Content validity of an instrument for motor assessment of youth with autism

Introduction: Children and adolescents with autism spectrum disorder (ASD) present motor disorders that have been the focus of physical therapy interventions. However, the standardized motor assessments available in the literature have important gaps, among them the complexity of the tasks evaluated and the absence of qualitative information about the subjects' performance. Objective: To develop and evaluate the content validity of the Gross Motor Assessment of Children and Adolescents with ASD checklist (GMA-AUT checklist). Methods: The GMA-AUT checklist was sent to a committee of experts for content validation. The content validity index (CVI) was used to assess the degree of agreement between the experts. To verify the content validity of the checklist, the minimal acceptable CVI was 0.80. Results: Based on the suggestions made, the GMA-AUT was reformulated and submitted to the same panel of experts for reassessment. In the second and final draft of the checklist, only two items had a CVI of 0.88, while all others reached a CVI of 1.00. Conclusion: The GMA-AUT checklist presents adequate content validity for assessing gross motor in children and adolescents with ASD according to experts in the field.

section varies from item to item, being the highest score correspondent to the one found in typical development.
The sum of the scores of each section ("performs or does not perform" + "how") represents the item's score.
To obtain the final GMA-AUT score, the scores of all items must be summed and then divided by the maximum score of the checklist. The final score of the individual is in percentage (Figure 2). The first draft of the checklist had 18 items, which are listed in Table 1.

Expert review
The group of experts included nine neuropediatric physiotherapists with experience in caring for children and adolescents with ASD, which were selected by convenience. According to Lynn,13 the definition of the number of experts is somehow arbitrary when determining the content validity of an instrument.
However, the author suggests that a minimum of five experts would provide a sufficient level of control for chance agreement, with a minimum number of three experts being acceptable if the content area is very restricted. Other researchers in the field of instrument development agree with the number of experts proposed by Lynn 13 and suggest five to ten professionals, 14,15 considering that a larger number of experts can provide more information about an instrument. 16 The initial search for experts was performed in scientific articles of the area and in posts on social networks. Professionals working in public rehabilitation institutions, private physical therapy clinics and physical therapy school clinics were selected.
After the initial contact with the experts, we sent the informed consent form (ICF) for those who accepted the invitation so that they could manifest their assent to participate in the research.

Data collection
After assenting to participate in the research, the   (4) With free arms, he/she changes to kneeling, half-kneeling and stands up.
(3) Supporting his/her weight on the arms against the body, he/she changes to kneeling, half-kneeling and stands up.
(2) Supporting his/ her weight on the arms against the floor, he/she moves into bear standing and then standing up.
(1) With the arms against furniture, he/ she pulls him/ herself for standing up. (5) mean expert proportion (MEP), which is the mean PRE. 15 The acceptable agreement rate among experts to verify the validity of a new instrument, in general, must be at least 0.80 and, preferably, higher than 0.90. 15,17

Results
During the process of content validity, the GMA-AUT instrument underwent two rounds of evaluation with the experts. After round 1, one expert withdrew consent to participate in the study, so the answers from this person were unconsidered in all rounds of assessment of the instrument. Then, eight experts participated in the whole process of content validity.
In round 1, the I-CVI ranged from 0.50 to 1.00 (Table   2), the S-CVI/Ave was 0.92, the S-CVI/UA was 72%, the PRE ranged from 83% to 100%, and the MEP was 91% (Table 2).     Note: = Item rated as 3 or 4 in the questionnaire of content validation of GMA-AUT checklist; x = item rated as invalid or less valid by the experts. Regarding motor stims/stereotypies, there was divergence among experts. Three professionals justified that such item should be excluded, as it did not concern motor skills directly. This disagreement also

Content validity
In this study, we evaluated the content validity of

Differentials
Motor abnormalities in ASD manifest early in childhood and often precede the emerging of primary deficits, besides presenting intrinsic relationship with central characteristics of ASD as they affect the learning of perceptual motor skills and limit social interactions. 3,4 as the unipodal balance assessment and the horizontal jump, for example. However, we believe that these skills are indirectly included in the evaluation of items such as kicking a ball (which requires unipodal support while the contralateral foot performs the kick) and vertical jump (which assesses the ability to provide impulse).
Content validation, according to the COSMIN criteria, 10 can also be performed on the target population.
However, we considered that this requirement would not be applicable as the target population in this study are children and adolescents with ASD. Another COSMIN's requirement that was not applied refers to recordings and transcripts of meetings and interviews, which did not occur since the questionnaire for content validity was closed-ended (but with empty space for writing suggestions), and it was replied by e-mail.

Conclusion
The content proposed by the GMA-AUT checklist was considered valid from the perspective of experts in the field. Nevertheless, it is noteworthy that it is still necessary to assess the reliability of the GMA-AUT to be used both in clinical practice and in research. Currently, the GMA-AUT checklist is available as a support material in the teaching-learning process in the academic context of physical therapy.