Infant Color Vision Research Articles

Infant color vision and color preferences: A tribute to Davida Teller

Almost 40 years ago, Davida Teller developed the forced-choice preferential looking method for studying infant visual capabilities and used it to study infant color vision. About 10 years ago, she used infant looking preferences to study infant color perception. Here, we examine four data sets in which the infant looking preference was measured using a wide range of saturated colors. Three of those data sets, from papers by Marc Bornstein and by Davida Teller and Anna Franklin and their respective collaborators, were fit successfully using MacLeod and Boynton's model of the equiluminant plane in color space, in spite of the varied luminances used in those studies. A fourth data set, from a paper by Zemach, Chang, and Teller, was less well fit by that model. Apparently, infants are able to ignore luminance, and pay attention just to the color of stimuli. These results are discussed in the context of Davida Teller's work on the philosophy of vision science.

Infant Color Vision: Spontaneous preferences versus novelty preferences as indicators of chromatic discrimination among suprathreshold stimuli

Infant Color Vision: Spontaneous preferences versus novelty preferences as indicators of chromatic discrimination among suprathreshold stimuli

Color vision in children and the Lanthony New Color Test

Much is known about color vision in infants, adolescents, and adults, but very few studies report the changes, which occur in color perception of children in their early schooling years. There is also a shortage of suitable color vision tests for children. This study investigated the changes in color vision of school students between 5-12 years old using the Lanthony New Color Test (NCT). Subjects of all ages were able to complete a shortened form of this test adequately. The Vingrys and King-Smith (1988) method of panel test analysis and Adams and Rodic (1982) color confusion score were adapted to analyze their performance of the test. This study confirmed that there are changes in color perception occurring in this age group. Color perception abilities increased as a function of age and there was also an improvement in the performance on the NCT with age. This can be attributed to both cognitive development and changes occurring to the color vision system.

Infant color vision: Infants’ spontaneous color preferences are well behaved

Stochastic transitivity (ST) is a property of preferences among pairs of objects formed from three alternatives, a, b, and c. In general, ST states that if a is preferred to b, and b is preferred to c, then a will be preferred to c. Stochastic transitivity can be weak, moderate, strong or strict (see text). In the present paper, we analyse the presence and degree of ST in the data from two experiments concerning 12-week-old infants’ spontaneous color preferences. In the first experiment (Triads), we tested five sets of three stimuli in pairs of two (a vs. b, b vs. c, a vs. c). In each case two stimuli were chromatic and one was White. Strict ST was seen in all cases. In Experiment 2 (Complementaries), we tested White against pairs of stimuli from opposite sides of the White point (red vs. blue-green, blue vs. yellow, and green vs. purple). The purities required for equal (50/50) preference between the two chromatic stimuli were consistent with the preferences for each of the two stimuli over White. In addition, 12 new triads were generated from the Complementaries experiment. Strict ST was seen in six out of 12 cases, and Moderate ST was seen in the other six. As discussed further in the accompanying paper [Zemach, I. K., Chang, S., & Teller, D. Y. (2007). Infant color vision: prediction of infants’ spontaneous color preferences], White was the least preferred stimulus in every triad tested. Although more extensive studies are needed, the data suggest that infants’ hue preferences are reasonably well behaved across different choices of stimulus pairs.

Infant color vision: Influence of surround chromaticity on spontaneous looking preferences

When infants are tested with stimuli of various chromaticities embedded in a dark or achromatic (white) surround, they show maximal preference for stimuli of maximal colorimetric purity, and minimal preference for achromatic stimuli. We investigated how this pattern of preferences changes with changes of surround chromaticity. Sixteen-week-old infants were tested in two experimental conditions. The surrounds in the first condition were red and white; and in the second condition green and white. The three test stimuli varied in colorimetric purity from white to red in the first condition, and from white to green in the second condition. A test stimulus that appeared achromatic to adults when viewed in the chromatic surround was included. Infant spontaneous looking preferences changed with changes of surround chromaticity. The changes were consistent with the conclusion that infant looking behavior is governed by a preference for the stimuli that differ maximally in purity from the surround. The implications of this pattern of results are discussed.

Infant photometry: are mean adult isoluminance values a sufficient approximation to individual infant values?

