Archetype Assignment

Archetype Assignment

Overview

Synthesize outputs from multiple prior analysis phases into a persona archetype classification by evaluating weighted evidence across dimensions, computing fit scores per reference archetype, assigning with confidence, and documenting supporting evidence. The core principle: an archetype is a behavioral pattern supported by converging multi-dimensional evidence, not a label derived from any single analysis -- assignment requires evidence from at least three independent analysis dimensions, and every assignment must carry an explicit confidence score and evidence trail.

When to Use

• Multiple prior analyses have been completed (taxonomy/interest classification, sentiment analysis, engagement scoring, temporal patterns, psycholinguistic profile, personality traits, network/social graph, or similar)
• Need to synthesize heterogeneous analysis outputs into a single persona classification
• Assigning user role archetypes (Subject Matter Expert, Lurker-turned-Leader, Community Gatekeeper, Emotional Support Seeker, Digital Nomad/Generalist, or custom archetypes)
• Building a composite user profile from multi-dimensional behavioral evidence
• Need confidence-scored archetype assignment with documented evidence trail
• Evaluating whether a user fits a reference archetype, a hybrid blend, or no established archetype

When NOT to use:

• Fewer than 3 prior analysis dimensions are available (see Insufficient Data Handling)
• Goal is to classify content rather than a user/account
• Attempting to psychologically diagnose or clinically profile an individual
• Assigning archetypes for evaluative, hiring, screening, or gatekeeping purposes
• Only a single analysis dimension is available (use that dimension's own classification instead)

digraph archetype_decision {
    rankdir=TB;
    node [shape=diamond, style=filled, fillcolor="#f0f0f0"];

    dims [label="How many analysis\ndimensions completed?"];
    one [label="1-2 dimensions", shape=box, fillcolor="#ffcccc"];
    three [label="3-4 dimensions", shape=box, fillcolor="#fff3cc"];
    five_plus [label="5+ dimensions", shape=box, fillcolor="#ccffcc"];

    one_action [label="STOP: Insufficient for\narchetype assignment.\nReport per-dimension\nfindings only.", shape=box, fillcolor="#ffcccc"];
    three_action [label="PROCEED WITH CAVEATS.\nAssign archetype but flag\nlow-confidence dimensions.\nHybrid assignment likely.", shape=box, fillcolor="#fff3cc"];
    five_action [label="PROCEED.\nFull archetype assignment\nwith confidence scoring.", shape=box, fillcolor="#ccffcc"];

    dims -> one [label="1-2"];
    dims -> three [label="3-4"];
    dims -> five_plus [label="5+"];
    one -> one_action;
    three -> three_action;
    five_plus -> five_action;
}

Quick Reference: Reference Archetypes

These five reference archetypes represent common behavioral patterns in online communities. They are NOT exhaustive -- users may fit none, one, or a blend of multiple archetypes.

Workflow

Copy this checklist and track progress:

Archetype Assignment Progress:
- [ ] Step 1: Inventory completed prior analyses and assess coverage
- [ ] Step 2: Extract key indicators from each completed analysis
- [ ] Step 3: Build the evidence matrix (archetype x dimension)
- [ ] Step 4: Compute fit scores per reference archetype
- [ ] Step 5: Evaluate hybrid/blended archetype possibilities
- [ ] Step 6: Assign archetype with confidence score
- [ ] Step 7: Document supporting and contradicting evidence
- [ ] Step 8: Assess temporal stability of the assignment
- [ ] Step 9: Write findings to docs/analysis/23-archetype-assignment.md

Step 1: Inventory Completed Prior Analyses

Before synthesizing, catalog which prior analyses are available and their confidence levels. Each analysis is one "dimension" of evidence.

Analysis dimension inventory:

Minimum requirement: 3 dimensions must be completed to proceed. If fewer than 3 are available, STOP and report that archetype assignment cannot be performed with sufficient confidence.

For each available dimension, record:

1. The analysis report location (e.g., docs/analysis/04-taxonomic-interest-classification.md)
2. The confidence level reported by that analysis (high/moderate/low/insufficient)
3. The key indicators that feed into archetype assignment

Step 2: Extract Key Indicators from Each Completed Analysis

For each completed prior analysis, extract the specific indicators relevant to archetype differentiation.

