Shine light on X stacked sheets of red colored cellophane/plastic sheet (probably can get at office supply or art supply store). Increase X from 1 to N(red) until no more light visible through paper. Repeat with same weight cellophane of other colors, record N(blue), N(green), etc. Compare N values. Control intensity of light source by placing a pinhole between light and cellophane stack - so you don't end up with a 100w bulb and $30 of cellophane for each color . . .
Another way would be to build a long cardboard box w/ light bulb at one end, then a series of baffles such that light has to reflect past each baffle to get to the next, other end closed. Pinhole above each baffle so that you can peer in and see if that baffle is illuminated. One side of box opens so that you can reach in and line the baffles w/ colored paper. Line all baffles w/ red paper, what is the last baffle illuminated, repeat for other colors. Also try "glossy" versus "matte" paper.
These might measure slightly different things, might be interesting to see if they agree.
Also interesting to see if you can detect the difference between different light sources - e.g. can you determine a diffierence between an incandescent bulb's color profile vs a daylight CFL vs a white LED, from the ratio of the various N's measured for each light source?
Ideally best to do a kid's science experiment w/o complicated test & measurement equipment.
Edit: Here is something that also looks fun, and more fun if replace the ammeter w/ a motor and spinning propellor, then count the rev/min. Or have the motor drive a pulley that lifts a weight and measure time to lift.
http://www.makeitsolar.com/science-fair-ideas/03-wavelength-light.htm