Dataset for "Arguing about Controversial Science in the News: Does Epistemic Uncertainty Contribute to Information Disorder?"

To gather news articles from the web that discuss the Cochrane Review, we used Altmetric Explorer from Altmetric.com and retrieved articles on August 1, 2023. We selected all articles that were written in English, published in the United States, and had a publication date <b>prior to March 10, 2023</b> (according to the “Mention Date” on Altmetric.com). This date is significant as it is when Cochrane issued a statement about the "misleading interpretation" of the Cochrane Review.

The collection of news articles is presented in the Altmetric_data.csv file. The dataset contains the following data that we exported from Altmetric Explorer:
- Publication date of the news article
- Title of the news article
- Source/publication venue of the news article
- URL
- Country

We manually checked and added the following information:
- Whether the article still exists
- Whether the article is accessible
- Whether the article is from the original source

We assigned MAXQDA IDs to the news articles. News articles were assigned the same ID when they were (a) identical or (b) in the case of Article 207, closely paraphrased, paragraph by paragraph. Inaccessible items were assigned a MAXQDA ID based on their "Mention Title".

For each article from Altmetric.com, we first tried to use the Web Collector for MAXQDA to download the article from the website and imported it into MAXQDA (version 22.7.0). If an article could not be retrieved using the Web Collector, we either downloaded the .html file or in the case of Article 128, retrieved it from the NewsBank database through the University of Illinois Library.

We then manually extracted direct quotations from the articles using MAXQDA.
We included surrounding words and sentences, and in one case, a news agency’s commentary, around direct quotations for context where needed. The quotations (with context) are the positions in our analysis.

We also identified who was quoted. We excluded quotations when we could not identify who or what was being quoted. We annotated quotations with codes representing groups (government agencies, other organizations, and research publications) and individuals (authors of the Cochrane Review, government agency representatives, journalists, and other experts such as epidemiologists).

The MAXQDA_data.csv file contains excerpts from the news articles that contain the direct quotations we identified. For each excerpt, we included the following information:
- MAXQDA ID of the document from which the excerpt originates;
- The collection date and source of the document;
- The code with which the excerpt is annotated;
- The code category;
- The excerpt itself.