B
member badge
Belinda Bailey
Enable Displayr’s Text Categorization to automatically propose and/or create Nets (overcodes) and Subnets from AI-discovered themes—so users can move from flat AI themes to a hierarchical codeframe in one click, keeping parity with classic manual coding workflows while retaining AI speed.
For example, this prompt will not produce the subnets:
You’re coding open-ended answers to:
Q11. “Main difference between cola drinkers.”
Please build a clear two-level codeframe:
WHAT TO CREATE
  • 4–8 broad parent groups (“Nets”)
  • 1–6 short child themes (“Subnets”) under each Net (2–5 words, business-friendly)
  • Aim for ~10 Subnets total across all Nets (not strict—prioritize clarity)
HOW TO CODE
  • Discover themes, then group related ones under a single Net.
  • Keep labels clear and non-overlapping; merge near-duplicates.
  • Assign each response to the Subnet(s) that fit (multi-coding is fine). Don’t assign directly to Nets.
  • Fold very tiny themes (<1% of responses) into “Other (low N)”.
  • Keep sentiment words out of labels.
OUTPUT
1) The hierarchy:
NET: <Net name>
- <Subnet A> (n=…, %…)
- <Subnet B> (n=…, %…)
2) A brief note on any merges/renames and why.