Supplementary Data for "What Constitutes Effective Step-by-Step Assembly Visualizations? A Systematic Scoping Review and Call for More Empirical Research"

The following publications were gathered using a systematic search strategy based on bidirectional snowballing, as proposed by Wohlin (2014). The methodology is described in detail in the publication, but is summarized here for context.

We identified a seed set of papers utilizing repeated queries on Google Scholar, Google Scholar Labs, and Perplexity, combined with our knowledge of key publications in the field. The following publications were selected as the starting point:

• Designing effective step-by-step assembly instructions
• Identification and Validation of Cognitive Design Principles for Automated Generation of Assembly Instructions
• Instruction Generation for Assembly Operations Performed by Humans
• How to Design Assembly Instructions

Starting from this seed set, we performed iterations of backward and forward snowballing. Forward snowballing was conducted by screening citations indexed in Google Scholar. For backward snowballing, we utilized the Google Scholar PDF Viewer plugin to facilitate rapid, ad-hoc screening of reference lists.

Publications were screened and reviewed based on the following inclusion criteria:

1. Must be written in English.
2. Must be a peer-reviewed publication (e.g., conference paper, poster, journal article, or a thesis).
3. Must be published before 2026.
4. Must explicitly investigate, propose, discuss, or evaluate specific visual design parameters contributing to the effectiveness of step-by-step assembly instruction steps.
5. Must explicitly address adults as the target audience of the instructions.
6. Must focus on instructions for non-immersive media (e.g., printed manuals, screens, tablets), excluding works focused on projector-based approaches, AR, or VR.

Starting from the 4 seed papers, we conducted exactly 2 iterations of backward and forward snowballing. In the first iteration, we scanned 82 references and 614 citations, identifying 19 and 13 relevant publications, respectively. Proceeding with these identified papers in the second iteration, we scanned 1115 references and 1543 citations, resulting in 13 and 19 additional publications. In total, we identified 68 publications that satisfied all initial inclusion criteria, which are listed in full detail in the interactive database below. After applying an additional empirical inclusion criterion, 21 of these publications remained eligible; the final paper then continued snowballing and reports a final corpus of 26 empirical studies. Importantly, the initial search represented by this dataset was not exhaustive or run until no further papers were found. The process was concluded after two bi-directional iterations, as we realized that a significant volume of literature discusses visual design parameters only tangentially or primarily relies on citations of prior work instead of presenting new empirical evidence, leading us to refine the scope towards empirical evaluations.