Lies, Damned Lies, And Facebook (Part 4 of ∞)

A statistical debunk (part 4 of the 'Lies, Damned Lies, And Facebook' series) of the viral claim that Vancouver's 'Don't Be That Guy' anti-rape poster campaign reduced sex crimes 10%. Scott shows it's cherry-picked (Edmonton's identical campaign saw +14%), the 'change in trend' (second derivative) is no bigger than random crime fluctuations (arson, B&E, murder all shifted more), the third derivative shows the 2011 shift merely continued a 2010 trend, and the 'sexual offenses' were mostly groping/stranger-crimes the posters couldn't affect. Concludes there's no evidence it works; the real function is tribal signaling ('fake consequentialism').

Quality 65 · Strong. 65 — low Strong. A careful, characteristic single-number-claim debunk with a decisive counterexample (Edmonton) and a good methods lesson (derivative noise, base-rate skepticism, fake consequentialism). Firm low-Strong as a single-claim takedown.

Claude’s paradigm shift 46 · Moderate. 46 — Moderate. The derivative-noise analysis + the fake-consequentialism/tribal-signaling read is a useful, mildly generative frame.

Real-world impact 1 · Negligible. 1 — negligible. A within-discourse debunk; no material reach.