Designing Scaffolded Interfaces for Enhanced Learning and Performance in Professional Software

TL;DR

The paper addresses the steep learning curve of professional software by introducing ScaffoldUI, a scaffolded-interface design that surfaces task-relevant tools, progressively discloses complexity, and ties tools to domain concepts to support learning and performance. A Blender-based implementation pipeline leverages LLMs to analyze workflows, map tools, and generate UI code, followed by expert refinement, demonstrating feasibility and transferability to other platforms. Two user studies (beginners N=32 and experts N=8) show that scaffolded interfaces reduce perceived task load, improve workflow clarity and task performance for both groups, and enhance concept learning for beginners, while experts provide nuanced feedback on flexibility and cross-platform applicability. The results offer a foundation for instruction-focused, productivity- and creativity-oriented scaffolded interfaces and highlight avenues for cross-application, cross-platform integration and adaptive, learner-centered design in professional tools.

Abstract

Professional software offers immense power but also presents significant learning challenges. Its complex interfaces, as well as insufficient built-in structured guidance and unfamiliar terminology, often make newcomers struggle with task completion. To address these challenges, we introduce ScaffoldUI, a method for scaffolded interface design to reduce interface complexity, provide structured guidance, and enhance software learnability. The scaffolded interface presents task-relevant tools, progressively discloses tool complexity, and organizes tools based on domain concepts, aiming to assist task performance and software learning. To evaluate the feasibility of our interface design method, we present a technical pipeline for scaffolded interface implementation in professional 3D software, i.e., Blender, and conduct user studies with beginners (N=32) and experts (N=8). Study results demonstrate that our scaffolded interfaces significantly reduce perceived task load caused by interface complexity, support task performance through structured guidance, and augment learning by clearly connecting concepts and tools within the taskflow context. Based on a discussion of the user study findings, we offer insights for future research on designing scaffolded interfaces to support instruction, productivity, creativity, and cross-software workflows.

Figures (6)

• Figure 1: ScaffoldUI implemented as a custom panel in Blender for the UV unwrapping task. Task-relevant tools are grouped into logical sections that align with workflow stages and incorporate domain concepts. Each tool includes a descriptive label and a tooltip explaining its function and how it relates to domain concepts. To help users connect the scaffolded interface with Blender’s native UI, the interface displays keyboard shortcuts and hints about where each tool can be found in Blender’s standard menus or toolbars. The interface manages complexity via user-selectable levels for progressive tool disclosure.
• Figure 2: ScaffoldUI technical implementation pipeline. (1) A user's task informs LLM-assisted workflow analysis to produce workflow stages that incorporate domain concepts. (2) These stages guide LLM-assisted tool selection, functional mapping, and complexity assessment, resulting in a candidate tool list. (3) The tool list is fed into LLM-based code generation, producing Python code and tool labeling for the UI implementation. (4) The UI code undergoes manual refinement, debugging, and integration as an add-on panel in Blender.
• Figure 3: Study 1 participant responses using Ours and Baseline interfaces for Tasks 1 and 2 (lower is better). Asterisks denote significant differences in post-hoc pairwise comparisons between interfaces within each measure ($*$: $p < .05$, $**$: $p < .01$, and $***$: $p < .001$).
• Figure 4: Study 2 participant responses for task performance, concept learning, and user interface experience comparing Ours and Baseline interfaces.
• Figure 5: ScaffoldUI implemented as a custom panel in Blender for the building walk cycle task. Task-relevant tools are grouped into logical sections that align with workflow stages and incorporate domain concepts. Each tool includes a descriptive label and a tooltip explaining its function and how it relates to domain concepts. To help users connect the scaffolded interface with Blender’s native UI, the interface displays keyboard shortcuts and hints about where each tool can be found in Blender’s standard menus or toolbars. The interface manages complexity via user-selectable levels for progressive tool disclosure.