Relationship Between Context and Sensory Processing in Children With Autism

Sensory Processing Concepts

Scholars have continued to advance the study of sensory processing since Sensory Integration Theory was first developed by Jean Ayres. Dunn’s (1997) Model of Sensory Processing is based in knowledge from neuroscience and behavioral science (Figure 1). The model conceptualizes sensory processing’s contribution to a child’s behavior. Dunn’s model hypothesizes that an interaction between neurological thresholds and behavioral responses exists (Dunn, 1997). In Figure 1, neurological thresholds (i.e., the vertical axis) indicate the amount of stimuli needed for a child to notice or react to the stimuli. The behavioral responses (i.e., the horizontal axis) indicate the manner in which a child responds to stimuli.

OPEN IN VIEWER

Figure 1.

Dunn’s Model of Sensory Processing.

According to Dunn’s Model of Sensory Processing, neurological thresholds and behavioral responses fall on a continuum and interact with each other. The intersection of the two continua results in four sensory processing patterns (i.e., registration, seeking, sensitivity, and avoiding) that offer possible interpretations of a child’s behavior (Dunn, 2006a):

The interaction of neurological thresholds and behavioral responses provides a method for explaining how children process sensory information and guidance for intervention planning (Dunn, 1999).

Sensory Processing and Children With Autism

One of the many difficulties children with autism face is that they respond to sensory experiences in an unusual way (CDC, 2007; Rapin, 1991; Strock, 2004). Children with autism may be sensitive and overreact to auditory stimulation and withdraw (Case-Smith & Bryan, 1999). They may also seek proprioceptive and vestibular input through self-stimulatory, repetitive behaviors such as rocking, spinning, or flapping their hands (Case-Smith & Bryan, 1999). The unique sensory processing patterns of children with autism are associated with dysfunction in attending, arousal, interactions with others, and goal-directed play (Case-Smith & Bryan, 1999; Greenspan & Wieder, 1997; Wieder, 1996).

Moreover, research has shown that sensory processing affects a child’s ability to learn (Ayres, 1979; Dunn, 2001; Dunn & Donaldson, 2001). Specifically, children with autism often fail to notice sensory input that is important; at other times, they are overly sensitive to sensory input and withdraw from stimuli (Ermer & Dunn, 1998; Kientz & Dunn, 1997; Ornitz, 1974; Rapin, 1991; Tomchek & Dunn, 2007; Watling, Deitz, & White, 2001). This pattern of responding makes it difficult for a child to learn because he or she misses important information needed to profit from instruction; this situation becomes critical in the school context. For these reasons, education teams need the appropriate tools to identify and design effective interventions specific to the educational context.

Evaluation Within the Educational System

With a cause for autism not yet understood, effective interventions for children with this disorder are in demand. Because of federal laws and the fact that children spend most of their day at school, the educational system is expected to provide much of the instruction and intervention for children with autism (Individuals With Disabilities Education Act of 1990; Individuals With Disabilities Education Improvement Act of 2004; No Child Left Behind Act of 2001). Public schools are required to provide supportive programming to facilitate children’s success in academics and social interactions. Including assessment of sensory processing in the educational assessment plans of children who have autism may facilitate individualized intervention planning.

A critical factor that the educational system faces when serving children with autism, however, involves conducting valid and reliable assessments. Assessments are used to determine children’s learning strengths and needs. Many authors have suggested that children’s reaction to sensation may interfere with both assessment and educational planning (Ayres, 1979; Case-Smith & Bryan, 1999; Case-Smith & Miller, 1999; Cook, 1991; Kientz & Dunn, 1997; Larson, 1982). If we are better able to understand children’s reaction to sensation, particularly their sensory processing patterns, then professionals will have important information to guide educational planning.

