REVIEW 2 major objections 5 minor 46 references
Developer Perceptions on Utilising Low-Code Approaches to Build Accessible and Adaptive Applications for Seniors
T0 review · 2 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash
Pith's one-line read Based on 18 developer interviews, this paper identifies the expectations practitioners hold for low-code tools that build accessible, adaptive applications, and derives design recommendations to make such tools industry-standard.
desk verdict A genuinely new 18-developer dataset with useful recommendations, but the video-only 'ease of use' findings overstate what was measured and the generalization claims need reining in. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The carrying object is AdaptForge, a low-code model-driven engineering tool built on the Eclipse Modelling Framework that takes two domain-specific languages as input, a Context DSL for modelling a user's context of use (impairments, preferences, platform, environment) and an Adapt DSL for conditional adaptation rules anchored to annotated Flutter widgets, and automatically generates adapted Flutter application instances from a base app. The evaluative machinery is a one-hour semi-structured interview built around a screen-recorded demonstration of the tool, followed by iterative thematic analysis whose top-level themes (positive reactions, negative reactions, improvement suggestions, related themes) were reinterpreted into the eight developer expectations. The DSL-plus-MDE pipeline is the mechanism that carries the argument: it is how accessibility work is delegated from manual UI engineering to automated code generation, and participants' expectations are measured against it.
What would settle it
Re-run the study with a second cohort who actually install and use AdaptForge (or an equivalent tool) on a real Flutter project, and compare their expectations and verdicts with the video-demonstration cohort; if hands-on users report materially different priorities, for example much higher friction in modelling or stronger doubts about generated-code maintainability, the transferability of the eight expectations would be refuted.
Extended reading notes
Core claim
The paper's central claim is that the 18 interviews reveal a stable set of developer expectations for low-code accessibility tooling across the whole software development lifecycle, and that these expectations can guide the design of industry-ready tools. Participants responded positively to the idea of generating personalised app instances automatically from context-of-use and adaptation models, motivated both by personal empathy with older relatives and by compliance pressures like the European Accessibility Act, but they consistently conditioned adoption on the tool not degrading the engineering process: generated code must stay readable and maintainable, multiple app variants must not produce merge conflicts, deployment must work within app-store constraints, and accessibility data collection must not drive users away. The paper claims that AdaptForge meets some of these expectations, particularly easy modelling via tree/JSON-style DSLs and fast generation of adapted Flutter instances, while falling short on others, most notably code maintainability, version control, and deployment. On that evidence it contributes eight developer expectations and a set of design recommendations for low-code tool developers generally, arguing the findings transfer beyond both AdaptForge and the senior-user context.
Load-bearing premise
All evaluative judgments in the study are perceptions of a screen-recorded demonstration, not of hands-on tool use, because the authors chose a video-based approach to avoid burdening busy practitioners (Section 4.1); if real use would produce materially different impressions, the findings about developer expectations would be weakened.
Editorial extensions
If this is right
- Low-code tool developers now have an empirically grounded checklist (DE1-DE8) that spans the whole software lifecycle, from accessibility data collection to deployment.
- Adoption of such tools will hinge as much on generated-code readability, maintainability, and version-control behaviour as on the modelling experience; tools that produce merge-prone or unreadable code will be rejected regardless of modelling convenience.
- Deployment support, including server-side generation, hot reload, or over-the-air updates that comply with app-store policies, is a concrete expectation that most accessibility-focused research prototypes currently do not meet.
- Participant testimony judged generative-AI code generation as not yet ready for production accessibility work, which positions MDE-based generation as the more pragmatic near-term route to adapted apps.
Reading between the lines
- A direct follow-up would be hands-on replication: only developers who actually install and use a tool like AdaptForge can confirm whether the video-based impressions reported here match real usage friction.
- The DE1-DE8 expectation set doubles as an evaluation rubric; applying it to other MDE accessibility tools or to generative-AI code generators would test its transferability and reveal coverage gaps.
- If these expectations generalize, the European Accessibility Act and similar regulation will push maintainability and deployment support to the top of accessibility tooling requirements, effectively making them market-entry criteria.
