|
Abstract:
|
“Lies, damned lies, and statistics” is a refrain every statistician has heard throughout their career. But how does one effectively use the tools of statistics and data science to lie and mislead? Studying examples of misleading data analysis can help us to spot intentionally misleading analyses and help us avoid unintentionally misleading others ourselves. The 2020 US presidential election combined extreme partisanship with abundant, variable-quality data creating a breeding ground for statistical misinformation. Here we examine how statistical analysis was misused to undermine faith in the electoral process. While the examples are specific to the 2020 election, the tactics are generalizable to other arenas from future elections to public health crises and climate denial.
|
