Productivity associated with visual status of computer users

Health care costs have escalated beyond the cost of inflation for some time. Coincidentally, fierce marketplace competition has forced employers to review total labor costs in an attempt to maintain or advance their market status. In this business climate, employers naturally have been reluctant to add potential health care costs to their menu of benefits.Concurrently, workers using computers make more- frequent reports about vision difficulties than any other problem. This project examines the hypothesis that these visual difficulties represent a drag on the efficiency of workers using computers, since lowered visibility of a computer monitor as a result of blur may extend necessary work time and/or cause increased mistakes. Likewise, reduced worker comfort from eyestrain or headaches may require more frequent breaks. Together, these detrimental effects may reduce work accuracy and volume, particularly if the worker is engaged using a computer for for an extended period of time.

The visual environment of a computer is different from that typically associated with hard copy.5 Major differences include increased susceptibility to refractive errors, altered ergonomic challenges as a result of the limitations and placement of the monitor, and increased susceptibility to glare. Typical placement of the monitor requires eyewear designed to achieve the most- efficient use for most presbyopic subjects.

No double-blind study has been completed that suggests that substantial productivity gains occur when optimal vision care is provided. An appropriately constructed study has the potential to determine whether optimizing the visual correction of workers affects the accuracy and/or the volume of their work. In addition,further analysis may allow an estimate of the cost-benefit of the provision of vision examinations and eyewear.

There is relatively sparse evidence to suggest the expected relationship between refractive error and productivity in computer use. If the relationship is subtle, the true effects may not be demonstrated because they would be washed out by uncontrolled secondary factors. Completing a study in the laboratory allows the refractive error of subjects to be changed without consideration of the marketplace. The present study, therefore, alters astigmatic refractive error in a laboratory setting where it is possible to critically assess productivity with substantial changes in refractive error —changes that would not be possible to make in an actual workplace setting.

The aim of this project is to examine the relationship between the astigmatic refractive corrections of subjects using computers and their productivity and visual comfort in the laboratory. We hypothesize that improving the visual status of subjects (by correcting astigmatic refractive errors) using computers in the laboratory results in greater productivity and improved visual comfort. This work also examines the cost-benefit ratio for these same factors. If the cost-benefit ratio is favorable, progressive businesses may achieve substantial productivity gains, in addition to improved worker satisfaction, by optimizing the visual corrections

Methods

Subject Inclusion / Exclusion Criteria

The subjects were required to be 19 to 30 years of age. This limitation was imposed to maximize the likelihood of detecting an effect since, accommodative dysfunction as a result of uncorrected astigmatism may contribute to discomfort and blur. Subjects were required to be willing to wear a single-vision spectacle correction in a trial frame during the study. Contact lens wearers were eligible to participate if they were willing to forego the use of contact lenses during the brief period of the study. Subjects were excluded from the study if they did not meet the age requirement or were unable to complete the experimental portion of the study. None of the subjects demonstrated any accommodative or vergence problems. No other factors were used to exclude subjects from the study. Subjects were paid an hourly rate to participate in the study.

Vision Examinations

All subjects were required to complete a comprehensive vision examination before enrollment in the study. The comprehensive vision examination included auto-refractometry and manifest refraction by experienced optometrists. A PRIO tester (PRIO Corporation, Beaverton, Oregon) was used to assess the best refractive correction for the 20 inches monitor distance using a subjective plus build-up technique (non-dilated).

Computer Set-up

The subjects used a Princeton Graphic Systems 15-inch computer display with 24-bit color and 800 x 100% pixels resolution under good lighting conditions without glare. The room was free of distractions, with no windows or other sources of glare. The experimental tests were presented using Windows 2000 and Microsoft Word 2000 or Excel 2000. The standard color settings for all software were used in the experiment. For MS Word 2000 and Excel 2000, the display was viewed in print layout view with the zoom setting at 100%. The monitor was placed on a table so that it was 20 inches away from the subjects, and the top of the screen was approximately at eye level so that the angle of gaze was slightly downward (approximately 10 to 20 degrees). The eye-to-monitor distance was set and maintained at 20 inches for each subject before the experiment commenced. A 3 x 11-and-three-quarters-inch piece of cardboard suspended from the ceiling indicated the 20-inch distance from the monitor and assisted in ensuring that the subjects maintained the appropriate distance from the monitor (within 1 to 2 inches).

