Have you tried this yet or is it an idea floating in your head for now?

Here are some things to take into consideration:
sample size, a sample size of 2 isn't enough to draw any conclusion. You should also test sativas, indicas and hybrids just to see if the effect is strain dependent or not.
I don't even think you can use a p value and the normal distribution curve with sample sizes less than 10.

Also if you cut the light to give the plants the strobe effect you're giving it less lumens of light. You're gonna be giving one plant 12 hours of HID light, the other plant is receiving only 4 hours, the other 8 hours it will be given the strobe light, which most likely doesn't have the right lumens, nor intensity that the other plant will be receiving.

If you don't cut the lights for the strobe light, you won't get a strobe effect. You could get a bulb similar in spect and intensity to the strobe light and use that for the control plant. You'd use that bulb alone for the same period that you're using the strobe light in the other room. But you'd probably want to have another room with another control plant in regular 12/12 HID to rule out the different bulbs having an effect on the outcome of the experiment.

Also I don't know how much effect a strobing effect will have on a plant. There's a lack of reasearch on the hormonal and physiological changes that occur when photoreceptive plants are put in the right environment.

Cannabis is a short day plant, but it'd probably make more sense to call it a long night plant. The plant produces a hormone that is rapidly degraded in the prescence of light. When it is dark for a long enough period of time, this florigen reaches a high enough level to trigger flowering in short day plants.

http://www.biologie.uni-hamburg.de/b-online/e30/30c.htm

here's another link
http://users.rcn.com/jkimball.ma.ult...html#mechanism
but I dunno about the circadian rhythm part. It doesn't seem to really matter if you do 12 hours light from 5am to 5pm, or 12 hours light from 5pm to 5am but if what they suggest is true, then this would matter cause it would change the amount of CONSTANS protein. Perhaps an idea for another study, actually it'd probably be easier to study this than what you propose.

peace