Visual symptoms are a common cause of discomfort among computer users.1,2 Visual and physical symptoms among workers should be minimized because they are a source of reduced performance, lost work time, employee turnover, and worker’s compensation costs. Performance at the workplace is commonly quantified by productivity (rate of work done) and quality or accuracy (fraction of work done correctly).
Daum et al.3 demonstrated that even minor vision problems, such as modest amounts of blur due to astigmatic refractive error, could result in lower performance of computer users. Other individual and environmental factors such as presbyopia, font size, and reflective glare may also affect performance. However, the effects of these factors on task performance have not been well quantified.
Many psychophysical studies have demonstrated that larger font sizes, up to a point, are associated with faster reading speeds.4–10 For printed materials, reading speed has an increasing, logarithmic relationship with visual acuity reserve (VAR).11,12 Visual acuity reserve is the ratio of the size of the font to the size where characters can just be accurately detected, that is, the visual acuity (VA) threshold. Words displayed at a VAR of 1 are difficult to read efficiently. If the VAR is 3, they are easier to read. However, there is little gain in reading speed when VAR is more than 4. In these studies of font size and reading speed, the viewing distance (VD) was constrained. It would be useful to verify these findings under more natural conditions where subjects are allowed to select their preferred posture and VD and where the characters are displayed on a computer monitor.
Most users think of font size in points or millimeter height on the screen. However, visual angle of font (VAF), which is estimated as the ratio of the font height to VD in units of arcminutes, has more practical applications. For print, a character size of 12 arcmin (e.g., 1.75 mm at 50 cm VD) is a consensus critical size value for normally sighted readers.13 Recommendations for the range of VAF for normally sighted users fall between 16 and 22 arcmin.14,15 Based on the default setting of the Safari web browser on a MacBook Pro (132 pixels per inch), the average character height of online newsletter text is 10.3 arcmin at 50 cm VD, which is smaller than the recommended values.
Lighting in the office can be an important environmental factor in comfort and workplace productivity. A simulated office lighting experiment found that workers prefer working in offices with windows and overhead, indirect lights compared with a darker room with task lights next to the monitor. In some settings, the demand for energy conservation has led to the replacement of incandescent and fluorescent luminaries with light-emitting diode (LED) lights. It has been suggested that LEDs, compared with fluorescent lamps, produce the same illumination levels with reduced glare, fewer subjective reports of eye fatigue, and higher productivity.16 Still, any light source can produce reflective glare, either localized on the screen or diffused evenly over the whole screen, and has the potential to affect visibility of characters on the screen and productivity. Indirect glare resulting from nearby light sources reflecting off the monitor is a frequent complaint related to computer use.17,18 However, the glare may not be immediately noticed and employees can work for long hours without conscious awareness of the glare. If the glare is strong or localized, it may be disabling (e.g., reduced VA) or may induce discomfort.19
Direct or indirect glare may interfere more with productivity for older than for younger computer workers. Bailey and Bullimore20 reported that older subjects (mean age, 64.9 years) have lower visual performance (i.e., lower VA) and may, therefore, be more susceptible to reduced visibility from glare than younger subjects (mean age, 28.4 years). Sheedy et al.19 reported that older subjects (mean age, 55.5 years) take longer to perform a visual task requiring repetitive transitions between brighter and dimmer areas than younger subjects (mean age, 27.9 years). These studies suggest that older people may have lower productivity on a computer task when viewing a screen with uneven luminance because of glare, but this theory has not been tested for reflective glare.
In addition to the possible higher susceptibility to glare, computer workers older than 40 years have presbyopia, an age-related reduction in the ability of the eyes to adjust focus (accommodate) because of the naturally occurring increase in size and stiffening of the lens of the eye with age. Typically, presbyopia presents as reduced ability to focus on near objects, but it also leads to a limited range of “in focus” VD even when corrected. Because presbyopes have a limited VD over which characters are in focus, their reaction to different font sizes may be different from younger, nonpresbyopes.
This study examines the effects of age, font size, and reflective glare on visual performance during visually intensive computer-based tasks. The null hypothesis is that font size and reflective glare do not affect productivity or accuracy; furthermore, these relationships are not affected by the presence or absence of presbyopia. Subjects were allowed to change their posture during the study because this is the usual situation when someone uses a computer. A change in posture may alter the distance to the screen and, therefore, the VAF, which is considered an intermediate variable in our modeling. The findings may provide guidelines for the setup of computers to improve productivity.