Study Conditions

A total of four refractive conditions were evaluated during the study. Although the four conditions were randomly presented to the subjects, the results consider the data in two parts. The first portion of the study evaluated three refractive error conditions. One condition involved a fully corrected refraction for the monitor distance (corrected) and two other conditions involved subjects wearing their best refractive correction with a superimposed error of either -1.0 or -2.0 diopter cylinder (DC) axis 90°. After that analysis was completed, another refractive error condition (the subject’s naturally occurring astigmatic error uncorrected; the spherical errors were corrected; uncorrected astigmatism) was used to examine the predictive capacity of the study. In this condition, the degree of uncorrected astigmatism was used as a measure of refractive error in the analysis of the uncorrected condition. Each of the conditions was applied in double-masked fashion with the order of presentation randomized.

We assessed refractive error using previously described vector methodology. Briefly, this methodology provides a mathematical technique to determine orthogonal refractive components of any refractive state. Since these components are orthogonal, tehy can then be added, subtracted, or averaged. In addition, this methodology allows different refractive states to be expressed as a single number (vector diopters or VD), thereby obviating difficulties that occur when refractive states with cylinders at different axes are compared. The vector system used in this article expresses refractive errors equivalent to the spherical errors and has a predictable effect on visual acuity.

Table 1. Refractive conditions used in the study
PART CONDITION DESCRIPTION
First Corrected-1DC x 90°-2DC x 90° Fully corrected refraction for the monitor distance -1DC x 90° superimposed on fully corrected refraction for the monitor distance-2DC x 90° superimposed on fully corrected refraction for the monitor distance
Second Uncorrected astigmatism Astigmatic error uncorrected & Spherical errors corrected; degree of uncorrected astigmatism and axis vary with the subject
DCDiopter cylinder

Eyewear

The lenses used in the study to correct the refractive error of subjects were trial lenses placed in a trial frame (Oculus UB-4), adjusted to be as comfortable as possible. The spherical portion of the best correction was placed in the sphere well of the trial frame and was not masked to the technician. For the cylindrical correction, a set of specially cut round CR-39 lenses was manufactured to fit the trial frame. These lenses included a set of 68 trial lenses (20 plano lenses and four sets of 12 cylindrical lenses, -0.25 to -2.75 DC in 0.25 DC steps) used in the refractive error correction portion of the study. These lenses were coded with labels on the minus cylindrical axis of the lenses or at a point specifying a placebo axis for the plano lenses. Coded lenses were labeled after the order was randomized, so that the investigator completing the trial was intentionally unaware of whether a cylindrical lens was being used. The design was intended to produce a methodology that allowed the correction of a subject’s astigmatic correction in the right and left eyes with unknown but appropriate cylindrical lenses. It also allowed the use of unknown and appropriate plano lenses that would leave the subject’s astigmatic correction uncorrected.

In addition to the lens set described, another set of eight lenses was used to further specify the desired refractive correction. The four refractive condition lenses were specially cut round polycarbonate lenses with antireflection coating (Crizal) and included pairs of -1 DC and -2 DC and two sets of plano lenses. For these pairs of cylindrical lenses, the axis was marked with a small label and a similar small mark was affixed onto a phantom axis of each pair of plano lenses.

Each subject completed the experimental session with spherical lenses correcting the spherical error in both eyes. In three of the refractive conditions, a set of cylindrical lenses corrected the astigmatism, while the pair of refractive condition lenses created either no error(best correction), or -1 or -2DC (both axis 90°). In the other refractive condition, two sets of plano lenses were placed in front of the spherical lenses, thus creating a condition in which the subject’s natural astigmatism uncorrected (uncorrected astigmatism).

Before a trial, an investigator randomly assigned the lenses for each subject for each condition and provided data protocol sheets to the technician (selected on the basis of being able to accurately complete the protocol), who conducted the experimental protocol. As a result, neither the subject nor the technician was aware of the lens condition being investigated during any of the trials.

Timing

A technician running the protocol timed all tasks with a stopwatch.

Practice

To minimize learning effects, all subjects completed a short practice period of one page for each condition before the experiment was begun. During this trial, subjects wore their habitual correction.

