Assistive Technology Research Institute
College Misericordia - Dallas, PA 18612
Founded and Sponsored by the Sisters of Mercy of Dallas


Patterns for Life


Patterns For Life: A Study of Young Children's Ability to Use Patterned Switch Closures for Environmental Control

Denis Anson, MS, OTR; Cheryl Ames, OTS; Lynn Fulton, OTS; Megan Margolis,OTS; Maria Miller, OTS

Communication regarding this article should be directed to



OBJECTIVE. The purpose of this study was to determine the age at which children are able to use patterns of switch closures to produce desired responses in the environment.

METHOD. Fifty-two typically-developing children, between the ages of 2 and 8, were instructed to activate specific toys by pressing two large, colored switches in indicated patterns. Testing involved 1, 2, and 3-step patterns.

RESULTS. By age 4, most children had both the cognitive and motor capacity to generate specific patterns for control of their environment. By the age of 5, all of the children were 100% successful.

CONCLUSION. The results of this study indicate that by the age of 5, children have the cognitive ability to use patterns of switch closures to produce desired responses in the environment.

KEY WORDS: Computer Access, Morse Code, Children, Development


One of the basic beliefs of occupational therapy is that purposeful activity influences human development. For the very young child, this early development is fostered through play as the child learns new skills through environmental exploration. By 6 months of age, infants have only a primitive understanding of cause and effect, but by 10 months they are able to recognize that their actions can make things occur in the outside world. With enhanced awareness of causes and their effects, the child also learns to repeat behavior in order to obtain desired responses (Berger, 1998). Later, as the child develops a behavioral repertoire, his/her exploration of the environment involves more possibilities. As young children continue to grow, learn and develop they need to participate in a multitude of purposeful activities. This participation enables them to explore their environment and begin to function independently in society.

Learned helplessness, or the perception that one cannot control the environments that one experiences, is commonly found in children with physical and cognitive disabilities (Sullivan & Lewis, 2000). It does not result directly from deficient ability, but is instead the result of a learned perception that one cannot control outcomes. Through assistive technology, self-efficacy might be gained and learned helplessness could be avoided (Abramson, Seligman, & Teasdale, 1978; Weisz, 1979, 1981).

Assistive technologies which provide control of the immediate environment are known, collectively, as electronic aids to daily living (EADLs). EADLs range from low to high technology and from basic to complex. One example of a basic EADL is a simple switch that activates a single device when pressed (Cook & Hussey, 1995; Sullivan & Lewis, 2000). High technology EADLs, through control strategies ranging from voice input to scanning, can provide individuals with very severe physical limitations a way to control their immediate environment.

When providing equipment to children with known disabilities, it is important for the children to have the ability to control the environment continuously as opposed to intermittently. Because it is difficult to unlearn helplessness once it is learned (Reeson & Ryan, 1988; Swinth, Anson, & Deitz, 1993), it is also necessary to introduce the use of switch toys at an appropriate age to increase the likelihood of success and avoid adding to the child's sense of frustration.

In their study, Swinth, Anson, and Deitz (1993) demonstrated the age at which typically-developing children could learn and then initiate cause-and-effect with the use of a simple switch. Children as young as 6 months were able to associate pressing a switch with the action on a computer screen, and then use that association to control the computer.

Although Swinth et al.'s study addressed learning behavior of typically-developing children, the researchers suggested that introducing computers to children with disabilities at early ages would encourage them to become competent and independent learners (Swinth, Anson, & Deitz, 1993).

Today, switch adapted toys are commonly used to provide control for children with physical limitations. Such toys allow the child to access experiences that are believed to lead to the acquisition of functional skills and abilities. Instead of becoming frustrated at the inability to retrieve a toy, a child would use switches to control desired toys, engage in the occupation of play, and ultimately achieve greater independence (Solano & Aller, 2002).

