This is a problem that often afflicts those who don't have a very good understanding of how statistics work. They assume that because two things have been shown to be correlated, they can say that factor A is the cause of factor B. If they find that pot smokers get lower grades on average, they jump to the conclusion that the pot is causing the bad grades, and to the layperson that's exactly what it looks like, so it makes for effective propaganda. But it must be understood that without data showing a direct link like that, it is equally plausible that the bad grades are causing the pot smoking, or that the bad grades and pot smoking both stem from a third factor of some sort.

For example, it could be that the kids who get better grades pay more attention to what their teachers say, so more of them have absorbed the teachers' anti-drug propaganda. Maybe the kids who get better grades are less likely to report their drug use for some reason. Maybe the kids who get lower grades come from poorer families, and poorer families are more likely to have drugs accessible to them. Maybe the kids who get lower grades have to put up with more bullshit from their teachers and feel more of a need to toke up.

Maybe it all boils down to Bill Hicks' theory:
They lie about marijuana. They tell you pot smoking makes you unmotivated. LIE!!! When you're high, you can do everything you normally do just as well ... you just realize it's not worth the fucking effort. "Sure, I could get up at seven, go to a job that I hate...for the rest of my life...*puff puff*...or I could get up at noon...and learn to play the sitar."