Intercoder Validation Dataset for Strategic Posture Coding of 10 NASDAQ-100 Firms (2024)
Description
This dataset presents an intercoder validation exercise for Ansoff’s Strategic Posture framework, applied to a 2024 sample of ten NASDAQ-100 firms. The purpose is to assess inter-rater reliability in measuring strategic alignment across three coders with distinct evaluation profiles: Coder A (aggressive): assigns higher turbulence (ETL ≈ 4.7) and emphasizes bold ratings for aligned firms, while moderating misaligned cases around 3.5. Coder B (conservative): assigns lower turbulence (ETL ≈ 4.1) and uses a restrained mix of 3s and 4s, with few 5s, reflecting a cautious interpretation of strategic behavior. Coder C (balanced): positioned between A and B, with ETL ≈ 4.6 and moderate ratings across aggressiveness and capability, providing a midpoint perspective. Each coder independently rated 65 items per firm, drawn from the Optimal Strategic Performance Positioning (OSPP) diagnostic: 25 Environmental Turbulence (ETL) items, 20 Strategic Aggressiveness (SA) items, 20 Capability Responsiveness (CR) items. Ratings are provided on a 1–5 Likert scale, anchored to observable firm behaviors and industry conditions. The ten firms include both strategically aligned exemplars (e.g., Apple, Microsoft, Amazon, Tesla, Costco) and misaligned cases (e.g., PayPal, Warner Bros. Discovery, Biogen, Kraft Heinz, Charter Communications). The dataset is organized in a single Excel file with three sheets: CODER_RATINGS – Long format item-level data (1,950 rows = 3 coders × 10 firms × 65 items). SUMMARY – Coder-firm means for ETL, SA, and CR, including alignment gaps (SA–ETL, SA–CR) and a simple posture score (X₁ average). METADATA – Documentation of coder profiles, coding logic, and data dictionary. This dataset enables reproducibility of intercoder reliability metrics (e.g., Krippendorff’s α, ICC), robustness checks of posture measurement, and sensitivity analyses of gap-based alignment scores. It complements the primary study on Ansoff’s Strategic Success Hypothesis by directly addressing methodological concerns about single-coder designs and demonstrating transparency in the coding process. All data are based on publicly available firm disclosures, industry reports, and observable strategic actions for 2024. No confidential or proprietary information was used.
Files
Steps to reproduce
Select firms Ten NASDAQ-100 firms were chosen for 2024, representing five strategically aligned (Apple, Microsoft, Amazon, Tesla, Costco) and five misaligned (PayPal, Warner Bros. Discovery, Biogen, Kraft Heinz, Charter Communications). Define coders and profiles Three independent coders participated, each with a distinct orientation: Coder A (aggressive, ETL ≈ 4.7), Coder B (conservative, ETL ≈ 4.1), Coder C (balanced, ETL ≈ 4.6). Apply diagnostic instrument Coders rated each firm on the Optimal Strategic Performance Positioning (OSPP) diagnostic, covering: 25 Environmental Turbulence (ETL) items, 20 Strategic Aggressiveness (SA) items, 20 Capability Responsiveness (CR) items. Each item was scored on a 1–5 scale, anchored to observable firm behaviors and industry conditions (e.g., frequency of new product launches, R&D intensity, cultural adaptability). Coding procedure Ratings were completed independently, without discussion between coders. ETL was held constant across firms within each coder to reflect year-level environmental turbulence. Notes fields were available for citing evidence, though most cells remain blank in this dataset. Aggregate ratings For each coder × firm × section, item-level scores were averaged to produce ETL, SA, and CR means. Alignment gaps were computed as SA–ETL and SA–CR. A simple posture score (X₁ average = mean of ETL, SA, CR) was calculated for quick comparison. Organize dataset The consolidated Excel file contains: CODER_RATINGS: All item-level ratings (1,950 rows = 3 coders × 10 firms × 65 items). SUMMARY: Section means, alignment gaps, and posture scores by coder–firm. METADATA: Documentation of coder styles, coding framework, and data dictionary. Replication Researchers can replicate intercoder reliability by importing the CODER_RATINGS sheet and computing Krippendorff’s α or ICC across coders for each section (ETL, SA, CR). Sensitivity to coder style can be tested by comparing X₁ scores and alignment gaps across coders.
Institutions
- Universita Ca' Foscari