux, 1989). In her early work identifying patterns of dysfunction in sensory integration and praxis, Ayres (1965, 1966) examined perceptual–motor dysfunction in children ages 4–8 yr through assessment of sensory processing, gross motor imitation skills, and fine motor object manipulation tasks. Specific motor tests and screenings involved motor accuracy in drawing on a line, replication of hand positions, imitation of arm and body postures, fine motor planning using a wire grommet device, and winding strings in a figure eight around two bolts. Findings from these studies resulted in identification of praxis deficits, which Ayres initially called “developmental apraxia” (1972a) and later called “developmental dyspraxia” (1979, 1985).

Of the various measures of motor performance that indicated dyspraxia, Ayres (1965)  determined that imitation of body postures and fine motor planning using the wire grommet device were the two most important ways to identify motor planning problems. The two tests she developed using these skills were highly correlated with each other and with other tests indicative of praxis deficits (Ayres, 1965, 1966). In fact, she identified the wire grommet maze as the single best indicator of dyspraxia. In spite of these findings, Ayres did not choose to use this test in further development of either the Southern California Sensory Integration Tests (Ayres, 1972b) or the Sensory Integration and Praxis Tests (SIPT; Ayres, 1989). However, many occupational therapists trained by Ayres continue to use this assessment observationally in clinical practice.

Currently, although assessments of fundamental gross and fine motor skills are available (Piek, Hands, & Licari, 2012), few praxis assessments are available for preschool children. The SIPT and the Miller Assessment for Preschoolers (MAP; Miller, 1982, 1988) are two standardized assessments used with younger children. The SIPT is normed for 4- and 5-yr-olds but not for children of younger ages and is lengthy and expensive to administer. Although it includes several tests that tap into facets of praxis, it does not include any fine motor planning or organization tests using objects similar to the wire grommet maze. The MAP has some individual praxis items (two postural imitation items and box and flat maze items analogous to the wire grommet maze) and can be used with children as young as 2 yr, 9 mo, of age. The MAP, however, is more than 20 yr old, and norms are becoming dated. In addition, one must administer the entire battery to obtain meaningful information on performance.

Thus, a need exists for additional assessments of praxis, especially motor planning tests, for younger children. In addition, screening tools are needed that may be administered quickly and inexpensively. The Motor Planning Maze Assessment (MPMA; May-Benson, 2006) was developed on the basis of the early work of Ayres to meet the need for a quick measure of motor planning ability that could be reliably administered and was inexpensive. At this time, little to no formal research has been published on the current mazes and format of the MPMA, and no information is available on the use of this tool with preschool children. The purpose of this study was to examine the applicability of the MPMA as a screening tool and the current scoring criteria for use with typical preschool children. The following questions guided this research:

This study was an observational quantitative study aimed at establishing developmental norms for preschool children (ages 3–5 yr) and interrater reliability on the MPMA. We obtained approval from the institutional review board at Virginia Commonwealth University and informed consent from parents for all participating children. This study was conducted as part of a larger study on assessment of praxis in preschoolers.

Eighty-five typically developing preschoolers (52 boys, 33 girls) ages 3–5 yr were recruited from three Richmond, Virginia, metropolitan area preschools (Table 1). Preschools were selected for convenience, because the authors had contacts at the schools, and because the schools represented a variety of approaches to learning (one is Reggio Emilia, one is Montessori, and one is a standard developmental preschool). As part of the consent process, parents completed a form that included information on the children’s age, handedness, and previous or current evaluations and diagnoses. Children who did not have a diagnosis were considered typically developing for the purposes of this study. Inclusion criteria stated that no child was to be excluded on the basis of race, ethnicity, health or medical condition, or education level; children with limited English proficiency were excluded because they needed to understand the directions of the assessment, which were given only in English. All children at each site whose parents consented to testing were included in the study.

The MPMA (May-Benson, 2006) was adapted from wire grommet mazes originally developed by A. Jean Ayres and used in her initial factor analyses of sensory integration. Currently, the MPMA is “an individually administered test that assesses the motor planning aspect of praxis in children from 4 years to teens” (May-Benson, 2006, p. 2). It has been used primarily as a clinical observation tool, and no normative data have been collected on any age group; it has not been used with preschoolers.

The test consists of three mazes: a simple wire rectangle with a single handle (Maze 1 Rectangle), a complex twisted wire maze with a single handle (Maze 2 Single-Hand), and a twisted wire maze with two handles (Maze 3 Two-Hand). Each maze has a small metal grommet on the wire maze, and the child is asked to manipulate the maze to move the grommet from one side of the device to the other. The test takes approximately 5 min to administer and score. Scoring criteria developed by May-Benson (2006)  consist of a combination of time completion and errors demonstrated. One point is given for each error demonstrated, for a total possible of 4 error points per maze. Errors include behaviors such as shaking the maze, using a hand to move the grommet along the wire, switching the maze to the opposite hand, and using excessive body movements to rotate the maze. Time scores are calculated by assigning points for bands of completion times, which vary by maze. We modified the original MPMA Time scores using data collected in this project to better reflect the young children’s performance (Table 2). Time scores and Error scores are added for a Total score for each maze. The maximum Total score for Maze 1 is 10, for Maze 2 is 12, and for Maze 3 is 9, resulting in a total MPMA score range of 0–31.