Participants were recruited from the university and community. Potential subjects were eligible if they were 18 to 35 or 55 to 65 years old, have at least 20/20 VA with lenses, have normal contrast sensitivity, and were experienced in using a mouse on a computer. The older, presbyopic subjects wore their usual multifocal or progressive addition lens during the experiment. All of the older subjects used reading lenses with prescriptions of +2.00 diopters (D) ADD or more (two with +2.00 D, five with +2.25 D, and one with +2.50 D). Exclusion criteria were recent musculoskeletal pain or eye diseases. The university institutional review board approved the study and subjects signed an informed consent form.
Participants in the age range of 55 to 65 years were more difficult to recruit; thus, there were unequal numbers of young (n = 19) and older (n = 8) subjects in the study. Normal vision of the participants was confirmed by VA and contrast sensitivity measurements. Visual acuity while seated at the computer workstation was measured at three distances (40, 63, and 100 cm) with the Colenbrander Mixed Contrast Letter Chart. Contrast sensitivity for small letters was measured using the Super Vision Test chart at 4 m (Precision Vision, La Salle, IL). There was no significant difference (p = 0.68) in the contrast sensitivity between the two age groups: the mean log CS for the younger group was 1.30 (SD, 1.06), and for the older group, it was 1.33 (SD, 0.10). Visual acuity for the two age groups was only significantly different at the distance of 40 cm: LogMAR was −0.09 (SD, 0.03) for the younger group and −0.03 (SD, 0.05) for the older group (p = 0.006). Visual acuity at 63 and 100 cm was −0.13 (SD, 0.06) and −0.16 (SD, 0.09) for the younger group and −0.11 (SD, 0.04) and −0.17 (SD, 0.08) for the older group (p = 0.39 and 0.96, respectively). The VA at 63 cm was used for the VAF and VAR calculation.
Experimental Design and Procedures
This was a full-factorial, repeated-measures experiment with two factors, font sizes and glare. Three levels of Arial font size were tested: small (S), 1.78 mm (capital letter height on screen, 8 pt); medium (M), 2.23 mm (10 pt); and large (L), 3.56 mm (16 pt). Small-medium, medium-large, and small-large font differences are 0.1, 0.2, and 0.3 log units apart, which is 1, 2, and 3 lines on the VA chart, respectively. The two levels of glare were with (G) and without (N) area glare produced by LED task light reflected off a matte liquid crystal display monitor. The order of testing for the six test conditions was randomized using a computer-based random number generator.
For each of the six test conditions, subjects performed three different text-based visual tasks that were similar to the tasks used in other research studies that evaluated visual demands with minimal cognitive demands. The tasks, presented as discrete paragraphs on separate screen pages, were a visual search task (V) and two matching tasks, the abridged MN task (A) and the hard MN task (H) (Fig. 1). The visual search task required subjects to identify the spaces, in a block of nonsense words, for which the same letter was on both sides of the space. The two matching tasks required remembering a template nonsense “word” consisting of some combination of the letters “M” and “N” at the beginning of a row and identifying the matching “word” from four options somewhere in the row. The average “word” lengths for the abridged and the hard MN tasks were three and five letters, respectively. Each block of text shown occupied an area approximately 16 cm by 5 cm at the center of the monitor. The visual search task produced a 50% matching letter rate for each page with an average “word” length of five letters.21,22 The visual search task was similar to the “manuscript editing” task of Daum et al.3 whereas the other two matching selection tasks were similar to their “county population listing” and “m and n word search” tasks. The tasks’ visual demands are similar to some Internet search and spreadsheet activities. The duration for each test condition (i.e., a task trial) was 7 minutes. After a session of three task trials (21 minutes), a 15-minute break was provided. Each test session started with a 5-minute baseline trial (performing the visual search task with the largest font without the presence of glare). After each test condition, a question was presented on subjective difficulty.
The experiment was controlled by a custom program (LabVIEW7, National Instruments) that generated the random trial order, controlled trial timing, displayed task images, and recorded the questionnaire responses. The task pages were saved as a screenshot file (.png) and imported into LabVIEW. Click locations and timestamps were recorded by LabVIEW and later analyzed in MATLAB. A correct click had to be within two pixels of the edge of the correct space. No feedback on accuracy was given during the experimental session; feedback was given only in a practice session to ensure familiarity with the precision requirements of the task. The subjects were instructed to right click to initiate the next paragraph of text and to work as rapidly and as accurately as possible. They were informed that their actions were recorded to calculate productivity and accuracy.