Traditionally used with individuals with physical disabilities, switch control systems are not often introduced to individuals with severe cognitive impairments. One reason given for this is based on Piaget's theory of human development. According to these theories, children at the age of two begin to develop the ability to identify objects and understand that objects are permanent, separate from their environment, and can be used to manipulate the environment. But it isn't until about the age of twelve, according to Piaget, that a child develops the ability to think abstractly and therefore would be able to make the secondary linkages required to master a switch control system. Piaget, however, used naturalistic observations to assess a child's abilities in various cognitive development stages. Behrmann (1984) argued that this method of classification failed to recognize that young children may be unable to express what they have learned and thus reveal their true capabilities.

Swinth et al.'s study demonstrated that children as young as 6 months were able to link the operation of one switch with an environmental response from something other than the switch. She also noted that the children in the older age groups (9-17 months) became disinterested with the simple cause-and-effect activity and sought more stimulating feedback (Swinth, Anson, & Deitz, 1993). It appears that an environment with only a single response to activity is little different from one that makes no response at all. Typically developing children quickly learn that different motor behaviors result in different environmental responses. Children with disabilities need to learn the same lesson, and develop a behavioral repertoire that produces a variety of outcomes. However, simply increasing the number of switches beyond a very few would be cumbersome, and would probably not provide an adequate solution.

Experience in text generation has shown that patterned closures of a relatively few switches (e.g. Morse code) can be used to allow individuals to use only a few switches to generate a large number of options. Early studies have shown that children with the ability to use language at the third grade level can use Morse for language (Beukelman, Yorkston, & Dowden, 1985). When used to generate language, Morse code allows patterned closures from one to three switches to select any character of the alphabet resulting in computer input and augmentative communication (Beukelman, Yorkston, & Dowden, 1985).

Although it has many applications, Morse code is not readily used with very young children because of its traditional utilization as a tool to generate language. Since young children have not yet developed written language skills, attempting to teach the patterns as written language only sets the up child for repeated failures. It seems possible that simpler, non-language applications of switch patterning could be usable at a younger age. Patterned switch closures could conceivably allow a wide range of options for a child with very limited motor control (Jarus, 1994; Anson, 1997; Wellings, & Unsworth, 1997). This technique might enable a large variety of toys and environmental responses to be controlled by one or two switches, operated in patterns. Engaging in play through switch patterns might maintain the motivation of the physically-limited child. With continued exposure and practice, a child might learn to generalize basic switch patterns to an environmental control system that would further encourage independence and self-efficacy.

Despite the many benefits of assistive technology, limited literature reflects the use of switch patterning or environmental control units with children with cognitive disabilities. The purpose of this study is to determine the age at which typically developing children are able to use patterns of switch closures to produce desired responses in the environment.



The convenience sample consisted of 52 typically-developing children divided into seven age categories. These children were all full-term children with no identified developmental disabilities. They had vision adequate to see the provided cues, and sufficient auditory acuity to follow spoken directions. (Initially, we had planned to test 10 children at each year of age from two years though eight years of age to identify the level of maturity required to use switch patterns to control objects in the environment. Once we determined an age of success, however, we ceased to test the older children.) Because of the rapid rate of growth and maturation of the brain in this age population, the children were grouped into twelve-month age intervals. During the study, ten two-year-olds, ten three-year-olds, eleven four-year-olds, eleven five-year-olds, five six-year-olds, two seven-year-olds, and three eight-year-olds were tested.

Data was collected in day care centers located in the New York and Pennsylvania areas.


Control System

Because no available EADL system provided Morse code control, we created this function for our study. The control system was based on an HP Pavilion n5420 notebook computer system with a color video display. This computer ran the X-10 ActiveHome Control software to control power modules used to control the stimulus devices (toys). Because the Home Control software does not have a keyboard interface, it could not be directly controlled by Morse input, so we also used a macro-program, mgSimplify , to translate keyboard commands to mouse movements and mouse clicks. On receiving a keyboard code, mgSimplify would move the mouse cursor over a specific X-10 controller in Smarthome, generate a mouse click to activate a toy, wait 5 seconds, and then click again to deactivate the toy. Keyboard input was provided by Darci USB , a Morse code device that connected to the laptop via the USB ports.

