Supplementary Engagement Analyses

Overview

Three complementary analyses that answer questions engagement metrics alone cannot: (1) does emotionally charged content actually correlate with high engagement, (2) which threads sustain influence long after the original poster leaves, and (3) what non-public data is exposed in the corpus. The core principle: engagement is multi-dimensional -- score counts measure attention, sentiment-engagement correlation measures resonance, thread depth measures influence persistence, and a privacy audit measures risk.

When to Use

• Corpus has both engagement metrics (scores, votes, likes, reactions) AND text content with measurable sentiment
• Threaded discussions with reply chains (comments referencing parents)
• Content export that may contain non-public data (IPs, emails, linked identities)
• Need to distinguish authentic voice from crowd-pleasing content
• Investigating whether high-scoring content is also the most emotionally charged
• Identifying "discursive catalysts" -- posts/comments that spark deep conversation threads