Trail Research On Sensory Integration Functions Assessment Scale (SIF)

According to review the studies of sensory integration function, and analyze the scales which are used to evaluate sensory integration function in mainland, we find that the number of scales is few and the content of scales can’t cover all aspects of the function, therefore, to evaluate the function in a more scientific and objective way, excellent scale with comprehensive content is needed.<The Sensory Integration Functions Assessment Scale> (SIF) is based on the sensory integration developmental theory proposed by Buddy, Lane and Murray in 2002. It is used to screen out children (3 to 10 years old) who ill with Sensory Integrative Dysfunction. It has seven sub-scales, and its content covers all aspects of sensory integration function, including posture and movement, bilateral action integration, sensory discrimination, sensory modulation, sensory seeking, attention and activity, emotional and behavioral responses. The SIF showed good reliability and validity in the usage of Taiwan. This research aim to analyze test its applicability in children of Chinese mainland, including examine its psychometric property, including item analysis, reliability and validity, and application analysis. To lay the foundation for the introduction of the scale.The research consist of five parts:First of all, it reviews the basic knowledge and related studies of sensory integration function.Secondly, it introduces the psychometric theories which are used to assess the quality of SIF——Classical Test Theory (CTT) and Item Response Theory (IRT).Thirdly, the first trial of SIR In this part, we use item analysis (difficulty and discrimination), reliability and validity to assess the quality of SIF. The result of reliability shows that the Cronbach a of all the sub-scales and the whole scale are over 0.834, the corrected split-half reliability of them are from 0.922 to 0.966. That means the internal consistency reliability of SIF and the sun-scales are excellent. The result of validity shows that the correlation between all the seven sub-scales and the whole scale are significant, and the correlation between the sub-scales are lower than the correlation between whole scales and the sub-scales, which means the construct validity is good. To sum up, the scale can be used to do the large-scale test.The forth part is to analysis result of SIF’s psychometric test in large-scales test. In this part, we not only use the CTT to analyze the data, but also use the IRT (Rasch Model) and case study, and we try to evaluate its property in different ways. The result of reliability shows that the Cronbach a of all the sub-scales and the whole scale are from 0.807 to 0.966, the corrected split-half reliability of them are from 0.894 to 0.989. That means the internal consistency reliability of SIF and the sun-scales are excellent. The test-retest reliability are from 0.434 to 0.828. All the reliability indicators are compliance with the requirement for a good scale of psychometric. The result of validity shows that the correlation between all the seven sub-scales and the whole scale are significant, from 0.729 to 0.860, and the correlation between the sub-scales are from 0.439 to 0.700, lower than the correlation between whole scales and the sub-scales, which means the construct validity is good. The result of confirmatory factor analysis shows that the seven dimension of the scale are three-order model. Taking the<Sensory integration checklist> made by Zheng Xinxiong as criterion, the result shows that the correlation between it and SIF is 0.718**. The result of Rasch model analysis demonstrated that: all the sub-scales are unidimensional. The infit and outfit of 95 items are from 0.5 to 1.5 which means they fit well with Rasch Model. And the SIF can provide more accurate evaluation for the level of ability from 0 to 2. What’s more, the result of case study shows that the scores of the two children in SIF can reflect their sensory integration function precisely. In conclusion, SIF has good property and can be used as evaluation tool to assess sensory integration function of 3-10 years old children in Chinese mainland.Finally, the study analyze prevalence rate of Sensory Integrative Dysfunction and compare the function of children in different gender, age, mature neonates or not, and parents’education level. The result show that, girls are outperform than boys on sensory integration as a whole. The sensory ability are improved with age increase, but it’s not always true with regard to sub- scales. Mature neonates are superior in all aspects of sensory integration to children who are not. Parents’ education level have no significant influence on their children’s sensory integration.Conclusion:SIF shows good property with different psychometric measures, so it can be used as a tool to screen out sensory integration dysfunction children in Chinese Mainland.