I just got this from hsc.csu.edu.au website:
Produce simulated starlight from the incandescent lamp in a ray box kit, commonly available in school science laboratories. This has the advantage that coloured filters mounted in 35 mm slide frames can easily be inserted in the light path. If this is not available, filters can be held by hand in front of any incandescent lamp.
Use a light intensity probe attached to a datalogger to measure the intensity of light at a set distance from the lamp. Set the datalogger to operate in manual or “snapshot” mode. A photographer’s hand-held light meter is a suitable alternative to measure light intensity.
Place different coloured filters, one at a time, between the lamp and the light probe. For each filter, measure the intensity of light with the datalogger. You should note that the filters used in photometry, unlike those in a ray box kit, transmit a carefully calibrated range of frequencies.
For each filter, also observe the light through a hand-held spectroscope to see qualitatively what effect the filter has on the spectrum of white light produced by the lamp. Use the in-built scale to measure the range of wavelengths transmitted.
Record all your observations systematically in a suitable table. Compare your qualitative and quantitative observations for different filters.
Use your observations to predict the effect of different filters on the measurement of apparent magnitude of stars of different spectral type.
