I am a bit stuck with this. Say for example there is a species of bird with variable colours, and they can present as red, blue, green or a combination (red-blue, red-green, blue-green, red-blue-green) and we are trying to determine if more birds have red colouring than blue (regardless of colour combinations, as long as it has the colour, it was counted), it does not seem right to run a simple t-test comparison, as the test assumes that the colours are exclusive right?

Another problem is that I do not have the raw data, and only have proportions!

Just using birds as an example, my dataset has nothing to do with birds!