Typically, when someone places a bet—be it on a sporting event, obscure trivia or where the ball in a roulette wheel will land—he or she is fully aware of the decision to do so. And now scientists at Oxford University in England have discovered that they can better determine if decisions in general are conscious (or subconscious) by having subjects literally gamble on, well, their decisions. The researchers, reporting in this week's Nature Neuroscience, found that study participants were reluctant to wager big bucks unless they were confident in their choices, indicating that they knew full well what they were doing.

Previous methods used by brain researchers to assess whether a subject was aware of a decision he or she had made—such as having them rate their confidence in their choices—were indirect at best, says researcher Navindra Persaud, a University of Toronto medical student who is on leave studying experimental psychology at Oxford. "That forces you to introspect and see whether or not that decision was more like other decisions that you are aware of or if it's more like guessing—and that's a rather difficult thing. You're not seeing if people are aware if they made the decision, you are seeing if they're aware of their awareness."

Persaud and his colleagues performed three experiments to determine whether gambling could better elucidate a person's awareness during the decision making process.

In the first, researchers asked a subject—whose visual cortex had been damaged in a childhood bicycle accident leaving him with only partial vision in his right eye—to determine whether there was a light pattern in his right visual field. After the subject gave his answer, he would wager the equivalent of either a dollar or two on his confidence in the veracity of his answer.

The participant performed well above chance, answering correctly 70 percent of the time. But he did not maximize his earnings—indicating that he wagered the higher sum only when he was confident he was correct.

Healthy volunteers were recruited to participate in the other two tests: a construct called an artificial grammar task and a gambling task. In the grammar task subjects were shown a string of letters made up of different combinations of X, T, V, R and told that all of these patterns (such as XXTVR) followed a particular rule—albeit not explicitly explained. Each participant was then shown new strings and asked to identify whether the new sequences followed the rule or not. They placed a wager of either $2 or $4 to back their confidence in their answers. Again, subjects performed better than chance, but they did not maximize their winnings because they did not always bet high even when they were correct.

The same results were replicated in the gambling task, where the subjects had to choose the top card of one of four decks and then bet either $20 or $40 on their decision; each card indicated either a monetary award or deduction—two of the decks had a positive net value and the other two had a negative net value.

"What is common to [the tasks] is that you have a binary decision and you can perform well on the task, but you're not aware that you're performing well—there's performance without awareness," Persaud says. "By having people wager on decisions that they've made, that gives you a way of measuring their awareness directly."

Christof Koch, a professor of neural computation at the California Institute of Technology, believes that the work done by Persaud's group could be "a vital hint" in teasing out the neural correlates of consciousness. "In 1995 Francis Crick and [I] hypothesized that the ability to plan is one of the key evolutionary advantages of consciousness," Koch says in a commentary on the recent study. "Persaud's method implies that computing expected gains and choosing the more profitable outcome—that is, placing a bet—has a close and intimate connection to awareness."