The tasks were performed under average office lighting (∼500 lux on the work surface, LX-1108-CAL CERT Lux/Footcandle Light Meter, D.A.S. Distribution, Inc) with no windows or extraneous sources of glare. The monitor was a 20-inch liquid crystal display flat-panel, matte surface monitor (Model 2007FPb, Dell) with a graphics card (Matrox Millennium G450 Dual Head LX) set at analog quality, 24-bit color, 1600 × 1200 resolution, and a 60-Hz refresh rate. The monitor was mounted on a stand that allowed height and tilt adjustments. The chair (Leap; Steelcase, Grand Rapids, MI) was adjusted such that the subject’s feet were on the ground and the elbows were supported on the chair armrests. The mouse was located to minimize shoulder flexion and abduction. The backrest was adjusted to a slightly reclined position. The monitor was positioned for an initial VD (eye-to-monitor center) of 63 cm and the monitor was positioned such that the vertical gaze angle to the center was 15 degrees below the horizon from the eyes. During the experiment, subjects could change the chair position and their posture but could not change the chair height or the monitor location or tilt.
The source of reflective glare was a luminaire with 72 LEDs (21.5 cm by 5.6 cm, Shanghai Raylin International Trade Co, Ltd, Shanghai, China; 5 mm round type; emitting white color; viewing angle, 15 degrees; luminous intensity, 13,500 millicandelas). The location of the luminaire and monitor was arranged to maximize the perceived glare in the central region of the monitor where the text was located (Fig. 2).
Luminance from the monitor with and without glare was measured with a photometer (CS-100, Minolta) for five 20-mm black circles (L min) and their adjacent white areas (L max) located at the center of the screen and the four corners of the task area. The measurements were made at 40 and 67 cm from the monitor at 0 and 15 degrees to the line of sight with auxiliary lenses for focal adjustments. The Michelson luminance contrast ratios were calculated (L max − L min) / (L max + L min) as were the percent reductions in contrast ratios due to glare. At the center of the monitor (at 0 degrees to the line of sight, where the reflective glare was strongest), the average contrast ratio was reduced from 0.83 to 0.28 (67.5% reduction) by the glare. Without the glare, the average luminance of the monitor was 103 candelas (cd)/m2; with glare, the luminance increased to 188 cd/m2, whereas discernable level of glare was 40 cd/m2.23 The four corner areas had a contrast reduction of less than 0.25 (30%) and little (<10 cd/m2) increase in luminance. The contrast ratio reduction was similar across the two distances from the monitor.
The three primary outcome measures were productivity, accuracy, and subjective ratings of task difficulty. For each test condition, productivity was calculated as the number of correct clicks per minute. Accuracy was the percentage of clicks that were correct clicks (i.e., the number of correct clicks divided by the total number of clicks). After each trial, subjects rated the task difficulty using a visual analog scale slider displayed on the screen with verbal anchors “Very Easy” and “Very Hard” on either end of the scale in response to the question “How Easy/Hard was the task?” As the font size decreased, the size of the pointing target decreased, and, therefore, according to the Fitt law, productivity might be influenced by the motor demands of the task. However, as the pointing target size decreased, the distance between the targets decreased proportionally. Therefore, for hand motor control, the mean time to move between targets should not be influenced by font size.
A secondary outcome measure was VD, which was continuously recorded at 30 Hz using an active motion capture system with two monitor banks (OptoTrak 3020; Northern Digital, Ontario, Canada). A cluster of three infrared-emitting diodes was secured just anterior to the subjects’ ear near the tragus. This was used to identify the representative point for the eyes, that is, the point on the ridge just above the nose between the two eyes. The calculated distance between the representative eye point and the center of the monitor was the VD.
The intermediate variables were the VAF and VAR. Visual acuity reserve is the ratio of VAF to the subject’s VA. Visual angle of font was estimated using:
Visual acuity reserve is the ratio of VAF to the subject’s VA. Because a 20/20 letter subtends 5 arcmin, the letter height at the VA of 20/X is X/4 arcmin. Therefore, the VAR is:
In our study, we used the VA at 40 cm for the analysis.
