e2

What it does

Agent E2 acts as an expert in systematic qualitative data coding, guiding researchers through codebook development, applying deductive, inductive, and hybrid coding approaches, and performing thematic analysis, grounded theory, content analysis, and saturation assessment. It provides structured steps, outputs, and examples for building and validating codebooks, conducting six-phase thematic analysis, and implementing grounded theory processes. It also includes considerations for coding quality, IRR, and reporting.

How it works

The skill outlines explicit procedures for multiple methodological pathways:

• Deductive coding: review literature, operationalize codes, pilot test, and assess reliability with outputs like a refined codebook and reliability metrics.
• Inductive coding: open coding, axial/focused coding, selective coding, and development of themes.
• Hybrid approach: combine deductive and emergent codes with tracking of sources and reporting of both a priori and emergent themes.
• Thematic Analysis (Braun & Clarke): six-phase process with familiarization, initial coding, theme searching, reviewing, defining, and reporting, including outputs such as annotated transcripts, a final theme map, and a findings structure.
• Grounded Theory: phases of open, focused, axial, and theoretical coding, with memos for codes, theories, and methodological decisions.
• Content Analysis: manifest and latent content coding with definitions, approaches, examples, and reliability notes across coding units (word to document level).
• Code quality: templates for high-quality code entries, including definitions, when_to_use, when_not_to_use, examples, counter-examples, related codes, and metadata.
• Inter-Rater Reliability (IRR): conditions under which IRR is required or optional, training steps, independent coding, calculation of Cohen’s kappa, percent agreement, and Krippendorff’s alpha, plus discrepancy resolution and reporting formats.

When to use it

Use when conducting rigorous qualitative coding projects that require structured codebooks, transparent coding procedures, and quality assurance across analytical paradigms (deductive, inductive, hybrid, thematic analysis, grounded theory, or content analysis). Trigger points include needing a formalized codebook, demonstrating reliability, or guiding saturation assessments and theoretical model development.

What it can touch

• Tools: CAQDAS and related software mentioned in outputs (e.g., CAQDAS, Excel, Index cards) aspects appear as accepted tooling in procedures.
• Methods: codebook elements, coding phases, thematic maps, memos, and reliability metrics are specified as outputs and processes.

Caveats

The material states detailed methodological steps and outputs but does not guarantee successful results; outcomes depend on data quality and researcher execution. It emphasizes standard metrics (e.g., Cohen’s kappa thresholds, percent agreement, Krippendorff’s alpha) and requires adherence to the described procedures for credibility. License and other metadata indicate NOASSERTION and versioning, with prerequisites and MCP checks defined for execution.