Methods for Assessing Sensory Processing

Behavior is influenced by context; therefore, it is important to consider a person’s context when considering his or her potential to learn (Dunn, Brown, & McGuigan, 1994). The Sensory Profile (Dunn, 1999), which is based on Dunn’s Model of Sensory Processing, has been used in many interdisciplinary studies to document sensory processing patterns in children and adults with and without disabilities (Brown, Tollefson, Dunn, Cromwell, & Filion, 2001; Dunn, 1994; Dunn & Bennett, 2002; Ermer & Dunn, 1998; Kientz & Dunn, 1997; Rogers, Hepburn, & Wehner, 2003; Watling et al., 2001). However, behaviors assessed by the Sensory Profile are not specific to behaviors that would occur in the classroom setting, where a school-age child spends most of his or her day.

Home and school contexts are unique, and contextually relevant assessments are therefore essential. Recently, the Sensory Profile School Companion (hereinafter referred to as the School Companion) has been published to answer this need (Dunn, 2006a). The School Companion is built on Dunn’s Model of Sensory Processing (Figure 1) and coordinates findings with the well-established Sensory Profile (Dunn, 1999). The School Companion reflects scores that are grounded in tested theoretical concepts from the literature (i.e., sensory processing patterns: seeking, avoiding, sensitivity, and registration), providing insights into the child’s sensory patterns in the context of school. The School Companion also offers sensory system scores (i.e., visual, auditory, touch, movement) and a behavior score. Because teachers complete this form, it also includes four school factor scores reflecting teacher and classroom perspectives.

The School Companion is designed for children ages 3–11. The child’s teacher uses a 5-point Likert scale to complete the 62-item questionnaire regarding the child’s responses to daily sensory experiences in the classroom. Additionally, the School Companion is designed to be used in conjunction with the Sensory Profile (Dunn, 1999), which solicits the caregivers’ input, providing a comprehensive view of the child’s sensory processing patterns. We used the Sensory Profile and School Companion in this study of home and school contexts because of their common conceptual basis.

Identifying the Best Source for Evaluation Information

Research has shown that when teachers and parents are asked the same questions about a child, the responses are only slightly correlated, suggesting that each informant has a unique view and that one view cannot be substituted for the other (Achenbach, McConaughy, & Howell, 1987; De Los Reyes & Kazdin, 2005; de Nijs et al., 2004; Kumpulainen et al., 1999). The low correlations between teacher and parent responses to the same questionnaire also suggest that home and school variables differ and indicate the need for contextually designed assessments and different interventions and goals (Achenbach et al., 1987).

Questions may be raised about how home and school versions of sensory processing assessments relate to each other. If assessment of sensory processing only concerns children’s reactions, then home and school assessments will be highly correlated. If, however, the contexts of children’s reactions are an important factor, then both similarities and differences will perhaps indicate the contribution of context to behavior. For this study, we hypothesized that the greatest difference between parent and teacher responses would be seen in the two active behavioral response quadrants from Dunn’s Model of Sensory Processing (i.e., avoiding and seeking). Perhaps the active behavioral response quadrant behaviors would be more noticeable, making it easier for parents and teachers to report behaviors.

In summary, a strong need for effective assessment and intervention for children with autism exists. The school system is expected to provide specialized instruction to meet the unique learning needs of all children. Research has shown that children with autism have unique sensory processing patterns affecting the way in which they respond in their everyday lives, including at home and at school (Ermer & Dunn, 1998; Kientz & Dunn, 1997; Ornitz, 1974; Rapin, 1991; Tomchek & Dunn, 2007; Watling et al., 2001). Being able to determine how specific sensory processing patterns might relate to participation at home and at school will help professionals support children, families, and teachers by demonstrating how therapists might adjust the context to better fit children’s needs.

The purpose of this study was to determine the relationship between sensory processing and context. To address this purpose, we examined Sensory Profile and School Companion findings for the same children, hypothesizing that each assessment would provide contextually specific information about the sensory processing responses of children with autism at home (Sensory Profile) and at school (School Companion). If context (home and school) and sensory processing patterns are related, then some aspects of sensory processing responses in children with autism will be unique to school. We hypothesized (1) a low relationship between seeking quadrant scores at home and at school and (2) a high relationship between avoiding quadrant scores at home and at school. For this study, we focused only on the seeking and avoiding quadrants. We hypothesized that the active quadrant behaviors from Dunn’s Model of Sensory Processing (i.e., avoiding and seeking) would be more noticeable, making it easier for parents and teachers to report them, and that these two quadrants would have the greatest difference in parent and teacher responses.

