Asosiy kontentga oʻtish
Boshqa

Replication data and code for "Measuring Internet Controls in Central Asia: Network Interference, Legal Frameworks and Digital Sovereignty in Five Post-Soviet States, 2020–2026"

Farxod Xushbakovich ShaymanovTermez State UniversityOlim Jo‘rayevTermez State UniversityMukaddas ZaripovaTermez State UniversityFakhriddin KhollievTermez State UniversityAsliddin NormurodovTermez State University
2026en
Ochiq kirishYashil

Litsenziya: cc-by

ABI

Annotatsiya

Replication data and code for the article "Measuring Internet Controls in Central Asia: Network Interference, Legal Frameworks and Digital Sovereignty in Five Post-Soviet States, 2020–2026". The deposit covers Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan and Uzbekistan, January 2020 – October 2026. All measurement data were retrieved passively from the public OONI aggregation API (web connectivity test and six service tests), with Germany, the Netherlands, the United States and Canada as reference countries. No probes were run and no personal data were collected. The deposit contains: Python scripts for collection, classification of blocking, sensitivity analysis, validation (evidence sample, manual review, Censored Planet comparison, domain-by-network rule), tables and figures; processed monthly aggregates by country, domain and network; the classification of every country–domain–month observation; raw OONI API responses, with SHA-256 hashes in the collection state files; and a manifest of the legal texts consulted. See README.md for the file list and the order in which to run the scripts. Citizen Lab test-list categories are not redistributed; the article uses commit 607fcb7523c3ed7bac9b24594f2b3291db6d005a. Version 2: legal-texts manifest extended (Tajikistan's Law No. 1808 on Countering Terrorism and Law No. 56 on Electric Communications, from the parliament's websites); second-coder protocol added. Data, classification and results are unchanged. Version 3: completed second-coder sheet and agreement results added (the two independent coders agree on 27/27 cases; each agrees with the automatic rule on 23/27, Cohen's kappa = 0.68); co-authors added. Data, classification and results are unchanged. Version 4: revision after an independent pre-submission review. The classification rule now treats every observation with fewer than k valid measurements as sparse, including block pages; Uzbekistan's median monthly share changes from 3.3% to 2.4% (new classification_output.zip; Table S1.3). Turkmen web measurements of October 2021 – July 2022 are excluded because OONI's control measurement failed (raw-measurement check, Table S2.7); control failures were checked in the other countries (Table S2.8). Added: bootstrap intervals and robustness checks, a screen of observations the rule did not accept, service robustness with Dutch and US reference data, measurement identifiers for the evidence sample, verify_hashes.py, reproduce.py, requirements.txt and the literature search strings. New archives: processed_data_part3.zip and raw_ooni_rawchecks_*.zip; code_and_results.zip and classification_output.zip replaced; the other raw archives are unchanged. Licence statement corrected: OONI data and files derived from them (raw, processed and classification archives, tables and figures) are licensed under CC BY-NC-SA 4.0, as OONI's own data licence requires; the code and documentation written by the authors are licensed under CC BY 4.0. Earlier versions stated CC BY 4.0 for the OONI data, which was incorrect.

Hali tarjima qilinmagan

Identifikatorlar

Iqtiboslar va manbalar

0 ta iqtibos0 ta foydalanilgan manba