*4.4. Strengths and Limitations*

A major strength of this study is that as far as we know, it is one of the few articles using video-coding to investigate the impacts of eye-gaze AT on communicative interaction between dyads. Using a system-based approach to coding guided by the foundation of previous research enables the meaning construction of the dyadic communicative behaviors in the EGAT and NEGAT conditions. The acceptable to good inter-rater reliability increased the credibility of this study. Higher kappa values were found in the EGAT condition compared to the NEGAT condition, indicating that the behaviors in communicative interactions were easier to observe and reach the consensus when the children/youths used EGAT compared to the other condition. Moreover, this study applied a three-tiered method, combining group, intermediate and individual analysis to validate group results and reduced the bias in that extreme cases could be hidden by group patterns [8]. Individual case studies enhanced understanding of typical and atypical individual patterns and provided a preliminary reference for further clinical implications.

Several limitations of the present study should be acknowledged. First, the small samples due to the low prevalence rates, the vulnerable health conditions [2,51], and accessibility to EGAT of our target group affect generalizability. The results should be interpreted with caution. In addition, the videos were collected from parents or teachers in order to observe their communication in natural contexts. Although instruction checklists for film clips were provided, we were finally able to use only 12 videos from six dyads for coding analysis. There were several reasons for excluding videos, for instances, technical issues, lack of information or visible behaviors in communicative interactions, or dissimilar activities in the EGAT and NEGAT conditions, which made comparisons difficult and meaningless. To reduce the likelihood of biased results from small samples, a three-tiered method was used to validate group results and provide clinically useful information. Future research could consider including trained researchers to collect film clips and, if possible, recruit more participants to consolidate the findings and shed more light on this

field of knowledge. However, the time expenditure required for video-coding analysis and inter-coder training might be necessary to take into consideration.

Another limitation is that we chose the best two videos in each dyad, which could represent the communicative interactions in a specific activity. The momentary observations based on videos might not be fully representative of the dyadic communicative interactions across various daily activities. Researchers could consider a combination of different methodologies to strengthen the research findings [51], for instance, integrating interviews with proxies to triangulate quantitative data and provide a comprehensive picture of the dyadic communicative interactions in daily life.

#### **5. Conclusions**

The results demonstrated that the children/youths increased initiations on communicative interactions, and the communication partners decreased dominance in communicative turns and made fewer initiations in the interactional structure when using EGAT compared to the NEGAT condition. Moreover, in communicative functions, children/youths tended to provide more information using EGAT, and the communication partners made fewer requests to direct children's responding behaviors, in contrast to the interaction in the NEGAT condition. The communication activities and the structure of the environment (e.g., play or school lessons), eye-control skills, and communication abilities could influence the dyadic interaction.

EGAT shows the potential to increase communication intelligibility, enable children/youths with complex needs gain power of communication, and facilitate the sharing of information in natural contexts. More research is needed, enrolling more participants and combining different research methods to improve generalization, and examining interactive strategies used by communication partners to support children's communication behaviors while using EGAT to guide future interventions.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/ijerph18105134/s1, Table S1: Turns, moves, communicative functions and modes of communication in communication partners, Table S2: Turns, moves, communicative functions and modes of communication in children and youths with complex needs.

**Author Contributions:** Conceptualization, H.H. and M.B.; Methodology, H.H., M.B., M.G. and Y.- H.H.; Formal analysis, Y.-H.H.; Investigation, M.B., D.G. and J.M.; Resources, H.H.; Writing—original draft preparation, Y.-H.H.; Writing—review and editing, H.H., M.G., A.-W.H., M.B., J.M., D.G. and Y.-H.H.; Supervision, H.H., M.G. and A.-W.H.; Funding acquisition, H.H. and M.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Vetenskapsrådet/Swedish Research Council (Dnr 2015-02427).

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the ethical review boards in Sweden (Dnr 2018/1809-32), Dubai (DSREC-11/2017\_10), and the USA (protocol ESSP-02).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data presented in this study are available in Supplementary Materials here.

**Acknowledgments:** We thank all participating children and youths, parents, teachers and therapists. We thank the chiefs and the teams in Swedish assistive technology centers, and special schools in Dubai and the USA for assisting data collection and making this study possible. We are grateful to Helena Vandin at Uppsala University for contribution to the inter-rater reliability, and to the Ministry of Education in Taiwan for internal funding to support the first author's work.

**Conflicts of Interest:** The authors declare no conflict of interest.
