Introduction
Cerebral Palsy (CP) is the most common chronic motor disability and a neurological complication in children, resulted from a non-progressive lesion in the developing brain which causes limitation in manual activity, impaired balance, decreased quality of life, and decreased participation in social activities [1, 2]. One area of participation for children with CP is attending school. Children with CP are often difficult to perform the expected functions in schools due to physical, cognitive, and social deficits affecting their ability to actively participate in learning activities and interaction with peers [3]. 
On the other hand, CP-related problems such as movement planning impairment, learning disorders, IQ and cognitive problems, vision and hearing impairments and difficulty communicating with peers, and emotional and psychosocial disorders, affect their level of participation and performance in school [2, 4]. Therefore, careful study of the factors affecting the performance of these children requires the use of a comprehensive and accurate tool. School Function Assessment (SFA) was first developed in the United States in 1998 by Coster et al., to measure a wide range of school-related functional skills. 
The test consists of three main parts: 1: to examine the student’s level of participation in different school activity settings; 2: To examine the support currently provided to the student when s/he performs school-related functional tasks; and 3: To examine the student’s ability to perform specific functional tasks, including physical and cognitive dimensions. These three parts identify strengths and limitations affecting child participation in a wide range of school-related tasks [5, 6]. Unlike traditional tests (e.g. Peabody developmental motor scale and motor skill test) that are designed primarily for normal children, SFA can be used in children with a wide range of functional disorders including sensory, behavioral, emotional, cognitive and motor deficits. The only disadvantage of this test is the high number of questions which has been somewhat solved by considering a separate score for each part and the possibility of reporting the results of only one part of the assessment.
In Iran, most of the used assessment tools measure the child’s academic skills and cognitive and physical dimensions, including hand function. Unfortunately, there is no exact statistics of assessments used in schools, but it seems that the most common assessment tool for school performance is the academic skills assessment tool. These tools are not able to examine all the problems of these children. Therefore, it seems necessary to have access to a test that, in addition to covering these skills, examines the non-academic school-related performance of SP children with a focus on rehabilitation. The aim of this study is to evaluate the content validity and reliability of the Persian version of the SFA in CP children.
Materials & Methods
This is a descriptive/correlational study with cross-sectional design. The study population consists of all children with CP studying in ordinary and exceptional schools in Tehran, Iran during 2016-2017. Of these, 120 were selected using a convenience sampling method. Inclusion criteria were: CP diagnosed by a neurologist, age between 7 and 12 years, CP without concomitant disability (blindness or deafness), and completion of a consent form to participate in the study by parents. All the questionnaires were answered by the students’ parents. In case of their unwillingness to continue participation, students were excluded from the study. A demographic form and SFA questionnaire was used. The SFA has 320 items examining participation (7 items), task support (21 items) and activity performance (292 items). Each question is scored numerically on a scale from 1 to 6, where score 1 indicates extremely limited participation and score 6 shows full participation [5, 6].
After obtaining the necessary permissions, to evaluate the reliability and validity of the Persian SFA test, the standard protocol of International Quality Of Life Assessment (IQOLA) and Lawshe method were used. The content validity of the test was assessed by eight experts and therapists in the field of children with CP in terms of simplicity, relevance, clarity and necessity. Scoring for each item was done according to Lawshe method [7-9]. To evaluate the test- retest reliability of the Persian SFA, we referred to three exceptional schools in Tehran named Soroush, Tavankhahan and Imam Ali, and the centers affiliated to the rehabilitation schools of Tehran University of Medical Sciences. After selecting the eligible students, a day was arranged as a briefing session for the participants in each school or center.
Due to the high number of questions and the possibility of being tired during completion, parents of students were explained so that if desired, the questionnaire can be provided to them for 24 hours to answer the questions or the questions can be read for them by the rater in one or two sessions. After providing the necessary explanations to them, written consent was obtained from the parents and the SFA questionnaire was completed based on the method selected by them. The questionnaires were completed through interview during 90-120 minutes. The entire interview process took place during the hours that parents were at schools. After four weeks, the SFA form was completed again by the parents of all 120 children and the results of both pretest and posttest stages were recorded. They were analyzed in SPSS V. 22 software.
