diff --git a/pages/@JSSOFTWARE-D-25-01931_reviewer.md b/pages/@JSSOFTWARE-D-25-01931_reviewer.md new file mode 100644 index 00000000..456a2eb4 --- /dev/null +++ b/pages/@JSSOFTWARE-D-25-01931_reviewer.md @@ -0,0 +1,91 @@ +tags:: [[#zotero]] +title:: @JSSOFTWARE-D-25-01931_reviewer +item-type:: [[document]] +original-title:: JSSOFTWARE-D-25-01931_reviewer +links:: [Local library](zotero://select/library/items/4AP8DKVV), [Web library](https://www.zotero.org/users/1039502/items/4AP8DKVV) + +- ### Attachments + - [PDF](zotero://select/library/items/3435E4YG) {{zotero-imported-file 3435E4YG, "JSSOFTWARE-D-25-01931_reviewer.pdf"}} +- ### Notes + - # Peer Review Report: Gao et al. (2025) + collapsed:: true + + ### Summary of the Work + + This manuscript presents a Systematic Literature Review (SLR) of **Low-Code Development (LCD)** research published between 2021 and early 2025. The authors analyzed 226 articles to map publication trends, definitions, and target user groups. The core of the work involves a thematic categorization of LCD benefits and challenges—spanning **speed, complexity, costs, versatility, quality, security, and compatibility**—evaluated through an evidence-based lens using **Wohlin’s evidence profile**. Additionally, the study applies a **Socio-Technical System (STS)** model to investigate the social and organizational dimensions of LCD adoption and concludes by identifying four key contextual factors and the evolving roles of AI within the low-code ecosystem. + + ### Recommendation: Minor Revision + + ### Detailed Comments and Suggestions for Improvement + + #### 1. Clarity of Novelty and Positioning + + While the paper provides a much-needed update by including work from 2021-2025, the unique contribution beyond being a "recent list" needs to be more forcefully articulated in the Introduction and Related Work sections. + + - + **Issue:** The current draft mentions existing surveys but does not sufficiently detail their limitations (e.g., lack of evidence-based strength assessment or neglect of social/organizational structures). + + - + **Suggestion:** Explicitly state how this paper overcomes the "flat mapping" of previous works. You should highlight the use of the **Wohlin evidence profile** and the **STS model** earlier as the primary methodological differentiators that provide depth rather than just breadth. + + #### 2. Early Emphasis on Key Findings + + The analysis of empirical evidence yields a significant insight: the most cited advantages (**speed, complexity, and quality**) are backed by strong empirical evidence, whereas **security and compatibility** remain under-supported. + + - + **Issue:** This critical finding, which validates the "viability" mentioned in the title, is currently buried deep in the results. + + - + **Suggestion:** This finding should be stressed early in the paper (Introduction or Abstract) to immediately establish the work's value. It serves as a strong hook by confirming which "promises" of low-code are actually scientifically grounded. (“JSSOFTWARE-D-25-01931_reviewer”) + + #### 3. Defining the Target Audience + + The paper provides diverse insights ranging from technical platform comparisons to organizational governance strategies. + + - + **Issue:** It is not entirely clear who the primary intended reader is—academic researchers, organizational IT managers, or platform providers. + + - + **Suggestion:** Add a dedicated paragraph or sub-section in the Introduction identifying the intended target users. For instance, researchers may benefit from the evidence-gap analysis, while practitioners will find the "four contextual factors" framework most actionable. (“JSSOFTWARE-D-25-01931_reviewer”) + + #### 4. Clarification on Replicability + + The manuscript claims that its findings enable "replication and advancing the field." + + - + **Issue:** It is unclear what specific elements are intended for replication. Is it the systematic review methodology itself, or the application of the identified contextual factors in experimental settings? + + - + **Suggestion:** Clarify this statement. If the authors refer to the **coding framework** or the **evidence assessment methodology**, they should specify how other researchers can use these tools to maintain a "living" review of the rapidly changing LCD landscape. (“JSSOFTWARE-D-25-01931_reviewer”) + + #### 5. Strengthening the "Social Aspects" Discussion + + The application of the STS model (People, Structure, Task, and Technology) is a strong point but could be more critical. + + - + **Issue:** The discussion of **IT governance** and **shadow IT** identifies challenges but could benefit from more specific suggestions on how organizations bridge the gap between "citizen" and "professional" developers. + + - + **Suggestion:** Use the "Relational Mechanism" findings to provide more concrete recommendations on how LCDPs act as a "shared vocabulary" to reduce friction, as this is a highly valuable insight for organizational practitioners. (“JSSOFTWARE-D-25-01931_reviewer”, p. 38) + - # LCD SLR Summary (2021–2025) + collapsed:: true + + This note summarizes Gao et al.’s systematic literature review on low-code development (LCD), covering research published from Jan 2021 to Jan 16, 2025 and analyzing 226 included articles drawn from an initial 1,006 records. The review’s goal was to map publication trends, definitions and target users, context-dependent benefits