Identity Archetypes

Document Orientation

Everythingist Research Dashboard — draft, May 2026.

This document specifies how the Everythingist Self-Space dashboard operationalizes self-schematic classification (Markus, 1977) and identity archetypes from the self-complexity tradition (Linville, 1985, 1987; Showers, 1992; McConnell, 2011). It is written for two readers:

• The developer implementing Phases 5J/5K/5L, who needs precise computational definitions, threshold values, and decision logic.
• The peer reviewer evaluating the framework's theoretical grounding, who needs to see how each operational choice traces to prior empirical work or is explicitly flagged as a novel proposal.

Sections are numbered for citability. Empirical anchors are flagged with their source. Novel proposals (operational definitions not previously published in this form) are flagged with [NOVEL] and accompanied by rationale.

1. Theoretical Anchors

The empirical literatures the framework rests on, with explicit attribution and the limits of each anchor.

1.1 Self-schemas (Markus, 1977)

Markus (1977) introduced self-schemas as "cognitive generalizations about the self, derived from past experience, that organize and guide the processing of self-related information." A person is schematic on a dimension when that dimension is highly self-descriptive AND highly important to their self-concept. Schematic individuals process schema-relevant information more efficiently, retrieve schema-consistent memories more readily, and demonstrate behavioral consistency with their schematic identities.

Markus's original operationalization used 11-point self-ratings on adjective pairs (e.g., independent/dependent), classifying participants as:

• Schematic on a dimension if they rated themselves at the extreme of the dimension (e.g., 8–11) AND rated that dimension as important to their self-concept (e.g., 8–11).
• Aschematic if their self-rating was in the middle range AND the dimension's importance was moderate or low.
• A separate counter-schematic classification (sometimes called "negative schematic") emerged in subsequent work: low self-descriptiveness on a dimension AND high importance ("this is NOT me, and that matters").

Markus's framework relied on multi-item measurement of both descriptiveness and importance for each schema dimension, with response-latency triangulation in laboratory studies.

1.2 Exercise self-schema and the Physical Activity Self-Definition Model (Kendzierski, 1988; Kendzierski et al., 1998; Kendzierski & Morganstein, 2009)

Kendzierski's work contributed two distinct extensions to the broader literature on identity and physical activity behavior:

(a) Exercise self-schema. Kendzierski (1988) transposed Markus's self-schema framework to the physical activity context, applying Markus's operationalization (multi-item ratings on descriptiveness and importance) to exercise-specific descriptors. This work demonstrated that exercise self-schematicity predicts behavioral engagement and self-regulatory effort, anchoring the broader literature linking exercise identity to behavior.

(b) Physical Activity Self-Definition Model. In separate work (Kendzierski, Furr, & Schiavoni, 1998; Kendzierski & Morganstein, 2009), Kendzierski developed her own model of physical activity self-definition that does not require cognitive-speed verification (the response-latency triangulation central to Markus's lab work). The PA Self-Definition Model uses binary self-categorization ("Do you see yourself as a runner / tennis player / [specific PA identity]?") supplemented by Likert-scale ratings of perceived criteria (e.g., commitment, competence, effort) and behavioral indicators. This is a measurement-feasible model designed for survey contexts where cognitive-speed measurement is impractical.

The Everythingist dashboard's framework is closer to Markus than to the PA Self-Definition Model in its threshold-based classification, but it differs from both in being applied to participant-generated self-aspects rather than researcher-provided dimensions or specific PA categories. [NOVEL framing.]

1.3 Self-complexity (Linville, 1985, 1987)

Linville's self-complexity (SC) construct describes the number of distinct self-aspects a person uses to organize self-knowledge and the degree of differentiation among those aspects. Higher SC has been associated with affective buffering — distress in one domain spills less into overall well-being. SC has been computationally operationalized in multiple ways ($H$ statistic, Sakaki's variants, alternative formulations); the Everythingist dashboard implements several variants for comparison (see the Self-Complexity Measurement Specification).

