Although placebo analgesia is popularly believed to be prompted by taking a "sugar pill" (an inert substance that have no pharmacological effects) while being led to believe that it is an effective painkiller, there are actually many different ways of prompting the placebo response. These include verbal suggestions that induce expectations of efficacy (do these exercises and you'll feel better), conditioning (repeated exposure to a treatment and its effectiveness), etc.

But Luana Colloca and Fabrizio Benedetti, two very prominent researchers in the area placebo effects research recently published an intriguing study investigating social influences on placebo analgesia.

What they did was randomly assign participants to one of 3 groups: social observation, conditioning, and verbal suggestion.

In all groups, participants experience a series of electric shocks preceded by either a red or green light. The placebo part of the study is a sham electrode affixed to the middle finger, which participants are told  delivers a nonpainful stimulation that has an analgesic effect; that is it can reduce the pain of the shocks. The green light is supposed to indicate that this analgesic stimulus has been turned on and that reduced pain should be anticipated.

Now, in the social observation group, participants learn from watching a "demonstrator" go through the series of trials first, that the green light precedes diminished pain relative to red light trials. They can see this because they can see the demonstrator rate the painfulness each pain shock.