Rudisill, M. (1952). Children's preferences for color versus other qualities in illustrations. Elementary School Journal, 52, 444-451.

(Conclusions.) The findings of this study seem to justify the following conclusions with respect to picture preferences of children of the ages here studied - kindergarten through Grade VI. Since the subject matter of all pictures used in the present study was realistic in character, the present findings and conclusions are not assumed to apply to other than realistic subject matter.
1. If two pictures are identical in all other respects, most children prefer a realistically colored one to an uncolored one.
2. If different pictures include the same subject matter and the same colors, most children prefer the one which is treated in such manner as to give the truest appearance of realism or lifelikeness.
3. If different pictures include the same subject matter, most children prefer an uncolored one which gives them an impression of reality above a colored one which does not seem to conform to reality.
4. If different colored pictures include the same subject matter, moste children prefer a less colorful one which gives a greater appearance of reality above more colorful ones which appear less lifelike.
5. There is an increase in unanimity of these preferences with increase in grade level up to Grade IV. This greater unanimity of preference of older children, as compared with younger children, is believed to be due to the former's greater capacity for discriminating reality.
6. Typical adult opinion overemphasizes the importance of color per se and underemphasizes the importance of other qualities in illustrations for children.
7. Photographs of excellent quality, both colored and uncolored, deserve much wider use than they are at present being given in illustrations for children.
8. These findings do not justify the statement of a general principle as to comparative preference between photographs and realistic colored drawings because the examples of the two types used in the present study were not of equal quality.
9. Consideration of the first four conclusions above, together with children's stated reasons for choosing or not choosing certain pictures, suggests: (a) In looking at a picture, a child apparently seeks first to recognize its content. (b) Any picture (assuming a certain content) proves satisfying to the child in proportion to its success in making that content appear real or lifelike. Whether it is colored or uncolored is less important than the appearance of realism. (c) A perfect visual representation of realism includes color, and color in pictures proves satisfying to the child in proportion to its success in increasing the impression of realism or lifelikeness.