Method

Results

The 49 students (43 boys and 6 girls) ranged in age from 3 yr 3 mo to 11 yr 11 mo (Table 1). Five of the children received one service (i.e., occupational therapy, physical therapy, speech–language therapy, special education, or counseling), 10 children received two services, and 34 students received more than two services (Table 2). Of the children, approximately 18% were from various ethnic backgrounds (4 African-American, 2 Asian, 1 Hispanic, 2 other–multiracial), and approximately 82% were White (n = 40).

Teachers’ experience ranged from 0 to ≥26 yr, and educational levels ranged from no degree to a doctoral degree (no degree = 1, bachelor’s degree = 22, master’s degree = 23, doctoral degree = 2, missing = 1; see Table 3). Seven teachers had contact with students 2 days a week, 17 teachers had contact with students 3–4 days a week, and 25 teachers had contact with students daily (Table 4). Most teachers had only 1 year of contact with students (n = 30), but some teachers had multiple years of contact (8 had 2 yr, 4 had 3 yr, 2 had 4 yr, 5 had ≥5 yr).

In addition, parents of the same children completed the Sensory Profile Caregiver Questionnaire. Mothers’ and fathers’ education levels (based on teacher report) ranged from ≤11 yr of school (mother = 0, father = 2) to ≥16 yr of school; with most parents had ≥16 yr of school (mother = 21, father = 21; see Table 5). Other education levels included 12 yr of school or a GED (mother = 12, father = 10), and 13–15 yr of school (mother = 16, father = 15); one parent was unreported (mother = 0, father = 1).

Table 6 summarizes the results of the correlations using the avoiding and seeking quadrant scores from the Sensory Profile (parent report) and School Companion (teacher report). The avoiding quadrants correlated at .59 (p = .01). The seeking quadrants correlation was also positive (.45, p = .01).

As a result of missing data, only 49 of the 56 pairs of teachers and students with autism were eligible for the study. In addition, we excluded cases pairwise in the correlations to minimize the effect of missing data. For this reason, the number of participants reported in Table 6 varies.

When data were plotted, both the seeking and avoiding results had outliers (Figures 2 and 3). When outliers were excluded, the correlation strength improved. The demographics, however, indicated no reason to exclude the outliers from the results; thus, the reported results include data from outliers.

OPEN IN VIEWER

Figure 2.

Scatterplot for seeking quadrant score.

OPEN IN VIEWER

Figure 3.

Scatterplot for avoiding quadrant score.

Discussion

In this study, we examined the relationship between sensory processing patterns at home and school in children with autism. Specifically, we correlated the seeking quadrant scores from the Sensory Profile (completed by parents) and the School Companion (completed by teachers). We also correlated the avoiding quadrant scores from each of the assessments. Data analysis revealed statistically significant correlations, with both good and fair correlations suggesting that sensory processing patterns have both universal qualities (i.e., the impact is the same everywhere) and context-specific qualities (i.e., the impact is specific to a situation or activity) in children with autism.

The avoiding quadrant had a moderate to good correlation coefficient (.59), suggesting that a child’s reactions of being overwhelmed by sensory experiences might sometimes be similar at home and at school. Perhaps children with autism are universally overwhelmed and avoid sensations across contexts. For example, both teachers and parents may report that students with autism hold their hands over their ears to protect against sound, avoid eye contact, or withdraw in response to changes in routine. The context has the potential to provide multiple sensory experiences. Children with avoiding patterns can become easily overwhelmed; stimuli that trigger avoiding reactions can occur anywhere and at any time. However, because the correlation between context and sensory processing was not perfect, data suggest that the home and school contexts also reflect sensory circumstances that are unique to each environment. For example, a family with one child may have a home context that is quieter than the school context. Such parents would likely observe fewer auditory reactions than would the child’s teacher, who observes the child amid a classroom full of students.