Input to the Darci USB was provided by two "Big Red" switches , one blue and one red. The "dah" switch was labeled with a large square corresponding to the visual cues provided to operate specific switch toys.

Controlled Devices

The study used a set of eight "power-adapted" toys. These toys were originally battery powered, but were adapted, in a manner similar to that for constructing "switch adapted" toys, to draw their power from an external power supply. These power supplies were connected through X-10 switching modules to allow the toys to be controlled remotely by the laptop computer.


Prior to a child entering the testing area, the computer was set up with the Darci USB connected, and the SmartHome and mgSimplify program running. This computer was positioned so that the child would not see the computer screen, or its controls. The two switches were positioned in front of the testing station.

In clear view, but out of reach of the testing station, the first two toys to be controlled were connected, via their power modules, to the X-10 controller. At no time during testing did the researcher or the child directly manipulate the toys.


The subjects were brought from their regular classroom to a quiet room as free from distraction as possible. The researchers then spent 5 to 10 minutes talking with each subject in the company of a caregiver or teacher to establish rapport with the child.

The child was then seated at a table and told that they were going to be playing a game with big buttons and toys. The child sat at the table in front of the two switches, which were positioned at a comfortable distance from the child and with the toys positioned out of reach of the child and directly in front of the switches, 12 inches apart. Each toy had a card showing the pattern for its activation, as a series of red circles and blue squares, placed immediately in front of the toy and clearly visible to the child.

The testing was preceded by a training session, in which the means of controlling the toys was described and demonstrated. The researcher first described and demonstrated to the child, "If you hit this button one time, it will make the toy go." For the next stage of training, the researcher guided the child's hand in the activation pattern. Each toy was activated until the child demonstrated the ability to activate the toys, or up to 5 times, prior to testing.

In each testing session, the child was asked to activate the target toys in a predetermined pattern. During the testing, the researcher pointed to the target toy and said, "Can you make the [elephant, robot, pig.] go?" If the child did not respond within 10 seconds, the second cue, "Make this [elephant, robot, pig.] go!" was given. If the child activated the correct toy, the trial was counted as a success. If the child did not respond within 10 seconds of the second cue, or activated the wrong toy, the trial was counted as a failure. At each level, the child was given up to 10 trials. If the child activated the correct toy at least 8 times (80% success), they advanced to the next level of training. If they did not achieve 80% success, they were dismissed from testing and given a small reward for their participation.

In the first level of testing, the child could activate one of two toys by pressing the appropriate switch one time. At the second level of testing, the child could activate one of three toys by pressing the switches in a two-step pattern selected for that toy (e.g. "circle-square" makes the elephant go). At the third level, the child could activate each of three toys using a three-step pattern.

Results and Discussion

Fifty-two children (19 boys and 33 girls) participated in this study. There was no difference in performance based on gender at any of the age levels. The length of practice time was child specific and was dependent on the child's attention span and interest level.

Table 1. Percentage of Subjects Meeting Criteria for Success at Each of the Three Levels

Age Level

No. of children tested

% Successful

Level I

Level II

Level III

2 years





3 years





4 years





5 years





6 years





7 years





8 years





As seen in Table 1, nearly all children in the study were able to master one-step control patterns. This was expected, based on Swinth et al.'s findings (1993), although the child had two options in this study as compared with a single choice provided by Swinth et al.

In our study, the placement of the visual-cue cards was very important, and most children younger than 5 required verbal cues throughout the testing to refer to them. The visual cues allowed the children to control the toys without the need to memorize the patterns.

