Research in social interaction and nonverbal communication among individuals with severe developmental disabilities also includes the study of body movements. Advances in analytical technology give new possibilities for measuring body movements more accurately and reliably. One such advance is the Qualisys Motion Capture System (QMCS), which utilizes optical markers to capture body movements. The aim of this study was to explore the practicality of measuring body movements in the nonverbal communication of a child with severe developmental disabilities. A preliminary case study has been undertaken. The social interaction between a boy with developmental disabilities and his teacher was analyzed (1) using observer ratings on psychological aspects of the social interaction and (2) measuring body positions, velocity, and angles of body movements using the QMCS. Associations between observer ratings and measured body movements were examined. This preliminary case study has indicated that emotional response and attention level during the social interaction corresponded with local, synchronized movements and face-to-face orientation. Measurement of motor behavior is suggested as being a potentially useful methodological approach to studying social interaction and communication development.
Social interaction and communication development among individuals with severe developmental disabilities (severe mental and/or physical impairments) has been a subject of research for decades (Hostyn & Maes, 2009) and is supposed to be fundamental for the development of social, language, and cognitive skills (Ogletree, Bartholomew, Wagaman, Genz, & Reisinger, 2012; Olsson, 2004; Wilder & Granlund, 2003). Supporting social interaction and communication can aid persons with severe developmental disabilities in reducing dependency or withdrawal and improving social connectedness, independence, and quality of life. Nevertheless, research has demonstrated that social interactions between people with severe developmental disabilities and their caregivers are not always rewarding (see Hostyn & Maes, 2009, for a review). The needs, thoughts, and emotions of the person with severe developmental disabilities are difficult to interpret and are often misunderstood or ignored (Grove, Bunning, Porter, & Olsson, 1999) due to his or her individual limitations (e.g., lack of motor control or vocalization/verbalization) (Olsson, 2004). Caregivers need to know how to read potentially communicative signals and notice different kinds of expressions while interacting with a child with severe developmental disabilities (Iacono, Carter, & Hook, 1998; Ogletree et al., 2012; Olsson, 2004). Caregivers may need to recognize subtle motor expressions, for example, movements of a leg or body position, as potential communication (Janssen & Rødbroe, 2007). Interpreting subtle expressions and nonverbal communication is often a difficult task. Only a small amount of research has previously been conducted concerning social interaction and communication development of individuals with severe developmental disabilities.
There are presently two challenges: (1) to become more knowledgeable about the social interaction process and how interactions can be improved (Hostyn & Maes, 2009) and (2) developing practical methods to explore these social interactions.
The Theoretical Frame: Inspiration From Mother-Infant Interaction Research
Theories explaining the social interaction of people with severe developmental disabilities have been influenced by research investigating mother-infant psychological interaction (Burford, 1988; Janssen & Rødbroe, 2007; Olsson, 2004). This research has demonstrated that infants show intentional communicative behavior from birth (Trevarthen & Aitken, 2001). Infants are not merely passive creatures; they actively engage in social interaction with their caregivers. Through movements of the eyes, face, hands, and legs, infants show sensitive responses when communicating with their caregivers (Trevarthen, 1985). These movements are ordered in a way that allows the caregiver to follow a turn-taking pattern (Schaffer, 1977a). Emotional engagement between infant and caregiver can be signified through the co-regulation of their body movements (Trevarthen, 1985). (See Trevarthen & Aitken, 2001, for a review of the psychological infant-mother interaction research and theoretical perspective.)
Theories and methods of study of communication concerning infant-mother dyads have been adapted to children with developmental disabilities. Social interaction and communication between children with developmental disabilities and their caregivers occur and develop during play activities and social routines as they do in nondelayed children. The caregiver responds to and expands upon the child's expressions by co-regulating contact, joint attention, turn-taking, rhythm, tempo, and emotions (Burford, 1990; Janssen & Rødbroe, 2007; Olsson, 2004). Co-regulation of movements, attention, and emotions between child and caregiver are an important issue to study for the development of communication (Janssen & Rødbroe, 2007).
