If you want to measure the effect of magnets on plants, you are thinking about two very important things: controls and sample size.

Controls allow us to be fairly sure that differences are due to the factor we're trying to measure, in this case, magnets. Just having several plants (a large sample size) is important, but how you treat the plants is critical in a fair, valid experiment. For example, you wouldn't want to put all your magnet plants in one place and all your controls in another, because the light, temperature, or something else might be different. You want several plants that you treat exactly the same except for having a magnet nearby. A really good control would be to have non-magnetized steel in place of magnets for the controls.

You might use 6-10 plants in each group. More is always better, but you may not have enough space or magnets. I don't know which plant would be better, but I would choose plants that grow quickly and aren't too delicate.

If you are growing plants outside, weather will definitely make a difference. Even indoors, plant will be influenced by light coming in windows, temperature differences in different areas of the room, and other factors.

Finding out whether there is an effect is part of the fun of the experiment. Remember not to let your hopes for a difference influence how you measure any difference. Think about what might be good to measure. Plant height? Number sprouted? Whether it grows in a certain direction? Number of leaves? There are a lot of choices. You can record more than one thing. Start thinking now about how you will show your results at the end. It may help you design a better experiment.

I majored in biology. I studied animals mostly, but good experimental design is the same in any branch of science.

Goodluck with your experiment.