For practitioners, these findings can be helpful in intervention planning. Both parents and teachers can implement strategies that could minimize sensory experiences to reduce a child’s need to use avoiding behaviors. Situations in both home and school contexts may be especially overwhelming (i.e., school assemblies, fire drills, shopping at a department store with a parent, or receiving a haircut). Therapists should ask additional questions about these high-risk situations when children have avoiding patterns so that strategies can be created to prevent challenging behaviors. Encouraging parents and teachers to discuss a student’s sensory preferences across contexts can provide additional information that will assist in determining which situations are most overwhelming for a student and in planning more effective interventions.

The seeking quadrant correlation (.45) was in the fair range. For seeking responses, expectations and limitations of a particular context may contribute to the fair correlation between home and school. For example, when a child sings out loud at home, such behavior might be considered delightful. This behavior would be considered disruptive when exhibited in the classroom setting, resulting in a difference in reporting by parents and teachers. Some parents may observe their children in multiple unstructured contexts in which seeking behaviors are acceptable and encouraged. Other parents might provide greater amounts of structure that guide the child’s responses. By contrast, teachers may be more likely to notice seeking behaviors during structured classroom activities (such as seatwork, a time when students are expected to be more controlled) and less likely to take note of seeking behaviors during recess when all students are more active. Alternatively, a child with seeking preferences may maintain self-control during the school day but exhibit more seeking behaviors when at home. As a result, the seeking patterns at home and school may require special attention to understand the impact of seeking on a child’s ability to participate in daily life. Discussion between parents and teachers provides a more comprehensive understanding of a student’s unique sensory processing needs related to seeking.

As shown in Table 6, the correlations using the corresponding seeking and avoiding quadrant scores were significant. However, we also found significant correlations between the School Companion avoiding quadrant score and the Sensory Profile seeking quadrant score (.42, p = .05), as well as between the School Companion seeking quadrant score and the Sensory Profile avoiding quadrant score (.46, p = .01). Both correlations fell in the fair range. Perhaps we found significant correlations across these quadrants because, according to Dunn’s Model of Sensory Processing (Figure 1), both avoiding and seeking patterns are active self-regulation patterns. Self-regulation is on a continuum and is defined as the strategy a person uses to manage his or her own needs and preferences (Dunn 2006a). Perhaps children’s need to control their own sensory input, whether it be to receive more (i.e., seeking) or less (i.e., avoiding), is reflected in these fair correlations.

The literature on multiple informants indicates that when parents and teachers are asked the same question, the answers are only slightly correlated, suggesting that each informant has a unique view (Achenbach et al., 1987; De Los Reyes & Kazdin, 2005; de Nijs et al., 2004; Kumpulainen et al., 1999). This literature also reiterates the importance of receiving information from both parents and teachers during the assessment process and the necessity of using context-specific assessments (Achenbach et al., 1987). In addition, having both home and school information provides an opportunity to identify successful strategies from parents and teachers that may be useful in the other context. Furthermore, using context-specific assessments (home and school versions) can provide a way to respectfully discuss what is happening at home and school without suggesting that either parents or teachers are providing an inaccurate perspective.

We know from the literature that children with autism have unique sensory processing patterns (Ermer & Dunn, 1998; Kientz & Dunn, 1997; Ornitz, 1974; Rapin, 1991; Tomchek & Dunn, 2007; Watling et al., 2001) when compared with those of typically developing peers. This study illustrates that for children with autism, behavioral responses to sensory experiences seem to be context related, suggesting that simply knowing a child’s sensory processing patterns without considering contextual factors would be insufficient information for comprehensive intervention planning.

Conclusion

This study’s findings provide initial evidence for the existence of a relationship between sensory processing and context for children with autism. Both similarities and differences in children’s reactions to sensory stimuli may indicate the influence of context on behavior. The relationships between home and school sensory processing patterns found in this study suggest that therapists must consider the context when assessing and treating the behaviors of children with autism. Comprehensive assessment should include information about behaviors both at home and at school to plan the most effective intervention strategies.