1.4 Compartmentalization vs. integration (Showers, 1992)

Showers introduced a distinct dimension orthogonal to SC: whether self-aspects are evaluatively compartmentalized (each aspect is uniformly positive or uniformly negative) versus integrated (each aspect contains a mix of positive and negative attributes). Compartmentalization vs. integration interacts with valence in predicting well-being outcomes.

1.5 Multiple selves and identity coherence (McConnell, 2011)

McConnell's Multiple Self-Aspects Framework (MSF) emphasizes that self-aspects vary in their importance, valence, and behavioral implications, and that identity coherence emerges from the network of relationships among aspects rather than any single aspect.

1.6 Single-item operationalization: the necessary compromise

The Everythingist self-map app collects descriptiveness and importance as single-item ratings on a 1–11 scale for each participant-generated self-aspect, plus per-attribute ratings within each aspect. This departs from Markus's multi-item measurement.

2. Single-Snapshot Classification: the Per-Aspect Status

Each self-aspect within a single journey snapshot is classified into one of four mutually exclusive single-snapshot statuses based on the participant's importance and descriptiveness ratings for that aspect.

2.1 Classification thresholds (1–11 scale)

Rationale for the "candidate" suffix on schematic / counter-schematic. A single-snapshot classification is necessary but not sufficient evidence of schematicity in Markus's sense, which entails durable cognitive structure. The "candidate" label honors this — a single-snapshot schematic candidate may reflect transient salience, motivational priming, expectancy effects, social desirability, or genuine schema, and the dashboard does not adjudicate among these from single-snapshot data alone. Aschematic and Unclassified do not carry the "candidate" suffix because they make weaker claims (the absence of high schematicity rather than the presence of it).

2.2 Mapping from app data

The self-map app collects, per aspect:

• importance (1–11)
• descriptiveness (1–11) — note: stored as descriptiveness in schema 1.7.0+; older snapshots may need migration
• certainty (1–11) — used for triangulation only; not part of the classification decision

The classifier reads aspect.importance and aspect.descriptiveness directly and assigns one of the four statuses. Aspects with missing or null ratings (e.g., aspects created but never rated) are assigned 'unclassified' with a sub-flag reason: 'incomplete-ratings'.

2.3 What this classification does NOT do

• It does not assert that the aspect is a "true schema" in Markus's cognitive-structure sense.
• It does not weight by certainty (consistent with the canonical Markus model; certainty is used elsewhere as a convergent indicator).
• It does not account for the number of attributes the aspect has (that is handled in §4 Elaboration).
• It does not distinguish between aspects that emerged at this snapshot vs. persisted from earlier (that is the longitudinal layer in §3).

2.4 Decision: the Markus "2 of 3 descriptors" operationalization

Markus's published operationalization uses a multi-descriptor pattern where schematic classification requires endorsement on ≥2 of 3 trait descriptors within a dimension (e.g., for the independence-dependence dimension, ratings on independent / individualistic / self-reliant must show ≥2 in the schematic range, with importance ratings paralleling the pattern). This multi-item structure is central to Markus's measurement model — it builds in redundancy that single-item ratings cannot provide.

The Everythingist app's data structure does not present 3 parallel descriptors per dimension; instead, it presents one descriptiveness rating and one importance rating per aspect (plus attribute-level data within). This is the single-item measurement limitation acknowledged in §1.6 — the framework cannot strictly implement Markus's 2-of-3 redundancy.

Decision (per author guidance, May 11): Apply Markus's classification thresholds (the rating ranges in §2.1) directly to each aspect's two ratings, treating each aspect as a single "descriptor" for classification purposes. This loses the redundancy Markus's multi-item operationalization provides but maps cleanly to the app's data shape. The compromise is partially offset by the longitudinal layer (§3) — persistence across snapshots functions as a different kind of redundancy, substituting temporal replication for parallel-item replication.

