Sometimes we can’t rely on anecdotes (cherry-picked) or summary statistics (too vague in aggregate), and in such situations random Sampling can help since you draw a random sample from the full issue.


References

  • Link to website, bibtex from Zotero or note with book/blog/etc summary