Differences in outcome measures between test conditions were initially evaluated by repeated-measures ANOVA (RANOVA) using the framework of general linear models with the PROC GLM command in SAS 9.0. The productivity outcome measures were averaged across the three tasks because there were no consistent effects of task on the main effects of font size and glare on the primary outcomes. Age was used as a grouping factor, whereas font and glare were trial factors. The model included the independent effect of each factor and the two-way interaction terms between these factors. Factors and interactions with significant F tests for fixed effects were followed up with the Tukey test for multiple comparisons. Font size was treated as a categorical variable. To investigate the effects of font size as a continuous variable, the dichotomous glare variable, and VAF and VAR on productivity, random-effects general linear models were used (xtreg command, account for repeated measures; STATA 10). The font-by-glare interaction term was dropped from the final model because the coefficient was not statistically significant and it did not improve the fit statistics.
There were no significant differences between the younger and older groups on productivity (p = 0.40), accuracy (p = 0.59), or perceived task difficulty (p = 0.93); hence, this grouping factor (e.g., age/presbyopia) was removed from the remaining analyses.
Font Size and Glare Effects on Task Performance
Font size had significant effects on productivity, accuracy, and perceived task difficulty (Table 1). Both speed and accuracy were improved with the large font condition compared with the two smaller font sizes. Tasks were perceived as easier with larger font sizes than the smaller font sizes. Neither glare by itself nor the font size-by-glare interaction term had significant effects on the performance measures or perceived task difficulty.
Font Size and Glare Effects on the Intermediate Variables: VD, VAF, and VAR
The VD for older subjects was greater than that for younger subjects but the difference was not significant (52.9 vs. 47.4 cm; p = 0.11). Subjects significantly reduced their VD when the font was smaller or the glare was present (Table 2). The reduced VD modified the effects of font size and glare on the intermediate variables VAF and VAR by definition. There was no significant effect of the font size-by-glare interaction term. Based on the Tukey follow-up tests, the effect of font size on VD was significantly different between each of the font size comparison pairs, indicating that the smallest font size difference that will affect these outcome measures is 0.1 log units, which corresponds to one line difference on a VA chart. Comparing the actual to the hypothetical VAFs (e.g., based on the subjects’ reduced VD at the smaller font conditions vs. hypothetical situation where the VD was not changed from the large font size condition), the actual VAFs were larger when the font was smaller (Fig. 3). For the medium and large font size conditions, the actual VAFs were greater than the ISO minimum recommendation of 16 arcmin (International Organization for Standardization, 1992, 2011).15,24 The relationship between productivity and VAF is similar to that of productivity and font size, where productivity was significantly higher in the largest font condition (Table 1). Although glare had no effect on productivity or accuracy, it did lead to a shorter VD and, therefore, an increased VAF and VAR.
To determine which font-related factor was a better predictor of performance outcome, general linear models were used to explore the relationships between font size, VAF, and VAR and the outcome measures productivity, accuracy, and perceived task difficulty (Table 3). Font size, VAF, and VAR were better predictors of productivity than they were of accuracy and perceived task difficulty. More of the variance in productivity (e.g., larger R 2) was explained with font size and VAR than with VAF. Adding glare to the model did not improve model prediction. Based on the β coefficients, each millimeter increase in font height increased productivity by 2.55 correct clicks/min and each arcminute of VAF increased productivity by 0.43 correct clicks/min. Similarly, each millimeter increase in font height reduced perceived task difficulty by 7.75%.
The ability to perform visually demanding tasks on a computer accurately and quickly is important for business and technical processes. This study found that small increases in font size can improve the speed and accuracy of visually demanding tasks. The text-based tasks used were similar in visual demands to Internet search, spreadsheet activities, and document editing. Increasing font size from 14.2 arcmin (1.78 mm) to 23.4 arcmin (3.56 mm) led to a 30% improvement in productivity and a 3% improvement in accuracy. This is a large productivity gain compared with other studies that have evaluated vision interventions for computer users. Daum et al.3 reported a productivity gain of 2.5 to 28.7% with astigmatic refractive correction. In their cost-benefit analysis, assuming the visual correction cost ($268 estimate) for an employee with a salary of $25,000, even a 2.5% increase in productivity led to a favorable cost-benefit ratio of 2.3.
