In order for data from experiments to be useful, there need to be experimental controls-- test conditions that make sure significant results are due to the factor you're testing, not anything else.
A 'vehicle control' is important because it makes sure a treatment, not its vehicle, is responsible for an effect. If a chemical dissolved in water is placed in a bacteria culture, for example, the vehicle control would be placing water without the chemical in another bacteria culture. Some vehicle controls are more interesting than others; in an experiment where plant fragments were added to bird nests, the vehicle control was going around and touching all the bird nests that didn't have fragments added, since the vehicle in this case was a human hand.
Does eating a balanced breakfast result in more alertness during the day? I looked at a few research papers on the topic and didn't find any with a good vehicle control, i.e. having people repeatedly place a metal spoon in their mouth. The awakening effects of oral contact with metal utensils could be a legitimate factor. Probably not, though.