An optional stricter alternative (Interpretation B) would apply the 2-of-3 pattern to each aspect's attributes (treating attributes as parallel descriptors of the aspect). This is reserved for future work; it is not part of the v1 framework. Researchers requiring closer fidelity to Markus's multi-item structure can compute Interpretation B from the same data offline.

3. Longitudinal Status: Persistence and Change

The single-snapshot status (§2) becomes meaningful for schematic inference only across snapshots. The longitudinal layer assigns a longitudinal status to each aspect that appears in at least 2 snapshots of a participant's journey.

3.1 Temporal requirements

• A longitudinal status requires ≥2 snapshots that include the aspect.
• The snapshots must be separated by ≥1 calendar month for the status to support schematic-persistence claims. This is the temporal-persistence threshold the literature broadly supports for distinguishing transient state from durable trait; specific anchor points vary by domain.
• Snapshots separated by <1 month receive single-snapshot status only, with an explicit flag 'longitudinal-status-deferred' (reason: temporal-gap-too-short).

3.2 Longitudinal status vocabulary

The vocabulary below distinguishes psychologically meaningful change patterns. Each status applies only when temporal requirements (§3.1) are met.

3.2.1 Persistent

Aspect was schematic candidate at prior snapshot(s) AND remains schematic candidate at most recent snapshot. The straightforward "still schematic" case. The longer the chain of persistent schematic snapshots, the stronger the persistence evidence; the dashboard reports persistenceChainLength for each persistent classification.

3.2.2 Emerging

Aspect was non-schematic (aschematic or unclassified) at prior snapshot AND is schematic candidate at most recent snapshot. Comes with cautionary framing that this transition may reflect any of: transient salience priming, motivational reactivity, expectancy effects, social desirability, OR genuine schema emergence. The dashboard does not adjudicate.

Sub-flag: emergingFrom: 'aschematic' | 'unclassified' | 'counter-schematic'. The 'counter-schematic' case is a polar reversal and is additionally flagged as Reversing (§3.2.4); this is rare and warrants strong cautionary framing.

3.2.3 Retreating

Aspect was schematic candidate at prior snapshot AND is now aschematic or unclassified at most recent snapshot. Quantitative loss of schematic status without polar reversal. Distinguish from §3.2.4 below.

3.2.4 Reversing

Aspect transitions across the schematic/counter-schematic axis between snapshots: schematic candidate → counter-schematic candidate, OR counter-schematic candidate → schematic candidate. Polar reversal. Always red-flagged. Psychologically distinct from Retreating: the participant is not merely no longer schematic, they have flipped to active oppositional repositioning (or its reverse).

Note: schematic → aschematic is Retreating (§3.2.3), not Reversing. The Reversing label is reserved for the polar transition specifically.

3.2.5 Oscillating

Aspect alternates between schematic and non-schematic candidate states across 3+ snapshots. Suggests instability of the construct in this participant. Cannot be detected from only 2 snapshots; the framework reports oscillation status only when ≥3 snapshots are present.

3.2.6 Stable non-schematic

Aspect is aschematic OR unclassified at multiple snapshots with no schematic phase between them. The "this isn't a core identity for this participant, consistently" case. Useful as a contrast condition; not all aspects need to be schematic.

3.2.7 Stable counter-schematic

Aspect is counter-schematic at multiple snapshots. The "consistent anti-identity" case ("I am consistently not an X, and this consistently matters to me"). Less common than other patterns but theoretically interesting.

3.3 What persistence does NOT establish on its own

Even with persistent schematic status across multiple snapshots, the dashboard does not assert "this is a self-schema in Markus's sense." Persistence in self-report ratings is one line of evidence; the convergent indicators framework (§4) provides additional lines. Researchers determine when accumulated evidence supports schematic inference for their specific question; the dashboard reports the evidence.

4. Convergent Indicators: Lines of Evidence Beyond Ratings

The persistence framework (§3) addresses temporal stability of single-snapshot ratings. The convergent indicators framework provides four additional lines of evidence that, when triangulated with persistent ratings, strengthen the case for a true schema in the cognitive-structural sense.