Extraction targets per dimension:

Taxonomy/Interest:

• Classification: Deep Specialist / Specialist / Moderate / Broad / Polymathic
• Interest shape: Peaked / Bimodal / Plateau / Flat
• Orthogonality score (0.0-1.0)
• Primary and secondary interest pillars
• Number of occupied categories

Sentiment:

• Mean compound score
• Sentiment distribution: % positive / negative / neutral
• Trajectory pattern: stabilizing / volatile / inflecting / plateau
• Sentiment skewness
• Tier mismatches (title vs body divergence)

Engagement:

• Composite engagement score (percentile within corpus)
• Component breakdown (which signals drive the score)
• Sensitivity analysis result (robust / moderate / sensitive)
• Temporal engagement trend (increasing / stable / decreasing)

Temporal Patterns:

• Activity regularity score (circadian consistency)
• Engagement style: steady / bursty / weekend-warrior / nocturnal
• Trend direction: increasing / stable / declining
• Burst frequency and characteristics
• Temporal inflection points (if any)

Psycholinguistic Profile:

• Dominant register: analytical-reflective / social-relational / expressive-positive / expressive-negative / achievement-oriented / authority-oriented / sensory-descriptive / process-heavy / content-dense
• Top 3 dimensions by baseline deviation
• Function word percentage (process vs content orientation)
• Summary variable approximations (Analytic, Clout, Authenticity, Tone)

Personality Traits (if available):

• OCEAN scores (normalized 0-100) with confidence bands
• Which traits have confident (non-indeterminate) readings
• Trait conflicts and resolutions

Network Position (if available):

• Degree centrality (how connected)
• Betweenness centrality (how much a bridge)
• Community membership (how many sub-communities)
• Network role classification (hub / bridge / peripheral / isolate)

Step 3: Build the Evidence Matrix

Construct a matrix mapping each reference archetype against each available evidence dimension. For each cell, record whether the evidence SUPPORTS, CONTRADICTS, or is NEUTRAL toward that archetype.

Evidence matrix structure:

                    | SME | Lurker-turned-Leader | Community Gatekeeper | Emotional Support Seeker | Digital Nomad |
--------------------|-----|----------------------|----------------------|--------------------------|---------------|
Taxonomy            | [S/C/N] | [S/C/N]          | [S/C/N]              | [S/C/N]                  | [S/C/N]       |
Sentiment           | [S/C/N] | [S/C/N]          | [S/C/N]              | [S/C/N]                  | [S/C/N]       |
Engagement          | [S/C/N] | [S/C/N]          | [S/C/N]              | [S/C/N]                  | [S/C/N]       |
Temporal            | [S/C/N] | [S/C/N]          | [S/C/N]              | [S/C/N]                  | [S/C/N]       |
Psycholinguistic    | [S/C/N] | [S/C/N]          | [S/C/N]              | [S/C/N]                  | [S/C/N]       |
Personality         | [S/C/N] | [S/C/N]          | [S/C/N]              | [S/C/N]                  | [S/C/N]       |
Network             | [S/C/N] | [S/C/N]          | [S/C/N]              | [S/C/N]                  | [S/C/N]       |

S = Supports (evidence aligns with archetype markers)
C = Contradicts (evidence conflicts with archetype markers)
N = Neutral (evidence neither supports nor contradicts, or dimension not available)

Scoring rules for each cell:

Example evidence evaluation for SME archetype on Taxonomy dimension:

• Taxonomy classification = "Deep Specialist" or "Specialist" --> Strong Support (+2)
• Taxonomy classification = "Moderate" with peaked shape --> Moderate Support (+1)
• Taxonomy classification = "Broad" or "Polymathic" --> Strong Contradiction (-2)

Step 4: Compute Fit Scores per Reference Archetype

For each archetype, compute a weighted fit score from the evidence matrix.

Weighting rules:

Dimensions with higher confidence get more weight. Dimensions that are primary differentiators for a given archetype get more weight than dimensions that are secondary.

Fit_score(archetype) = SUM(evidence_score_i * dimension_weight_i * confidence_weight_i) / SUM(dimension_weight_i * confidence_weight_i)

Dimension weight guidelines:

Confidence weight mapping:

Fit score interpretation:

Step 5: Evaluate Hybrid/Blended Archetype Possibilities

Users rarely fit a single archetype perfectly. Evaluate whether a hybrid classification better represents the evidence.

