Coded analytical matrix of Environmental Impact Assessment reports for managed free-roaming cat colonies in protected natural areas of Gran Canaria, Spain (2025)
Description
This dataset contains the complete analytical coding matrix used in the study: Structural bias outweighs ecological evidence in assessing the environmental impacts of managed free-roaming cat colonies in protected natural areas (Manzanares-Fernández et al., 2026, Environmental Science & Policy). The dataset covers the full corpus of 72 Environmental Impact Assessment (EIA) reports produced between June and October 2025 by the consultancy ECOS Estudios Ambientales y Oceanografía SL on behalf of 11 municipalities in Gran Canaria (Canary Islands, Spain), under the mandate of the 2024 Canary Islands Regional Resolution on the management of community cat colonies in protected natural spaces. The reports evaluated registered managed free-roaming cat colonies located within the Canarian Network of Protected Natural Spaces and Natura 2000 sites (Rural Parks, Protected Landscapes, Natural Reserves, and Natural Monuments). Each of the 72 reports was coded independently across 19 variables organised in five analytical dimensions: (1) impact construction (definition mode, evidence type, empirical grounding); (2) risk framing and burden of proof; (3) animal representation and welfare incorporation; (4) procedural standardisation (structural, textual, and measures-level); and (5) assessment quality indicators (internal coherence and normative silences). Coding was applied holistically at the document level following a critical discourse analysis framework. Inter-coder reliability was assessed on a 14% sub-sample; Cohen's kappa values ranged from 0.71 to 0.82. The workbook comprises four sheets. The README sheet provides dataset metadata, variable definitions, and citation information. The Dataset sheet contains the full 72 × 19 coded matrix, with one row per report and anonymised report identifiers (RPT_001–RPT_072). The Codebook sheet documents each variable's data type, description, category labels, and coding notes. The Summary_statistics sheet provides cross-tabulations of all categorical variables by conclusion type (Incompatible, n = 34; Compatible, n = 21; Deferred, n = 17).
Files
Steps to reproduce
The coding matrix was produced through a qualitative-documentary analysis of 72 Environmental Impact Evaluation (EIE) reports, constituting the complete corpus of reports commissioned by 11 municipalities of Gran Canaria (Canary Islands, Spain) between June and October 2025 under the 2024 Canary Islands Regional Resolution on the management of community cat colonies in protected natural spaces. Corpus compilation. Reports were obtained through formal public information requests submitted to each commissioning municipality under Spain's Law 19/2013 on Transparency, Public Access to Information and Good Governance. All reports were produced by a single consultancy (ECOS Estudios Ambientales y Oceanografía SL) and followed a standardised 8-section structure. The corpus represents the complete set of reports produced under the mandate — not a sample — which eliminates selection bias within the case. Coding procedure. Each report was read in full and coded across 19 variables organised in five analytical dimensions: (1) impact construction; (2) risk framing and burden of proof; (3) animal representation and welfare incorporation; (4) procedural standardisation; and (5) assessment quality indicators. Coding was applied holistically at the document level using a critical discourse analysis framework to capture logical relationships between field observations, intermediate reasoning, and final conclusions. Variable definitions, category labels, and decision rules are documented in full in the Codebook sheet of this workbook. Reliability assessment. Inter-coder reliability was assessed on a randomly selected sub-sample of 10 reports (14% of corpus) coded independently by two researchers. Cohen's kappa values exceeded 0.79 for structural variables and ranged from 0.71 to 0.82 for interpretive variables, consistent with accepted thresholds for systematic qualitative content analysis (Krippendorff, 2019). Anonymization. Original internal report reference codes were replaced with sequential anonymous identifiers (RPT_001–RPT_072) to remove residual identifiability. Municipality names are retained as they are public-record administrative units named in the associated publication. No personal data relating to colony caretakers or individual animals are included in this dataset.
Institutions
- Universidad de Las Palmas de Gran CanariaCanary Islands, Las Palmas de Gran Canaria