4.1 The four indicators

4.1.1 Markus ratings (required)

The single-snapshot classification (§2) at the most recent snapshot. Required: a schematic classification cannot be reported without meeting the Markus rating thresholds, regardless of other indicators. This anchors the framework in the original construct.

4.1.2 Salience prominence (researcher-selectable)

Where in the participant's voluntary self-listing did this aspect appear? An aspect that the participant listed first or near-first when generating their self-aspects shows higher salience than one listed last. Operationalization: salienceRank (1 = listed first; integer rank thereafter). Threshold for "high salience prominence" is researcher-selectable; suggested default: top quartile of the participant's listed aspects. (See the Salience Measurement Specification for the full salience-stream methodology.)

4.1.3 Dwell pattern (researcher-selectable)

Behavioral evidence of cognitive elaboration. Operationalized from three dwell-derived metrics (per Phase 5K; see §5):

• High active dwell on this aspect (top quartile of within-participant aspect dwells).
• Low naming latency (the participant settled quickly on the label — $committedAt - createdAt$).
• Moderate-to-high revision count (the participant returned to refine the aspect over the session).

These can move in opposite directions in interesting ways: high dwell + low naming latency suggests confident elaboration, whereas high dwell + high naming latency + high revision count suggests deliberation. Both patterns can support schematicity but indicate different psychological signatures. Threshold for "supportive dwell pattern" is researcher-selectable.

4.1.4 Integration (researcher-selectable)

A hybrid construct combining within-aspect elaboration and between-aspect embeddedness — full operationalization in §4.2.

4.2 Integration as a hybrid construct [NOVEL operationalization]

The "integration" construct in self-aspect literature has been operationalized in multiple ways that emphasize different facets. The Everythingist framework computes and reports two sub-scores plus a composite.

4.2.1 Elaboration (within-aspect)

How deeply has the participant articulated this aspect? Operationalized as the count of attributes attached to this aspect whose importance ≥ 8 OR descriptiveness ≥ 8, normalized by the maximum attribute count across the participant's aspects.

where $\mathrm{strong\_attr\_count}$ = number of attributes on aspect $i$ with importance ≥ 8 OR descriptiveness ≥ 8, and $\mathrm{max\_attr\_count}$ = maximum attribute count on any aspect in the participant's journey. Yields a 0–1 score. Higher = more elaborated.

Theoretical grounding. Schematic identities are characterized by detailed cognitive elaboration in Markus's original framework. Aspects with many richly-rated attributes show this elaboration directly.

4.2.2 Embeddedness (between-aspect)

How connected is this aspect to other aspects in the participant's self-network? Operationalized as a blend of two computable signals:

1. Attribute overlap (Jaccard): pairwise Jaccard similarity between this aspect's attribute set and the attribute sets of other aspects, averaged. Aspects that share attributes with many other aspects are more embedded.
2. Identity-network centrality: treat aspects as nodes and shared-attributes as edges in a co-occurrence network. Compute the aspect's eigenvector centrality (or degree centrality if eigenvector is unstable for small $N$).

Yields a 0–1 score. Higher = more embedded.

Theoretical grounding. Identity coherence emerges from interconnections among self-aspects (McConnell, 2011); embedded identities show structural integration into the broader self-system.

4.2.3 Integration composite

The geometric mean of Elaboration and Embeddedness:

The geometric mean penalizes lopsidedness: an aspect strong on one dimension but weak on the other receives a lower composite than one moderate on both. This captures the theoretical requirement that genuine integration involves both elaboration AND embeddedness.

Rationale for geometric over arithmetic mean. An aspect with $\mathrm{elaboration} = 1.0$ but $\mathrm{embeddedness} = 0.1$ (deeply elaborated but isolated — the compartmentalized identity profile) should not be labeled "highly integrated"; the geometric mean correctly assigns it a composite of ~0.32 rather than the arithmetic mean's 0.55.