Results
Of the 120 students participating in the study, 53% were boys and 47% were girls with an mean age of 9.4 years (ranged 7-12 years). The distribution of CP was as follows: 10% hemiplegic CP, 27.5% diplegic CP, 5.8% monoplegic CP, and 56.7% quadriplegic CP. Table 1 shows the demographic characteristics of children and their parents. According to the interpretation of the CVI scores, a score above 0.79 is appropriate; a score between 0.70 and 0.79 is questionable and needs to be corrected or revised; and a score less than 0.70 is unacceptable [8]. The results showed that the CVI value was between 0.87 and 1, which is an acceptable value. Therefore, all parts have good content validity. 
According to Lawshe [8], CVR >0.62 is required for eight experts to retain the items. The results of this study showed a CVR of 0.7-1, which indicates an acceptable CVR. Cronbach’s alpha calculation was used to check the internal consistency of the three parts and ICC was used to check the test-retest reliability, the results of which are shown in Tables 2 and 3. Cronbach’s alpha >0.9 indicate excellent reliability; 0.8-0.9, good; 0.7-0.8, acceptable; 0.6-0.7, questionable; 0.5-0.6, poor; and α <0.5, unacceptable reliability. Regarding ICC values, scores >0.9 indicates excellent reliability; 0.75-0.9, good; 0.5- 0.75, moderate; and <0.5 shows poor reliability [10]. 
The Cronbach’s alpha coefficients, shown in Table 2, show an excellent reliability for the first and second parts and a good-to-excellent reliability for the third part, indicating that the Persian SFA has excellent internal consistency. The ICC values, shown in Table 3, show an excellent reliability for the first part and a good-to-excellent reliability for the second and third parts of the Persian SFA.
Discussion
Many of the assessment available tools have been developed based on the skills of normal children in ordinary schools. As a result, they are not well able to assess the challenges facing children with CP. This study examined the validity and reliability of SFA-Persian version in school-aged children with CP. Hwang et al. translated and developed a Chinese version of SFA for cross‐cultural use in Taiwan [11]. Due to differences in the linguistics structure between Chinese and English, incomprehensible sentences were seen in the translated version. Therefore, to increase the readability of the translation, a number of items were adapted by giving examples or using synonymous words. 

For example, due to the fact that the presence of cafeterias is not common in Taiwanese schools, the item of eating in the cafeteria was replaced by eating in the classroom, and items such as card games and ground activities were removed. In the end, a total of 14 items were matched, 13 were deleted, and 3 were added [11]. In the study by Shojaei et al. in Iran, some of the items were culturally adapted. For example, since the Iranian schools do not have a bath for children in the school, this word was changed to the word “toilet”; the dining area replaced the cafeteria; children’s football replaced kickball; Vasati game replaced dodgeball; classroom equipment replaced classroom utensils; and USB flash drives or CDs replaced tape or floppy disks. Moreover, the unit of measurement was changed from 20 feet to 6 meters, and direction was changed from “left to right” to “right to left” [9]. The present study investigates the validity and reliability of this developed Persian version. 

The results showed that all items in the Persian version of this test have good content validity for children with cerebral palsy (CVI >0.7). Content validity of the Persian version of this test is appropriate for normal children in Iran, except in 5 items (CVI <0.7) [9]. For children with movement disorders, its content validity was reported to be acceptable except in 23 items (CVI <0.42) [12]. The content validity of the Chinese version of SFA was also suitable for children with CP (CVI >0.7) [13]. Hwang et al. [14] investigated convergent and known-groups validity of SFA on 64 children: 29 in general education without disabilities, 18 with learning disabilities, and 17 with CP. 

The convergent validity was assessed by the Wayland Adaptive Behavior Scale (VABS). The results showed a moderate to high relationship (0.5-0.7) between comparable areas of the SFA and VABS. The results of known-groups validity showed a significant difference in all three parts of SFA between the three groups. There was no significant difference between normal children and those with learning disabilities in participation and physical tasks, but a significant difference in cognitive-behavioral tasks of these children was observed. Their results, moreover, showed that there was a significant difference in both physical and cognitive-behavioral domains between children with CP and normal children [14]. Our study was performed only to evaluate the content validity of the Persian SFA on 120 children with CP, and a good content validity was obtained.
