DOCUMENT RESUME ED 370 559 IR 055 012 AUTHOR Pett, Dennis TITLE White Letters on Colored Backgrounds: Legibility and Preference. PUB DATE 93 NOTE 6p.; In: Visual Literacy in the Digital Age: Selected Readings from the Annual Conference of the International Visual Literacy Association (25th, Rochester, New York, October 13-17, 1993); see IR 055 055. PUB TYPE Reports - Evaluative/Feasibility (142) -- Speeches/Conference Papers (150) EDRS PRICE MF01/PC01 Plus Postage. DESCRIPTORS *Color; Slides; Visual Learning; *Visual Literacy; *Visual Perception; *Visual Stimuli ABSTRACT This paper reviews two studies of color preference and the relationships between color and legibility. The Gustin study in 1991 dealt with the legibility of and preference for projected slides with colored backgrounds and white text. The order of background color preference was cyan, blue, green, yellow, magenta, and red. The follow-up study by Cuttill in 1991 focused on the same variables viewed on a cathode ray tube. The order of background color preference was blue, cyan, magenta, red, green, and yellow. Based on the overall results of these studies, letter size and adequate contrast between lettering and background are the two most important criteria for producing materials with white lettering on colored backgrounds. (Contains 6 references.) (JLB) *********************************************************************** Reproductions supplied by EDRS are the best that can be made from the original document. *********************************************************************** U.S. DEPARTMENT Of EDUCATION Office ot Educational Research end Impronnuent EDUCATIONAL RESOURCES INFORMATION CENTER (ERIC) 0 INS document has been reproduced as mowed from the person or organization originating it 0 Minor chenpos hav been made to improve reproduction duality Points of view or opinions statod sn this docu- ment do not necesunly represent official 0E91 position or policy White Letters on Colored Backgrounds: Legibility and Preference by Dennis Pett RR 2, Box 120 Brattleboro, VT 05301 "PERMISSION TO REPRODUCE THIS MATERIAL HAS BEEN GRANTED BY Alice Walker 2 TO THE EDUCATIONAL RESOURCES BEST COPY AVAILABLE INFORMATION CENTER (ERIC)." White Letters on Colored Backgrounds: Legibility and Preference Dennis Pett Introduction There have been numerous studies of color lettering and background. These dif- preference and the relationships between ferences were not indicated. Snowberg color and legibility that have been reported also reported color preferences which, in the literature. Although there is evi- from most to least preferred, were blue, dence that color preferences change with green, red, and yellow. These preferences age and are influenced by cultural dif- were based on viewing colored chips. ferences and individual characteristics, studies show that cool colors are generally Sanner (1973) conducted a study using preferred. It is difficult to compare color blazk letters on colored backgrounds. The preference studies because methodologies order reported from most legible to least have varied and the colors used have not legible was white, green, red, yellow, and been adequately specified ( De Santis and blue. Pett, 1980). In recent years, white letters on colored In addition, it is important to recognize the backgrounds have teen increasingly used fact that color preferences, as measured in for slides; however, no studies of legibility most controlled research studies, may not or prefeLence for this configuration have be related to the preferences that would be been reported. Similarly, no studies of indicated in real life conditions whete white letters on colored backgrounds colors are seen in a context. A color that viewed on a cathode ray tube (CRT) have is preferred in one situation may not, and been reported. A study completed by probably would not, be preferred in Gustin in 1991 dealt with legibility of, and another situation. preference for, projected slides with colored backgrounds and white text. A In general, studies have tended to show follow-up study by Cuttill in 1991 focused that acuity is greatest for colors in the on the same variables viewed on a CRT. middle of the spectrum. Gustin Study However, most studies did not deal with projected images. Two studies dealing Fifty adults with normal color vision with projected images were found in the participated in the experiment To test literature. Snowberg (1971) reported v that legibility, slides with white lettering on when black letters were used on colored red, green, blue, cyan, magenta, and backgrounds the order of legibility from yellow backgrounds with carefully most to least was white, yellow, green, controlled low, medium, and high red, and blue. These results probably brightness levels were produced on a slide relate to brightness differences between duplicator (Figure 1). It should be noted Subjects viewed the slides and responded that yellow of a low brightness would by reading the letters from left to right and usually be called brown. The colored from top down. The top six slides, in backgrounds were produced by using order of legibility, are noted in Figure 3. sharp cutting primary color filters, A graph of the interactions between hues red #25, green #58, blue #47B, cyan and background brightness levels is #44A, magenta #32, and yellow #12. shown in Figure 4. 4Hue, Slides B.G. I Density j Legibility Rankings .--Hue, B.G. c .44/.45 I Y1 1 [-hi] Ili] I B I R Density+ 1 ri rn .44/.45 R 1 I R I B NI I .95/.97 I I 171 nil P51 FI:1-1 4 5 2 3 6 roi Fin R1 .95/.97 G I R 1.67/1.69 G 1 1 Y I Figure 1 67/1.69 rf71] Lettering was produced on a Macintosh Figure 3 computer using the bold Helvetica typeface in the MacDraw software program. The selected letters were limited to the ten letters that make up the Sloan test letters LOW IMMO:LW Ofticinti55 LE VEL 35 '''' le RIDIUM0ACKOROU1130111811111(55 LEVU (Sloan, 1951). From the viewing 711011134(701131.1430111014DESS LEVEL 31.72 position, the letter sizes corresponded to 31 46 1 rim 31.42 30 90 31.10 .... 