STATISTICS PROJECT 12
Todetermine the rate at which people’s eyes were deteriorating due toexcessive usage of computers, an experiment was carried out. Thesubjects involved were people with different levels in terms of hoursof computer usage. It involved two simple questions. Do you have anitchy irritation in your eyes? Do you have a problem seeing clearlywithout any aid? The questions were selected as such because theseare the symptoms of problematic eyes affected by the light emitted bythe computer monitor. The frequency of the itchy irritation was alsoconsidered in terms of whether it occurred often or rarely. The kindof eye aid looked out for were spectacles, contact lenses, andsunglasses.
Tocarry out this experiment, five experimenters were used, three ladiesand two gents. They were to approach people randomly in the streetoutside a busy mall and question people of different genders and agebrackets. These experimenters had some optical kits with them in casethe subjects of the experiment were not aware of whether they couldsee clearly. These had some cards printed with letters of differentsizes. The experimenter would hold the card towards the subject andask them to make out the letters on the cards at close range. Thenthey would move away at different distances and request the subjectsto read a different card with different letters of different sizes.The distance at which the subject failed to make out the exactletters or started to see blurry images was checked against theaverage distance of sight of a person of their age but with goodeyesight. If the distances were nearly the same, then the subject hadgood eyesight, otherwise their eyes were problematic.
Cardswith different printed letters were used to avoid the memory effecton the subjects. After seeing a letter, it is easy to make it outagain even without seeing it clearly. To avoid this error in theexperiment, different prints were used.
Ahypothesis that was to be tested during the experiment wasformulated. It stated that people with more hours of computer usageper day develop eyesight problems.
Toget the data that was to be analyzed to determine whether thehypothesis was null or needed to be dropped, a sample had to bedetermined. People from different genders were used although the datawas later filtered into the number of computer hours one utilizes perday. A randomized test sequence was preferred because there is nospecific group of people with the same kind of optical impairment. Itwas preferred that a young age group of early 20s to early 30s beused. It was because of the effect of age on eyesight. Using a sampleof the older age groups would bring about misinformation sincelong-sightedness is common in old people regardless of whether theyuse computers or not. For reliable results, a sample size of 10participants per range of hours was deemed appropriate for theexperiment. The range of hours was selected as over 16, 14-16, 12-14,10-12, 8-10, 6-8, 4-6, 2-4 and less than 2 hours per day. Consideringthese 9-hour brackets, the total participants were 90, and eachexperimenter was to approach and investigate an average of 18participants. A standard deviation of 0.5 and a margin of error of+/- 5% were selected and a confidence interval of 95 applied.
Itwas preferred that the participants be asked whether they werewilling to participate in the brief experiment, after which it wouldbe explained to them the purpose of the experiment. Any questionsthey had would also be answered by the experimenters. In fact, advicewas given to those who required some. If there was a respondent withcompletely sick eyes and did not know what to do, they would be askedto visit the nearest opticians. Some leaflets on proper eye care werealso issued to the participants who volunteered to answer ourquestions.
Thehypothesis was quite appropriate for this experiment and the modewith which it was tested was bound to yield good results.
Thedata collected was recorded in tables indicating the range of hoursand average hours used in front of a working computer monitor. Thenumber of the affected participants was also recorded per range ofhours. The number of respondents was also divided into groups asthose with itchy eyes only, those with poor eyes only and those withboth itchy eyes and poor eyesight.
Thedata collected was as in the table below:
Table1: Data collected
Fromthe table above, it is evident that the number of people affected byboth poor eyesight and itchy eyes increases as the average hours perday increases. The opposite is true of the number of people notaffected. Their number is highest for the hour bracket of less thantwo hours per day and zero for the high hour brackets. Data from thetable is statistically analyzed to give various results.
Thedescriptive analysis of the different eye defects in relation to thenumber of hours the subjects are exposed to light from a computermonitor is performed. The results are as shown below:
Thetwo eye defects are compared to the hours worked in a correlation togive the output shown below:
Itcan be observed that the P-value for the two correlations differs,indicating a varying relationship between those affected and thosenot affected by light from a computer monitor. The variation isbrought about by the number of hours each individual has to sit infront of a computer.
Therelationship between exposure to light from a computer monitor andits effects on the human eyes can also be seen in a regressionanalysis of statistics. By the use of Minitab software, theregression model can be developed to indicate the variance andregression error. It is shown in the processed information below:
Theabove is the regression analysis of average hours per day versus bothitchy eyes and poor eyesight.
Belowis the regression analysis of average hours per day versus those whoare not affected by light as a result of little or no exposure atall:
Toget a clear and comprehensive view of the regression analysis(Edwards, 1969), residual plots of average time versus those withboth itchy eyes and poor eyesight and average time versus those notaffected are made. The comparison is represented graphically asbelow:
Figure1: Residual plots for itchy eyes and poor eyesight.
Figure2: Residual plot for those not affected
Theresidual plots give a comparison of different kinds of plots, amongthem box plots and slopes. The comparison is beneficial infiguratively explaining the variation of eye problems among people ofvarying hour groups. It is evident from the graphs that the longerthe time one works in front of a computer with minimal or no lightshield at all, the more they are likely to end up having eyeirritations and defects.
Toemphasize on this, a plot comparing average hours per day and thenumber of people affected was done.
Figure3: Average hours versus number with both itchy eyes and poor eyesight
Theline of the perfect fit is a clear indication that there is indeed adirect relationship between the number of hours one is exposed tolight to the effect this has to their eyes. From the graph, peoplewith less than two hours are rarely affected while those with sixteenhours of exposure are highly affected and irritated by their eyes.The findings are emphasized by an individual value plot.
Figure4: Individual value plot for the different effects of exposure
Fromthe individual value plot, it can be seen that majority of the hourbrackets have zero as the number of the people not affected, one forthose with poor eyesight only, two for those with itchy eyes only andsix for those with both poor eyesight and itchy eyes. It indicatesthat the majority of the subjects of the experiment had one or moreeye defect caused by working on a computer for too long.
Finally,the ANOVA analysis was performed. The general linear model was usedand the results are as below:
Fromthe data collected and the analysis done as per the graphs, tablesand statistical analysis above, it can be concluded that there is adirect relationship between the length of time a person is exposed toa working computer monitor and the defects and irritations they getin their eyes. The longer the time one was exposed to light from acomputer, the more their eyes were defective. The experiment was,therefore, successful in proving the hypothesis that people with morehours of computer usage per day develop eyesight problems. It isrecommended that further studies be carried out on ways in whichlight emitted by a computer monitor can be reduced to eye-friendlylevels to curb the increasing cases of eye defects.
Edwards,A. (1969). Statisticalanalysis(3d Ed.). New York: Holt, Rinehart and Winston.