Studying body movements may be a useful approach when researching social interaction and nonverbal communication of individuals with severe developmental disabilities. Clinical practices and research of individuals with congenital deafblindness have used a movement-based therapeutic approach to support development and improve communication for these individuals (Janssen & Rødbroe, 2007; Wheeler & Griffin, 1997; Writer, 1987). How others respond to and co-regulate motor behavior exhibited during social interaction is the key element in the child's communicative development according to movement-based approaches. Therefore, the caregiver's sensitivity to the child's physical expressions is vital. Co-regulation of body orientation, physical proximity/contact, and movements are some of the variables emphasized in a movement-based approach (Janssen & Rødbroe, 2007).
A study by Burford (1988) is one example of research in psychological infant-mother interaction. Burford investigated rhythmic actions (groups of cyclical movements) exhibited by children with severe developmental disabilities during social interaction with a partner. A frame-by-frame analysis (.01 s) of video recordings was used to label the movement cycles. Regularity within each category of action cycle was found to indicate the communicative purpose of the child.
Technological Development of the Measurement of Body Movements
Empirical research investigating body movements in mother and infant social interactions expanded in the 1960s facilitated by new recording technologies. Social interactions between infant and caregiver were the subject matter of investigation and microanalyses of video observations were used as methods of analysis (Beebe, Knoblauch, Rustin, & Sorter, 2005; Schaffer, 1977b). Burford's 1988 study, as introduced above, is one example of using an electronic time counter in video sequences, making it possible to execute a frame-by-frame behavior analysis. Analyses were performed by trained observers who rated segments of the video sequences identifying behaviors, such as seconds of eye contact or number of smiles, and rated emotional response or attention level.
Some of the problems with this methodological approach were that to carry out these observer ratings was very time consuming and that acceptable reliability of the ratings could be difficult to obtain. Some of these problems, for example, obtaining exact quantitative measurement, were partly solved toward the end of the 20th century due to advances in digital video recording and computer technology. The Qualisys Motion Capture System (QMCS), as used in this study, is one among a number of systems that can measure and track body movement. The QMCS provides accurate data on body position and movements by tracking retro-reflective markers attached to the test participant. Cameras capture the markers' positions, and computer software analyzes the data (see description of the QMCS in the method section and at www.qualisys.com). No other studies were found using the QMCS or another similar program to study social interaction among individuals with developmental disabilities.
Body movements between two people may be used as an indicator of the quality of interaction between them. Exploring advances in analytical technology, the present study intends to pursue the objectives of earlier research, investigating how the measurement of body movements may be useful in studying communication and social interaction.
The aim of this study is to explore the usefulness and effectiveness of measuring body movements in social interaction. This will be conducted by an investigation of how observer ratings of social interaction correspond to the QMCS's measurement of body movement parameters.
A case study design was deemed appropriate for this preliminary study, exploring advances in analytical technology in researching social interaction. A four-and-a-half-year-old boy with severe developmental disabilities, moderate hearing loss (average three frequency hearing loss of 50 dB), low vision (6/18), and spastic cerebral palsy due to congenital cytomegalovirus syndrome was the participant in this study. Using the Bayley Scale III (Bayley, 2006), the boy's developmental age was estimated to be 13 months. This estimation may not be wholly accurate as many of the items on the Bayley Scale III presume no disabilities, neither perceptual (vision and hearing), nor motor functioning.
The boy reacted to stimuli in his surroundings and changes in the environment and routines. He employed hearing and visual aids but was capable of using his residual vision (e.g., looking at objects and faces). He was unable to walk alone and mastered only a few manipulative skills, for example, reaching for objects. He was dependent upon others to eat and drink, and mobility was restricted to being pushed in a wheelchair. Expressions, such as body movements, eye contact, and smiles, were interpreted as forms of communication by his caregivers. He responded vocally to voices and sounds but was unable to form words. The boy responded to and recognized known objects (toys with light and sound) and familiar persons. He expressed this by smiles and vocalizations. Initiating communicative interaction was often a difficult task for the caregiver and subsequently required the ability to understand and respond to the boy's expressions of communication. His communication was found to be intentional but not symbolic.