Experimental Tasks

Subjects were required to perform three tasks requiring inspection of written materials. The first involved searching a listing of county populations; the second required identifying nonsense words that match a nonsense test word (m and n test); the third required performing an editing task on a manuscript. None of the tasks involved any typing or manual manipulation of a mouse by the subject. The subjects were informed that both speed and accuracy were important and were instructed to complete the exercises as quickly as possible without errors. The exact order in which the tasks were performed was randomly generated for each individual subject. All sessions for a subject were completed during a single data session. The data sessions lasted a total of about one-and-a-half to 2 hours per subject, including any necessary breaks. All subjects completed a version of each of these tests with the masked lens combinations described in 

County population listing

In this portion of the experiment, subjects viewed a master list of randomly sorted counties and their associated populations in a spreadsheet. The data were drawn from the first seven alphabetical states in the 2000 U.S. Census. The counties were listed in randomized order in Excel. Subjects examined a single screen of data that contained 25 pairs of counties and their populations . Each subject examined three sets of counties with Arial font sizes of 12-, 10-, and 8-point type. For each font size, at a signal to begin, the subjects were required to state the county population of 10 counties randomly selected from the list as quickly and as accurately as possible. The technician verbally stated the county immediately upon completion of the previous one and immediately as the experimental portion began. This process was repeated for the three fonts, so that the maximum possible productivity was 30 county populations for each of the refractive condition . Different lists of counties and populations were used for each of the four viewing conditions. An error was tabulated each time an incorrect population was stated. The time required to complete each task and the number of errors were used as measures of task performance for each viewing condition.

The data collected included the length of time required to complete the exercise; the number of errors as functions of the viewing condition; and the font. Data analysis included the Kruskal-Wallis non-parametric analysis of variance (ANOVA) test for the time of completion, and the number of errors as a function of viewing condition, and the size of the font.

m and n word search

Subjects were required to search rows of nonsense words comprised of the letters “m” and “n” in Excel. The test sheets are shown in . Subjects identified which of the eight listed “words” on a single row exactly matched the sample provided by stating the column number of the matching “word.” The length of the sample words varied from three letters to five letters for a single page comprised of eight words by eight columns. The font sizes used were 12-, 10-, and 8-point Times New Roman. These correspond to approximate visual acuity of 20/63, 20/51, and 20/39 for the monitor distance of 50.8 cm (20 inches). The time to completion and the number of errors as a function of viewing condition and font size were measured and recorded.

The data collected included the length of time required to complete the exercise and the number of errors. Kruskal-Wallis non-parametric analysis of variance was used to examine the hypotheses that the time of completion or the number of errors was a function of viewing condition and/or font size.

Manuscript Editing

A manuscript of about 400 words was presented on the monitor in MS Word . Subjects were required to search through the manuscript and find the single-quotation symbol (‘) wherever it was inserted throughout the text. This symbol was selected because it is commonly used and represents a subtle difference in text. The symbol was placed somewhere within a word (e.g., hippo’potamus). When the symbol was identified, the subject was required to state the word. Therefore, subjects did not have to read the text, but scanned the text for the symbol in question. Approximately 15 words with the symbol in question were embedded in each of the text sections. Any punctuation that could have been mistaken for the marks was omitted from the text (such as regular, “normal” apostrophes). The text for each section was drawn from an online version of World Book Encyclopedia.

The data collected included the length of time required to complete the exercise and the number of errors. Kruskal-Wallis non-parametric ANOVA was used to examine hypotheses that the time to completion or the number of errors was a function of teh viewing condition and/or font size.

Comfort Levels

A modified, previously validated scale was used to assess visual comfort at the end of each session with a refractive condition . This questionnaire includes nine questions, which address the following potential issues while using the computer: visual problems, clarity, episodes of blurred or double vision, limitations, losing place, lighting, discomfort, headaches and frustration with vision. Subjects completed the Modified Vision Quality of Life Questionnaire immediately upon completion of the protocol for each refractive condition . The scale for each question allowed an assessment of the significance of each item for that refractive condition and resulted in a score ranging from 100 (very comfortable, with no limitations or other problems) to 0 (extremely uncomfortable, with many problems).

Productivity Measures

Productivity is a measure of output per hour. Although the assessment of productivity can be complex, this work uses the length of time required to complete these exercises (of equivalent length) and the number of errors as measures of productivity. Presently, no standard protocol to evaluate workplace productivity exists.

Analysis

The data were examined to reflect performance (the elapsed time or the number of errors, the dependent variable vs. refractive error (the independent variable). The corrected condition provided an undisturbed assessment of the effect of astigmatic refractive error on performance. The uncorrected condition allows an examination of whether “naturally present” astigmatic errors cause changes in productivity. In one portion of the analysis, we forced the data to be categorical by assigning the trial to the closest 0.25D of VD (vector dioptric) error. The created conditions are designed to provide a controlled assessment of refractive errors for each of the subjects. This analytic technique provides a method of examining the relationship between the effects of undisturbed vs. created refractive error. The study used non-parametric ANOVA to examine whether there was evidence of differences in productivity between the lens conditions over the experimental periods.

