You are faced with a problem that many human health professionals are faced with every day - they must make decisions based on anecdotal and correlational (epidemiological) studies as opposed to a "real experiment" (a clinical trial) in order to scientifically test the treatment.

In the specific case of food dyes/additives, I think you could do a basic study that looks at "an average diet" (maybe you could survey your classmates over the course of 2-4 weeks, or even just use your school's lunch menu) and then you would have to look up how much dye and what kind was in each type of food. This is probably do able using the internet and ingredient labels. This would give you a baseline for average diets. It's not an experiment though, just a data collection exercise.

Correlating this with some sort of human behavior is much more difficult (ethically as well as practically). Perhaps you can do some research to see if any studies have been conducted to determine if food dyes have any effects on the behavior of rats, a common model test organism (you would have to look into the actual scientific literature) and then think about extrapolating to humans.It is also possible that some studies have been done directly on human populations (not necessarily related to autism). I suggest that you check out the National Institutes of Health website (nih.gov.org) and perhaps even www.clinicaltrials.gov to see if there are some specific studies that have been conducted in humans.

There might be some experiments you could do to look at the effect of dyes on bacterial growth or cell culture growth. Again, these may already have been conducted (you would have to check the scientific literature). But you could extend or tweak the design to do your own experiment if you have the resources available.

You can do scientific literature searches on the Pub Med database.

When you get on the database, you can search with specific words - I tried "food dyes human behavior" and a whole bunch of papers popped up, so maybe this will help you.

See what you can come up with, and if you want more help or have questions, get back to us. Good luck!

I really like your idea for your science experiment. Since you can not experiment on people very easily my first thought was that you might want to experiment on an easily obtained backyard insect. You could take different dyes and expose the insects to these dyes. For the experiment to be interesting, you would hope that the dyes do have some effect on the insect. I did a quick Google search and found thatin fact this has been done! So you are in luck if you want to duplicate this experiment. What there researchers did was feed thecommon house fly all the commercially available dyes in a milk mixture.Apparently, the dyes combined with light exposure caused up 100%mortality in the flies! As well as some behavioral changes .Obviously flies are not people but this experiment shows that some food dyes are not be benign in insects and therefore warrant further research in humans as well. Here is a link to a summary of the results of the experiment done in the 1970's:

Anyhow, this experiment outlines some of the key points you will need for whatever experiment on dyes that you decide: a living organism that will react to the dyes in a short period of time (few days). Don't forget to have an experimental control - some insects that are not exposed to any dye.

Chemistry is not my area of expertise, but if you need some help understanding the results of the experiment or need help with designing your experiment feel free to email me.

Food dyes are compounds that we humans have not eaten in the past and have not evolved with the need to digest or otherwise deal with them chemically. As a consequence, some of them may be carcinogenic, or poisonous in some other way.