Eyes Glazed Over: Using Eye Tracking And FMRI To Measure .
Eyes Glazed Over: Using Eye Tracking and fMRI to MeasureHabituation to Warnings over a WorkweekAbstractA major inhibitor of the effectiveness of securitywarnings is habituation: diminished attention due tofrequent exposure to warnings. Although this problemis widely recognized, previous security studies havelargely inferred or indirectly measured the occurrenceof habituation. Moreover, although habituationdevelops over time, previous studies have examinedhabituation only within a single experimental session. Ittherefore remains unclear how habituation to securitywarnings evolves over longer periods of time.We address this gap by conducting a longitudinalexperiment that examines how habituation to securitywarnings develops over the course of a five-dayworkweek. In addition, we measure the occurrence ofhabituation using two neurophysiological methodssimultaneously: fMRI and eye tracking. Our resultsshow a dramatic drop in attention in terms of neuralactivity and eye fixations after only the secondexposure to a warning, with further decreasesthroughout the workweek. We also find thatparticipants’ attention partially recovers betweenworkdays when there was no exposure to the warningstimulus. Finally, as a potential cost-effective measureto mitigate habituation, we test a polymorphic warningthat updates its appearance with each repetition. Wefind that such warnings are substantially more resistantto habituation across the workweek as compared toconventional warnings.1IntroductionUsers often represent the last line of defense betweenattackers and organizations. User response to securitywarnings is thus a critical aspect of behavioral security[23]. A major inhibitor of the effectiveness of securitywarnings is habituation: diminished attention due tofrequent exposure to warnings [31]. Through thisprocess—also known as warning blindness [50] orfatigue [2]—users’ attention to warnings can attenuateto the point where they hardly see the warning anylonger. Although this problem is widely recognized[e.g., 18; 29; 35; 44], few studies have examinedhabituation empirically. Moreover, the few empiricalstudies that do exist either infer habituation or measureit indirectly through a behavioral proxy [2; 9; 10; 29;45; 51]. An exception is [4], which used fMRI toexamine how habituation to warnings develops in thebrain.Another major limitation of prior research is that it isbased on cross-sectional experimental ological phenomenon that evolves over time[37]. Therefore, past research on habituation to securitywarnings has provided only a static snapshot of adynamic problem. Our first research question istherefore:RQ1. How does habituation evolve in the brain inresponse to security warnings over time?We addressed this question by extending the work of[4] in two key respects. First, we performed alongitudinal experiment that examined user habituationto security warnings over the course of a five-dayworkweek. This experimental design allowed us tomeasure not only the attenuation of user warningresponse over the course of the workweek, but alsoanother core characteristic of habituation: responserecovery, that is, the increase in user response after arest period in which the stimulus is absent [37]. Giventhat past work has been based exclusively on crosssection experimental designs, this paper is the first toexplore how users recover from habituation effectsbetween exposures to warnings.Secondly, [4] used fMRI and mouse cursor tracking tomeasure habituation. However, neither of thesemethods directly measures visual processing. fMRImeasures cognitive activity that lags visual processingby 3–5 seconds. Mouse cursor tracking provides asurrogate measure of attention as the mouse cursorhovers over UI elements, but this too follows the eyes’inspection of visual elements, and only provides anincomplete view of attention. Because securitywarnings are mainly graphical in nature, it is importantto understand how visual attention to the warningschanges over time. In this paper, we measuredhabituation using two neurophysiological methodssimultaneously: fMRI and eye tracking, using an fMRIcompatible, long-range eye tracker. This allowed us tomeasure how both cognitive processing and visualinspection of a warning habituate over time.Our second research question is:RQ2. How can security warnings be designed to bemore resistant to habituation over time?Previously, [4] showed that polymorphic warnings thatrepeatedly change their appearance can be effective in1
maintaining attention during a single experimentalsession, but left unresolved whether this novelty fadeswith time. We extend their study by testing theirpolymorphic design in our longitudinal experiment,hypothesizing that the polymorphic warning willexhibit less attenuation and greater recovery across thefive-day workweek as compared to conventionalwarnings.Our results showed a dramatic drop in attention interms of neural activity and eye fixations after only thesecond exposure to a warning, with further decreasesthroughout the workweek. We also found thatparticipants’ attention partially recovered betweenworkdays when the stimulus was absent. Interestingly,we found that the polymorphic warning design wassubstantially more resistant to habituation as comparedto conventional static warnings, and that this advantagepersisted throughout the five-day experiment. Thepolymorphic design may thus be a cost-effectivesolution which can easily be put into practice.2Literature ReviewHabituation is widely recognized as “the simplest andmost basic form of learning” [36, p. 125]. It is believedto be ubiquitous in the animal kingdom, having beenfound “in every organism studied, from single-celledprotozoa, to insects, fish, rats, and people” [13; 37, p.125]. Habituation is an important survival mechanismbecause it allows organisms to filter out irrelevantstimuli in the environment, and to thus conserve energyfor response to stimuli which are relevant for survival[46]. Not surprisingly, humans also exhibit habituationto a wide variety of stimuli—visual, auditory, andothers—and this is evident as early as infancy [14].Given its strong security implications, habituation isfrequently cited as a key contributor to users’ failure toheed warnings. However, many studies infer thepresence of habituation, rather than empiricallyexamine it. For example, Egelman et al. [23] found acorrelation between user disregard for warnings anduser recognition of warnings as previously viewed, andattributed this correlation to habituation. Sunshine et al.