Figure 1
Figure 1. Mean time to completion for the three task conditions. Bars indicate best correction, -1 DC (or 0.71 VD error), and -2 DC (or 1.41 VD error) and lines indicate the uncorrected astigmatism condition. All refractive errors are plotted as a function of vector dioptric errors (VD).

Significance Levels, Sample Size

p value equal to or less than 0.05 was considered significant for the a significance level. The Þ significance levels were calculatedpost hoc. Before the study, the data from a pilot study of five subjects were used to estimate the necessary sample size. None of those subjects was included in the subsequent experimental trial.

Randomization

The study refractive conditions (corrected, uncorrected [natural astigmatism], -1 DC and -2 DC), experimental tasks (County population search, m and n search, and manuscript editing), and the font size for each experimental task were randomly applied using a random number generator in MS Excel. After a refractive condition was determined, all nine experimental tasks (three tasks in theree font sizes) were completed before progressing to the next refractive condition.

Masking

The subjects were informed of the objectives of the study. The subjects were not informed of the specific lens conditions. They were informed only that four different lens conditions were to be evaluated during the study and that these reflected different refractive states of interest to the study.

The technician who arranged the lenses in the trial frame was unaware of which pair of lenses was being used by the patients in each trial. A different individual-who did not otherwise participate in the collection of data-marked the pairs of lenses and provided a random assignment of the lens conditions an the order of testing. To protect the identity of cylindrical or spherical lenses, the individual placing the lenses in the trial frame was instructed not to look through the lenses. The principal investigator processed the data after all data were collected from the subjects. Therefore, all participants in the study were masked as to the lens condition during the phases of the study. Subjects in the study served as their own controls. The lens pairs were switched according to the randomization process.

Figure 1
Figure 2. Mean errors for the three task conditions. Bars indicate best correction, -1 DC (or 0.71 VD error), and -2 DC (or 1.41 VD error) andlines indicate the uncorrected astigmatism condition. All refractive errors are plotted as a function of vector dioptric errors (VD).

Cost-Benefit Ratio

Following data collection and analysis, a cost-benefit ratio was derived. The cost was determined by calculation of an average cost of a vision examination and hypothetical eyewear (with the best correction) that would be provided to subjects if they were in a workplace setting (the UAB Faculty practice). The benefit was determined by calculating the worth of the net change in productivity for the average of the workers over a 1-year period. We assumed that the productivity–as determined during the study–would remain constant over that year. The cost-benefit ratio was determined by dividing the cost by the benefits likely to accrue over a 1-year period.

For example, assume a given worker is provided a vision examination and a pair of glasses with a total cost of $268 ($80 vision examination; $88 pair of lenses; $100 frame-all prices are hypothetical and are not based on any actual prices). If the mean change in productivity of the worker is 10% (i.e., the decrement in productivity might be 2 words per minute [wpm] to 18 wpm from 20 wpm) and the worker earns $25,000 per year, then this error costs 10% of the worker’s salary. And if one further assumes that the change in productivity is 10% for a year’s period, then a cost-benefit can be calculated. Ten percent of a $25,000-per-year employee salary equals a loss of $2,500 in productivity over that year. For this result, the cost-benefit ratio would be favorable; a gain of $2,500 in productivity would result from an investment of $268–a ratio of 9.3 to 1. For every $1 invested in workers in this manner, the employer would receive $9.33. All things being equal, if the company provides health insurance that already provides for a portion of the vision examination or eyewear, the cost-benefit ratio would be favorably and correspondingly altered.

IRB approval

This protocol received Institutional Review Board (IRB) approval before its initiation. All potential subjects underwent informed consent before enrollment in the study. All subjects were free to withdraw at any point in the study, without penalty.

Figure 3
Figure 3. Box plot of symptom scores for subjects in three conditions: best correction, -1 DC and -2 DC (box is interquartile range and asterisks“*” indicate outliers).

Results

Thirty-nine subjects participated in the study (12 men and 27 women; mean age, 24.9 yrs; SD 2.4; range, 21 to 30 years). Twenty-six subjects had a dominant right eye, whereas the remaining 13 were left eye dominant. Mean spherical refractive error for the test population was -1.85 DS (SD 2.44; range, -10.50 to 1.75 D). The subjects; mean astigmatic correction was -0.36 DC (SD 0.59; range 0.00 to -2.75 DC).

