Figure 2
The flowchart is arranged in three sections with rectangular boxes connected by downward arrows, starting at the top with two parallel columns labeled “Phase 1: Expert interviews (Hybrid thematic analysis)” on the left and “Phase 2: Policy programs (Directed analysis plus rubric scoring)” on the right. In Phase 1, stacked boxes read “Transcription and familiarisation”, “Thematic coding: Codebook versions 1 and 2”, “Focused and axial coding: Codebook version 3”, “Two-level analysis: (1) Within-sector; (2) Cross-sector (see Results section)”, and “Summarise expert recommendations under each element (see Table 4)”. In Phase 2, stacked boxes read “Eligibility checking and full-text reading”, “Capture objectives and instruments”, “Mapping each policy to the three elements of stepwise retrofitting”, “Codebook version 4”, and “Compute the overall support level of policies for each element (see Table 5)”. The arrow from the last block of phase 1 leads to the fourth box of phase 2. Arrows from the final boxes of both columns converge downward into a central section labeled “Phase 3: Integration and framework development”, which contains vertically arranged boxes reading “Comparing codes generated from experts interviews and policy analysis”, “Thematic coding: Codebook version 5”, “Developing the application framework (see Figure 6)”, and “Contextualising findings against the broader international literature”, indicating a merged analytical workflow. Within each phase, arrows connect each box to the next box.The flowchart of data analysis. Source: Developed by authors