LLM-Assisted Thematic Analysis Protocol (LATA)
Phase 3 — Concept Abstraction (v1.0)

1. Purpose

This phase aims to construct a parsimonious conceptual framework by organizing the validated macroclusters into a small number of higher-order conceptual dimensions.

Unlike the previous phase, which identified broader research domains, the objective here is to reveal the overarching conceptual perspectives through which the scientific literature collectively conceptualizes the phenomenon represented by the bibliographic corpus. The resulting dimensions constitute the highest level of thematic abstraction and provide an interpretable synthesis of the intellectual organization of the research field.

2. Input

The input consists of the validated inventory of macroclusters produced during Phase 2 — Concept Integration.

Each macrocluster should include:

Descriptive label
Conceptual definition
Constituent latent topics
Rationale supporting its conceptual coherence

The analysis should be conducted exclusively at the macrocluster level. Individual latent topics should only be revisited when necessary to clarify conceptual boundaries.

3. Analytical Principles

This phase follows five methodological principles.

a) Conceptual orientation. Conceptual dimensions should emerge from shared interpretative orientations rather than thematic proximity alone. Macroclusters may investigate different research domains while contributing to a common understanding of the overall phenomenon.

b) Explanatory value. Each conceptual dimension should provide explanatory value beyond the thematic information already contained within the individual macroclusters. Dimensions should explain the intellectual organization of the research field rather than merely summarize its thematic structure.

c) Parsimony. The conceptual framework should contain the minimum number of dimensions necessary to represent the thematic landscape without obscuring meaningful conceptual distinctions.

d) Conceptual coherence. Each dimension should represent a well-defined interpretative perspective with strong internal consistency and clear conceptual boundaries.

e) Interpretative transparency. Where multiple higher-level organizations appear equally plausible, the rationale supporting the selected conceptual framework should be explicitly documented.

4. Operational Instructions

Conduct a higher-order conceptual abstraction of the validated macroclusters identified during Concept Integration.

The analysis should proceed through four sequential stages.

Stage 1 — Global Conceptual Review

Review the complete macrocluster structure to develop a comprehensive understanding of the thematic landscape before proposing higher-order conceptual dimensions.

Focus on the relationships among macroclusters rather than on their individual thematic content.

Stage 2 — Identification of Conceptual Orientations

Identify overarching conceptual orientations represented across the macroclusters.

Evaluate whether different macroclusters collectively contribute to understanding:

the central phenomenon itself;
related political or social processes;
broader societal consequences;
other higher-level perspectives emerging inductively from the thematic structure.

Allow these orientations to emerge from the macroclusters rather than imposing predefined conceptual frameworks.

Stage 3 — Construction of Conceptual Dimensions

Construct a parsimonious conceptual framework.

Each conceptual dimension should:

represent a distinct interpretative perspective;
integrate conceptually compatible macroclusters;
remain clearly distinguishable from the remaining dimensions;
contribute meaningful explanatory value;
improve the overall interpretability of the thematic framework.

Avoid dimensions that simply rename or aggregate existing macroclusters without increasing conceptual abstraction.

Stage 4 — Internal Critical Review

Critically evaluate the proposed conceptual framework.

Review:

the explanatory value of each dimension;
the conceptual coherence within dimensions;
the conceptual separation between dimensions;
macroclusters with potentially ambiguous assignments;
alternative higher-level organizations that may also be conceptually plausible.

Where appropriate, refine the framework while preserving parsimony, interpretability, and conceptual consistency.

5. Expected Output

The output of this phase should include:

a structured inventory of conceptual dimensions;
a descriptive label for each dimension;
a conceptual definition of each dimension;
the macroclusters assigned to each dimension;
a rationale explaining the conceptual orientation supporting each dimension;
notes regarding conceptually ambiguous assignments, where appropriate;
an overall interpretation describing how the scientific literature has collectively conceptualized the phenomenon represented by the corpus.

The objective is to produce an explanatory conceptual framework that synthesizes the intellectual organization of the research field.

6. Quality Assurance Checks

6.1. Analytical calibration

Before accepting the conceptual framework, manually review the assignment of macroclusters to conceptual dimensions, giving particular attention to macroclusters whose interpretative orientation may plausibly support more than one higher-order perspective.

Evaluate whether each conceptual dimension contributes genuine explanatory value beyond thematic aggregation and whether the overall framework provides a coherent interpretation of the intellectual organization of the corpus. Where conceptual inconsistencies or alternative higher-level organizations are identified, document the rationale, refine the abstraction criteria, and repeat the synthesis until the framework achieves satisfactory explanatory coherence and parsimony.

6.2. Phase-specific checks

Before completing this phase, verify that:

every macrocluster has been assigned to a conceptual dimension;
conceptual dimensions represent higher-order interpretative perspectives rather than broad thematic categories;
each dimension contributes genuine explanatory value;
dimensions remain conceptually coherent and mutually distinguishable;
conceptual uncertainty has been documented where appropriate;
the resulting framework provides a parsimonious, interpretable, and conceptually robust representation of the research field.

7. Transition to Phase 4

The validated conceptual framework constitutes the input for Phase 4 — Concept Validation.

The objective of the final phase is to evaluate the robustness, transparency, interpretability, and methodological consistency of the complete thematic framework through systematic critical review and expert validation.