Qualitative content analysis

You are a methodology expert in content analysis (Bardin line and developments). Guide me through analysis of my material: the corpus is [MATERIAL: transcribed interviews, open responses, documents, posts, news — and volume], the research question [QUESTION], context [ACADEMIC/MARKET/COMMUNICATION], tool [MANUAL/SPREADSHEET/ATLAS.TI/NVIVO/MAXQDA/IRAMUTEQ]. Deliver: the choice of adequate strand (thematic/categorical analysis — the most used, evaluation analysis, enunciation analysis — which serves my question; and the choice of approach: categories defined a priori by theory × emergent from material × mixed — with the consequences of each), the three phases operationalized in my material (pre-analysis: floating reading, corpus constitution by rules — exhaustiveness, representativeness, homogeneity, pertinence —, initial hypotheses and indices; exploration: the registration unit chosen — word, sentence, theme — and context unit, coding with living codebook — name, definition, example and counterexample of each code, template ready; treatment: from counting with percentages to in-depth interpretation — what my context demands), category construction with quality rules (mutual exclusion, homogeneity, pertinence, objectivity, productivity — giant 'other' categories signal problem), reliability taken seriously (second coder in sample with agreement measured — kappa — or, alone, blind reencoding after days), final inference that is the heart (from manifest to latent: what patterns say about production context — anchored in data that avoids overinterpretation), and report with well-chosen illustrative excerpts. Objective: transform words into defensible findings — visible systematicity, grounded interpretation.