Data associated with the publication: DIY Cruelty: The Global Political Micro-Practices of Hateful Memes

The main component of the DIY Cruelty dataset is approximately 500 image-text memes and their associated comment threads. Each post within the comment thread is an observation. The total number of observations is approximately 15,000. All the observations of image-text memes and many of the comment thread posts include codings and textual descriptions that explain the researchers’ interpretation of the level (or absence) of cruel meaning. The collection also includes the raw data (only textual posts and only image-based posts in addition to the image-text memes), as test samples of meme posts and comment threads from other subreddits and from another social media platform.