Replication Data for: Bunea, Adriana; Sergiu Lipcean; and Christian Rauh (2026) \"Responsive to What? Explaining the Information Quality of Public Comments on Bureaucratic Policymaking Using a Text-as-Data Approach\", forthcoming in Governance

Replication files for Bunea/Lipcean/Rauh (2026) “Responsive to What? Explaining the Information Quality of Public Comments on Bureaucratic Policymaking Using a Text-as-Data Approach”, forthcoming in Governance Guide to the repository: The replication archive is organized as an RStudio project, please start from BuneaLipceanRauh-Governance-Replication.Rproj to have all relative paths working in your environment All additional R packages used are indicated with version number and source close to the top of each script. 0_MainAnalyses_BuneaLipceanRauh_2026_InformationQuality_Governance.Rmd replicates the results reported in the main article text (output in 0_MainAnalyses_BuneaLipceanRauh_2026_InformationQuality_Governance.pdf) 1_Appendix_BuneaLipceanRauh_2026_InformationQuality_Governance.Rmd replicates the supplementary appendices (output in 1_Appendix_BuneaLipceanRauh_2026_InformationQuality_Governance.pdf) X_TextMeasureValidation.R replicates the validation of the text indicators against human codes (Appendix A) X_Simex.R and X-Simex-DV-function.R are needed to replicate the robustness check in Appendix F (note very long run times) Folder Data holds all large data files on the input and intermediary output side Folder Figures holds all plots reported in the main text and the appendices as png files Folder HumanCoderResults holds the human ratings of sampled text snippets used in the validation of the text-as-data measures Folder ValidationPlots holds the visual results of the human validation exercise (generated by X_TextMeasureValidation.R) Folder AdditionalDocumentation holds documentation of prior steps not directly replicated here (a.o. extraction of the text-based information quality measures and construction of the human coder samples) If you are missing anything, please contact Christian Rauh. The corpora on EC consultations and feedback texts is currently being expanded in the CONSULATIONEFFECTS project. If you require access to full text variables, please contact Adriana Bunea.

本数据集为Bunea、Lipcean与Rauh（2026）的论文《响应何为？基于文本数据方法（Text-as-Data Approach）解析官僚决策中公众评论的信息质量》的复刻文件，该论文即将发表于《治理》（Governance）期刊。本仓库使用指南如下：
本复刻归档采用RStudio项目结构搭建，请先打开`BuneaLipceanRauh-Governance-Replication.Rproj`以确保环境中所有相对路径均可正常运行。所有用到的额外R包均会标注版本号与来源，相关信息位于各脚本文件的顶部附近。
`0_MainAnalyses_BuneaLipceanRauh_2026_InformationQuality_Governance.Rmd`用于复刻正文中报告的研究结果，输出文件为`0_MainAnalyses_BuneaLipceanRauh_2026_InformationQuality_Governance.pdf`。
`1_Appendix_BuneaLipceanRauh_2026_InformationQuality_Governance.Rmd`用于复刻补充附录内容，输出文件为`1_Appendix_BuneaLipceanRauh_2026_InformationQuality_Governance.pdf`。
`X_TextMeasureValidation.R`用于复刻基于人工编码的文本指标有效性检验（对应附录A）。
`X_Simex.R`与`X-Simex-DV-function.R`用于复刻附录F中的稳健性检验（请注意：运行耗时极长）。
各文件夹功能说明如下：
`Data`文件夹存放所有输入数据与中间输出的大型数据文件。
`Figures`文件夹存放正文中及附录中所有报告的可视化图表，格式为PNG文件。
`HumanCoderResults`文件夹存放文本数据方法有效性检验中用到的抽样文本片段的人工评级结果。
`ValidationPlots`文件夹存放人工有效性检验的可视化结果（由`X_TextMeasureValidation.R`生成）。
`AdditionalDocumentation`文件夹存放未在此直接复刻的前置步骤文档（包括但不限于：基于文本的信息质量指标提取流程，以及人工编码样本的构建流程）。
若存在任何缺失内容，请联系Christian Rauh。
欧盟委员会（EC）咨询与反馈文本语料库目前正在`CONSULATIONEFFECTS`项目中进行扩充。若您需要获取全文变量的访问权限，请联系Adriana Bunea。