Conclusions
 

Impact for practitioners

Color selection methods, or color pickers, offer a standardized, direct-manipulation interface for the user to select a color based on a variety of models, regardless of the underlying software and hardware representations. The practitioner (most likely a software engineer), should offer a range of color pickers, ranging from the user-oriented Crayon, or discrete palette picker, and HSV picker to hardware-based RGB and CMYK pickers. The user then has the freedom to select the color picker to the task at hand.

The Crayon, or discrete palette, picker is most appropriate when the user does need a continuous range of colors. The user thinks in terms of, for example, blue, and its immediate variations dark blue, and light blue. This may fulfill the needs of users who use word processors and use color just to create contrast among the text. Younger users might also benefit, as the discrete palettes are easier to use based on subjective surveys, and younger users may not be able to disciminate among the subtle shades of a color. And what child would not be able to comprehend the familiar box of crayons?

The HSV color picker is an excellent user oriented continuous color selection method. User can accurately select a color by clicking in its general area on the color wheel, and then fine tune the color's brightness. HSV was subjectively rated easy to use and is faster and more acurate than RGB and CMYK. HSV is perfect for use in software that requires a fine control over color such as in image editing software.

The RGB and CMYK color picker color models are based on hardware characteristics, and should be used when the hardware needs to be taken into consideration. RGB is used by computer graphics hardware to represent color on the computer monitor. Decreasing the value of red of an RGB value with accordingly only decrease the value of the red pixel on the monitor. Similarly, CMYK is often used in publishing industry, as cyan, magenta, yellow, and black are used as inks in printers and typesetters. Like RGB, maipulating CMYK values will have a direct, one-to-one correspondence with the cyan, magenta, yellow, and black inks in the printers.

From the results, it appears that the lack of a darkness slider slows down color selection and reduces accuracy. Include a darkness control independent of the hue controls.

An ideal solution to color selection, would be to combine the ease, speed, and accuracy of the discrete palette with the control of a continuous color picker. A small window can function as a "favorites" discrete color palette that holds freqently used colors. These colors can be reselected with exact accuracy from this palette, or a new color could be added by double-clicking on it to summon a continuous color picker.

Suggestions for future researchers

Several refinements can be made to this experiment. The change most influential on the results, increasing its accuracy, would be to have finer control over the Color Picker API (Application Programming Interface). A minimal API was used to merely summon the Color Picker dialog box. Using the extended Color Picker API, more accurate data can be collected by using the API's to eliminate the distracting picker menu on the right side of the dialog box, and presetting the picker type for the user, instead of relying on the user to switch pickers, which may lead to longer times, and errors.

Further physical refinements can also be made, such as calibrating the computer monitor, use consistent lighting, and using a large mousepad.

More attention can also be given to variations in time, accuracy, and ease of selecting various, distinguished colors among the treatments, as was lightly touch upon in this experiment.

And of course, this experiment can be expanded to include other color pickers.

Refine the theory or develop a new one

The determining factor in minimizing time and maximizing accuracy seems to be minimizing mouse operations. The Crayon Picker leads because it requires one mouse operation to select a color from its domain. HSV is second because it requires two mouse operations. We attribute CMYK exceeding our expectations to our subjective observation that the subjects seem to select color by determining hue and brightness. CMYK offers hue in terms of cyan, magenta, and yellow, and offer brightness in terms of black (independent of CMY). RGB fails in minimizing time and maximizing accuracy because hue and brightness are both determined by the same red, green, and blue sliders. RGB earns 3 mouse operations when the user first determines the hue, and earns 3 more when the user determines the brightness, adding up to 6 mouse operations minimum for an arbitrary color.

Department of Computer Sciences Direct questions and comments to the student editorial team