4.3 Composite framing for schematic inference

The framework's epistemic stance: the dashboard reports all four indicators (Markus ratings, salience prominence, dwell pattern, integration composite) plus directionality and timing flags. Researchers decide how many converging indicators are required for schematic inference for their specific analytical question.

The dashboard surfaces a convergentSupportCount (0–3, since Markus is required and the other three are researcher-selectable supplements). It does NOT impose a threshold; that choice is the researcher's.

Suggested defaults for researcher orientation (not enforced):

• Strong schematic evidence: Markus required + 3 of 3 supplementary indicators converge.
• Moderate schematic evidence: Markus required + 2 of 3 supplementary indicators converge.
• Weak / single-source evidence: Markus required alone, no supplementary support.

5. Dwell-Derived Metrics and Their Role

Phase 5K (dwell visualizations) implements the visualization layer; this section specifies the underlying computations that the visualizations expose. The metrics below are derived from the per-aspect dwell fields in schema 1.7.0+ and the per-session sessionIntegrity block.

5.1 Per-aspect dwell metrics

5.1.1 Active dwell

Sum of milliseconds the participant was actively interacting with the dashboard while this aspect was the focal selection. From aspect.dwell.activeDwellMs. Captures cognitive engagement specifically attributable to this aspect.

5.1.2 Naming latency

The interval between when the participant created the aspect row (uncommitted, dwell.createdAt) and when they finalized its label (dwell.committedAt). Reflects deliberation effort in identifying and labeling this identity dimension.

Low naming latency on a schematic candidate suggests automaticity — the participant knew immediately what to call this aspect. High naming latency suggests deliberation — the participant had to think about how to describe this identity. Both can be consistent with schematicity but indicate different psychological signatures.

5.1.3 Revision count

Number of post-commit edits to the aspect's name or ratings, from aspect.dwell.revisionCount. Reflects iterative refinement of the aspect's articulation.

Moderate revision count on a schematic candidate (e.g., 2–4 revisions) suggests careful articulation; very low (0) suggests automaticity; very high (10+) may suggest instability or uncertainty.

5.2 Per-session metrics

From the sessionIntegrity block (schema 1.7.0+):

• Total elapsed time, active dwell, idle duration — session-level timing.
• Idle warnings triggered at 30-min / 2-hr / 12-hr thresholds.
• Session cleanliness enum: 'clean' | 'review-confirmed' | 'review-pruned' | 'contaminated' | 'unknown' — derived from the participant's session integrity behavior.

These are reported descriptively at the per-participant and per-cohort levels (Phase 5K) but are not part of the schematic classification decision logic.

5.3 Current dashboard behavior (carry-forward from v0.19.x)

• Per-cohort aggregate mean active dwell per aspect (Metrics → Cohort scope, period-table at line 17389 in v0.19.3).
• Session-level activeDwellMs in supplemental CSV export.
• Tooltip on the integrity filter showing total active dwell per participant.

5.4 What Phase 5K adds (new visualization scope)

• Individual dwell-per-aspect visualization (intra-individual).
• Naming-latency view.
• Revision-count visualization.
• Dwell as a primary metric in CohortComparisonRenderer (cross-cohort inferential statistics).
• Session-cleanliness composition breakdown.
• (Stretch) Dwell × importance and dwell × certainty scatter plots.

6. Identity Archetypes: Macro-Level Profiles

Beyond per-aspect classification, the framework identifies six identity archetypes that describe macro-level patterns across all of a participant's aspects within a journey snapshot. Archetypes are computed from the constellation of aspect-level classifications and convergent indicators.

6.7 Archetype assignment: exclusivity and confidence

Each participant is assigned one primary archetype and optionally one secondary archetype if the journey is borderline. Both come with a confidence score.

