In this episode of Life, the Universe & Everything Else, Gem Newman talks about the evidence (or lack thereof) for ghosts with Greg Christensen, Ian Leung, and Mark Forkheim. Also on this episode, Gem chats with Hemant Mehta of Friendly Atheist.
Warning: In the first few minutes of the podcast, Greg asks Gem to give a primer on Bayes’ Theorem and conditional probability (and why this is at all relevant to ghosts). It involves a little bit of math, so brace yourself.
Links: Solstice Party (22 June 2013) | Drinking Skeptically (11 June 2013) | Bayes’ Theorem | Sleep Paralysis | Apophenia | Pareidolia | Friendly Atheist | Steinbach Pastor Voices Opposition to Bill 18 | I Sold My Soul on eBay | The Young Atheist’s Survival Guide
P(A|B) = P(B|A) • P(A) / P(B)
The probability that it is raining, given the fact that it is cloudy:
P(Raining|Cloudy) = P(Cloudy|Raining) • P(Raining) / P(Cloudy)
The probability that ghosts exist, given the fact that you heard a weird noise:
P(Ghosts|Noise) = P(Noise|Ghosts) • P(Ghosts) / P(Noise)
In this instance, P(Noise|Ghosts) is the probability that you’d hear a weird noise, assuming ghosts exist, P(Ghosts) is the prior probability that ghosts exist, and P(Noise) is the probability that you’ll hear weird noises (generally speaking).
The example numbers plugged in during the podcast:
P(Ghosts|Noise) = 0.95 • 0.05 / 0.60 = 0.08
So in this case, hearing a strange noise might increase your belief in ghosts from 5% to 8%; definitely not a smoking gun.