Font Size Effect on VD
The mean self-selected VD for the large font (3.56 mm; 23.4 arcmin) for all subjects was 53.9 cm, which is close to the typically preferred VD of 50 cm for computer monitors.6 For the smaller fonts, subjects self-selected to reduce this VD by 5.5 cm for the medium font (2.23 mm; 16.4 arcmin) and by 9.1 cm for the small font (1.78 mm; 14.2 arcmin). This matches findings from a prior study in which subjects reduced the visual distance when the VAF was less than 16 arcmin.22 A study evaluating font size and VD with electronic paper displays required subjects to use a chin rest and hold the display with the arms.25 For the 2.4-mm font, their mean VD of 48.3 cm was very close to the VD that we observed for the 2.23-mm font. However, for the 3.7 mm font, their mean VD was 49.8, which was 4.1 cm less than what we observed for the 3.56-mm font. This suggests that use of a chin rest while holding the display with the arms may have placed limits on the selected VD.
Overall, there was no significant difference in the average VD between the two age groups (18 to 35 years vs. 55 to 65 years) in this study. In a study involving reading Chinese characters on a handheld electronic paper display, the preferred VD for the young group (21 to 35 years) was significantly greater than that for the middle-age (36 to 49 years) or older (50 to 70 years) groups: 50.3, 45.5, and 44.4 cm, respectively.26 The differences in findings may be due to differences in prescriptions for the older group. The prescriptions for the older subjects were not given but the author suspected that overcorrection of farsightedness in the older groups may have been partially responsible for the VD finding. Reading glasses affect the refractive state of older people, and assuming normal 0.5 D tolerance of blur, +2.25 D ADD gives a 36- to 57-cm range of clear vision. In our study, the prescription powers for the older, presbyopic subjects were mostly +2.25 D ADD.
Font Size and VAF
Standards and textbooks provide recommendations for character size for reading in VAF not in millimeters. The Sanders and McCormick Human Factors in Engineering and Design handbook14 recommends a range of 16 to 22 arcmin. The Kodak ergonomic handbook27 recommends 14 to 22 arcmin. The BSR/HFES 100 draft standard (2002) recommends 16 to 18 arcmin. The ISO 9241-3 standard15 recommends 20 to 22 arcmin (capital letter size). The ISO standard for electronic visual displays (ISO 9241-303:2011[E])24 states that the minimum Latin character height shall be 16 arcmin and the upper limit height should be 30 arcmin when readability is critical. Of course, to select a font size in millimeters to achieve a recommended VAF, the visual distance must be known.
The VAF range in our study (14.2 to 23.4 arcmin) fell largely within the range of the ISO 9241-3 standard. In previous studies, when subjects were allowed to adjust the visual distance for a visual search task, they selected a visual distance that led to a VAF of 22 or 23 arcmin for 3.3-mm22 (VD, 50 cm) or 5-mm28 characters, respectively. Our results suggest that when individuals can freely adjust their posture and chair, they also select a VD that increases the VAF toward 23 arcmin.
However, for the smallest font size tested, 1.78 mm, the mean VD was 44.8 cm, which corresponds to a VAF of 14.2. There are several reasons why subjects may have not moved closer to the screen to increase the VAF further. A closer distance may have increased the visual accommodative and convergence demands to an uncomfortable degree. For the older subjects, as mentioned before, there may have been more blur when the distance is less than 50 cm, or it may have been more uncomfortable to lean further forward.
Font Size and VAR
Whittaker and Lovie-Kitchin10 recommended that monitor placement should consider VA when selecting font size and distance to the monitor. They report that the VAR should be at least 3 to achieve a high fluent reading rate for words with four to six characters, and it should be 6 for optimal reading. Our study supports this recommendation. A 6.1 VAR was achieved only for the large font size (23.4 arcmin), and this condition consistently produced the highest speed and accuracy compared with the 4.3 VAR of the middle font size (16.4 arcmin).
Effect of Glare on VD
In this study, we investigated the effects of reflective glare on the monitor, as opposed to direct glare from light sources surrounding the monitor,19,20 on preferred VD and productivity. Although we found a significant effect of reflective glare on the VD, the effect on productivity was not significant. The finding that subjects moved toward the screen when reflective glare is present has not been previously reported. Subjects likely moved forward to mitigate the glare-induced reduction of contrast of characters. To reduce the effects of the glare and to maintain productivity, subjects moved forward and increased the VAF. On the other hand, subjects may have moved forward to change their viewing geometry to try to move the reflective glare off the reading area. In the experimental setup, it was not possible for the subject to shadow the glare source by moving the head forward and it was difficult to move the glare off the reading area with forward head movement.