Hybrid assignment criteria:

digraph hybrid_decision {
    rankdir=TB;
    node [shape=diamond, style=filled, fillcolor="#f0f0f0"];

    gap [label="Gap between top two\nfit scores?"];
    large [label="> 0.25 gap", shape=box, fillcolor="#ccffcc"];
    small [label="<= 0.25 gap", shape=box, fillcolor="#fff3cc"];

    large_action [label="Single archetype assignment.\nClear dominant fit.", shape=box, fillcolor="#ccffcc"];
    small_check [label="Both scores > 0.30?", shape=diamond];
    both_yes [label="Hybrid assignment:\nPrimary + Secondary archetype.\nDocument which dimensions\nsupport each.", shape=box, fillcolor="#fff3cc"];
    both_no [label="Weak overall fit.\nConsider 'Unclassified'\nwith trait description.", shape=box, fillcolor="#ffcccc"];

    gap -> large [label="> 0.25"];
    gap -> small [label="<= 0.25"];
    large -> large_action;
    small -> small_check;
    small_check -> both_yes [label="yes"];
    small_check -> both_no [label="no"];
}

Hybrid classification format:

• Primary archetype: The highest-scoring archetype (if score > 0.30)
• Secondary archetype: The second-highest (if within 0.25 of primary AND score > 0.30)
• Blend description: One sentence describing which traits from each archetype are present
• Example: "SME-Gatekeeper hybrid: deep domain expertise (SME) combined with cross-community bridging behavior (Gatekeeper)"

When NO archetype fits (all scores < 0.30):

• Do NOT force an assignment
• Classify as "Unclassified / Emerging Pattern"
• Describe the user's behavioral profile in terms of the individual dimension findings
• Note which reference archetypes came closest and why they did not reach threshold
• Consider whether the evidence suggests a novel archetype not in the reference set

Step 6: Assign Archetype with Confidence Score

Confidence score computation:

The confidence in the archetype assignment depends on three factors:

1. Fit strength: How high is the top fit score?
2. Evidence coverage: How many dimensions contributed evidence?
3. Evidence agreement: How consistent is the evidence (few contradictions)?

Confidence = (fit_strength_factor + coverage_factor + agreement_factor) / 3

Factor definitions:

Confidence interpretation:

Step 7: Document Supporting and Contradicting Evidence

For the assigned archetype (and secondary archetype if hybrid), document:

For each supporting dimension:

• Which specific indicator(s) support the assignment
• The indicator value and what makes it supportive
• The confidence level of that prior analysis

For each contradicting dimension:

• Which specific indicator(s) conflict with the assignment
• Why the contradiction was overridden (e.g., outweighed by stronger evidence, low confidence on the contradicting dimension)

For neutral dimensions:

• Why the dimension did not contribute (not available, value was ambiguous, dimension is not relevant to this archetype)

This evidence trail is the most important output. An archetype label without its evidence trail is an unsupported assertion.

Step 8: Assess Temporal Stability

If temporal data is available, evaluate whether the archetype assignment would differ across time periods.

Temporal stability checks:

• If temporal analysis showed a clear activity trend (increasing, declining), the archetype may be transitional
• If sentiment trajectory showed inflection points, the affective component of the archetype may have shifted
• If engagement patterns changed over time, the user may have transitioned between archetypes

Stability classification:

Step 9: Write the Report

Write all findings to docs/analysis/23-archetype-assignment.md. See the report template below.

Report Output Template

The final report MUST be written to docs/analysis/23-archetype-assignment.md with this structure:

# Archetype Assignment

## Methodology
- **Approach:** Multi-dimensional evidence synthesis across [N] completed analysis phases
- **Reference archetypes:** Subject Matter Expert (SME), Lurker-turned-Leader, Community Gatekeeper, Emotional Support Seeker, Digital Nomad/Generalist [+ any custom archetypes]
- **Minimum dimensional threshold:** 3 independent analysis dimensions required
- **Scoring method:** Weighted evidence matrix with dimension confidence weighting
- **Date of synthesis:** [date]

