ElectionRumors2022: A Dataset of Election Rumors on Twitter During the 2022 US Midterms

TL;DR

This paper addresses how rumors about election administration spread on Twitter during the 2022 U.S. midterms. It builds the ElectionRumors2022 dataset, a high-recall, low-noise collection of 1.81 million posts tied to 135 rumors, enriched with geographic, domain, and user-partisan annotations. The authors describe a multi-stage workflow—data collection via Twitter's API, rumor-lead identification, pool construction, post-hoc rumor coding, and tweet-quality assurance—and present five descriptive quantitative analyses plus three Arizona-focused case studies, benchmarked against a 2020 dataset. The work highlights AZ-centric rumor dynamics, shifts in external link sharing, and persistent retweet concentration, offering a valuable resource for studying online rumor dynamics, misinformation, and disinformation in election contexts while acknowledging ethical and methodological limitations.

Abstract

Understanding the spread of online rumors is a pressing societal challenge and an active area of research across domains. In the context of the 2022 U.S. midterm elections, one influential social media platform for sharing information -- including rumors that may be false, misleading, or unsubstantiated -- was Twitter (now renamed X). To increase understanding of the dynamics of online rumors about elections, we present and analyze a dataset of 1.81 million Twitter posts corresponding to 135 distinct rumors which spread online during the midterm election season (September 5 to December 1, 2022). We describe how this data was collected, compiled, and supplemented, and provide a series of exploratory analyses along with comparisons to a previously-published dataset on 2020 election rumors. We also conduct a mixed-methods analysis of three distinct rumors about the election in Arizona, a particularly prominent focus of 2022 election rumoring. Finally, we provide a set of potential future directions for how this dataset could be used to facilitate future research into online rumors, misinformation, and disinformation.

Figures (12)

• Figure 1: A diagram showing the process of curating this dataset
• Figure 2: A diagram of the data and feature relations contained in our dataset. Columns that are joinable across tables are linked via arrows.
• Figure 3: A timeline plot showing the number of rumor tweets per day in the dataset. The start of Election Day (November 8, 2022, at 7:00 UTC) is denoted with a red vertical dashed line.
• Figure 4: Relative distribution of tweets about U.S. states in 2020 and in 2022 by both direct references through substring searches and rumor-level coding. Each row is populated by its eight most frequently referenced states, with the rest bucketed into other. Tweets mentioning multiple states were counted for each state. Note, the percentages for direct reference searches are relative to the total number of tweets referencing a state, not the entire dataset.
• Figure 5: A timeline showing the distribution of tweets authored by users in each partisan category. For readability, we represented left-leaning account activity with a downward-facing line, and right-leaning account activity with an upward-facing line. The vertical dashed line represents the start of Election Day.