At age 4, all but one child were able to master the three-step control sequence, and by age 5, all children were able to master the control sequences. Age-specific behavioral characteristics and learning styles emerged during the study.


As an infant approaches 6 months, he or she acquires the ability to discriminate simple forms and shapes. At age two, children begin to think through mental and physical actions (Berger, 1998). In this study, two-year-olds consistently demonstrated an understanding of cause and effect interaction with objects and a one-to-one correspondence between controls and devices, but their attending or motor skills were inadequate to achieve success past the first level. Also , at this age the children appeared to prefer to activate favorite toys rather than activating the toys identified by the researcher.

While 90% of the ten two-year-olds tested demonstrated an understanding of cause and effect, as evidenced by their successful completion of level one testing, none of them were successful beyond the first level. The two-year-olds appeared to grasp the concept of the two-step patterns during training, but once testing began they reverted back to simple cause and effect. They also did not follow the verbal cues to refer to the visual cue cards. This may reflect the inability to organize complex motor behaviors at this age rather than the inability to conceptualize motor patterns as control strategies.

In addition, w ith the younger children there was a critical period between providing adequate practice time and exceeding the child's attention span. If this critical period was surpassed, it resulted in a loss of interest in the toys.

Finally, the two-year-olds had difficulty pressing and then releasing the switch in time to activate the toys. They also tended to wait too long between two and three-step patterns, which is common when individuals are learning Morse for language. The "end of character" time (the amount of time the Darci would wait between a switch press and generating a character) for the Darci USB had to be adjusted to the rate of the children. However, if set too long the delay between completing a pattern and the toy activating became a distracter for the child. This distraction was also evident throughout the first level of testing as the two-year-olds needed to have a very short end-of-character interval (.5 seconds) to provide them with immediate feedback of toy action.


The three-year-olds continued to require training to properly depress the switch. It was often helpful to first ask them to "slap me five" (i.e. child slapping the palm of researcher's outstretched hand) and then repeat this motion on the switch. This supports the conjecture that the issue is at least in part one of motor control, as "slap me five" provides a motor cue, rather than solely a pattern cue. The children in this age level demonstrated an emerging understanding of patterns, but still failed to consistently utilize the cue cards. Of the ten children tested, 50% were successful with two step control sequences and 20% were successful with three step commands. For these children, the end-of-character interval had to be increased to two seconds to accommodate the delay in producing the two and three-step patterns. At this age, the children demonstrated a preference for specific toys by spontaneously activating their favorites without waiting for instructions.

As stated by Julesz (as cited in Pedretti & Early, 2001), pattern recognition requires the ability to identify the features that distinguish an object from its surroundings. In this study, although the cue patterns for each toy included both the shape on the switch and the color of the switch, by the age of three the children appeared to use shape preferentially over color as a distinguishing characteristic. This preference was shown as the children verbalized the cue patterns as "square-square-circle" rather than "blue-blue-red."

At age three, while the concept of patterns rather than switches being used for control was beginning to emerge, the children continued to have difficulty with the motor control required to produce the patterns.


Between the ages of two and six, the most significant developmental changes involve the maturation of the brain and central nervous system. Although Piaget believed that abstract reasoning did not begin to emerge until the age of 12, research suggests that he underestimated the cognitive ability of younger children (Berger, 1998). The results of this study indicate that a four-year-old, using modern technology and Morse code, has the cognitive ability to acquire a higher level of independence by effectively controlling their environment. In addition, by this age the children had the motor capacity to generate specific patterns for control of their environment and switch-control was no longer an issue.

Of the eleven four-year-olds tested, 91% were successful in producing two step commands and 82% were successful with three step control sequences. By this age, the children consistently looked at the cue cards for guidance. Several of the children began to memorize the two-step patterns and they were better able to wait for the researcher's direction before activating toys. Additionally, the children who repeated the patterns out loud appeared to master the switches more easily.