## Prior Analysis Inventory

### Completed Analyses
| Dimension | Report Location | Confidence | Key Indicators Extracted |
|-----------|----------------|------------|-------------------------|
| [dimension] | [path] | [high/moderate/low] | [brief summary of extracted indicators] |
| ... | ... | ... | ... |

### Analyses Not Available
| Dimension | Impact on Assignment |
|-----------|---------------------|
| [dimension] | [how absence affects confidence or differentiability] |

## Evidence Matrix

### Raw Evidence Scores
| Dimension | SME | Lurker-turned-Leader | Community Gatekeeper | Emotional Support Seeker | Digital Nomad |
|-----------|-----|----------------------|----------------------|--------------------------|---------------|
| [dim] | [+2/+1/0/-1/-2] | ... | ... | ... | ... |
| ... | ... | ... | ... | ... | ... |

### Evidence Justification
[For each non-zero cell in the matrix, a one-sentence explanation of why that evidence score was assigned]

## Fit Scores

| Archetype | Raw Fit Score | Weighted Fit Score | Interpretation |
|-----------|--------------|-------------------|----------------|
| [archetype] | [score] | [score] | [strong/moderate/weak/poor fit] |
| ... | ... | ... | ... |

## Archetype Assignment

### Primary Assignment
- **Archetype:** [name]
- **Fit score:** [value]
- **Confidence:** [score] ([High/Moderate/Low/Very Low])

### Secondary Assignment (if hybrid)
- **Archetype:** [name]
- **Fit score:** [value]
- **Blend description:** [one-sentence hybrid description]

### Confidence Breakdown
| Factor | Score | Rationale |
|--------|-------|-----------|
| Fit strength | [0-1] | [explanation] |
| Evidence coverage | [0-1] | [N of M dimensions contributed] |
| Evidence agreement | [0-1] | [N contradictions out of M contributing dimensions] |
| **Overall confidence** | **[0-1]** | **[label]** |

## Evidence Trail

### Supporting Evidence
| Dimension | Indicator | Value | Why It Supports |
|-----------|-----------|-------|-----------------|
| [dim] | [indicator] | [value] | [explanation] |

### Contradicting Evidence
| Dimension | Indicator | Value | Why It Contradicts | Resolution |
|-----------|-----------|-------|--------------------|------------|
| [dim] | [indicator] | [value] | [explanation] | [why overridden] |

### Neutral / Non-Contributing Dimensions
| Dimension | Reason |
|-----------|--------|
| [dim] | [not available / ambiguous / not relevant to assigned archetype] |

## Temporal Stability Assessment
- **Stability classification:** [Stable / Mostly stable / Transitional / Unstable]
- **Temporal evidence:** [summary of temporal patterns affecting stability]
- **Trajectory (if transitional):** [from archetype X toward archetype Y, based on evidence Z]

## Profile Summary
[2-3 paragraph narrative synthesizing all findings into a cohesive behavioral portrait. This should read as a description of how the user engages with the community, NOT as a personality assessment or value judgment. Reference specific evidence from the dimensions.]

## Limitations and Caveats
- Archetypes are behavioral pattern descriptions, not personality types or fixed identities
- The reference archetype set is not exhaustive -- users may exhibit patterns not captured by any reference archetype
- Assignment confidence is bounded by the confidence of the input analyses
- [Specific dimensional gaps and their impact]
- [Evidence contradictions and how they were resolved]
- Archetype assignment reflects behavior observed in THIS corpus during THIS time period
- The same user may exhibit different archetype patterns in different communities or at different times
- These results should NOT be used for evaluative judgment, gatekeeping, or any decision that affects the individual

## References
- Ferrara, E., et al. (2025). Characterizing User Archetypes and Discussions on Social Hypernetworks. *MDPI Big Data and Cognitive Computing*, 9(9), 236.
- Akar, E. & Mardikyan, S. (2018). User Roles and Contribution Patterns in Online Communities. *SAGE Open*, 8(3).
- Welser, H.T., et al. (2011). Finding Social Roles in Wikipedia. *iConference*.
- Tu, N. & Dong, X. (2010). Using cluster analysis in Persona development. *IEEE/ACIS International Conference*.
- Stachl, C., et al. (2022). The kernel of truth in text-based personality assessment. *Psychological Bulletin*.