- The three-persona approach (vision, dexterity, cognition) is reusable for other demographics with the same impairment classes, so the recommendations likely carry over to accessibility tooling beyond seniors even though the study did not test that directly.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper reports an interview-based empirical study with 18 professional Flutter developers who watched a screen-recorded demonstration of AdaptForge, a low-code model-driven engineering (MDE) tool that generates accessible and adaptive Flutter applications for senior users. From these interviews, the authors derive eight developer expectations (DE1-DE8) organized across the software development lifecycle and use them to formulate recommendations for low-code tool designers. The central claim is that these insights provide 'empirically grounded recommendations' for designing low-code tools that support accessible and adaptive software development for seniors and other user groups.
Significance. If valid, the findings fill a genuine gap: developer perceptions of MDE-based low-code tools for accessibility are rarely studied empirically, and the DE1-DE8 framework covers a broader range of SDLC concerns (maintainability, versioning, deployment, documentation) than typical tool-evaluation papers. The authors are transparent about participant demographics and provide verbatim quotes, and the qualitative analysis is described in enough detail to be followed. However, the significance is curtailed by two validity concerns: the video-only evaluation method and the narrow, self-selected sample. These issues limit the strength of the 'empirically grounded' and 'industry-standard' claims, but they do not eliminate the value of the interview data as a starting point for hypothesis generation.
major comments (2)
- [Section 4.1, Table 3 (DE3)] The study protocol asks participants to evaluate 'the practicality and usefulness of the context DSL tool, its ease of use' after watching a screen-recorded demonstration, and the same is repeated for the Adapt DSL. No participant ever used the DSL editors. Ease of use is an experiential property that depends on factors absent from a video, such as syntax-error handling, editor responsiveness, error recovery, and the friction of integrating generated code into a real codebase. DE3 ('Ease of use in modelling tools') is load-bearing: it directly justifies the recommended tree hierarchies, icons, and dual textual/graphical representations in Table 3. As it stands, the evidence does not support DE3 as an empirically grounded expectation about actual tool use; it supports at most a perception formed from a scripted demonstration. Please either add a hands-on component to the study or reclassify DE3 as 'perceived ease of use based on a demonstration' and soften the associated design recommendations accordingly.
- [Abstract; Section 8] The abstract and summary claim 'empirically grounded recommendations' and relate the findings to adopting tools as 'industry-standard solutions.' The dataset is 18 self-selected Flutter developers (Table 1), 15 of whom are aged 21–34, recruited via Reddit, LinkedIn, and the authors' networks, with no incentives and no control over prior familiarity with accessibility. There is no 'Threats to Validity' section to bound these claims. The sample cannot support generalization to the broader low-code community. I recommend adding an explicit limitations subsection and replacing 'industry-standard' and 'empirically grounded' with more cautious language, such as 'candidate recommendations' or 'hypothesis-generating findings.'
minor comments (5)
- [Section 3.1] The phrase 'we will emphasise' should be 'we emphasize' for stylistic consistency, and the Figure 2 caption reads 'an illustration of the the abstract architecture' with a doubled article.
- [Throughout] Several symbol placeholders appear instead of intended content, e.g., 'an�-tuple' in Section 3.2, '��� attributes' in Section 3.2, and '� � 18' in Section 4.3. These appear to be PDF-extraction artifacts and must be fixed in the final version.
- [Table 1] The 'Developer profile taxonomy' values mix capitalization styles ('Pragmatist' vs. 'pragmatist'); use consistent capitalization for all taxonomy labels.
- [Section 4.3] The percentages reported for years of experience (50%, 16.7%, 11.1%, 22.2%) do not sum exactly to 100%, and the counts (9, 3, 2, 4) sum to 18; please adjust rounding so the text and table are consistent.
- [General] The paper would benefit from a dedicated 'Threats to Validity' subsection addressing the video-based method, self-selection bias, and the narrow age range of participants.
