EA Survey 2017 Series: Distribution and Analysis Methodology

Abstract

The 2017 Effective Altruism (EA) Survey outlines its methodology for data collection, distribution, and analysis. Data was collected using LimeSurvey, incorporating a “Donations Only” version for returning participants. The distribution strategy involved extensive outreach to over 300 individuals and groups across diverse digital channels, including mailing lists, social networks, and forums, from April to June 2017, with referrer data tracked via unique URLs. Analysis was conducted in R using anonymized data, with scripts and raw data made publicly available. Subpopulation analysis identified variations in demographics and cause affinities among respondents from different sources, raising concerns about potential sampling bias and representativeness of the broader EA community due to the unknown “true” EA population. A comparison with an independent SlateStarCodex survey indicated reasonable capture of that specific subpopulation, albeit with some demographic and belief differences. – AI-generated abstract.