Mean vector dioptric (VD) errors15-16 for the three refractive conditions were 0, 0.71 and 1.41 for the best correction, -1 DC, and -2 DC conditions, respectively. The second part of the study that used the subjects’ uncorrected astigmatism had an overall VD mean error of 0.25 D (SD 0.42).

For the first part of the study, non-parametric ANOVA (Kruskal-Wallis) provided evidence that supported a hypothesis of differences of completion time and errors as a function of the refractive condition (Time: H = 62.90; df = 2; p ,< 0.0001; errors: H = 85.44; df = 2; p < 0.0001). Figures 1 and 2 show the mean times and errors for the three groups . Converted to percentage differences, subjects in the -1 and -2 DCcorrection conditions completed their tasks 8.9% and 28.7% slower, respectively, than the best correction condition and makde 38.1% and 370.0% more errors, respectively. The 8-point font was associated with increased errors and extended completion time (Kruskal-Wallis non-parametric ANOVA, Time: H = 68.77; df = 2; p < 0.0001; errors: = 232.59; df= 2; p< 0.0001).

Likewise, non-parametric analysis of variance (Kruskal-Wallis) provided evidence that supported a hypothesis of differences of symptom scores as a function of the refractive condition (H = 33.12; df = 2; p < 0.0001). Figure 3 shows a box plot of symptom scores for the three groups. Converted to percentage differences, subject in the -1 and -2 DC correction conditions had symptom scores 76.3% and 42.9% (less comfortable) than the best correction condition (100%), respectively.

Analysis of the group whose astigmatism errors were not corrected suggests a complex relationship between errors made, time to ompletion, the symptom score, and the vector dioptric error group (see Figure 4). Median errors increased with uncorrected astigmatism in a relatively smooth fashion. Although there was a tendency to require more time with greater uncorrected astigmatism, the relationship was neither linear nor consistent. Likewise, symptoms reported with increased degrees of uncorrected astigmatism were erratic, possibly indicating that the presence of symptoms may not be a reliable method of identifying the presence (and the consequent effects) of uncorrected astigmatic refractive error. Uneven sample sizes probably introduce variability into attempts to determine the relationship in this portion.

Discussion

The idea that poor vision may result in poor visual comfort is hardly surprising and has been previously reported. In earlier placebo-controlled, double-masked trials, even small differences of refractive error have been shown to be reliably detected and result in poorer reported visual comfort.5-6

Likewise, although we could not locate any work directly aimed at relating possible effects of visual correction on visual performance, the idea that poor vision could hamper the ability to perform a task is also predictable. The results of the present work indicate that, under the conditions of this experiment, uncorrected astigmatic refractive errors are likely to produce poor performance, along with reduced visual comfort, and that the degree to which each is affected is related to the degree of poor refractive correction in non-presbyopic subjects.

“Productivity isn’t everything, but in the long run it is almost everything.”21 As has been widely stated, productivity is an influential factor in determination of the success or failure of a company or the economy.19-2022 Increased productivity, if it outstrips labor and other production costs, is an extraordinarily beneficial factor. Much of the recent ascent in the markets throughout the 1990s was at least partially a result of increased productivity.

Figure 4a
Figure 4b
Figure 4c
Figure 4. A, Median errors; B, time to completion; and C, total symptom score as a function of vector dioptric (VD) errors for the uncorrected astigmatism group. Subjects are grouped to the nearest 0.25 D VD error; N=39 subjects.

Assessing the productivity of a company (or an economy) can be extraordinarily complex.19-2022-24 Factors that ordinarily affect the yield of a business include skill of the workers, management of the company, technology involved in the process, availability of necessary input materials and a wide variety of other factors.25Although common productivity assessments include single-factor and multi-factor assessments, the present work involves a simple assessment of productivity affected by astigmatic refractive error alone. The precise impact of this single factor in a business setting remains to be explored. Since other factors also play a role, the impact on productivity is likely to be a blend of the effects of refractive error combined with other factors that influence task performance. Furthermore, the relative effect of the refractive correction of a workforce on productivity probably depends on the activity. More-visual activities, such as data entry or word processing, may be most susceptible to refractive effects, while “cruder” activities may be less likely to demonstrate effects.

Over the 10-year period from 1992 to 2001, non-farm business yearly productivity increases in the United States averaged 2.0%.26Over this period of time, the productivity in the U.S. ranked at a high level, as compared to most of the rest of the world economies.2224High-tech industries demonstrated the highest productivity changes, on the order of 9.5% per year from 1987 to 1999.25 Considering these numbers, the possible productivity increase as a result of the correction of refractive error estimated as a result of this work–2.5%–is probably a significant factor.