The classifier evaluates the journey against all six archetype criterion sets and returns:

• primaryArchetype: <archetype label>
• primaryConfidence: 0–1 (proportion of criteria fully met for the primary archetype)
• secondaryArchetype: <archetype label> | null
• secondaryConfidence: 0–1

Confidence interpretation: $> 0.85$ = strong assignment; 0.60–0.85 = moderate; 0.40–0.60 = borderline (researcher attention warranted); $< 0.40$ = 'unclassified' archetype.

6.8 What archetypes do NOT do

• They do not label the participant (people are not "Specialists"; their journey at a snapshot has a Specialist pattern).
• They do not assert stability — the same participant may have a Generalist journey at $T_1$ and a Multi-Anchored journey at $T_2$; this transition is informative and is captured separately as Transitioning.
• They do not carry evaluative weight; no archetype is "better" or "more developed" than another.

6.9 Archetype detection vs. archetype generation

The same operational criteria serve dual purposes:

• Detection (Phase 5L): given a real participant's journey, the classifier evaluates against the criteria and assigns archetype + confidence.
• Generation (Phase 5J): given a target archetype, the procedural cohort generator produces fixtures whose data meet that archetype's criteria.

This dual purpose ensures the generator and detector are tightly coupled: any participant generated by the generator should pass the detector, and the detector's behavior on real data can be validated against the generator's known-truth fixtures.

7. Identity-Content Vocabulary

The procedural generator (Phase 5J) draws aspect names from a controlled vocabulary spanning the identity categories specified by Mullen in the design conversation. This is the content layer — orthogonal to the archetype (structural) layer.

7.1 Categories and target prevalence

7.2 Generator behavior

The generator samples aspect names from this vocabulary with archetype-aware weighting:

• Generalist journeys draw aspects across many categories with roughly even distribution.
• Specialist journeys draw 1 strongly-anchored aspect from a primary category (sampled with researcher-relevant prevalence) plus peripheral aspects from elsewhere.
• Multi-Anchored journeys draw 2–4 anchoring aspects from different categories with realistic pair-correlations (e.g., work + family is more common than work + hobby).
• Transitioning journeys show category shifts between snapshots (e.g., the "retiree" category appears in the most recent snapshot but was absent at $T_1$).
• Compartmentalized and Ambivalent journeys have category-neutral content distribution; the archetype shows in valence and elaboration patterns.

7.3 Names within categories

The generator uses neutral-sounding generated names (drawn from a name pool) for fictional participants, NOT names appropriated from any specific demographic group. This avoids inadvertent demographic-pattern signaling in the demo data.

8. Architecture: Dashboard Responsibilities and Researcher Responsibilities

This section makes the epistemic stance explicit.

8.1 What the dashboard computes and reports

• Single-snapshot status for each aspect (§2).
• Longitudinal status for each aspect with ≥2 snapshots (§3).
• All four convergent indicators per aspect (§4).
• Archetype assignment with confidence (§6).
• Cautionary flags for acute change, polar reversal, post-intervention timing concerns.

8.2 What the dashboard explicitly does NOT do

• Adjudicate whether a schematic candidate is a "true schema" — this is a researcher judgment based on accumulated evidence.
• Set the threshold for required convergent indicators — this is a researcher choice for their specific question.
• Interpret an Emerging classification as schema emergence — the dashboard reports the transition; researchers interpret in study context.
• Make claims about identity authenticity, validity, or pathology — the framework is descriptive, not evaluative.

8.3 What the researcher provides

• Threshold choices for convergent-indicator counts.
• Domain expertise for interpretation.
• Awareness of intervention timing and contextual factors that the dashboard does not see.
• Decisions about when accumulated evidence supports schematic inference.

8.4 Versioning and reproducibility commitments

For methodological reproducibility (especially for the planned Social and Personality Psychology Compass manuscript):

• The operational thresholds in §2 are stable across dashboard versions unless an explicit major version increment occurs.
• Any change to thresholds is documented as a methodological revision with version stamp.
• Researchers reporting results from a specific dashboard version cite the version stamp; subsequent reanalysis with later versions should be explicit about which version's operationalizations apply.
• The procedural generator's deterministic seeds are documented so demo fixtures are exactly reproducible across runs and versions.