The boy's teacher had a special-education degree and was skilled in communication and social interaction with children with developmental disabilities and congenital deafblindness. The boy and the teacher had gotten to know each other fairly well over a period of three years.
Information about the procedure and ethics was carefully given to all persons involved. Written consent was obtained from both the teacher and the boy's parents.
Interactions between the boy and his teacher were captured using the QMCS and recorded using two video recorders at the Babylab of The Department of Psychology, University of Copenhagen. The pair was observed and recorded over two episodes of 10 minutes of free play activity in a face-to-face position. The boy sat in his wheelchair, and the teacher sat in a chair in front of him. The first two minutes of sustained interaction from the second episode were selected as data for analysis. The duration of two minutes of interaction has been found to be a sufficient parameter for nonverbal communicative analysis and is a frequently used time interval in similar research designs studying infant-mother interaction (Beebe et al., 2005).
Selected measurements were inspired by psychological theory drawn from research in infant-mother interaction (Trevarthen & Aitken, 2001) and devised keeping in accordance with the subject matter of this study: children with dual sensory loss and developmental disability (Janssen & Rødbroe, 2007; Rødbroe & Souriau, 1999).
The QMCS is an optical marker–based system that captures and calculates 2-D and 3-D coordinates for each marker. By placing retro-reflective markers on the test person, the system is able to record the markers' positions in 3-D space. Data is captured by the Qualisys Track Manager software, which is used to analyze the data (www.qualisys.com).
The QMCS is found to provide accurate and reliable data tracking body movements. To try out different kinds of measurements that reflect important aspects of the social interaction as introduced, three parameters were selected before the interaction was recorded: (1) distances between body parts (short distance may indicate physical closeness), (2) differences in velocity of body parts (small differences may indicate synchronized movements), and (3) angles between body parts (may indicate body orientation and position). Markers were placed on the hands and heads of the child and the teacher. The unit for measuring was one frame, and the duration of one frame was 1/60 of a second. All movements were measured by the QMCS and analyzed frame by frame by the Qualisys Track Manager software.
Distance between the two active hands
One optical marker on the boy's left hand and one on the teacher's right hand were used. The distance was measured in decimeters.
Difference of velocity of the two active hands
The speed of movement of the child's left hand minus the speed of movement of the teacher's right hand were measured. The velocity was measured in decimeters per second.
Angle between heads
Three optical markers were used on the heads. Two markers were placed toward the front of the head (above the eyes) of the boy and one on the front of the head (in line with the nose) of the teacher. The three markers created a triangle. The angle by the marker on the teacher's head was used in this study (see Figure 1). The angle was measured in degrees. Face-to-face orientation and/or shorter distances between the heads increased the angle.
Two variables (attention and emotional response) were selected before the observation to represent relevant aspects of early stages of social interaction and communication development among individuals with severe developmental disabilities (McLean & Snyder-McLean, 1987). The two variables resembled part of a rating system used in a former study on children with congenital deafblindness (Dammeyer, 2009), and similar rating variables have been used in similar mother-infant interaction experiments (Beebe et al., 2005). Inter-rater reliability was evaluated for the two variables in the study by Dammeyer (2009). The level of agreement was considered to be good or excellent (kappa values with quadratic weighting were 0.79 and 0.86) (Fleiss, 1981). In both studies, ratings were made on children with deafblindness and/or developmental disabilities in social interaction with a teacher and completed by the same coders. The coders were educational consultants working with children with deafblindness and/or developmental disabilities experienced in working with the theoretical frame of mother-infant interaction. The coder used in this study did not know the child. To maintain independence from the QMCS parameters, the observer ratings were completed before the measurements of the QMCS parameters were analyzed.