and challenges, evidence strength, and social/organizational adoption factors. (“JSSOFTWARE-D-25-01931_reviewer”, p. 3) + + Methods — The authors followed Kitchenham and PRISMA guidelines, searched multiple databases with the string ("low code" OR "no code") AND ("development" OR "application" OR "platform" OR "software"), applied inclusion/exclusion and quality filters, validated selection with inter-rater reliability (Cohen’s kappa ≈ 0.83), and coded papers in ATLAS.ti using a pretested codebook; data analysis combined evidence-based profiling and context-based coding. (“JSSOFTWARE-D-25-01931_reviewer”, p. 9) + + Publication trends (RQ1) — LCD research is growing and diverse: many papers are general-purpose but domain-focused work is strong in AI, manufacturing, healthcare, and IoT; five research streams were identified — (1) proposals of new LCD platforms, (2) using existing LCDPs for real applications, (3) methodologies that incorporate LCD, (4) empirical studies of practitioner perspectives, and (5) LCD-related techniques (language design, quality, collaboration, AI support, citizen-development support). Interaction modalities extend beyond GUIs to conversational, voice, mixed reality, etc. (“JSSOFTWARE-D-25-01931_reviewer”, p. 14) + + Definitions and target users (RQ2) — Definitions vary: LCD is described in terms of language/notation (visual/model-driven), development environment and deployment (often cloud/PaaS), and societal/organizational aspects (democratization, productivity). Target users range from citizen developers and end-users to professional developers and domain experts; definitions of “citizen developer” are inconsistent across studies. The lack of uniform definition complicates comparisons but common traits (visual interfaces, minimal hand-coding) recur. (“JSSOFTWARE-D-25-01931_reviewer”, p. 17) + + Benefits/challenges by theme (RQ3) — The review groups findings into seven themes: speed, complexity, costs, versatility, quality, security, and compatibility. Typical benefits: faster development, lower entry barriers (ease-of-use), cost/resource savings, prebuilt functionality, and often acceptable functional quality. Typical challenges: limited customizability, integration and vendor lock-in, testing/maintainability gaps, security/privacy concerns, and platform-specific limitations—many of these depend strongly on project complexity and platform maturity. (“JSSOFTWARE-D-25-01931_reviewer”, p. 20) + + Evidence strength (RQ4) — Evidence is uneven: claims about speed, reduced complexity, and comparable quality have notable empirical support (including real-world and experimental evidence), whereas security, compatibility and some contextual claims rely more on circumstantial or first-party reports. Overall, benefits are reported more frequently than challenges, but contextual/qualitative evidence often underpins the reported limitations. (“JSSOFTWARE-D-25-01931_reviewer”, p. 30) + + Social/organizational aspects (RQ5) — Organizational adoption hinges on motivations (digitalization, prototyping, resource shortages), platform selection, availability of developers/citizen-developers, governance (IT oversight, access control), workflows, role definitions, and knowledge/culture (training, resistance, empowerment). Effective adoption requires governance to avoid shadow IT, clear role division, training, and knowledge management. (“JSSOFTWARE-D-25-01931_reviewer”, p. 34) + + Contextual factors & AI (discussion) — The authors distill four contextual adoption factors: project needs, platform choice, developer background, and governance strategy. AI is a prominent domain: LCDPs are used to build AI apps, integrate AI features, apply LLMs for natural-language interaction and model/code generation, and support data-driven insights. The AI–LCD interplay is highlighted as a major growth area. (“JSSOFTWARE-D-25-01931_reviewer”, p. 41) + + Implications & recommendations — Researchers should specify platform types and contexts, and strengthen empirical evidence for understudied risks (security, compatibility). Practitioners should evaluate fit against the four contextual factors, choose mature platforms, plan governance, and provide training. Platform vendors are encouraged to embed governance and transparency (versioning, access control, monitoring) to reduce lock-in and adoption risk. (“JSSOFTWARE-D-25-01931_reviewer”, p. 43) + + Limitations & conclusion — The review acknowledges limits (search in two rounds, evolving terminology, selection constraints) but provides a comprehensive, context-aware synthesis showing LCD’s promise for rapid, democratized development while flagging integration, governance, and evidence gaps that need further empirical study. (“JSSOFTWARE-D-25-01931_reviewer”, p. 45) \ No newline at end of file diff --git a/pages/JSSOFTWARE-D-25-01931.md b/pages/JSSOFTWARE-D-25-01931.md index 3113a0c3..a943d7fa 100644 --- a/pages/JSSOFTWARE-D-25-01931.md +++ b/pages/JSSOFTWARE-D-25-01931.md @@ -1,17 +1,17 @@ +collapsed:: true type:: [[REVIEWS]] tags:: -year:: -venue:: -full-title:: +year:: 2026 +venue:: [[JSS]] +full-title:: date-start:: [[17-01-2026]] - 12:24 date-submitted:: external-links:: status:: [[DOING]] deadline-submission:: -file:: +file:: [[@JSSOFTWARE-D-25-01931_reviewer]] parent:: -todoist:: -collapsed:: true +todoist:: https://app.todoist.com/app/task/jssoftware-d-25-01840-6fP8pRMvc4XPq97c - ### [[Comments]] - #.tabular