43. From the desk of James Randi (Podcast)

[Chris Noble]

Originally Posted by Larry Boy

I find it strange to say that just because an effect is small, it's likely not to exist. We see similarly small effects in medicine, but those results are still seen as valid. Where's the difference? Theory? That's funny because I recall reading that there are cases in medicine where effects are stumbled upon by chance, verified experimentally, and not until after that explained theoretically.

The protective effect of aspirin against heart disease could not have been discovered by stumbling across it. It is necessary to conduct large scale double blind randomised trials in order to pick it up.

By contrast the antipyretic and analgesic effects of salicylic acid are strong enough that they could be stumbled across.

I am definitely not saying that small effect sizes are not real but I am extremely skeptical of claims involving small effect sizes that do not have any basis in established science especially where very minor problems in experimental design can account for the observed effect.

[Larry Boy]

Originally Posted by Chris Noble

The protective effect of aspirin against heart disease could not have been discovered by stumbling across it. It is necessary to conduct large scale double blind randomised trials in order to pick it up.

By contrast the antipyretic and analgesic effects of salicylic acid are strong enough that they could be stumbled across.

I am definitely not saying that small effect sizes are not real but I am extremely skeptical of claims involving small effect sizes that do not have any basis in established science especially where very minor problems in experimental design can account for the observed effect.

That's fair enough, I guess.