Circularity Check
No significant circularity: the interview-based developer expectations are independently grounded, with self-citations only providing tool context.
full rationale
The paper's central claims are empirical, not derivational. The developer expectations (DE1-DE8, Table 3) and the associated recommendations emerge from thematic analysis of 18 interview transcripts (Sections 4.2, 5, 6), not from an equation, fitted parameter, or definition that presupposes the conclusion. The authors' self-citations ([41], [42], [43]) supply the AdaptForge tool design, personas, and earlier user-study context, but none of these is load-bearing for the present findings: the developer perceptions were independently elicited and are reported as such. The video-based demonstration is a genuine evidentiary limitation for judgments like 'ease of use' (Section 4.1 states that participants watched a screen-recorded demonstration rather than installing and using the tool), but this is a validity threat, not circularity—no result is equivalent to its input by construction, and no prediction is statistically forced. The paper does not import a uniqueness theorem, does not rename a known result as a new derivation, and contains no fitted-input-called-prediction step. Therefore no circular step can be identified.
Assumptions & free parameters
assumptions (3)
- domain assumption Self-reported perceptions of volunteer interviewees reflect real practitioner beliefs and needs.
- ad hoc to paper A screen-recorded demonstration of AdaptForge is an adequate proxy for hands-on tool use when forming evaluative judgments.
- domain assumption The personas and accessibility findings from the authors' earlier focus-group study [41] are valid design inputs for AdaptForge and for the evaluation.
Cite this review
Pith. "Pith review of Developer Perceptions on Utilising Low-Code Approaches to Build Accessible and Adaptive Applications for Seniors." pith.science (2026). https://pith.science/paper/6MJEVSHX
@misc{pith2026250802968,
author = {Pith},
title = {Pith review of: Developer Perceptions on Utilising Low-Code Approaches to Build Accessible and Adaptive Applications for Seniors},
year = {2026},
howpublished = {\url{https://pith.science/paper/6MJEVSHX}},
note = {Machine review of arXiv:2508.02968}
}
read the original abstract
The global ageing population presents a growing societal challenge, creating an urgent need for inclusive technologies that promote autonomy among older adults. Software practitioners can address this by delivering digital services that enhance seniors' independence and reduce reliance on routine support from family members and healthcare infrastructure. However, traditional development practices, constrained by time and resources, often result in applications with major accessibility and personalisation barriers. Increasing pressure from regulatory requirements, such as the European Accessibility Act (EAA), and the personal empathy many developers feel toward supporting their older loved ones and their own future selves have created a demand for tools that support the development of accessible and adaptive software. To address this demand, this paper presents an interview-based empirical study with 18 software practitioners, evaluating AdaptForge: a low-code model-driven engineering (MDE) tool that enables the efficient creation of accessible and adaptive applications for senior users by mitigating development constraints through automated code generation. Based on these insights, we identify developer expectations for adopting such tools as industry-standard solutions and provide empirically grounded recommendations for designing low-code tools that support accessible and adaptive software development.
Reference graph
Works this paper leans on
-
[1]
United States General Services Administration. 2018. Information and Communication Technology Revised 508 Standards and 255 Guidelines. https://www.section508.gov/develop/universal-design/
work page 2018
-
[2]
Bilal Ahmad, Ita Richardson, and Sarah Beecham. 2020. A Multi-method Approach for Requirements Elicitation for the Design and Development of Smartphone Applications for Older Adults. In 2020 IEEE First International Workshop on Requirements Engineering for Well-Being, Aging, and Health (REWBAH) . IEEE, Zurich, Switzerland; Online, 25–34. doi:10.1109/REWBA...
work page Pith review arXiv 2020
-
[3]
Pierre Akiki, Arosha Bandara, and Yijun Yu. 2016. Engineering Adaptive Model-Driven User Interfaces. IEEE Transactions on Software Engineering 42, 12 (12 2016), 1118–1147. doi:10.1109/TSE.2016.2553035 ACM Trans. Softw. Eng. Methodol., Vol. 1, No. 1, Article . Publication date: September 2025. 30 Shavindra Wickramathilaka, John Grundy, Kashumi Madampe, and...
