There are many ways that folks interested in the art and science of color — folks like me — test for color aptitude. One way usually involves esteemed institutes of knowledge and quantifiable standards — the Farnsworth-Munsell color discrimination test, for example. Yet another approach is a little more experimental, inspired by sheer curiousity and a “what-if” attitude. The Kristi Test clearly falls in the latter camp. The results of each can be equally fascinating.
The Discriminating Color Scientist
The Farnsworth-Munsell color discrimination test was created by Dean Farnsworth, a color scientist who worked for the Munsell Division of the Kollmorgen Corporation in the 1940s and ’50s (Munsell is now a division of GretagMacbeth.) His work took him in a number of directions, one of which was the study of human visual perception and color recognition.
Also called the Farnsworth-Munsell 100 Hue Test, the test is a fascinating device as well as a methodology for measuring the ability of an individual to see the differences between colors. The Farnsworth-Munsell test will certainly show color blindness when it exists, but its purpose is to measure more-subtle perceptual capabilities — the ability to discern color differences and put colors in order. The test can be learned, and those who take it more than once usually earn better scores.
The test comprises four wooden boxes containing a total of 100 small cylinders, each of which has a colored cap. The idea is to arrange the caps in color order, from pale green to blue, or green to yellow.
The person being tested is given as long as 10 minutes to make the arrangement of cylinders in each of the four boxes of 25. When complete, the box is closed, and sorting begins with another. When the cylinders in each box have been sorted, the administrator turns the box upside-down and reads the numbers of the cylinders through a slot in the bottom of the box. Color transpositions are noted on a special chart, and a score is assigned for the completed test.
Each simple transposition is counted as 4 points, and a “perfect” score is 2. Colors out of order by more than one position take on a higher point value. The higher the score, the poorer your color discrimination.
According to Farnsworth’s research, about 16 percent of the population will have very good color discrimination, with scores in the range of 0 to 20. Another 16 percent of the population will score quite badly on the test, with scores in the 100-plus range. The majority of the population will have scores in the 20-to-100 range.
Students Against Color Discrimination!
I had a chance to put the Munsell-Farnsworth test through its paces earlier this year when I taught color management in the Graphic Communication Department at Cal Poly University, San Luis Obispo. I had 41 eager students in my class, some of whom are involved in extracurricular activities such as working on the daily “Mustang” newspaper and at University Graphic Systems, where some of the university’s printing is done. In the 10 weeks of the course, we had an opportunity to profile the behavior of the university’s Goss newspaper press and its 4-color Heidelberg sheet-fed machine.
I required that each student take the Farnsworth-Munsell color discrimination test as a part of my class. I was interested in having the students experience it, and with the results I hoped to show how color discrimination affects color quality in prepress.
The students paired up, and over a few weeks they all took the test (the procedure requires a student and an administrator). Results were recorded on the official Farnsworth-Munsell test sheets. I had an opportunity to evaluate the results, and I enjoyed a surprise with these 41 students: All but one scored very well on the test.
Only two students had transpositions of more than two colors, and seven of 41 had perfect scores. Statistically, 87 percent of the group scored better than 74 percent of the general population would, with scores of 16 or lower. The one student whose score was high (20) came to me after the midterm and sheepishly told me that he had gone out drinking the night before taking the color test! Though he offered to take it again (and it would be very interesting to see the results), he has yet to do so. He is still enrolled, so I might have a chance to follow up on this aspect of the test procedure.
Pulling an All-Nighter?
Having a body of color data like this is a real treat for me. It’s difficult to line up 40 people to take a color test. I did find on the Internet an instance where the test had been given to a general population of Thai citizens — 421 men and women — and the average of their scores was 47.7. In the Thai test, at least, older people scored worse than younger people, and men scored better than women — all interesting statistics.
So having students who were willing and capable was a real bounty. But, curiously, they exceeded the “average” by such a significant amount that I questioned the validity of my body of data. Could it be that Cal Poly students in Graphic Communication have superior color vision? Is is possible that my students “crammed” for the test and somehow skewed the results? Is it possible that they had taken the test before?