Testing was conducted with individual preschoolers from February to April in 2012 and 2013. For each session, two researchers, including two trained occupational therapy student research assistants or a research assistant and a lead researcher (author Ivey or Lane), were present; one researcher administered the MPMA, and the other videotaped the session. All researchers received training on administration of the MPMA and on all aspects of the research proposal (including consent, methods, confidentiality, and privacy procedures). Testing occurred in an enclosed room free of distractions with an outlined 6- × 8-ft testing area. Participants were handed three mazes in their numbered order as outlined in the MPMA manual (May-Benson, 2006)  and told, “I want you to hold this maze and use [one hand or both hands] to move this little circle from this point to this point without touching the circle.” Initial scoring was completed by research assistants and lead researchers. When scoring was revised, it was necessary for the two lead researchers to score the MPMA for assigned participants. Time and Error scores were added for a Total maze score. Maze scores were then added to obtain a total MPMA score, with lower scores indicating better performance.

We used IBM SPSS Statistics, Version 21 (IBM Corporation, Armonk, NY) to analyze data. Preliminary analysis involved computation of means and standard deviations and exploratory stem and leaf and box plots to identify outliers. Means and standard deviations were calculated to examine the typical performance of the preschool children. Differences in age trends and gender differences were calculated using a one-way analysis of variance (ANOVA); an a priori α level was set at p < .05. Bonferroni post hoc analysis was used to determine which means differed. Interrater reliability was examined for all of the Total scores and individual item scores using the intraclass correlation coefficient (ICC; Shrout & Fleiss, 1979).

Preliminary means and standard deviations were computed for Time, Error, and Total scores for each of the three mazes by age group. Exploratory stem and leaf and box plots showed five outliers (i.e., >3 standard deviations from the mean for their age group) across the three age groups and three mazes. We eliminated the outliers from further analyses, leaving a final study group of 80 children: 30 were 3-yr-olds (37.5%), 29 were 4-yr-olds (36.3%), and 21 were 5-yr-olds (26.3%; Table 1). Means and standard deviations were then recalculated for each Age × Maze group (see Table 3).

We found no statistically significant differences by gender, race, or type of preschool. A one-way ANOVA revealed an overall age effect for Maze 1 for Error score, F (2, 77) = 5.255, p = .007; Time score, F (2, 77) = 5.166, p = .008; and Total score, F (2, 77) = 6.736, p = .002. Older children completed Maze 1 with significantly fewer errors in significantly less time than younger children. Bonferroni post hoc analysis revealed significant age effects between 3- and 5-yr-olds for Error score (p = .006), Time score (p = .006), and Total score (p = .001); comparisons between 3- and 4-yr-olds and between 4- and 5-yr olds were not significant.

A one-way ANOVA revealed an overall age effect for Maze 2 for Error score, F (2, 77) = 6.857, p = .002; Time score, F (2, 77) = 8.808, p = .000; and Total score, F (2, 77) = 10.434, p = .000. Older children completed Maze 2 with significantly fewer errors in significantly less time than younger children, resulting in a lower Total score with each age group. Bonferroni post hoc analysis revealed significant age effects for Error scores between 3- and 4-yr-olds (p = .012) and 3- and 5-yr-olds (p = .004). Significant age effects were also seen for Time scores on Maze 2 between 3- and 4-yr-olds (p = .006) and 3- and 5-yr-olds (p = .000) and for Total score between 3- and 4-yr-olds (p = .006) and 3- and 5-yr-olds (p = .000). No significant differences were seen between 4- and 5-yr-olds in Error, Time, or Total scores.

A one-way ANOVA revealed an age effect for Maze 3 for Error score, F (2, 77) = 13.968, p = .000; Time score, F (2, 77) = 16.233, p = .000; and Total score, F (2, 77) = 18.371, p = .000. Older children completed Maze 3 with significantly fewer errors in significantly less time than younger children, resulting in lower Total scores with each age group. Bonferroni post hoc analysis revealed significant age effects for Error scores between 3- and 4-yr-olds (p = .002) and 3- and 5-yr-olds (p = .000). Significant age effects were also seen for Time scores on Maze 3 between 3- and 4-yr-olds (p = .000) and 3- and 5-yr-olds (p = .000) and for Total score between 3- and 4-yr-olds (p = .000) and 3- and 5-yr-olds (p = .000). No significant differences were seen between 4- and 5-yr-olds in Error, Time, or Total scores.