-
[4]
Firas Bacha, Kathia Oliveira, and Mourad Abed. 2011. A model driven architecture approach for user interface generation focused on content personalization. In 2011 FIFTH INTERNATIONAL CONFERENCE ON RESEARCH CHALLENGES IN INFORMATION SCIENCE. IEEE, Zurich, Switzerland; Online, 1–6. doi:10.1109/RCIS.2011.6006839
-
[5]
Lorenzo Bettini. 2016. Implementing domain-specific languages with Xtext and Xtend . Packt Publishing Ltd, Livery Place, 35 Livery Street, Birmingham B3 2PB, UK
work page 2016
-
[6]
Sara Bongartz, Yucheng Jin, Fabio Paternò, Joerg Rett, Carmen Santoro, and Lucio Davide Spano. 2012. Adaptive User Interfaces for Smart Environments with the Support of Model-Based Languages. In Ambient Intelligence, Fabio Paternò, Boris de Ruyter, Panos Markopoulos, Carmen Santoro, Evert van Loenen, and Kris Luyten (Eds.). Springer Berlin Heidelberg, Ber...
-
[7]
Amina Bouraoui and Imen Gharbi. 2019. Model driven engineering of accessible and multi-platform graphical user interfaces by parameterized model transformations. Science of Computer Programming 172 (2019), 63–101. doi:10.1016/j.scico.2018.11.002
-
[8]
Amani Braham, Maha Khemaja, Félix Buendía, and Faiez Gargouri. 2022. Towards a Model-Driven Ontology-Based Architecture for Generating Adaptive User Interfaces. In Ambient Intelligence – Software and Applications – 12th International Symposium on Ambient Intelligence , Paulo Novais, Joao Carneiro, and Pablo Chamoso (Eds.). Springer International Publishin...
Show all 46 references
-
[9]
Marco Brambilla, Jordi Cabot, and Manuel Wimmer. 2017. Model-Driven Software Engineering in practice: Second edition (2 ed.). Springer Cham, Gewerbestrasse 11, Cham, Ch 6330, Switzerland. doi:10.1007/978-3-031-02549-5
2017 doi
-
[10]
Gaëlle Calvary, Joëlle Coutaz, Laurent Bouillon, M Florins, Quentin Limbourg, L Marucci, Fabio Paterno, C Santoro, N Souchon, David Thevenin, and Jean Vanderdonckt. 2002. The CAMELEON reference framework. Technical Report. CAMELEON Project
2002
-
[11]
Penny Cheek, Linda Nikpour, and Heather D. Nowlin. 2005. Aging Well With Smart Technology.Nursing Administration Quarterly 29, 4 (October 2005), 329–338. BSN, RN
2005
-
[12]
Felipe Dias, Lianna Duarte, and Renata Fortes. 2021. AccessMDD: an MDD approach for generating accessible mobile applications. In Proceedings of the 39th ACM International Conference on Design of Communication (Virtual Event, USA) (SIGDOC ’21). Association for Computing Machin...
2021
-
[13]
José-Manuel Díaz-Bossini and Lourdes Moreno. 2014. Accessibility to mobile interfaces for older people. Procedia Computer Science 27 (2014), 57–66
2014
-
[14]
Dodd, Rukshan I
Connor T. Dodd, Rukshan I. Athauda, and Marc Thomas Philipp Adam. 2017. Designing User Interfaces for the Elderly: A Systematic Literature Review. In Australasian Conference on Information Systems (ACIS 2017) . Association for Information Systems (AIS), Hobart, Australia. http...
2017
-
[15]
Lesly Elguera Paez and Claudia Zapata Del Río. 2019. Elderly users and their main challenges usability with mobile applications: a systematic review. In International Conference on Human-Computer Interaction . Springer, Springer, Cham, Orlando, FL, USA, 423–438
2019
-
[16]
European Union. 2019. Directive (EU) 2019/882 of the European Parliament and of the Council of 17 April 2019 on the accessibility requirements for products and services (Text with EEA relevance). http://data.europa.eu/eli/dir/2019/882/oj PE/81/2018/REV/1, OJ L 151, 7.6.2019, p...