These data were achieved under double-masked, randomized conditions, suggesting that there is a real relationship between astigmatic refractive error and productivity (as defined here). The second part of the experiment, in which productivity was related to the uncorrected “natural” astigmatism of the subjects, again suggests a relationship between refractive error and productivity. Adaptation of the subjects to their uncorrected “natural” astigmatism probably accounts for a portion of the difference. In addition, the effect of the astigmatic mis-corrections on the print on the computer monitor is probably quite different from that of the smaller, uncorrected “natural” astigmatism, so that adaptation may account for only a portion of the difference. The small degree of uncorrected “natural” astigmatism, averaging 0.25 D, further demonstrates the likelihood of an effect, since a subtle relationship probably would not reach statistical significance. In addition, in Figures 1 and 2, the mean effect on productivity of the uncorrected astigmatism group (the lines in the figures) roughly parallels the data from the experimental groups, which also suggests the effect.

These data suggest that employers may achieve a beneficial effect on productivity by the provision of refractive corrections to their non-presbyopic employees. The appropriate cost-benefit ratio for this work depends on the identification of the overall change in productivity as a result of correcting refractive error. In the workplace, the cost-benefit ratio for a given employee could be calculated as described in this article, provided appropriate productivity data were available. For the present work, cost-benefit estimates may vary according to the outcome being assessed. In this regard, the cost-benefit ratio would vary according to font size, task, and whether errors or time to completion is used as the primary productivity index. For the present work, we have selected the most conservative of these estimates, the percent difference between the time to completion of best-corrected status to the subjects’ own uncorrected astigmatism, 2.5%. Other estimates increase from there for time to completion, depending on the refractive error (-1 mis-corrected cylinder: 8.9% increase or -2 mis-corrected astigmatism: 28.7% increase). Errors suggest larger estimates of the cost-benefit ratio, beginning with a 17% decrease with the patients’ own astigmatism uncorrected. A cost-benefit ratio as small as 2.5 suggests that even employees who make modest salaries would more than make up for the cost of vision care in increased production. Further examination of the relationship of refractive error (and other parts of visual performance) to the critical factor of productivity should continue. These should continue to provide impetus to the economy for the mutual well being of employees and employers alike.

Acknowledgements

We gratefully acknowledge funding from the PRIO Corporation, Mr. Jon Torrey, CEO. The authors do not have any financial interest in PRIO Corporation. No restrictions on the content of this manuscript were imposed by the funding agreement.

 


 

References

1. Jaklevic MC. Runaway costs. Healthcare inflation at highest level in 7 years. Modern Healthcare. 2002; 32(3):12-3.2. McCarthy M. US health-care inflation accelerates, according to government report. Lancet 2002; 359(9302):240.

3. Strunk BC, Ginsburg PB, Gabel JR. Tracking health care costs. Health Affairs, 2001; Suppl Web Exclusives:W39-50.

4. Sheedy JE. Vision problems at video display terminals: a survey of optometrists. J AM OPTOM ASSOC 1992; 63:687-92.

5. Wiggins NP, Daum KM, Snyder C. Effects of residual astigmatism in contact lens wear on visual discomfort in VDT use. J AM OPTOM ASSOC 1992; 63:177-181.

6. Wiggins NP, Daum KM. Visual discomfort and the astigmatic refractive errors in VDT use. J AM OPTOM ASSOC 1991; 62:680-4.

7. Nakazawa T, Okubo Y, Suwazono Y, et al. Association between duration of daily VDT use and subjective symptoms. Am J Industrial Med 2002; 42:421-6.

8. Daum KM, Good GW, Tijerina L. Symptoms in video display terminal operators and the presence of small refractive errors. J AM OPTOM ASSOC 1988; 59:691-7.

9. Foye PM, Cianca JC, Prather H. Industrial medicine and acute musculoskeletal rehabilitation. 3. Cumulative trauma disorders of the upper limb in computer users. Arch Phys Med Rehab 2002; 83(3 Suppl 1):S12-5, S33-9.

10. Shackel B. People and computers–some recent highlights. Appl Ergon 2000; 31:595-608.

11. Anonymous. Computer workstation ergonomics guidelines. Occup Health Saf 2000; 69(10):214, 216, 218-20.

12. Saito S, Piccoli B, Smith MJ, et al. Ergonomic guidelines for using notebook personal computers. Technical Committee on Human- Computer Interaction, International Ergonomics Association. Industrial Health 2000; 38(4):421-34.