Some difficulties specific to this age level also emerged. Because they had not yet begun to read, many of the four-year-olds had not developed the habit of following the patterns from left to right rather than right to left. In addition, several of the children grouped similar shapes together resulting in, for example, the circle-square-circle pattern becoming circle-circle-square. It appeared that at this age, a child was better able to master the control pattern for a "favorite toy" than for other toys. This supports the idea that activities that are personally meaningful are more easily learned than those with limited appeal.


At age five, brain maturation brings important gains in physical abilities and higher-order cognition (Berger, 1998). Due to their advancing cognitive and motor skills, all children in the study of at least five-years of age were able to successfully master all three control strategies. In agreement with Piaget's stages of development, these children were able to understand basic concepts of classification. It was noted that the five-year-olds appeared less enthusiastic in the first two stages, but became excited by the challenge presented with the three-step patterns.

Six, Seven, and Eight-year-olds

Five six-year-olds, two seven-year-olds, and three eight-year-olds also participated in the study and they all successfully completed the third level. The tasks were very easy for them and they appeared bored throughout the testing. The 100% success rate at the five-year-old level predicted the success of children above this age.

Study Strengths and Limitations

Training Period

One limitation of this study was that the training and testing occurred on the same day, and over a short time span. The researchers believe that given additional training periods and possibly other switches the majority of three-year-olds are capable of following the patterns.

It is not clear from the results of this study if the pattern skills learned on one day would persist to a later test time.

Design Constraints

The timing of the switches was very child specific and continuous adjustments were necessary when moving between children. If the "end-of-character" delay was too short, the switch pattern produced by the child would be interpreted as a set of shorter, unintended patterns. If it was too long, the child would become impatient, and either lose interest in the task or try to repeat the pattern resulting in a failure.

These findings suggest some of the design constraints that might apply to a switch-pattern controlled EADL. As with Morse code interfaces for text generation, the switch timing must be carefully set to allow success. Both the time that an individual switch must be held down to be detected, and the time-delay before a pattern is considered to be complete must be adjustable. A switch-pattern controlled EADL device would need to have the ability to easily set the switch timing. In our study, switch patterns were designed to correlate with the physical location of the toy on the table. The ability to assign specific codes to specific devices also seems important to the design of a pattern-controlled device.

Training Strategies

This study provides insight into training strategies for switch control of the environment. For most of our subjects, shape was a much stronger cue than color. This suggests that the color of switches should not be used as the primary cue for patterns, but that the shapes associated with the switches should be clearly identified. (It is not clear that a square switch would be a stronger cue than a round switch with a square on it.) It is, however important to not confuse switch control issues with pattern issues. When an individual is having difficulty using switch patterns to control the environment, it may be that the problem is in organizing and executing motor patterns, but it may also be that the demands of a specific switch exceed the capacity of the individual. Before abandoning this control strategy, a variety of switches should be assessed.

While specific design constraints identified in this study will be helpful for future product development and training strategies, the purpose of this study was to determine the age at which a child is capable of using patterns to operate the toys, not the child's ability to operate the switches.

Future Direction

This study supports the use of patterns rather than devices to control EADL devices for individuals with severe motor impairments and for those with significant cognitive impairment. An individual with the physical capacity to produce only two volitional movements could still use those very limited movements to control a wide array of devices in the environment. Such a control strategy would be accessible to individuals who are too impaired to use speech control, but who want or need a faster or more flexible control strategy than single switch scanning.

The findings of this study suggest that children can learn to use patterns of behavior rather than simple behaviors to control their environment. The study establishes the viability of using patterns for control, but does not address how many patterns can be used? How many patterns or cues could an individual use before they were not able to absorb additional control strategies? In this study, many of the younger children failed to refer to the visual cues. Would their performance have improved if the toys were directly labeled with their corresponding patterns rather than the pattern being on a card in the foreground?