9. Cautionary Framing and Methodological Limitations

Explicit limitations researchers should hold in mind when interpreting dashboard outputs.

10. Build-Phase Implementation Map

This section maps the specification to concrete dashboard build phases.

10.1 Phase 5J — Procedural cohort generator + archetype criteria

In scope:

• IdentityArchetypeCriteria module: pure functions implementing the operational definitions in §2 (single-snapshot), §3 (longitudinal), §6 (archetypes). Each function takes a journey or aspect and returns a structured classification result.
• SyntheticFixtures module rewrite: procedural generator producing 6 archetypes × 20 participants = 120 fixtures, schema 1.7.0+, with realistic dwell distributions per archetype.
• Each fixture validated against its target archetype criteria (deterministic — if the generator says "this fixture is a Generalist", the criteria function confirms it).
• Cohort Comparison guardrail: default ≤4 active cohorts in any comparison; researcher can opt more in via existing UI.
• Help modal explaining demo cohort design.

Out of scope (deferred to 5L):

• Running the criteria against real participant data (the detector role).
• Visualization surfaces showing archetype labels on the UI.
• Per-aspect schematic status visualization.

10.2 Phase 5K — Dwell visualizations

In scope (as planned, with this spec as anchor):

• Individual dwell-per-aspect visualization.
• Naming-latency view.
• Revision-count visualization.
• Dwell as primary metric in CohortComparisonRenderer.
• Session-cleanliness composition breakdown.
• Per-participant QA descriptive statistics table at import time.

Connected to schematic framework: the dwell visualizations expose the data underlying §4.1.3 (dwell pattern as convergent indicator); they do not yet drive classification.

10.3 Phase 5L — Identity Archetype Detector + Schematic Classifier

In scope:

• Detector module: runs IdentityArchetypeCriteria (from 5J) against real participant journeys.
• Schematic classifier: single-snapshot and longitudinal status assignment with full reporting.
• Per-aspect schematic status visualization on Identity Space, Metrics, and Individual Report surfaces.
• Cohort-level archetype distribution visualization.
• Convergent indicator reporting per aspect.

10.4 Phase 5M — Methodological documentation surface

In scope:

• In-dashboard "Research Design" tab or modal containing this specification's user-facing summary.
• Citation panel for the empirical anchors.
• Version stamp visibility for reproducibility.
• Researcher-configurable threshold panel (for the supplementary convergent indicators in §4.3).

10.5 Phase 5N — Methods paper validation work

Out of dashboard scope; in research-program scope. Running the framework against existing lab data; preparing the SPPC manuscript.

11. Open Methodological Questions (for redline)

The following questions are deferred for Sean's review and resolution before code commitment.

1. Aschematic threshold precision (§2.1). Current threshold is descriptiveness 5–7 AND importance 1–7. Is this Markus-canonical or a working approximation? (Author noted on May 11 that the precise published aschematic threshold should be confirmed.)
2. Persistence temporal threshold (§3.1). 1-month gap proposed; is this literature-anchored or a working default? Alternative anchors: 2 weeks (some intervention literatures), 3 months (most longitudinal personality literatures).
3. Salience prominence threshold (§4.1.2). "Top quartile" suggested as default; literature anchor? Alternative: top tercile, or rank position 1–3 specifically.
4. Dwell pattern thresholds (§4.1.3). Currently "researcher-selectable"; should the dashboard suggest concrete defaults for one-click researcher onboarding?
5. Embeddedness centrality choice (§4.2.2). Eigenvector centrality vs. degree centrality vs. betweenness — which is theoretically most defensible? Computational stability with small aspect counts ($N=3$–5) is a concern.
6. Compartmentalization vs. Ambivalence boundary (§6.5–§6.6). The operational boundary between Compartmentalized and Ambivalent is currently set at within-aspect valence SD (0.5 vs. 0.7). Are these thresholds defensible or working approximations?
7. Transitioning sub-pattern detection (§6.4). Consolidation and Fragmentation require detecting "aspects merging" or "aspects splitting" across snapshots. The operational logic for this is non-trivial (attribute-set overlap thresholds, name similarity, etc.) and is not fully specified in this draft.
8. Counter-schematic in voluntary self-listing (§9.2). A philosophical question — does it make sense to classify an aspect as counter-schematic when the participant nonetheless chose to list it? Or is voluntary listing itself evidence against counter-schematicity? This may require explicit discussion in the manuscript.