Indeed, when I asked, about half of the students said they had taken the test in other classes, and thus were familiar with it and its scoring method. Farnsworth himself indicated that repeated taking of the test yields better scores. I think also that my students, being fourth-year students of graphic arts, have color in their blood. They work with it every academic day. They design and build documents, then prepare and print them as a routine part of their Cal Poly education. Thus, they have a familiarity with color that the general population lacks.
Introducing the Kristi Test
After the first week of class, when I had announced that they would each take the Farnsworth test, one of my students raised her hand and asked a question that stumped me. Kristi Murray, a graduating senior with many talents and skills to her credit, raised her hand and said: “I wear colored contact lenses. What impact will that have on the test?”
Her question really threw me, for I had never considered it, read about it, or heard anyone mention it (but I would be naive to assume that it has never been discussed). I did not know the answer to her question, but I assumed that her blue-colored contact lenses would filter the light reaching her eyes and cause a variation in her perception of colors on the test. (Kristi is brown-eyed, and when I asked her why she wears blue contacts, her answer was simple: “So I can see them in water.”)
Kristi was more than willing to be a human guinea pig. I had her take the test with her contacts in (and after having them in all day so she was well-adjusted to them). She aced the test, with only one transposition of two cylinders — numbers 28 and 29, which are pale yellow-green in color. (See Kristi’s test results.)
About a week later, Kristi removed her contact lenses and took the test again. Once more, she aced the test with only one transposition of colors, this time missing cylinders 41 and 42, which are pale blue-green in color.
What was going on here? Her Focus Monthly Toric Visitint contacts are made by Ciba Vision Corporation of Duluth, Georgia, whose product literature, not surprisingly, does not describe adverse effects on color discrimination. Upon closer inspection, I noted that the color of the contacts is constant, not gradient, so the color covers her pupil rather than leaving a clear area. The contacts unquestionably act as blue filters to all the light reaching her eyes. In fact Ciba’s literature describes the lenses as having a 93-percent transmittance, which means that the lenses and their colorant absorb a small amount of light as it passes through.
Complicating the results of the “Kristi Test” was another factor: Of the 35 students who took the test and made any transpositions, 18 of them (51 percent) transposed cylinders 28 and 29, just as Kristi had in her first, contact-enhanced test! Looking at the distribution of other transpositions, this is evidence of something wrong, because no other pair of cylinders shows up more than twice. It may be that these cylinders are not clean, that they have faded, or that the lighting in the controlled lighting booth where the tests were given has a quality that fails in the range of those two colors.
It’s hard for me to accept the possibility that 51 percent of the students have the same color discrimination weakness. This is statistically unlikely. Obviously more study is needed, and I will have a chance when I teach the class again in January. On the next round, I think I will conduct the test in two different lighting booths by splitting the class population into groups to determine if one light booth is the source of this anomaly. Farnsworth also makes a suggestion for conducting the test in north-source daylight, which might also be possible.
If I assume that the 28-29 transposition is an error of lighting or another factor, then Kristi scored perfectly with her contact lenses in, and with just one transposition of colors with them out. Since she removed the lenses just before taking the test the second time, her vision may have been skewed toward the complement of the lenses’ color (yellow), which might account for an error in discrimination of blue-green shades in cylinders 41 and 42. This is certainly explainable, and to prove it I must ask Kristi to take the test again after not wearing her contacts for a longer period of time.
Unfortunately for my testing, Ms. Murray is now at work in her new profession in San Francisco, and I will have to track her down to have her take the test again under different circumstances. Or else I will have to find another student who wears tinted contact lenses.
Kristi Murray’s discrimination of colors is extraordinarily good, and she will have no trouble in the execution of her work as a print buyer. I am impressed by the tests she and her classmates took, and will have bragging rights about my “superior” students when it comes to color vision and discrimination.
And, I’ll report back later on the effect a hangover has on color discrimination, once I have a chance to gather more information on that.