13. Daum KM. Anomalies of accommodation. In: Rutstein RP, Daum KM. Anomalies of Binocular Vision. Mosby-Yearbook Publishing Co., 1998.

14. Saladin JJ. Convergence insufficiency, fixation disparity, and control systems analysis. Am J Optom Physiol Optics 1986; 63:645-53.15. Harvey EM, Miller JM, Dobson V, et al. Measurement of refractive error in Native American preschoolers: validity and reproducibility of autorefraction. Optom Vis Sci 2000; 77:140-9.

16. Thibos LN, Wheeler W, Horner D. Power vectors: An application of Fourier analysis to the description and statistical analysis of refractive error. Optom Vis Sci 1997; 74:367-375.

17. Raasch TW. Spherocylindrical refractive errors and visual acuity. Optom Vis Sci 1995; 72:272-5.

18. McKeon C, Wick B, Aday L, et al. A case-comparison of intermittent exotropia and quality of life measurements. Optom Vis Sci 1997; 74(2):105-10.

19. “Productivity Measures: Business Sector and Major Subsectors,” Chapter 10 of the BLS Handbook of Methods, BLS Bulletin 2490; April 1997: 89-102.

20. “Industry Productivity Measures,” Chapter 11 of the BLS Handbook of Methods, BLS Bulletin 2490; April 1997: 103-109.

21. Wessel D. CAPITAL Column. Productivity: Too much of a good thing? Wall Street Journal 2002; (December 19):A2.

22. “Foreign Labor Statistics,” Chapter 12 of the BLS Handbook of Methods, BLS Bulletin 2490; April 1997:110-21.

23. Sherwood MK. Difficulties in the measurement of service outputs. Monthly Labor Review (March 1994):11-19.

24. Dean ER, Sherwood MK. Manufacturing costs, productivity, and competitiveness, 1979-93. Monthly Labor Review (October 1994):3-16.

25. Kask C, Sieber E. Productivity growth in ‘high-tech’ manufacturing industries. Monthly Labor Review (March 2002):16-31.

26. U.S. Department of Labor, Bureau of Labor Statistics. Major Sector Productivity and Costs Index, Output per hour, non-farm business, 1992-2001. (http://data.bls.gov/cgi-bin/surveymost?pr).

Corresponding author:Kent M. Daum, O.D., Ph.D.School of Optometry

The University of Alabama at Birmingham

1716 University Boulevard

Birmingham, Alabama 35294-0010kdaum@uab.edu

Appendix 1:

Kings CountyMiami County

Medina County

Adams County

Stanislaus County

Merced County

Hardin County

Crawford County

Santa Cruz County

Yolo County

Shasta County

Auglaize County

Butler County

12946198868

151095

27330

446997

210554

31945

46966

255602

168660

163256

46611

332807

Solano CountyEl Dorado County

Marin County

Athens County

Sacramento County

Los Angeles County

Madera County

Lassen County

Champaign County

Hocking County

Delaware County

Carroll County

Mahoning County

394542156299

247289

62223

1223499

9519338

123109

33828

38890

28241

109989

28836

257555

Appendix 2:

Examples of the M & N Test of Visual Sequential Attention

Responses

1 2 3 4 5 6 7 8
1 mmn mmn mnn mmm mmn mnn mmm mmn mnn
2 nmm nmm nmm nnm nmn nmn nmm mmn nmm
3 nmnm nnm nmnm nmmm nmnm mm nnmm nmnm nmnm
4 nnmm nmmm nnnm nnmm nmnm nnmm nnmm nmnm nnmm
5 mmnmn mmnmn mmnmn mmmnn mmmmn mmnmn mmnmn mmmmn mmnnn
6 mnmnm mnmnm mnmnm mnnnm mnmnm mnmnn mmmnm mnmnm mnmnm
7 nmnnmm nmnnnm nnnmm nmnnmm nmnnmm nmmnmm nmnnmm nmnnmm nmnmmm
8 mnmmnm mnmmnm mnmmnm mnmmnm mnmmmm mnmmnm mnmnmm mnnmnm mnmmnm
9 nnmmnmn nmmmnmn nnmnnmn nnmmnmn nnmmnmn nnmmnnn nnmmnmn nmmmnmn nnnmnmn
10 mnnmnnm mnnmnnm mnnmnnm mnnnnnm mnmmnnm mnnmnnm mnnmnmm mmnmnnm mnnmnnm

Appendix 3:

Example of text edited by subjects

If you look at the n’ight sky without a teles’cope or a pair of binoculars, you will see what the ancients saw. If the night is clear and you are far from city lights, you will see ab’out 3,000 stars. Stretching across the sky will be a splotchy band of bright and dark areas called the Milky Way. In addition, a few fuzzy spots will be visible.