Swinth et al. (1993) found that once a task was mastered, children preferred a more challenging form of play. Similarly, the children in this study appeared, at times, to lose interest in the toys. In particular, the five year olds were noted to have limited interest in one-step and two-step control sequences, but were engaged and challenged by the three-step commands, although the toys being controlled were similar. In both studies, the children were taken from a rich, challenging environment to a quiet environment containing a relatively simple task, and sought return to the more challenging setting. The researchers believe that, for children with profound disabilities, this loss of interest could be minimized. First, the tasks to be controlled should be challenging to the child. Second, the alternative, for the child with a profound disability, is inactivity rather than a return to an enriched environment. If a child with physical or cognitive disability is provided control of their environment, and the tasks are constantly challenging and enriching, might they avoid learned helplessness?


Swinth, et al. (1993) provided the foundation for this study by suggesting that very young children are capable of understanding cause and effect and thereby able to utilize switches to control their environment. The current study went a step further and tested the ability of young children to correctly reproduce a multi-step switch pattern in order to use Morse code patterning to interact with the environment. The results of this study indicate that a person with a cognitive level of at least four years of age, using modern technology and Morse code, would be able to effectively control their environment. As a result, the individual might achieve a level of independence, and avert learned helplessness.


Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. D. (1978). Learned helplessness in humans: Critique and reformation. Journal of Abnormal Psychology, 87 , 49-74.

Anson, D. K. (1997). Alternative computer access a guide to selection. Philadelphia : F. A. Davis.

Behrmann, M. M. (1984). A brighter future for early learning through high technology. The Pointer, 28 (2) , 23-26.

Berger, K. S. (1998). The developing person through the lifespan (4 th ed.). St. Louis , MO : Mosby.

Beukelman, D. R., Yorkston, K. M., & Dowden, P. A. (1985). Communication augmentation: A casebook of clinical management . San Diego , CA : College Hill Press.

Cook, A. M., & Hussey, S. M. (1995). Assistive technologies: Principles and practice. St. Louis , MO : Mosby.

Jarus, T. (1994). Learning Morse code in rehabilitation: visual, auditory, or combined method? British Journal of Occupational Therapy, 57, 127-130.

Pedretti, L. W., & Early, M. B. (Eds.). (2001). Occupational therapy practice skills for physical dysfunction (5 th ed.). St. Louis , MO : Mosby.

Reeson, D. J., & Ryan, M. (1988). Computer microtechnology for a severely disabled preschool child. Child: Care, Health, and Development, 14, p. 93-104.

Solano, T., & Aller, S. K. (n.d.). Tech for tots: A rationale for assistive technology for infants and young children. Retrieved February 6, 2001 , from

Sullivan, M. & Lewis, M. (2000). Assistive technology for the very young: creating responsive environments. Infants and Young Children, 12(4), 34-52.

Swinth, Y., Anson, D., & Deitz, J. (1993). Single-switch computer access for infants and toddlers. The American Journal of Occupational Therapy , 47 , 1031-1038.

Weisz, J.R. (1979). Perceived control and learned helplessness among mentally retarded and nonretarded children: A developmental analysis. Developmental Psychology , 15, 311-319.

Weisz, J.R. (1981). Learned helplessness in black and white children identified by their schools as retarded and nonretarded: Performance deterioration in response to failure. Developmental Psychology, 17 , 499-508.

Wellings, D. J., & Unsworth, J. (1997, August 16). Environmental control systems for people with a disability: An update. British Medical Journal, 315 , 409-413.

SMARTHOME, Inc., 16542 Millikan Avenue , Irvine , CA 92606 , United States ,

mgSimplify, 22298 Davenrich, Salinas , CA 93908 , United States ,

The Darci Institute of Rehabilitation Engineering, 810 W. Shepard Lane, Farmington , UT 84024 , United States ,

Ablenet, Inc., 1081 Tenth Ave SE, Minneapolis, MN, 55414, United States,