Individual differences in isoluminance values were studied in infants and adults using a motion nulling paradigm. Two luminance-modulated sinusoidal grating components (spatial frequency=0.25 cpd, temporal frequency=5.6 Hz, speed=22.4 deg/s) were superimposed and moved in opposite directions across a color video screen. The contrasts of the two components were traded off to determine motion nulls. Two conditions were used: red/black vs. green/black, and red/black vs. blue/black grating components. An eye movement based response measure was used for infant subjects, and an average of 308 trials per infant were obtained. As observed in earlier studies, the mean motion null values for infants and adults were highly similar in each condition. The standard errors of motion null values for individual subjects were very small. Individual differences among infants were also small, and were clearly measurable only in the red/black vs. blue/black condition. The close similarity of mean null values, combined with the small individual differences among infants, supports the idea that under the right circumstances mean adult isoluminance values can be used as a sufficient approximation to individual infant isoluminance values in studies of infant color vision. These circumstances are discussed and evaluated in detail.

Infant color vision: sharp chromatic edges are not required for chromatic discrimination in 4-month-olds

In our previous demonstrations of chromatic discrimination in infants, we have used test and surround fields of different chromaticities that abutted each other at sharp chromatic edges. In order to see whether sharp chromatic edges are necessary for infants to make chromatic discriminations, 16-week-old infants were tested with three stimulus configurations in which sharp chromatic edges were eliminated. The three edge manipulations involved black borders, a dark surround, or blurred edges around the chromatic test field. In each case red, green, and violet test fields were used. Although performance decreased when sharp chromatic edges were eliminated, observers’ percent correct scores remained clearly above chance for eight of the nine discriminations (three colors×three edge manipulations). We argue that all three edge manipulations reduce the likelihood of mediation of chromatic discrimination by M (magnocellular) cells. These data thus provide evidence that young infants have functional P (parvocellular) pathways, and use them for making chromatic discriminations.

M- and L-cones in early infancy: III. Comparison of genotypic and phenotypic markers of color vision in infants and adults

Genetic analyses were performed on five male children (∼3 years), two suspect color-normals and three suspects for congenital color vision deficiencies. These classifications were based on visually-evoked potential (VEP) responses to M- and L-cone-isolating stimuli obtained in a previous study when each subject was either 4- or 8-weeks old. The present analyses were performed in a blind study to characterize the genotypes of these subjects. Four male adults with various color vision phenotypes were also tested as a control. DNA was isolated using a non-invasive technique followed by polymerase chain reaction (PCR) amplification and restriction enzyme analysis to examine the genomic DNA of each subject. The genetic analyses confirmed the VEP identification of two color defective infants, and were consistent with the diagnosis of two other infants as color normal. A third infant was predicted by VEP analysis to have a protan defect, but he did not have a gene array typically found in protan observers.

Spatial and temporal aspects of infant color vision

The present paper constitutes a review of the literature on young infants’ chromatic discrimination capabilities. A series of early studies showed that infants as young as two months postnatal can make at least some chromatic discriminations between stationary, homogeneous fields of different wavelength compositions. Current studies of spatial and temporal contrast sensitivity functions (CSFs) for red/green isoluminant stimuli suggest that spatial chromatic CSFs show developmental changes in sensitivity and spatial scale, but not curve shape; while temporal chromatic CSFs (tCSFs) show developmental changes in sensitivity and curve shape, but not temporal scale. Infants can also code the direction of motion of moving isoluminant red/green gratings, for both continuous and quadrature motion. The possible mechanisms that underlie infants’ chromatic discriminations are discussed.

A Modified Card Procedure for Measuring Human Infant Color Vision

To improve test efficiency, we modified our previously introduced contrast/color card test by including a patterned test stimulus and reducing the number of stimuli in both experimental phases. Compared with the prototype, completion rate improved substantially (79 vs. 37%) but test time decreased only modestly (19 vs. 21 min). Achromatic contrast discrimination improved threefold (mean, 0.06 vs. 0.20 log units), but the percentage of 2-month-old infants who discriminated (from gray) 660-nm red (86 vs. 80%) and 580-nm yellow (52 vs. 55%) was consistent. In addition, 48% discriminated 574-nm yellow-green. Moreover, because 88% of infants' failures included the respective adult brightness/luminance match, a small range of relative luminances is adequate for testing infant color vision.