The ICC for the participation part of the Persian SFA was obtained 0.90; for the second part, it was 0.91 in physical tasks and 0.84 in cognitive-behavioral tasks; and for the third part it was between 0.83 and 0.97, which indicates its good reliability. The test-retest reliability using ICC for the original version of SFA on children with a wide range of movement disorders (including 6 children with CP) for three parts of participation, task support, and activity performance was reported 0.95, 0.80-0.97, and 0.90-0.99, respectively with a total value ranged 0.82-0.98 [5, 6]. 
The test-retest reliability of Chinese version of SFA on 14 children with CP aged 6-18 years using ICC was reported 0.83 for participation, 0.78-0.84 in physical tasks and 0.69-0.82 in cognitive-behavioral tasks related to the second part, and 0.49-0.97 for the third part, and “moderate” in total [13]. The test-retest reliability of the Persian version of the SFA was evaluated by 40 teachers of normal children with an interval of two weeks, which showed that the ICC was in the range of 0.75-0.97 which was good in overall. 
Although the ICC value in cognitive-behavioral and physical tasks was in the good range, but perhaps the reason for the low ICC score in these two domains (0.84 in maintaining and changing position as a cognitive-behavioral task, and 0.89 in functional communication as a physical task) can be related to large motor involvement (due to the quadriplegia prevalence of 56.7% among students) and consequently, low self-confidence in movement and communication.
In assessing the internal consistency of the Persian SFA, Cronbach’s alpha coefficient for the participation part was reported 0.95; physical tasks, 0.99; cognitive-behavioral tasks, 0.95; and activity performance, 0.83-0.99; and 0.92, in total (excellent). This is while the internal consistency using the Cronbach’s alpha calculation method for the Persian version of SFA on 80 children with various movement disorders is 0.96 (excellent). The results of examining the internal consistency of the original version of SFA during its design and development by Rush Analysis method on 363 children with various mobility impairments reported a Cronbach’s alpha of 0.92-0.93 for the participation part, 0.94 - 0.96 for task support, 0.93-0.98 for activity performance, and 0.82-0.92 in total [5, 6]. 
For its Chinese version, Cronbach’s alpha coefficient for the participation part was reported 0.92; physical tasks, 0.94 and cognitive-behavioral tasks, 0.93 both related to task support section; and activity performance, 0.91-0.97; and 0.92, in total (excellent) [13]. Although their reported Cronbach’s alpha coefficient is at good level for the task support part, but perhaps the reason for its lower score compared to that reported in our study, especially in safety item (0.84), can be attributed to environmental problems and insufficient environmental facilitators such as educational facilities and equipment in exceptional schools in Tehran.
Conclussion
Overall, it can be concluded that the Persian version of SFA has good content validity and reliability for children with CP in Iran, and therapists can use it to determine and advance treatment goals to improve the work performance of CP children. The limitation of this study was the length of the interview time, which reduced the parent’s willingness to retake the test. Moreover, due to the lack of access to similar articles on the validity and reliability of this test in children with CP, it was not possible to provide more comparisons. Studies are suggested to evaluate the reliability between therapists and caregivers of children with CP and children with other defects. Furthermore, it seems that structural validity and factor analysis can complement this study.
Ethical Considerations
Compliance with ethical guidelines
Ethical considerations were observed in this study. Ethical approval was obtained from the Research Ethics Committee of the University of social welfare and Rehabilitation Sciences (IR: USWR.REC.1396.29).
Funding
This study was extracted from the MSc. thesis of first author in the Department of Occupational Therapy, University of Social Welfare and Rehabilitation Sciences, Tehran.
Authors' contributions
Conceptualization: Nazila Akbarfahimi, Mehdi Rassafiani; Methodology: Farzaneh Amiri Ebrahim Mohammadi, Samaneh Hosseinzadeh; Investigation: Farzaneh Amiri Ebrahim Mohammadi; Writing-original draft: Nazila Akbarfahimi, Farzaneh Amiri Ebrahim Mohammadi; Writing-review & editing: All author
Conflicts of interest
The authors declared no conflict of interest.


Acknowledgements
Authors officials of exceptional schools (Soroush, Tavankhahan and Imam Ali) who participated and cooperated in this study.