The New Yorks Times ran a story on the psychology of false confessions. Here’s an excerpt, but the whole thing is worth reading:
If you have never been tortured, or locked up and verbally threatened, you may find it hard to believe that anyone would confess to something he had not done. Intuition holds that the innocent do not make false confessions. What on earth could be the motive? To stop the abuse? To curry favor with the interrogator? To follow some fragile thread of imaginary hope that cooperation will bring freedom?
Yes, all of the above. Psychological studies of confessions that have proved false show an overrepresentation of children, the mentally ill and mentally retarded, and suspects who are drunk or high. They are susceptible to suggestion, eager to please authority figures, disconnected from reality or unable to defer gratification. Children often think, as Felix did, that they will be jailed if they keep up their denials and will get to go home if they go along with interrogators. Mature adults of normal intelligence have also confessed falsely after being manipulated. […]
In experiments and in interrogation rooms, adults who are told convincing fictions have become susceptible to memories of things that never happened. Rejecting their own recollections through what psychologists call “memory distrust syndrome,” they are tricked by phony evidence into accepting their own fabrications of guilt — an “internalized false confession.”
That is what happened to a shaken Martin Tankleff, and although he quickly recanted, as if coming out of a spell, he was convicted and drew 50 years to life. He spent 17 years in prison before winning an appeal based on new evidence that pointed to three ex-convicts. But they have never been tried. Whoever killed the Tankleffs remains at large.