Discussion

The goal of this experiment is to compare four color selection methods in quantitative terms of speed and accuracy, and in qualitative terms of ease of use and preference. The color selection methods were selected for this experiment because of their dependence on different color models.

The Crayon color picker can be considered a 1-dimensional color selection method since only one degree of intereaction (or mouse-click) can determine any color within the domain of the color picker. Although the Crayon picker is limited to 60 discrete colors, this was only because of its implementation. Any of the 60 colors can be selected with a minimum of one click.

The HSV color picker is a 2-dimensional color selection method in that an arbitrary color within its domain can be selected with a minimum of two mouse clicks (disregarding the chance that the color is initially present). Although the HSV model uses three coordinates to describe a color (hue, saturation, and brightness), the three coordinates are presented to the user in terms of a color wheel and a slider. The HSV color wheel allows users to select hue and saturation simultaneously with one mouse-click (by using polar coordinates; hue is determined by the angle, and saturation is determined by the distance from the origin). A slider determines value, or brightness, requiring another mouse click. Manipulating the wheel and the slider result in a total of two mouse clicks.

The RGB color picker is a 3-dimensional color selection method in which the user manipulates three sliders representing the intensity of the values of red, green, and blue. A minimum of three mouse-clicks must be made to select an arbitrary color.

The CMYK color picker is a 4-dimensional color selection method. The user must manipulate four sliders, representing intensities of cyan, magenta, yellow, and black.

We hypothesized that color selection time would be shortest for Crayon, followed by HSV, RGB and CMYK, for two reasons. The first reason is that we believe that dimension of the color picker affects the time to select a color. Intuition leads that performing one mouse operation is faster than 2 mouse operations, 3 mouse operations, and 4 mouse operations. Another reason for our hypothesis on color selection time is the cognitive color calculation that the user must perform to select a color. For example, the Crayon picker would require nearly no speculation to select a color, just finding it. The HSV picker requires the user to relate the hue and saturation with the value (brightness). RGB and CMYK require even more cognitive processes in that each attribute is selected independently (unlike the HSV color wheel where two attributes are selected at once with one mouse operation, or a crayon.), and must require more speculation of the user to predict what values must be selected to get the desired color.

According to our One-way ANOVA test on Color Selection Time, our result for color selection is statistically significant.

Our hypothesis for color selection time was proven to be nearly correct based on the mean color selection times for each treatment. The Crayon color picker had a mean value of 688 ticks (60ths of a second), which is 2.8 times faster than the second fastest HSV picker with a mean of 1939 ticks. CMYK, surprisingly, was the third fastest color selection method, with a mean of 2242 ticks. RGB follows with a mean of 2584 ticks. A possible reason why CMYK was faster than RGB despite being of a higher degree, may be the existence of the black (darkness) slider. This provided direct control (one mouse operation) over the darkness of the selected hue, whereas for RGB, changing the darkness of the current hue requires careful adjusting of all three RGB sliders proportionally (three additional mouse operations).

For color accuracy, or color distance (the distance between the desired and selected color, as represented in 3-dimensional space, as RGB values), we hypothesized that color selection accuracy would be best for Crayon, followed by HSV, RGB, and CMYK. We thought that as the complexity (degree) of selecting a color increased, the accuracy would decrease due to the increased cognitive load and decreased intuitiveness of managing increasing variables to arrive at a focus color value.

According to our One-way ANOVA test on Color Selection Accuracy, our result for color selection is statistically significant.

Our hypothesis, was nearly correct. Like the results for color selection time, Crayon was most accurate with a mean distance of 8.81, followed by HSV with 32.43 which was 3.6 times less accurate than Crayon, CMYK with 38.56, and RGB with 43.11. CMYK was more accurate than RGB, to our surprise. Again, we attribute this to CMYK's black slider, used to control darkness with one mouse operation, without disrupting the current hue. Adjusting the darkness value with RGB requires adjusting all three sliders, while also increasing the possibility of disrupting the current hue.

According to our One-way ANOVA test on Color Selection Accuracy, our results for the subjective survey is statistically significant

The actual test results were somewhat representative of the subjective survey results. The opinions of the subjects on the speed and ease of each treatment was in line with the actual times. The survey has Crayon as the fastest, followed by HSV, CMYK, and RGB, as the quantitative results show.

However the subjective survey did not accurately depict color accuracy as Crayon, CMYK, HSV, and RGB, when quantitative results show it as Crayon, HSV, CMYK, and RGB.

Additionally, all subjects indicated in the pre-experiment survey that they were familiar with using a mouse, more than 75% of the subjects indicated that they were familiar with using a computer graphics program, and nobody indicated that they were color blind.

When referring to the comments made by the subjects on the questionnaire we found that some people preferred the crayon picker because it was the quickest and easiest to use, although other people referred to the HSV picker as being the most accurate. Also other subject comments seem to agree on the fact that the RGB picker is the slowest and hardest picker to use in the experiment

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