Interrater and Test–Retest Reliability of the Sensory Integration Clinical Observations

BACKGROUND AND PURPOSE: Ayres (1972) developed structured clinical observations of postural and motor tasks to assist in assessment of sensory integration and sensory processing difficulties. These tasks assess client factors and performance skills, such as soft neurological signs, motor performance, and quality and duration of motor and sensory performance, that may support or impede ability to engage in occupational performance areas such as activities of daily living, play, or academics. Areas assessed are postural reflexes, cocontraction of muscles, muscle tone, extraocular muscle control, vestibular system function, integration of function of two sides of the body, praxis skills, and quality of movements.

Some researchers have developed similar tests (Dunn, 1981; Mutti et al., 1974; Wilson et al., 1994); however, limitations of these include lack of psychometrics, limited number of observations, and lack of standardization. The Sensory Integration Clinical Observations (SI-COs), adapted from Ayres (1972), was developed to address these problems by manualizing the administration of 20 commonly used clinical observations to ensure uniform administration, including a 4-point rating scale to increase sensitivity as an outcome measure, and providing formal scoring criteria for each item. Preliminary test–retest reliability suggested the measure was reliable over a 1-wk period, but formal examination of interrater reliability and test–retest reliability was needed before further investigation and use of the measure.

METHOD: This study used a quantitative interrater and test–retest reliability design. We tested two hypotheses: (1) The SI-COs will demonstrate test–retest reliability of at least .70 for total score, and (2) the SI-COs will demonstrate interrater reliability of at least .70 for total score. Participants were 17 children (6 girls, 11 boys) aged 4–10 yr (M = 6.2) and 7 occupational therapists (experience: 5 yr, n = 5; 10 yr, n = 1; 35 yr, n = 1) from a convenience sample at an occupational therapy sensory integration clinic. Inclusion criteria were as follows: age 4–12 yr; no parent-reported cognitive or intellectual disability (estimated IQ >70), uncontrolled seizure disorder, diagnosis of neurological motor coordination problem (e.g., cerebral palsy), or mental health diagnosis involving psychosis (e.g., bipolar disorder or schizophrenia; children with sensory processing disorder, high-functioning autism, and comorbid attention deficit disorders were eligible); sufficient understanding of English to follow simple directions; and no completion of the clinical observations as part of routine clinical care within the past 6 wk. Institutional review board approval and informed consent were obtained.

Participants were administered the SI-COs by a trained occupational therapist who scored the test during administration. Each testing session was videotaped for later independent interrater reliability scoring. Retesting occurred with the same occupational therapist 2–4 wk later for participants who agreed to this.

RESULTS: Interrater reliability for both initial testing and retesting sessions (n = 18 sessions) was ICC (2, 1) = .96. Test–retest reliability was ICC = .96.

CONCLUSION: The SI-COs had excellent interrater and test–retest reliability over a 2- to 4-wk period. With manualized administration and scoring, this tool may be used reliably by novice occupational therapists. Excellent reliability supports clinical and research use as a potential proximal outcome measure of change for sensory integration intervention.

IMPACT STATEMENT: Development of reliable, valid, and sensitive outcomes measures at both proximal and distal levels is important to support the efficacy of Ayres Sensory Integration® intervention. This study is a first step toward development of a reliable outcome measure of sensory–motor factors that support or hinder participation.

References

Ayres, J. (1972). Sensory integration and learning disabilities. Torrance, CA: Western Psychological Services.

Dunn, W. (1981). A guide to testing clinical observations in kindergartens. Rockville, MD: American Occupational Therapy Association.

Mutti, M., Sterling, H. M., & Spaulding, N. (1974). Quick Neurological Screening Test. San Rafael, CA: Academic Therapy Publications.

Wilson, B. N., Pollock, N., Kaplan, B. J., & Law, M. (1994). Clinical Observations of Motor and Postural Skills. Tucson, AZ: Therapy Skill Builders.