The unit for rating was one communicative act. A communicative act was defined as any verbal or behavioral action or reaction as part of the social interaction between the child and the caregiver (Janssen & Rødbroe, 2007). The child was rated for the full 120 seconds, and all communicative acts were documented. Here follows a description of the two variables used in this study (see also Dammeyer, 2009).
Attention was rated on a scale ranging from 1 to 5. The score was high if the child was observed paying attention to the social interaction and low if the child was not. Attention included listening, looking, a verbal response, arousal, and other observed behaviors interpreted to be linked to the occurring social interaction. The concept of attention was adapted from theories drawn from research on mother-infant interactions (Janssen & Rødbroe, 2007). A score of 1 indicated that “the child does not pay any attention to the social interaction” while a 5 was rated if “the child pays full attention to the social interaction.” A score of 2 was given if the child “paid only little attention,” 3 indicated “neither paid nor did not pay attention,” and 4 meant “paid some attention.”
Emotional response was also rated on a scale ranging from 1 to 5. The score was high if the emotional response was observed to be in line with and responsive to the social interaction. Emotional response included such expressions as being sad, angry, surprised, happy, etc. and was interpreted from the child's facial expressions, body position, verbal response, movements, etc. A score of 1 indicated that “the emotional response was not at all in line with the social interaction.” A 5 indicated that “the emotional response was very much in line with the social interaction.” A 2 was given if the emotional response was “only a little in line with,” 3 indicated “neither in line with nor not in line with,” and 4 meant “in line with” the social interaction.
The body movement parameters measured by the QMCS and the observer ratings of the social interaction were gathered in one Excel 2003 file and one SPSS 17.0 file. Analysis of the body movement parameters and the social interaction variables were made by graphic representation in line charts and by nonparametric gamma-tests. The gamma-test tests the association of cross-tabulated data, and test values range from −1 (100% negative association) to +1 (100% positive association). Tests were completed separately for all pairs of variables. To avoid Type I errors on account of the large number of frames (7,200), the level of significance was set to p < 0.001 (Howell, 2002).
In the first 30 seconds of the two-minute episode of free play, the teacher established contact with the boy by calling his name, obtaining eye contact, and by touching his hands. After the first 30 seconds, they had contact and started a familiar game of putting the thumb into the mouth and having the partner pull it out. The game went on with increased intensity during the next 90 seconds. The number of communicative acts was 22.
Figure 2 shows the graphical representation of the three QMCS parameters of the body movements and both observer ratings of the social interaction. After 30 seconds (1,800 frames), the contact between the boy and the teacher was established, and the thumb game began. A descriptive analysis of the graph shows that after the 30 seconds the distance between the two active hands decreased, indicating a higher degree of physical proximity of the hands; the difference of the velocity of the two active hands decreased, indicating a higher degree of synchrony of movements; and the angles between the heads increased, indicating more and closer face-to-face orientation. The observer ratings of attention and emotional response increased after the first 30 seconds.
The association between measured body movements and observer-rated psychological variables of the social interaction was measured and found to be significant with a p level of p < 0.001 (see Table 1). High scores on attention and emotional response were significantly associated with short distance between hands, little difference in the velocity of the hands, and a greater angle between heads (closer and more face-to-face orientation), respectively.
Results show that face-to-face orientation, synchronized movements, and closeness of active body parts (the active hands) may be associated with the observer ratings of the social interaction.