Good Patterns

• Require multi-dimensional evidence -- never assign an archetype from a single analysis dimension
• Weight evidence by confidence -- a high-confidence taxonomy classification matters more than a low-confidence personality inference
• Allow hybrid archetypes -- most users blend traits from multiple archetypes; forcing a single label loses nuance
• Document the evidence trail -- every archetype assignment must trace back to specific indicators from specific analyses
• Report confidence explicitly -- an assignment with 0.35 confidence is fundamentally different from one with 0.85 confidence
• Evaluate temporal stability -- if evidence suggests the user's behavior is changing, a static archetype label is misleading
• Treat archetypes as behavioral descriptions -- they describe what someone DOES in a community, not who they ARE as a person
• Allow "Unclassified" -- forcing every user into a reference archetype when evidence is mixed produces false precision

Anti-Patterns

Boundaries

This skill SHOULD:

• Synthesize outputs from 3+ prior analysis dimensions into a persona archetype
• Evaluate evidence per archetype using the weighted evidence matrix
• Assign archetypes with explicit confidence scores
• Document supporting AND contradicting evidence for every assignment
• Allow hybrid/blended archetype classifications
• Report "Unclassified" when evidence is insufficient or contradictory
• Assess temporal stability of the assignment
• Write findings to docs/analysis/23-archetype-assignment.md

This skill should NOT:

• Treat archetypes as fixed personality types or stable identities
• Force classification when evidence is insufficient (all fit scores below 0.30)
• Use archetypes for judgment, evaluation, hiring, moderation, or gatekeeping
• Assume the five reference archetypes are exhaustive (they are a starting framework)
• Override high-confidence evidence with low-confidence evidence
• Assign archetypes from fewer than 3 completed analysis dimensions
• Present assignments without confidence scores and evidence trails
• Claim that archetype assignment reveals who someone "really is"
• Make predictions about future behavior based on archetype assignment
• Compare archetype assignments across different communities without recalibrating

Insufficient Data Handling

digraph insufficient_data {
    rankdir=TB;
    node [shape=diamond, style=filled, fillcolor="#f0f0f0"];

    start [label="How many analysis\ndimensions completed?"];
    zero_two [label="0-2 dimensions", shape=box, fillcolor="#ffcccc"];
    three_four [label="3-4 dimensions", shape=box, fillcolor="#fff3cc"];
    five_plus [label="5+ dimensions", shape=box, fillcolor="#ccffcc"];

    zero_action [label="STOP: Cannot assign archetype.\nReport available dimension\nfindings individually.\nState minimum not met.", shape=box, fillcolor="#ffcccc"];
    three_action [label="PROCEED WITH CAVEATS.\nAssign archetype but widen\nconfidence bands.\nHybrid/Unclassified more likely.\nDocument missing dimensions.", shape=box, fillcolor="#fff3cc"];
    five_action [label="PROCEED.\nFull archetype assignment\npipeline. All features\navailable.", shape=box, fillcolor="#ccffcc"];

    start -> zero_two [label="0-2"];
    start -> three_four [label="3-4"];
    start -> five_plus [label="5+"];
    zero_two -> zero_action;
    three_four -> three_action;
    five_plus -> five_action;
}

Common Mistakes

References

• Ferrara, E., et al. (2025). Characterizing User Archetypes and Discussions on Social Hypernetworks. MDPI Big Data and Cognitive Computing, 9(9), 236.
• Akar, E. & Mardikyan, S. (2018). User Roles and Contribution Patterns in Online Communities. SAGE Open, 8(3).
• Welser, H.T., et al. (2011). Finding Social Roles in Wikipedia. iConference.
• Tu, N. & Dong, X. (2010). Using cluster analysis in Persona development. IEEE/ACIS.
• Stachl, C., et al. (2022). The kernel of truth in text-based personality assessment. Psychological Bulletin.
• NN/g: Personas vs. Archetypes.
• Multi-dimensional clustering in user profiling (Westminster/CORE).
• Data-driven Personas: Constructing Archetypal Users with Clickstreams and User Telemetry (CHI 2016).
• An, J., et al. (2021). A Survey of 15 Years of Data-Driven Persona Development. International Journal of Human-Computer Interaction.
• Discourse Blog: Understanding Lurkers in Online Communities.