Companion Specifications

Related framework documents that extend or interface with this specification.

How to Cite

This is a draft, living specification. Always include the version number when citing.

Archetypes specification

@misc{mullen2026archetypespec,
  author       = {Mullen, Sean P.},
  title        = {Identity Archetypes \& Self-Schematic Status: An Operational Specification},
  year         = {2026},
  version      = {0.1.1},
  note         = {Draft},
  publisher    = {Self-Complexity Research Network},
  url          = {https://selfcomplexityresearch.org/docs/identity-archetypes.html}
}

Forthcoming manuscript

Citation note

This specification builds on:

• Markus (1977) — self-schemas and the schematic / aschematic / counter-schematic typology
• Kendzierski (1988) — exercise self-schema as transposition of Markus to physical activity
• Kendzierski, Furr, & Schiavoni (1998); Kendzierski & Morganstein (2009) — Physical Activity Self-Definition Model (distinguished from self-schema; see §1.2)
• Linville (1985, 1987) — self-complexity
• Showers (1992) — compartmentalization vs. integration of valenced self-knowledge
• McConnell (2011) — Multiple Self-Aspects Framework

Novel operational definitions in this specification (§4.2 integration composite; §6.3 Multi-Anchored parametric formulation; §6.4 Transitioning sub-patterns; §6.6 Ambivalent as distinct archetype) developed by Sean P. Mullen, PhD, University of Illinois Urbana-Champaign.

References

1. Kendzierski, D. (1988). Self-schemata and exercise. Basic and Applied Social Psychology, 9(1), 45–59. https://doi.org/10.1207/s15324834basp0901_4
2. Kendzierski, D., Furr, R. M., Jr., & Schiavoni, J. (1998). Physical activity self-definitions: Correlates and perceived criteria. Journal of Sport & Exercise Psychology, 20(2), 176–193. https://doi.org/10.1123/jsep.20.2.176
3. Kendzierski, D., & Morganstein, M. S. (2009). Test, revision, and cross-validation of the physical activity self-definition model. Journal of Sport & Exercise Psychology, 31(4), 484–504. https://doi.org/10.1123/jsep.31.4.484
4. Linville, P. W. (1985). Self-complexity and affective extremity: Don't put all of your eggs in one cognitive basket. Social Cognition, 3(1), 94–120. https://doi.org/10.1521/soco.1985.3.1.94
5. Linville, P. W. (1987). Self-complexity as a cognitive buffer against stress-related illness and depression. Journal of Personality and Social Psychology, 52(4), 663–676. https://doi.org/10.1037/0022-3514.52.4.663
6. Markus, H. (1977). Self-schemata and processing information about the self. Journal of Personality and Social Psychology, 35(2), 63–78. https://doi.org/10.1037/0022-3514.35.2.63
7. McConnell, A. R. (2011). The multiple self-aspects framework: Self-concept representation and its implications. Personality and Social Psychology Review, 15(1), 3–27. https://doi.org/10.1177/1088868310371101
8. Showers, C. J. (1992). Compartmentalization of positive and negative self-knowledge: Keeping bad apples out of the bunch. Journal of Personality and Social Psychology, 62(6), 1036–1049. https://doi.org/10.1037/0022-3514.62.6.1036