Table 1 Variable information.
From: The power of protest in the media: examining portrayals of climate activism in UK news
Variable | Definition (example) | Coding |
|---|---|---|
Partisanship | Is the paper politically conservative (Daily Mail), neutral (BBC), or liberal (The Guardian)? | −1 to 1, with −1 representing liberal (e.g. The Guardian) and 1 representing conservative (e.g. The Daily Mail) We selected newspaper partisanship based on the ratings of an Oxford study detailing the endorsements that the party makes in general elections (Black and White, 2023) which we cross referenced with YouGov audience ratings by party registration for each source. For example, The Daily Mail both endorses a majority of Conservative candidates in general elections and has a majority of its voters identify as Conservative. |
Word Count | Numeric word count of the article | Integer |
Length XR | Length of coverage devoted to XR What percentage of the article is devoted to XR, rounded to the nearest percentile? | Coded 1–4, where 1 = 0–25%. and 4 = 76–100%. |
Shares | Number of times the article has been shared on social media. | Integer |
Comments | Number of comments the article has received 1+ weeks after publication | Integer |
Word Count | Numeric word count of the article | Coded in integer form, equaling the number of words in the article |
Sentiment | What is the overall sentiment of the article? | Coded −1–1, where −1 is negative, 0 is neutral, and 1 is positive. Calculated using hand dictionary coding based on the language used in the article. We used manual coding following findings from (Van Atteveldt et al. 2021) that manual coding of media coverage specifically is more reliable than automated or ML coding systems. The author assembled his codebook (specific to coverage of environmental protest in the UK) based on best practices from (Van Atteveldt et al. 2021) and especially (Boukes et al. 2020), who find off-the-shelf sentiment analysis tools wanting and urge the usage of smaller lexicons specialized to the context, language, and domain of the research project at hand. The codebook is available upon request from ES. |
Legality | Was the action legal or illegal at the time it was performed? | 0 = illegal (e.g., throwing paint on trade union buildings, which violated laws against vandalism) 1 = legal (e.g. The Big One, which did not break any laws at the time as it was coordinated with the government in advance) Internal information of XR actions, cross-listed with public court documents/legal resources when necessary. |
Target | Who was the action targeting? | Categories: Public, Elite, Government, Industry. |
Scale | Where did the action take place? How many sites was the action taking place at and where were those sites? | Categories: National (coordinated), Local, London. |
Accuracy | How accurate was the article? How many major factual errors did the article make (who, what, when, where, why). Measured against internal XR documents which outlined details for each action carried out. | Measured 0 and up, with a higher score being more inaccurate, 1 point awarded per major factual error |
Action | Was the article covering an action or something else such as a press release? “Action” here defined as an act of protest or disruption carried out by one or more people against a target. | Action = 1 Not an action = 0 |
Action Code | A unique numeric code assigned randomly by action | Integer |
Category Code | A code classifying each action into a category by target and legality. | Press Releases = 1 Legal Government = 2 Illegal Government = 3 Illegal, Elite = 4 Legal Public = 5 Illegal Public = 6 Illegal Industry = 7 Legal Industry = 8 |
Day of Week | What day of the week was the article published on? | Used as a control variable to account for possible variations in the publishing cycle |
Month | What month was the article published in? | Used as a control variable to account for changes in coverage over time |
National Event | If targeting a sporting event, was the event in question broadcast nationally or not? | National Event = 1 Non-national event = 0 |
Is.Xr | Was XR officially included in the action? Internal Documents | 1 = yes, 2 = no |
mention.xr | Does the article mention XR? | Yes = 1 No = 0 |
after.we.quit | Did the article come out before or after the Jan 1 2023 “We Quit” statement from XR? | Yes = 1 No = 0 |
Source | Which news source published the article? | Categorical variable |
Press Coverage | What percentage of the article is devoted to XR? Multiply the amount of the article, in percentage terms, that is about XR by the number of words in the article. | Decimal, from 0-1, where 100% = 1 and 0% = 0 |