Previous studies primarily characterized the luminance of the light source without quantifying the glare perceived by the viewers.19,20 Our study measured actual luminance over the reading area. The glare in this experiment came from a rectangular LED luminaire positioned to create a reflected image on the screen, and thus the reflection of the luminaire was optically superimposed in the center region of the visual tasks. The position and the luminous characteristics of the reflected glare patch changed somewhat with changes in the observer’s viewing position. Here, the glare is a localized “hot spot” reflection from the screen with only a modest effect on the total light flux impinging of the eye. In the experiments of Bailey and Bullimore20 and Sheedy et al.,19 the glare came from lighting surrounding the task. The total luminous flux falling upon the eyes was substantially greater, and although there was some increase in the luminance of the test target displays resulting from the elevation of the ambient illumination within the room, there was little change in the apparent evenness of brightness across the displays. Our finding, that there was no significant effect on productivity, might be expected from the prior direct glare studies.19,20 At the monitor luminance level measured in our study (~300 cd/m2 with the glare source on), Bailey and Bullimore20 reported that VA reduction was minimal for older subjects (50 to 82 years) compared with younger subjects (15 to 41 years). Similarly, Sheedy et al.19 reported that the main luminance effect on VA was not significant for either young (23 to 39 years) or older (47 to 63 years) subjects.
Aging eyes may have less tolerance to glare; however, self-reported glare discomfort rating is mixed.29 In our study, we did not find an age-dependent effect of glare. This may be because the contrast sensitivity for older subjects was normal and similar to that of the younger subjects.
Limitations of the Study
Several limitations of this study should be noted. First, the sample size, especially for the older group, was relatively small; therefore, the nonsignificant findings related to age should be interpreted with caution. Furthermore, the findings may not be applicable to those older than 65 years. Second, the tasks used were selected because they have been used by other researchers and are visually demanding, not cognitively demanding. Some computer-based reading tasks use common words that are readily recognized; therefore, the productivity reported here might not be generalizable to less visually demanding tasks. Third, there is a tradeoff between font size and the amount of content that will fit in a window. Larger font size, up to a point, will improve reading productivity, but a larger font will also reduce the number of words or characters visible in a window of fixed size and may limit other dimensions of productivity. Fourth, the monitor used had a matte surface. A glossy screen would have produced a brighter reflection with visible patterns that may have interfered more with the task. These factors should be considered in future studies of font size, glare, and reading productivity.
This study demonstrates that productivity is improved when using a large font size (3.6 mm) compared with smaller font sizes (1.8 or 2.2 mm) under natural viewing conditions for computer users performing visually demanding tasks. Surprisingly, there was a trend for this effect to be more pronounced in the younger than in the older group. Overall, for each 1-mm increase in font size, mean productivity improved by 3 correct clicks/min, accuracy improved by 2%, and perceived task difficulty was reduced by 8%. The presence of glare led subjects to lean forward and reduce their VD. The larger font size of 3.6 mm was associated with a VAR of 6.1 and a VAF of 23.4 arcmin, which is just above the upper bound of ISO 9241-3 recommendations (e.g., 20 to 22 arcmin). The findings may be useful in setting up workstations and computer monitors, selecting the default font size shipped with applications software, and the training of computer users on the font size to select for their work. Users can easily check the font height on their monitor using the “rule of the 20 dollar bill”; the green serial number on a bill is about 3.5 mm in height.
David M. Rempel
Ergonomics Graduate Program
University of California Berkeley
UC Richmond Field Station
1301 S 46th Street, Bldg 163
Richmond, CA 94804
This study was supported in part by a Graduate Training Grant from the National Institute for Occupational Safety and Health (NIOSH T42 OH008429).
Received November 13, 2013; accepted March 31, 2014.
1. Brewer S, Van Eerd D, Amick BC 3rd, Irvin E, Daum KM, Gerr F, Moore JS, Cullen K, Rempel D. Workplace interventions to prevent musculoskeletal and visual symptoms and disorders among computer users: a systematic review. J Occup Rehabil 2006; 16: 325–58.
2. Sheedy JE. Vision problems at video display terminals: a survey of optometrists. J Am Optom Assoc 1992; 63: 687–92.
3. Daum KM, Clore KA, Simms SS, Vesely JW, Wilczek DD, Spittle BM, Good GW. Productivity
associated with visual status of computer users. Optometry 2004; 75: 33–47.
