In a famous scene from the Coen Brother's cult classic "The Big Lebowski," Jeffrey "The Dude" Lebowski is cruising behind the wheel of his car, tunes blasting as he sips beer and smokes pot. In a classic stoner move, he tries to flick his lighted roach out the window - except that it's closed. The burning roach ricochets onto his lap, causing him to wildly swerve and then crash into a garbage can.

Lebowski doesn’t get hurt, but there are those, such as State Assemblyman Tom Lackey (R-Palmdale), who fear an increase in real-life risky driving if on Nov. 8 Californians pass Prop. 64, the initiative that would legalize recreational marijuana. 

Lackey, a retired California Highway Patrol officer, wants the state to adopt a legal limit for THC - the mind-altering ingredient in cannabis. Without it, he says, patrol officers must rely mostly on their own subjective judgment as to whether a driver is impaired. That, he says, makes many officers uncomfortable.

"In my own experience I’ve seen this phenomenon," he says. "I’ve seen somebody clearly impaired and I’ve seen officers still hesitant to remove [that person] from the roadway because they’re not a drug recognition expert."

Lackey says drug recognition experts have advanced training in identifying impairment levels caused by marijuana and a slew of other drugs; most highway patrol officers don't.  So he co-sponsored legislation - modeled after drunk driving laws - that would have made it easier for officers by creating  a legal blood limit for THC in drivers.

But that bill failed this spring, in large part because THC levels are not good indicators of intoxication. 

"Unlike alcohol - which has a generally linear relationship between the amount of alcohol you consume, your breath alcohol content and driving performance – the THC route of metabolism is very different," says Tom Marcotte, co-director of the Center for Medicinal Cannabis Research at UC San Diego.

THC is fat soluble. Unlike alcohol, it doesn’t absorb uniformly into the body’s fluids and tissues. So that means the measure of THC in one’s breath, blood or urine doesn’t correspond to impairment in the same predictable way that alcohol does. 

And that’s why it’s a bad idea to adapt drunk driving laws to driving while high, says Prof. Mark Kleiman, a drug policy expert at New York University.

"You can be positive for THC a week after the last time you used cannabis," he says. "Not subjectively impaired at all, not impaired at all by any objective measure, but still positive."

And the reverse is true as well, says Dale Gieringer, director of  California NORML,  the National Organization for the Reform of Marijuana Laws.

"You can be very, very stoned and not have a very high level of THC actually apparent in the blood," he says.

And that poses a quandary that researchers, including Marcotte and his colleagues, are hoping to solve. The state legislature has awarded UCSD's cannabis research center $1.8 million to study THC impairment and to develop a scientifically-sound field sobriety test for marijuana intoxication.

UCSD's research center is in a downstairs office-turned-lab, where a desktop driving simulator is stationed. It’s here that 60 yet-to-be recruited volunteers will undergo a battery of tests this fall - first while sober and then while high.

"They’ll smoke the cannabis and the rest of the day they’ll be doing simulation assessments each hour," Marcotte says. 

The tests will include their ability – before and after smoking - to perform basic tasks on an iPad.

Research associate Kevin McShea demonstrates the apps for a visitor on a recent afternoon. One, featuring a square that moves around the screen at varying speeds, requires keeping one's fingers on the square at all times - a test of "critical tracking." It's a skill that's not easy to perform when a person is stoned, he explains.

"We have you do another task to ask how much time has lapsed, ten seconds? Twenty seconds?" says Marcotte. "Because we know that time perception is affected by cannabis."

So, too, is the ability to learn something new when high, which another app the team has developed will measure.

All of the test volunteers will also have their ability to switch from driving to other common tasks measured - before and after smoking. 

"Your task is to find the larger target circle and touch as close to the center of it as possible with your finger as soon as you can," McShea says as he holds up a screen populated with dozens of circles. 

It's an app designed to mimic the type of  distracted driving one might experience while tuning the car radio; while adjusting the GPS or – in the case of the Big Lebowski – grabbing for the burning roach that landed in his lap.  

The studies at the cannabis center are funded for three years. If Californians approve Prop. 64, the research will be extended. The ballot initiative provides for more funding to help researchers develop better ways of measuring when drivers are stoned.