Supplementary data for "Who Hosts the Ritual? Platform-Native Festive Participation and the Spring Festival Gala on Weibo"

Résumé fourni par la source

This deposit contains the data underlying the article "Who Hosts the Ritual? Platform-Native Festive Participation and the Spring Festival Gala on Weibo", prepared for the Special Issue "The Ritual Functioning of Online Media" of Journalism and Media. A link to the published article will be added to this record on publication. The article examines the participatory campaign that Sina Weibo ran under licence around the China Media Group's 2026 Spring Festival Gala (Chunwan), and in particular its horse-raising metric game (「午」福临门,马力全开), which ran on the campaign house account @春晚大赏 from 6 February to noon on 15 February 2026. Contents Contents A single workbook of 15 sheets, together with the coding manual. A README sheet inside the workbook documents every sheet, column and coding decision, and carries the version history of the deposit. Census (455 posts) — every post of the horse-raising module published by the campaign account inside the module window, with displayed and Beijing timestamps, a genre code, hashtags, engagement counts, and a provenance note recording how each post was located. Census correction log — every correction made to the census during the duplicate, arithmetic and live-verification passes, with the original and corrected values. Genre mapping — the consolidation of the 23 first-pass genre labels into the 10 genres of Table 1 of the article. POST-GENRE reliability — the 50-post genre sub-sample with the assignments of all three coders. Comment coding — 546 unique top-level comments drawn from 27 posts, each carrying its codebook v4.1 practice and referent codes, the superseded v3 codes, and an adjudication note wherever a rule was invoked or a machine assignment was overridden by hand. Practice distribution — practice-layer frequencies recomputed from the coding sheet. Reliability Round 1 and Reliability Round 2 — two sub-samples of 125 items each, giving the codes of all three coders side by side so that Krippendorff's α can be recomputed independently. ECHO formulas — the campaign formula set used in the subtraction test that separates echoed liturgy from residual content. Tier A collection — the full collection record for the anchor posts, including each comment's position in the field and the view, algorithmic or chronological, in which it was captured. Post level Tier A and Post level Tier B — post-level attributes for the 33 sampled comment fields: genre, referent, engagement, the silence typology, and the reading of the comment-zone marker taken on 8 September 2026. TOP100 name classification — the pseudonymised classification of the published individual leaderboard by name type, with the rule outcome for each account. Programme-list verification — the check of the campaign's leading celebrities against the broadcast's published programme.Codebook v4.1 (English) — category definitions, inclusion and exclusion criteria, the decision ladder, and worked examples. Coding and reliability First-pass coding was machine-assisted under the written codebook, and the machine pass is entered into the reliability computation as a third coder alongside the two authors, so that human–machine agreement is measured rather than assumed. On the second round the practice layer reaches Krippendorff's α = 0.811 and the celebrity-referent layer α = 0.896. The procedure is described in full in Section 3.8 of the article. Ethics ccount identifiers have been replaced with stable pseudonyms (475 accounts, A001–A476; the published leaderboard separately as T001–T100) and comment texts are not redistributed. Post-level material comes from institutional and public campaign accounts, and is reproduced with its original URLs so that the census can be verified. The ethical framework is set out in Section 3.9 of the article.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.