2019
-
[17]
International Organization for Standardization. 2008. Ergonomics of Human System Interaction - Part 171: Guidance on software accessibility (iso 9241-171:2008 ed.). International Organization for Standardization, Geneva, Switzerland. https://www.iso.org/standard/39080.html
2008
-
[18]
Eclipse Foundation. [n. d.]. Acceleo. https://www.eclipse.org/acceleo/ Last accessed 20 November 2024
2024
-
[19]
Nadia Ghaibi, Olfa Dâassi, and Leila Jemni Ben Ayed. 2017. A Tool Support for the Adaptation of User Interfaces Based on a Business Rules Management System. In Proceedings of the 29th Australian Conference on Computer-Human Interaction (Brisbane, Queensland, Australia) (OZCHI ...
2017
-
[20]
Google. [n. d.]. Flutter - Build apps for any screen — flutter.dev. https://flutter.dev/. [Accessed 22-08-2024]
2024
-
[21]
Richard Harte, Leo R Quinlan, Liam Glynn, Alejandro Rodríguez-Molinero, Paul Ma Baker, Thomas Scharf, and Gearóid Ólaighin. 2017. Human-Centered Design Study: Enhancing the Usability of a Mobile Phone App in an Integrated Falls Risk Detection System for Use by Older Adult User...
2017 doi
-
[22]
Apple Inc. 2024. Accessibility. https://developer.apple.com/design/human-interface-guidelines/accessibility
2024
-
[23]
Elmar Krainz, Johannes Feiner, and Martin Fruhmann. 2016. Accelerated Development for Accessible Apps – Model Driven Development of Transportation Apps for Visually Impaired People. In Human-Centered and Error-Resilient Systems Development, Cristian Bogdan, Jan Gulliksen, Stef...
2016
-
[24]
Elmar Krainz, Klaus Miesenberger, and Johannes Feiner. 2018. Can We Improve App Accessibility with Advanced Development Methods?. InComputers Helping People with Special Needs: 16th International Conference, ICCHP 2018, Linz, ACM Trans. Softw. Eng. Methodol., Vol. 1, No. 1, Ar...
2018
-
[25]
Jan Köhnlein and Cédric Brun. 2017. Xtext / Sirius - Integration: The Main Use-Cases . White Paper. Obeo Designer. https://www.obeodesigner.com/resource/white-paper/WhitePaper_XtextSirius_EN.pdf Accessed: 2025-05-20
2017
-
[26]
2008-2025
Legislative Assembly of Ontario. 2008-2025. Accessibility for Ontarians with Disabilities Act (AODA). https: //www.aoda.ca/
2008
-
[27]
Raúl Miñón, Fabio Paternò, Myriam Arrue, and Julio Abascal. 2015. Integrating adaptation rules for people with special needs in model-based UI development process. Universal Access in the Information Society 15, 1 (01 2015), 153–168. doi:10.1007/s10209-015-0406-3
2015 doi
-
[28]
Morey, Rachel E
Stephanie A. Morey, Rachel E. Stuck, Amy W. Chong, Laura H. Barg-Walkow, Tracy L. Mitzner, and Wendy A. Rogers
-
[29]
United Nations. 2006. Convention on the Rights of Persons with Disabilities (CRPD). https://social.desa.un.org/issues/ disability/crpd/convention-on-the-rights-of-persons-with-disabilities-articles
2006
-
[30]
Government of France. 2016. General Accessibility Framework for Administrations (RGAA). https://disic.github.io/ rgaa_referentiel_en/rgaa-companion-guide.html
2016
-
[31]
World Health Organization. 2024. Ageing and Health. https://www.who.int/news-room/fact-sheets/detail/ageing-and- health. https://www.who.int/news-room/fact-sheets/detail/ageing-and-health Last accessed 08 April 2025
2024
-
[32]
Fortes, Herbert Kuchen, Felipe Dias, and Lianna Duarte
Christoph Rieger, Daniel Lucrédio, Renata Pontin M. Fortes, Herbert Kuchen, Felipe Dias, and Lianna Duarte. 2020. A model-driven approach to cross-platform development of accessible business apps. In Proceedings of the 35th Annual ACM Symposium on Applied Computing (Brno, Czec...