Infant color vision: Temporal contrast sensitivity functions for chromatic (red/green) stimuli in 3-month-olds

In order to investigate the development of temporal contrast sensitivity functions (tCSFs) for chromatic (red/green) stimuli, we obtained chromatic contrast thresholds from 3-month-old infants and adults using behavioral techniques. Stimuli were moving or counterphase-reversing sinusoidal gratings of 0.25 c/deg. Five temporal frequencies were used: 0.7, 2.1, 5.6, 11 and 17 Hz (corresponding speeds = 2.8, 8.4, 22, 44 and 67 deg/sec). In order to compare chromatic results with those obtained under luminance-defined conditions, luminance tCSFs were also obtained from adults, and previously obtained infant luminance tCSFs were used (from Dobkins & Teller, 1996a). In accordance with previous studies, adults exhibited bandpass luminance tCSFs with peaks near 5 Hz and lowpass chromatic tCSFs that declined rapidly at temporal frequencies greater than 2 Hz, and the two curves crossed one another near 4 Hz. By contrast, infants exhibited bandpass rather than lowpass chromatic tCSFs with peaks near 5 Hz. These chromatic curves were quite similar in peak frequency and general shape to previously obtained infant tCSFs for luminance stimuli. Moreover, both chromatic and luminance tCSFs in infants were found to be quite similar in peak and shape to luminance tCSFs observed in adults. These findings point to the possibility that, for 3-month-old infants, both chromatic and luminance stimuli are detected by the same underlying mechanism under these conditions. We propose that such a mechanism is probably a physiological pathway dominated by magnocellular input. Earlier studies of infant color vision are discussed in this context.

Infant Color Vision: Moving Tritan Stimuli do not Elicit Directionally Appropriate Eye Movements in 2- and 4-month-olds

The purpose of the present study was to investigate the capacity of infants to code the direction of motion of moving tritan-modulated gratings. Infant and adult subjects were tested with 0.2 c/d sinusoidal gratings moving at a speed of 20 deg/sec. Three conditions were tested: luminance-modulated gratings, tritan-modulated gratings, and luminance- vs tritan-modulated gratings superimposed and moving in opposite directions in a chromatic motion nulling paradigm. Two-month-old infants were tested in all three conditions, while 4-month-olds were tested in only the first two conditions. For infant subjects, an adult observer reported the direction of the slow phase of the infant's eye movements; adult subjects judged the perceived direction of motion of the stimuli. Luminance-modulated gratings produced directionally appropriate eye movements (DEM) in all age groups. Tritan gratings presented alone did not produce DEM in either 2- or 4-month-olds, but did so in adults. Mean equivalent luminance contrasts were near zero in 2-month-olds, and small but reliably above zero in adults. In sum, the present study provides no evidence that infants can code the direction of motion of moving tritan gratings. © 1997 Elsevier Science Ltd. All rights reserved.

Infant color vision: Motion nulls for red/green vs luminance-modulated stimuli in infants and adults

Four-week-olds, 9-week-olds, and adult subjects were tested with low spatial frequency sinusoidal gratings moving at a speed of 25 deg/sec. Luminance-modulated and red/green gratings were presented either separately, or superimposed and moving in opposite directions in a chromatic motion nulling paradigm. An adult observer judged the direction of the slow phase of the subject's eye movements. Luminance-modulated gratings elicited directionally appropriate eye movements in all three age groups, with contrast thresholds decreasing markedly with age. For red/green gratings alone, 4-week-olds responded only marginally, but 9-week-olds and adults produced consistent directionally appropriate eye movements. In the motion nulling condition, 15 % contrast luminance-modulated gratings were about equally effective in nulling the motion of the red/green gratings in all three age groups. A formal model of the motion nulling paradigm, separating threshold and equivalent luminance contrast parameters, was developed and applied to the data. Model fits showed that equivalent luminance contrast was constant or nearly constant across age groups. This outcome is consistent with the hypothesis that, with respect to adults, infants show a uniform rather than a differential loss of sensitivity to moving red/green vs luminance-modulated stimuli.

Infant luminance and chromatic contrast sensitivity: Optokinetic nystagmus data on 3-month-olds

Infant color vision is poor, and most psychophysical experiments agree that infant color vision emerges between ages 3 weeks and 3 months. Presumably, the color vision of infants is poor during the months immediately after it has emerged. We have tested two alternative explanations for the poor color vision of infants: (1) there is a special critical immaturity within the color pathways of infants; (2) infants have poor infant color vision because they are insensitive to contrast. Luminance and chromatic contrast thresholds were measured on 3-month-olds using optokinetic nystagmus (OKN), and adult luminance and chromatic contrast thresholds were measured using OKN and two forced-choice methods: direction-of-motion discrimination and grating detection. The infant chromatic-to-luminance contrast threshold ratio shows that infants are as sensitive or even more sensitive than adults to color, depending on the testing method used on adults. This result suggests that the general contrast insensitivity hypothesis is correct. Conservative “worst-case” quantitative analysis strongly suggests that this result is not the consequence of a luminance artifact.

