Tip:
Highlight text to annotate it
X
The only two major measurements
that we made where there was a consistent effect across all subjects
was the average speed they travelled which was slower
and their tracking ability, their ability to stay within the lanes.
The simulator is set up to measure how accurately
people stay in the lane that they're in
and drive as normally as they can.
And the more they weave about in that lane,
this is the measure of the effect of cannabis
on what we call their psychomotor performance.
And cannabis has a significant effect on that.
That was a consistent effect across all subjects.
The tracking test shows cannabis does affect hand-eye co-ordination.
The simulator test measures the more complex decision-making processes
involved in driving.
The computer records all this data 60 times a second
and turns it into a graph for analysis.
Here it took the participant 1.3 seconds to react
by braking in response to the programmed event.
The conclusions here, generally speaking,
were there were some instances of differences in behaviour
in the various measures being used
which suggest that the higher the dose of cannabis,
the more cautious perhaps they're going to be as a driver.
More cautious on the one hand
but on the other hand they can't react as quickly.
So anything sudden that happens
or if there's any control skills that are needed fairly urgently,
then they'll have more difficulty.
So, overall, alcohol makes you more confident.
Cannabis, it seems, makes you more cautious
but not necessarily a safer driver.
So how reliable are these conclusions?
And if they're reliable, how relevant are they
to what happens in the real world?
We've conducted a lot of trials over the years now
using the driving simulator
and an important element of that
has been to validate the test tool itself.
We've done that in a variety of ways.
Mostly we focus on speed choice, distances from other vehicles
and lane position.
And we've observed those in the driving simulator
and out there in real life.
Those results have shown two things:
firstly, that the driving simulator produces results
which are very close to real life, however they are different
and most importantly, they're systematically different
to those which you observe in real life.
So we don't just get more variation,
we tend to get, for example with speed,
people will drive slightly quicker in the simulator
than they will on the real road,
particularly in motorway environments.
However, it's systematic, that means it's repeatable
and it's always in the same direction.
When we know that, basically we can take the sort of data
which we get from the simulator and scale it
and make our predictions about what's going to happen
out there in the real world.
The testing techniques, the tracking task and the driving simulator
prove to be valid and reliable measures of performance.
What's been less successful is the premise that participants
who regularly smoke cannabis can be given different doses of the drug
or none at all and not notice the difference in the effect.
It's probably impossible in any sort of trial
when something as strong as alcohol or cannabis
is having an effect on you
not to know whether it's a placebo or not
but I don't think in any sort of experimental situation
we could have actually done better than we've done.
Even in a carefully designed experiment such as this
there are some factors you simply cannot control.
In this case it's the fact that the participants
know how it feels to take the drug.