1 1 30,3 20/30, 20/20, 20/15, 20/10, and 20/7.5. I 1 t ..-.. 30 The position of each letter on each slide 28 00 20.16 ING11 I I t was randomized. Figure 2 shows the 29 24 26.78 29 00 29.20 lettering of a sample slide. LOW 1 2590 25.90 25.60 251.52 25 16 ZOHKVDNRS 24.72 D V K HNR OC S YELLOW MAGENTA EUX CYAN GREEN RED RE RE NJE ILE V I-INCRZ 0 D S Figure 4 II fl 0 K C D N S Z Hu zsutt The numbers represent the average number G of correct responses for each of the 18 cells (six hues and three brightness levels). A medium background density was Figure 2 significantly better than either a high or low density and a high background density was significantly better than low density in facilitating legibility. Yellow, blue, green, cyan, and magenta at medium background 4 densities were significantly better in significantly better than cyan and green, facilitating legibility than any hue at low or and red was significantly better than green. high background densities. Across the three background densities, yellow and cyan were significantly better in facilitating legibility than red or blue, and green was C. R. T. 1 1 significantly better than blue. Legibility Rankings 4Hue B.G. To teEt color preference, slides were I Brightness produced with split halves of R, G, B, 4 C, M, Y. This provided 15 pairs of Idl High colors. Each pair was produced in two R G I M 1 versions with the right-left position alternated for a total of 30 pairs. All slides 5 3 1 s were produced with the medium bright- c I Medium LGI 1 ness range used in the legibility portion of the study. Subjects viewed the pmjected m 2 6 pairs and reported their preference for one fdl P1 tml Low 101 color in each pair. The order of back- ground color preference was cyan, blue, green, yellow, magenta, and red (Figure 5). Cyan and blue were significantly Figure 6 preferred ever yellow, magenta, and red. Green was significantly preferred over The order of background color preference red. was blue, cyan, magenta, red, green, and yellow. Blue and cyan were significantly preferred over red, green, and yellow. Magenta was significantly preferred over Slides green and yellow (Figure 7). 4 Hue 0. B. G. Density [C. R .T. IGI a .95/.97 4-- H ue-0. B.G. Preference 6 3 2 5 4 Brightness Rankings 1 Idl ImI B I Medium EFT 1 Preference Figure 5 4 2 5 6 3 1 Rankings Cuttill Study The Cuttill (1991) study followed Figure 7 procedures of the Gushn (1991) study. Background colors were adjusted to Observations match, as nearly as possible, the colors used by Gustin. These would vary It is interesting to note that white letters on slightly, depending on the monitor used. a yellow background of medium density The six colors that ranked best in faci- ranked most legible in both slide and CRT litating legibility are shown in Figure 6. formats. However, the yellow back- Across the three brightness levels, ground was significantly less preferred magenta, blue, and yellow were than cyan or blue in either format. . w . of these In looking at the overall results On slides, all six hues were more legible said that letter size two studies, it can be on medium density backgrounds than on lettering and adequate contrast between backgrounds of high or low brightness, and background are the two most while on a CRT lettering on medium and important criteria for producing materials low brightness backgrounds was sig- with white lettering on colored nificantly more legible than lettering on a backgrounds. background of high brightness. This suggests that producers should be careful There seems to be no adequate explanation to maintain medium to high contrast for the differences reported in the Gustin between lettering and background while (1991) and Cuttill (1991) studies. avoiding extreme contrast for slides. This Although the hues were slightly different would also be helpful for color-deficient in the two studies, these differences would viewers who depend on value differences not seem to be sufficient to cause either rather than hue differences to differentiate preference or legibility choices to differ. between lettering and its background. Replication of the studies might shed light on the cause of the differences. It is also interesting to note that for both slides and CRT, the preferred background colors were blue and cyan. Green and cyan ranked high for both prefecence and legibility as background colors for white lettering on slides, but ranked low for legibility on a CRT. References Sanner, J. W. (1973). The relationship of on The effects of color Cuttill, W. (1991). color to the legibility of projected high legibility and preference with video contrast negatives. Unpublished doctoral computer displayed text. Unpublished dissertation, Indiana University, manuscript, Indiana University, Center Bloomington, IN. for Media and Teaching Resources, Bloomington, IN. Sloan, L. L. (1951, June). Measurement of visual acuity. Archives of DeSantis, L, & Pett, D. (1980, October). Opthalmology, 45, 704-725. Toward an understanding of color. Visual Literacy Newsletter. Snowberg, R. L. (1971). The relationship benveen color preftrence and readability Gustin, K. W. (1991). The effects of of pmjeaed black characters with a color on legibility and prrftrence with colored background under conditions of projected negative ter slides. controlled luminance and _transmission. Unpublished doctoral dissertation, Indiana Unpublished doctoral dissertation, Indiana University, Bloomington, IN. University, Bloomington, IN.