Ancient people in the Western world noticed that certain stars are arranged in patterns shaped somewhat like human beings, animals, or common objects. Man’y ancient civilizations associated such patterns, called constellations, with mythology. Many names of constellations have come to us fro’m Greek myths. In one myth, Artemis, goddess of hunting, was greatly saddened by the death of a human hunter nam’ed Orion. In her sorrow, she placed Orion in the sky as a constellation.

If you m’ap the location of several stars for a few hours, you will observe a regular motion. The stars move relative to Earth because of Earths rotation ab’out its own axis. All stars move in circles around a point in the sky known as a celestial pole. Stars rotate coun’terclockwise ar’ound the celestial north pole and clockwise around the celestial south pole. Stars that are far from the pole rise from below the horizon in the east, move upward, and then set in the west.

One bright star, Polaris, is so cl’ose to the celestial north pole that it m’oves very little. Bec’ause of its location, Polaris is also known as the North Star. There is no “South Star,” but the constellation Octans (Octant) is close to the celestial south pole.

The sun, like the other stars, rises in the east and sets in the west. But the sun also moves eastward relative to the other stars about 1° each day. By noting which stars are visible above the horizon just before sunrise and just af’ter sunset, people have mapped the path of the sun among the stars for thousands of years. This path is known as the ecli’ptic, and the band of constellations near this path is called the zodiac.

 


 

Appendix 4:

Name _________________________________________ Date __________________________________

Lens combination _______________________________(Technician completes)

Modified Vision Quality of Life Questionnaire

University of Texas School of Public Health/University of Houston Optometry Clinic

Modified March 26, 2002

Kent M. Daum, O.D., Ph.D.

School of Optometry/ University of Alabama at Birmingham

Instructions

For each of these questions, please circle the number that best represents your answer. Your answer should be made in reference to the lens combination and tasks you’ve just completed on the computer.

1. In general, would you say that with this lens combination you have problems with your eyes:

a. All of the time

b. Most of the time

c. A good bit of the time

d. Some of the time

e. A little of the time

f. None of the time

2. How would you rate the clearness of your vision with this lens combination?

a. Excellent

b. Very good

c. Good

d. Fair

e. Poor

3. How often have you had episodes of blurred vision and/or double vision with this lens combination?

a. All of the time

b. Most of the time

c. A good bit of the time

d. Some of the time

e. A little of the time

f. None of the time

4. To what extent did problems with this lens combination limit your ability to do tasks on the computer or the amount of time that you needed to do them (for example, because you became tired, lost concentration or were not able to see well enough to complete the task)?

a. Extremely

b. Quite a bit

c. Moderately

d. Slightly

e. Not at all

5. How often did you lose your place, reread the same line or skip lines when you were completing the tasks on the computer with this lens combination?

a. All of the time

b. Most of the time

c. A good bit of the time

d. Some of the time

e. A little of the time

f. None of the time

6. To what extent did the lighting affect your ability to complete the tasks on the computer with this lens combination?

a. Extremely

b. Quite a bit

c. Moderately

d. Slightly

e. Not at all

7. How much did your eyes hurt, watered, burned, itched or become red or swollen in completing the tasks on the computer with this lens combination?

a. All of the time

b. Most of the time

c. A good bit of the time

d. Some of the time

e. A little of the time

f. None of the time

8. How much did you have headaches when completing the tasks on the computer with this lens combination?

a. All of the time

b. Most of the time

c. A good bit of the time

d. Some of the time

e. A little of the time

f. None of the time

9. To what extent were you embarrassed or frustrated when you were unable to complete the tasks on the computer because of this lens combination?

a. Extremely

b. Quite a bit

c. Moderately

d. Slightly

e. Not at all

References

Kent M. Daum, O.D., Ph.D.a, Katherine A. Clore, O.D.a, Suzanne S. Simms, O.D.a, Jon W. Vesely, B.S.a, Dawn D. Wilczek, B.S.a, Brian M. Spittle, B.S.a, Greg W. Good, O.D., Ph.Db.

aSchool of Optometry, The University of Alabama at Birmingham, Birmingham, Alabama and bCollege of Optometry, The Ohio State University, Columbus, Ohio