2020
-
[33]
Dave Steinberg, Frank Budinsky, Marcelo Paternostro, and Ed Merks. 2009. EMF: Eclipse Modeling Framework (2nd revised ed.). Addison-Wesley, 221 River Street, Hoboken, NJ
2009
-
[34]
Constantine Stephanidis, Demosthenes Akoumianakis, Margherita Antona, and Liam Bannon. 2001. User Interfaces for All: Concepts, Methods, and Tools (1 ed.). Human factors and ergonomics, Vol. 1. CRC Press, Boca Raton
2001
-
[35]
Vladimir Viyović, Mirjam Maksimović, and Branko Perisić. 2014. Sirius: A rapid development of DSM graphical editor. In IEEE 18th International Conference on Intelligent Engineering Systems INES 2014. 233–238. doi:10.1109/INES.2014.6909375
2014
-
[36]
Minh Hieu Vu, Joshua Wyman, John Grundy, and Anuradha Madugalla. 2022. Better understanding diverse end user website usage challenges with browser-based augmented reality approaches. In International Conference on Evaluation of Novel Approaches to Software Engineering . Spring...
2022
-
[37]
W3C. 2015. Authoring Tool Accessibility Guidelines (ATAG) 2.0. https://www.w3.org/TR/ATAG20/
2015
-
[38]
W3C. 2024. Web Content Accessibility Guidelines (WCAG) 2.2. https://www.w3.org/TR/WCAG22/
2024
-
[39]
Willian Massami Watanabe, David Fernandes Neto, Thiago Jabur Bittar, and Renata P. M. Fortes. 2010. WCAG conformance approach based on model-driven development and WebML. In Proceedings of the 28th ACM International Conference on Design of Communication (São Carlos, São Paulo,...
2010
-
[40]
Ivan Watkins, Bill Kules, Xiaojun Yuan, and Bo Xie. 2014. Heuristic Evaluation of Healthy Eating Apps for Older Adults. Journal of Consumer Health on the Internet 18, 2 (2014), 105–127. doi:10.1080/15398285.2014.902267
2014
-
[41]
Shavindra Wickramathilaka, John Grundy, Kashumi Madampe, and Omar Haggag. 2025. Accessibility Recom- mendations for Designing Better Mobile Application User Interfaces for Seniors. arXiv:2504.12690 [cs.SE] https: //arxiv.org/abs/2504.12690
2025 arXiv
-
[42]
Shavindra Wickramathilaka, John Grundy, Kashumi Madampe, and Omar Haggag. 2025. Adaptive and Accessible User Interfaces for Seniors Through Model-Driven Engineering. Automated Software Engineering (2025). Accepted for publication. Preprint available at https://arxiv.org/abs/2502.18828
2025 arXiv
-
[43]
Shavindra Wickramathilaka and Ingo Mueller. 2023. Addressing Age-Related Accessibility Needs of Senior Users Through Model-Driven Engineering. In 2023 IEEE/ACM 16th International Conference on Cooperative and Human Aspects of Software Engineering (CHASE) . IEEE, Melbourne, Aus...
2023
-
[44]
Enes Yigitbas, Ivan Jovanovikj, Kai Biermeier, Stefan Sauer, and Gregor Engels. 2020. Integrated model-driven development of self-adaptive user interfaces. Software and Systems Modeling 19, 5 (09 2020), 1–25. doi:10.1007/s10270- 020-00777-7
2020 doi
-
[45]
Bendaly Yousra, Lamia Zouhaier, and Leila Ben Ayed. 2018. Model driven approach for adapting user interfaces to the context of accessibility: case of visually impaired users. Journal on Multimodal User Interfaces 13 (11 2018). doi:10.1007/s12193-018-0277-z ACM Trans. Softw. En...
2018 doi
-
[2019]
Ergonomics in Design 27, 4 (2019), 4–13
Mobile Health Apps: Improving Usability for Older Adult Users. Ergonomics in Design 27, 4 (2019), 4–13. doi:10.1177/1064804619840731 arXiv:https://doi.org/10.1177/1064804619840731
2019 doi
Reviewed August 6, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.