4. Lee DS, Shieh KK, Jeng SC, Shen IH. Effect of character size
and lighting on legibility of electronic papers. Displays 2008; 29: 10–7.
5. Legge GE, Pelli DG, Rubin GS, Schleske MM. Psychophysics of reading—I. Normal vision. Vision Res 1985; 25: 239–52.
6. Legge GE, Bigelow CA. Does print size matter for reading? A review of findings from vision science and typography. J Vis 2011; 11.
7. Pelli DG, Tillman KA, Freeman J, Su M, Berger TD, Majaj NJ. Crowding and eccentricity determine reading rate. J Vis 2007; 7: 20.1–36.
8. Chung STL, Mansfield JS, Legge GE. Psychophysics of reading. XVIII. The effect of print size on reading speed in normal peripheral vision. Vision Res 1998; 38: 2949–62.
9. Latham K, Whitaker D. A comparison of word recognition and reading performance in foveal and peripheral vision. Vision Res 1996; 36: 2665–74.
10. Whittaker SG, Lovie-Kitchin J. Visual requirements for reading. Optom Vis Sci 1993; 70: 54–65.
11. Bowers AR, Reid VM. Eye movements and reading with simulated visual impairment. Ophthalmic Physiol Opt 1997; 17: 392–402.
12. Yager D, Aquilante K, Plass R. High and low luminance letters, acuity reserve, and font effects on reading speed. Vision Res 1998; 38: 2527–31.
13. Legge GE. Psychophysics of Reading in Normal and Low Vision. Mahwah, NJ & London: Lawrence Erlbaum; 2006.
14. Sanders MS, McCormick EJ. Human Factors in Engineering and Design. New York, NY: McGraw-Hill; 1993.
15. International Organization for Standardization. Ergonomic Requirements for Office Work with Visual Display Terminals (VDTs). Part 3: Visual Display Requirements. Geneva, Switzerland: International Organization for Standardization; 1992.
16. Liu KS, Chiang CM, Lin YS. Influences of visual fatigue on the productivity
of subjects using visual display terminals in a light-emitting diode lighting environment. Arch Sci Rev 2010; 53: 384–95.
17. Bernecker CA, Davis RG, Webster MP, Webster JP. Task lighting in the open office: a visual comfort perspective. J Illum Eng Soc 1993; 22: 18–25.
18. Blehm C, Vishnu S, Khattak A, Mitra S, Yee RW. Computer vision syndrome: a review. Surv Ophthalmol 2005; 50: 253–62.
19. Sheedy JE, Smith R, Hayes J. Visual effects of the luminance surrounding a computer display. Ergonomics 2005; 48: 1114–28.
20. Bailey IL, Bullimore MA. A new test for the evaluation of disability glare. Optom Vis Sci 1991; 68: 911–7.
21. Jaschinski W. The proximity-fixation-disparity curve and the preferred viewing distance
at a visual display as an indicator of near vision fatigue. Optom Vis Sci 2002; 79: 158–69.
22. Rempel D, Willms K, Anshel J, Jaschinski W, Sheedy J. The effects of visual display distance on eye accommodation, head posture, and vision and neck symptoms. Hum Factors 2007; 49: 830–8.
23. Veitch JA, Newsham GR. Preferred luminous conditions in open-plan offices: research and practice recommendations. Lighting Res Technol 2000; 32: 199–212.
25. Shieh KK, Lee DS. Preferred viewing distance
and screen angle of electronic paper displays. Appl Ergon 2007; 38: 601–8.
26. Wu HC. Electronic paper display preferred viewing distance
and character size
for different age groups. Ergonomics 2011; 54: 806–14.
27. Eastman Kodak Company. Kodak’s Ergonomic Design for People at Work, 2nd ed. Hoboken, NJ: John Wiley and Sons; 2004.
28. Jaschinski-Kruza W. Eyestrain in VDU users: viewing distance
and the resting position of ocular muscles. Hum Factors 1991; 33: 69–83.
29. Shi W, Lockhart TE, Arbab M. Tinted windshield and its effects on aging drivers’ visual acuity and glare response. Saf Sci 2008; 46: 1223–33.
Keywords:© 2014 American Academy of Optometry
video display terminal; electronic display; character size; viewing distance; visual angle of font; visual acuity reserve; office ergonomics; usability; productivity