A one-way ANOVA revealed an age effect for the total MPMA score, F (2, 77) = 20.709, p = .000. The total MPMA score decreased (improved) with each age group, with post hoc analysis revealing significant age effects for each age group: 3- and 4-yr-olds (p = .000), 3- and 5-yr-olds (p = .000), and 4- and 5-yr-olds (p = .048).

The two lead researchers scored the MPMA for assigned participants, and for 41.3% of the participants (33 of 80), both lead researchers scored the same participants. Excellent interrater reliability was found for the total MPMA score (ICC = 0.96) and for the Total score for each maze (Maze 1 ICC = 0.90, Maze 2 ICC = 0.98, Maze 3 ICC = 0.96). Interrater reliabilities for individual maze scores (Error and Time scores) were at the upper range of adequate to excellent (Table 4). Analysis of individual errors for each maze revealed adequate to excellent reliability as well, with the exception of “uses whole arm or body” for Mazes 2 and 3.

Results of this study provide evidence that the MPMA is a reliable, objective screening of a preschooler’s ability to plan and execute motor movements. The total MPMA score can distinguish developmental differences among preschoolers ages 3, 4, and 5 yr. No differences were seen by gender, race, or educational approach. Our preliminary score guidelines for use of this screening at the preschool ages are now available, but continued research in a larger sample across socioeconomic levels, geographic location, and ethnicity will strengthen these results.

This study extended previous work on the MPMA to include revised, expanded scoring criteria for preschoolers and revision to the test administration procedure regarding the order of presentation of mazes, at least for preschool children. Original administration procedures started with the easiest maze (the Maze 1 Rectangle) and proceeded to the more complicated Maze 2 Single-Hand before asking children to complete Maze 3 Two-Hand. Analysis of data from this study on time to complete the mazes and number of errors suggest that the two-handed maze, Maze 3, was easier to understand and master, followed by Maze 1, then Maze 2. Thus, we recommend that for preschool-age children, the order of maze presentation should be changed to Maze 3, Maze 1, and Maze 2. We also recommend that the order of presentation be examined in older children. These findings do not suggest that bilateral skills are easier to master or develop earlier in preschoolers; instead, for this screening, it was easier for the children to understand the directions, specifically the instruction to not use their hand to “touch the circle” when both hands were holding the maze handles. In addition, in this task, the two hands work together in a symmetrical bilateral task rather than a more complex asymmetrical activity.

Results indicate that with initial training, interrater reliability on the MPMA is high. Although the MPMA is quick and easy to administer, rater training is needed not only to ensure reliability and correct administration of the MPMA but also to ensure that examiners understand the scoring parameters, particularly those related to error definitions. This understanding is essential for clinical use when scoring is done in real time instead of later using videotapes, as we did for this study. Overall, the MPMA can be a useful, reliable screening tool if properly administered and scored.

This study was limited by use of a convenience sample. A larger sample is needed to broaden the generalizability of the findings. Future research should examine test–retest reliability in the preschool population. Normative data are also needed on school-age children. The construct validity of this screening, although initially examined in Ayres’ (1965, 1966) original work, should be explicitly examined with both preschool- and school-age children. The discriminative ability of this screening to distinguish children with motor planning problems from typically developing peers needs to be examined as well. Next, praxis skills have traditionally been examined using assessments of gestures and imitation of postures. The relationship of the MPMA to these types of traditional praxis assessments needs to be examined to fully validate the MPMA as a screening tool of motor planning skills. Last, the MPMA involves considerable problem solving and cognitive planning; it will be important to examine the relationship of this screening with other measures of executive planning skills (McCormack & Atance, 2011).

The results of this study have the following implications for occupational therapy practice:

Children with motor planning problems, such as dyspraxia and developmental coordination disorder, engage less frequently in activities requiring motor planning and choose activities that are quieter or more socially isolating than their peers without motor problems (Magalhães, Cardoso, & Missiuna, 2011). Participation restrictions affect play, classroom tasks, handwriting, and self-care tasks for elementary school–age children, but evidence suggests that these skills are compromised from an early age (Green et al., 2011). Therefore, early identification is critical, and screening and evaluation tools are needed to help identify children at an early age for intervention. This study describes a screening tool that allows for easy and reliable examination of the motor planning aspect of dyspraxia, which may prove useful in helping occupational therapy practitioners with early identification of and intervention with children who have participation restrictions.

We are grateful to the preschool directors for allowing recruitment of their students, to the families that allowed their children to participate, and especially to the children for their participation. We extend special thanks to our Virginia Commonwealth University master of science in occupational therapy research students and research assistants for data collection and scoring.