The two observed psychological variables and the three measured body movement parameters included in the present study may reflect different aspects of co-regulation: synchronicity of movements, body orientation, joint attention, emotions, engagement, and intentionality. Co-regulation has been emphasized as a prominent aspect of infant-mother interaction and is understood as the dynamic alteration of actions with respect to the mother's actions (Fogel, 1993). Among children with congenital deafblindness and/or severe developmental disabilities, the co-regulation of different aspects of body movement interaction has been presented as being important for the child's cognitive, social, and communicative development. Co-regulation of attention and emotion as explored in this study and co-regulation of physical proximity, rhythm, tempo, and intensity have all been reported to be important issues (Janssen & Rødbroe, 2007). This study indicates that co-regulation may be studied successfully using the QMCS or a similar motor capture system.
The association between body movement and social interaction indicates that body-movement measurement may be a valuable technique when studying social interaction and communication. As exemplified in this study, this technique can be especially useful when studying interactions, based on nonlingual communication, between individuals with severe developmental disabilities and their caregivers. The QMCS is found to provide accurate and reliable data, which are objective and quantitative, thereby making the method a valuable contributor to future research in this area. It expands the procedural possibilities when researching individuals with severe developmental disabilities by measuring reliable data and providing opportunities for studying social interaction and communication in greater detail. Body movements may provide valuable information to a caregiver that may improve the quality of communication with a person with severe developmental disabilities. This information may aid in recognizing subtle motor expressions made by the child to help the caregiver to engage in and maintain regulation of mutual contact, joint attention, turn-taking, rhythm, tempo, and emotions (Janssen & Rødbroe, 2007; Olsson, 2004). Other aspects of interest may be engagement and intentionality. Different movements patterns may indicate if the movement is intentional or not. More research is needed on these issues.
Methodological Limitations and Perspectives
This study is a preliminary case study exploring the implementation of a new methodological concept and techniques of studying the social interaction and nonverbal communication of children with severe developmental disabilities. There are several limitations present in this study, but it shows one way in which technology-based, quantitative measurements and ratings of social interaction can be combined to improve methods of research. But still much more work is needed to make this tentative approach practical in developmental disability research. It is important to elaborate on both the theoretical and methodological approach to the measurement of body movements associated with social interaction and communication. This study acts as a first step into this field. The QMCS is an accurate and reliable tool, but the data interpretation needs a theoretical frame. The theories and concepts from psychological research in infant-mother interaction (Beebe et al., 2005; Trevarthen, 1993; Trevarthen & Aitken, 2001) offer one solution.
This study is a preliminary case study involving one child and one teacher in a single environment. This limits the ability to generalize the results until further investigation is undertaken. Other kinds of quantitative and qualitative evaluations of social interaction among children and their caregivers need to be explored in combination with the measurement of body movements.
The aim of this study was to explore and compare two different methodological approaches: observer ratings and objective measurement of social interaction. This study cannot fully explain the association found between the observer ratings of social interaction and the measurements of body movements. One question is to what degree are the observer ratings of social interaction based on the rater's interpretation of body movement. The emotional response might, for example, be rated highly if the observer noted that the child was more animated. More research is needed to explore how body movements and social interaction may have any meaningful relationship.
It was found that during social interaction with his teacher, the body movements of the boy with severe developmental disabilities, tracked by the QMCS, were associated with observer ratings of the social interaction. This preliminary study illustrates that capturing body movement parameters, like body positions, velocity, and angles, may be a practical approach to studying social interaction and communication between individuals with severe developmental disabilities and their caregivers.
Microanalysis of body movement may be a methodological window to better understand how individuals with severe developmental disabilities using nonverbal communication can be supported. Trevarthen (in Hendriks-Jansen, 1996) illustrated that studying human movements and actions before the use of film was as difficult as studying the planets before the invention of the telescope. Technologies such as QMCS may open up new opportunities to expand upon the way in which developmental disability research on social interaction and communication can be implemented.
Editor-in-Charge: Glenn T. Fujiura
Jesper Dammeyer (email: Jesper.Dammeyer@psy.ku.dk), University of Copenhagen, Department of Psychology, Øster Farimagsgade 2A, 1353 Copenhagen K, Denmark; and Simo Køppe, University of Copenhagen.