Supplementary materials 2 - Online survey results

This workbook contains the raw, un‑cleaned responses that were exported directly from the on‑line questionnaire described in the companion “Survey_full.docx”. Each row is one questionnaire attempt and each column reproduces a single survey item or an internal platform field. 1. The survey was run between July and September 2023 by COST Action BETTER and the Federation of Veterinarians of Europe to map ongoing projects and training in livestock biosecurity and herd‑health management. Participants were recruited through professional and academic networks; no personal identifiers were captured by the platform. 2. File structure one worksheet: Raw Data 166 submissions × 672 variables system variables: Response ID, status (Started/Completed), timestamp, IP, duration, duplicate flag substantive variables grouped exactly as in the questionnaire: Introductory self‑assessment sliders & experience tick‑boxes Profession branch (Veterinarian/Farmer/Student/Other) and species‑specific follow‑ups Project inventories (Biosecurity & HHM) – up to 5 free‑text rows each Training inventories (Biosecurity & HHM) – up to 5 free‑text rows each Perceived training‑gap matrices (6‑point Likert or slider) for different stakeholder roles Final open‑comment cell Column headers are the full question texts produced by the export function; multi‑select questions generate one Boolean column per option and appear as “Unnamed: ##” where the export inserted placeholders. Numeric answers use 0–100 for sliders and 1‑6 for Likert grids; text answers are free text. 3. Response volume & completeness 73 rows are marked Completed, 92 Started, 1 missing status. 110 respondents filled the biosecurity knowledge slider. Duplicate flag = 0 for 165 rows. 4. Using the data Cleaning: shorten verbose column names; recode “Unnamed” Booleans into tidy long format; drop platform fields that are not required for analysis (IP, duplicate, etc.) to protect privacy. Filtering: restrict to Response Status = Completed when exact denominators are needed; or keep all rows for item‑level analyses. Missing values: blank cells are genuine non‑responses. Treat string “I do not know” as explicit category, not NA. Aggregation: many training‑gap items are parallel across stakeholder groups—after reshaping they can be compared by profession or species. Confidentiality: IP addresses are not present. 5. Interpretation tips High slider scores do not necessarily translate into long experience—cross‑tab them with the categorical “years of experience” item before drawing competency conclusions. For Likert blocks, note that “1” is labelled “Very big gap” and “6” “No gap”, so lower numbers mean more training need. 6. Translations The questionnaire shown to each participant was available in Albanian, German, Italian, Portuguese, Spanish and Turkish in addition to English; responses from all language versions were pooled into this single file. Coding conventions are identical irrespective of language.本工作簿（workbook）包含直接从配套文档《Survey_full.docx》中所述的线上问卷导出的原始未清洗响应数据。每一行对应一次问卷作答，每一列则复刻了单条问卷题目或平台内部字段。 1. 调研背景 本调研由COST行动BETTER（COST Action BETTER）与欧洲兽医联合会（Federation of Veterinarians of Europe）于2023年7月至9月期间开展，旨在梳理畜禽生物安全与畜群健康管理（herd-health management, HHM）领域的在建项目及培训情况。调研对象通过专业与学术网络招募，平台未采集任何个人身份标识信息。 2. 文件结构 仅含1个工作表：原始数据（Raw Data） 共166条作答记录 × 672个变量 系统变量：作答ID（Response ID）、状态（Started/Completed，即已开始/已完成）、时间戳（timestamp）、IP地址、作答时长、重复标记（duplicate flag） 实质性变量与问卷中的分组完全一致： - 入门自我评估滑块题与经验勾选框 - 职业类别（兽医/养殖户/学生/其他）及物种专项追踪问题 - 项目清单（生物安全与畜群健康管理）—— 每项最多支持5条自由文本行 - 培训清单（生物安全与畜群健康管理）—— 每项最多支持5条自由文本行 - 针对不同利益相关方角色的培训感知差距矩阵（采用6级李克特量表或滑块作答） - 最终开放式评论单元格 列标题为导出功能生成的完整题目文本；多选题会为每个选项生成一列布尔值，当导出系统插入占位符时，该列会显示为"Unnamed: ##"。滑块题的数值作答范围为0–100，李克特量表题的数值作答范围为1–6；文本类题目则为自由文本作答。 3. 作答量与完整性 其中73条记录标记为已完成，92条标记为已开始，1条未标注作答状态。 共有110名受访者填写了生物安全知识滑块题。 165条记录的重复标记值为0。 4. 数据使用规范 数据清洗：简化冗长的列名；将"Unnamed"列的布尔值重构为规范的长格式数据；删除无需用于分析的平台字段（如IP地址、重复标记等）以保护受访者隐私。 数据筛选：若需精确计算分母，可将样本限制为"作答状态=已完成"的记录；若进行题目级分析，则可保留全部行数据。 缺失值处理：空白单元格代表真实的未作答情况。将字符串"我不知道"视为明确的作答类别，而非缺失值（NA）。 数据聚合：多数培训差距题目在不同利益相关方群体中具有平行性—— 经数据重构后，可按职业类别或物种维度进行对比分析。 保密性保障：本数据集未包含任何IP地址信息。 5. 解读提示 滑块得分较高并不必然代表从业年限更长—— 在得出能力相关结论前，需将其与"从业年限"分类题目进行交叉制表分析。对于李克特量表区块，需注意"1"代表"差距极大"，"6"代表"无差距"，因此数值越低代表培训需求越高。 6. 问卷语言说明 本调研面向参与者的问卷除英文外，还提供阿尔巴尼亚语、德语、意大利语、葡萄牙语、西班牙语及土耳其语版本；所有语言版本的作答均合并至本单一数据文件中。无论使用何种语言，编码规则均保持一致。