Galaxy Zoo: morphologies derived from visual inspection of galaxies from the Sloan Digital Sky Survey
The main conclusion of the paper is that an education in astronomy no more qualifies the ability of a person to classify galaxies than does any other personal possession. Second, the accuracy of classifying galaxies is rather low, at just ten percent. Third, the paper concludes that the data provided by the public obviously shows differences between color and shape on a graph. Finally, the paper concludes that classification by shape is important because a large number of galaxies of one color are the same shape, yet organization by shape don't provide "traditional" classification based on shape. This might be trying to say that, when sorted by color, galaxies are not sorted by shape in the standard/expected sorting. I guess the conclusion is that color and shape of galaxies are not correlative because you cannot sort galaxies by color and produce the same organization as you would, were you to sort them by shape.
Descriptions of Graphics:
Figure nine, a scatter plot of quantity and color of galaxies, shows elliptical galaxies as being more common than spiral galaxies, cross all color variations.
Figure ten, a scatter plot of color and quantity of spiral versus elliptical galaxies, shows no correlation between color and shape, apparently.
Key Terms:
Full paper available online at http://mnras.oxfordjournals.org/content/389/3/1179.full.pdf
Morphology: Shape or form
Fidelity: Maintaining accuracy in respect to an original version
Bimodalities: two differences in shape on a graph
Lintott et al.