Visual function deficits in children.

In the past year, new work on sensory testing in preverbal infants has expanded our knowledge in three areas: 1) visual acuity testing in infants in varying levels of illumination; 2) visual evoked cortical potential-evidence for binocular function in infantile esotropia; and 3) a new time-efficient method to evaluate color vision in newborns and infants. A newly reported sequel to loss of binocularity is the development of A and V pattern strabismus. In another study, loss of fusion, as with acute concomitant esotropia, was shown to be a sign of serious intracranial disease. Regarding motor visual function deficits, new treatments of infantile esotropia syndrome and nystagmus with botulinum toxin type A are presented. Regarding social issues, visual function deficit of abused children and infants born to drug-abusing mothers are presented as a timely reminder that child abuse is still an enormous social problem, and that the visual system is far from exempt from damage.

Contrast/Color Card Procedure: A New Test of Young Infants?? Color Vision

We have developed a new test which can rapidly evaluate basic color vision in individual infants. The test consists of a series of large cards constructed with Munsell Hues. It uses a modified preferential looking procedure (FPL) and, to control brightness cues, incorporates a two-phase systematic variation of luminance. First, we evaluate an infant's ability to discriminate 9.5 by 16 degrees achromatic patches of varying luminance from a 26 by 65 degrees achromatic background of midrange luminance. In the second phase the test patch is chromatic and its luminance, relative to the background, is varied over a range of about 1.0 log cd/m2. The number of relative luminances chosen for each infant depends upon his/her performance in phase 1. Seventy 2- and 3-month-olds were tested with 4 broad-band chromatic patches, a red (dominant lambda = 660 nm), a yellow (dominant lambda = 580 nm), a green (dominant lambda = 520 nm), and a blue (dominant lambda = 475 nm). Results showed that 3-month-olds had little difficulty making any of the chromatic-achromatic discriminations but many 2-month-olds appeared to fail to discriminate the yellow and green from the background at relative luminances close to an adult brightness match. Most importantly, the test shows promise as a relatively simple, time-efficient, and portable tool for the assessment of early color vision.

Development of visual sensitivity to light and color vision in human infants: A critical review

The recent literature on test threshold and color vision in human infants is critically reviewed. Test thresholds are higher in infants than in adults at the absolute threshold and at all adapting luminances, but approach adult values rapidly over the first six months of life. The spectral luminous efficiency function of infants is similar to V (λ) above 1.0 log phot, cd/m 2. For lights below about 1.0 log scot, cd/m 2 and wavelengths shorter than 590 nm, the luminous efficiency function is simlar to V′(λ). The luminous efficiency of any given stimulus may differ markedly between infants and adults, especially when the adult data depend on the temporal and spatial parameters of the stimuli. Color vision improves greatly over the first three postnatal months, and most normal 3-month olds have at least some color vision. The overall insenstivity of infants to contrast is likely to provide a satisfactory explanation of the poor color vision of infants. The critical immaturity primarily responsible for the high thresholds and poor color vision of infants is probably after the site of visual adaptation, although lower-level factors may also play a role.

Unsettled issues in infant color vision

The purpose of this paper is to clarify several issues in infant color vision that, despite previous reviews in this journal, remain in dispute. We argue that the data are still insufficient to conclude that 2-3-month-old infants are adult-like trichromats, but they are at least dichromats. How closely infant hue perception parallels adult hue perception cannot be discerned from existing data.

Human infant color vision and color perception

Werner and Wooten write that the basic premise of their 1979 review was that current evidence did not justify the conclusion that infants (2-3 months of age) possess three cone receptors and the visual apparatus that give rise to normal color vision. They asserted: “It remains possible that infants are dichromats and that one or both of their cone photopigments is anomalous” (p. 262). In 1985, they reiterate their conclusions and repeat commentary on two of my 1976 papers. However, Werner and Wooten ignore 6 years of research (19781984) following the publication of my own 1976 studies that provides compelling evidence that infants even younger than 3 months possess the visual apparatus that gives rise to normal, three-dimensional color vision. Here, I review conclusions we may safely reach today about infant color vision, and refute again specific commentary of my 1976 papers.

Two perspectives on infant color vision research

Two perspectives on infant color vision research