[51] observed that participants remembered theirresponses to previous interactive security warnings andapplied them to new warnings—even if the level of riskor context had changed—and likewise pointed tohabituation as the probable cause. Akhawe and Feltfound that the most common browser SSL error had thelowest adherence rate and the shortest response time,and noted that this result was “indicative of warningfatigue” [2, p. 268].Bravo-Lillo et al. [8; 9] empirically measuredhabituation, albeit indirectly. For example, theymeasured habituation in terms of the percentage ofusers who immediately recognized that the contents of adialog message had changed after a rapid habituationperiod. Only 14% of the users in their studyimmediately recognized the change in the dialogmessage [9]. A follow-up study examined four differentlevels of warning exposure frequency. They found thatincreasing the frequency with which a dialog wasdisplayed caused a threefold decrease in the proportionof users who immediately recognized a change in thedialog message [8].In contrast to the above studies, Anderson et al. [4] usedfMRI to measure habituation in the brain in response towarnings. Their results showed a large drop in activityin the visual processing centers of the brain after onlythe second exposure to a warning, and found furtherdecreases with additional exposures.However, all of these prior studies share a majorlimitation in common: they are based on singleexperimental sessions. This is a problem becausehabituation is a fundamentally neurobiological processthat occurs over time [37]. Consequently, crosssectional experiments that observe habituation at asingle moment in time are unable to capture howhabituation evolves over several days. Furthermore,they are unable to measure response recovery after awarning stimulus has been withheld. As a result, ourunderstanding of how habituation evolves and of howto address the problem is limited. This is the primaryresearch gap addressed in this study.3HypothesesWe develop our hypotheses around the two mostprevalent characteristics of habituation: (1) responsedecay—an attenuation of a response with multipleexposures—and (2) response recovery—the increase inresponse after a rest period in which the stimulus isabsent [37]. Hypothesis 1 explores how user responseto security warnings weakens over the course of severalrepeated viewings, and how polymorphic warnings(described below) can deter this effect. Hypothesis 2explores how user response to warnings recovers afterthe warning is withheld, and how polymorphicwarnings enhance this recovery. Our hypotheses restupon two prominent theories of habituation fromneurobiology: the stimulus-model comparator theory(SMCT) [49] and the dual-process theory (DPT) [28].Although their mechanisms differ, both modelsdescribe a consistent process of habituation (see Figure1).2
Figure 1: Graphical depiction of the stimulus-model comparator theory (unique terminology in red) and dual-processtheory (unique terminology in blue).3.1 Response DecaySMCT [49] explains that the brain creates a mentalmodel when exposed to a stimulus (e.g., when seeing awarning). When people see the same stimulus again,they automatically and unconsciously compare thestimulus to this model. If the model and stimulus aresimilar, an blocking system in the brain inhibitsbehavioral responses to the stimulus—e.g., people payless attention to the stimulus [52].DPT [28] describes this reduced response to stimuli ashabituation. In the context of security warnings, usersunconsciously compare subsequent warnings to themental model of warnings they have seen previously. Ifusers unconsciously determine that a warning is similarto others they have seen before, they pay less attentionto it. This automatic, subconscious mechanism becomesmore ingrained with each successive repetition of thewarning.We predict that this habituation will occur both whenviewing repeated warnings within a single computingsession, and when viewing repeated warnings incomputing sessions over consecutive days [27]. Whenviewing repeated warnings within a single computingsession, the brain creates a robust mental model of thesecurity message, which results in habituation duringthat session. However, these mental models can alsopersist across several days and even for much longerperiods. Over successive days, users will thus rely ontheir mental models rather than actively process thewarning [37]. In summary, we hypothesize:H1a: Users habituate to warnings in computingsessions over consecutive days.We hypothesize that users will habituate more slowly topolymorphic warnings—warnings that change theirappearance with each repetition [4]—than to staticwarnings. Wogalter states that, “habituation can occureven with well-designed warnings. . . . Where feasible,changing the warning’s appearance may be useful inreinvigorating attention switch previously lost becauseof habituation” [56, p. 55]. Changing the appearance ofa warning creates novelty. The orienting reflex,described by SMCT as the primary reaction of the bodyto a novel stimulus, is influenced by a comparison ofthe current stimulus with a mental model of thestimulus as it was previously experienced. If a new orchanged stimulus is experienced that does not match themental model, then response strength will recover—(e.g., people will pay more attention to the warning)[49]. DPT describes this process as sensitization, anenergizing process that strengthens the orienting reflexand thereby the attention span [28]. Sensitizationcounterbalances habituation [37]. Consequently, bychanging the appearance of a warning, users’ orientingreflexes are unconsciously sharpened, and thus userswill habituate less to polymorphic warnings on both theneural and behavioral levels [5].We predict that polymorphic warnings will engendersensitization, reducing habituation within a singlecomputing session as well as between computingsessions over multiple days. When users encounter apolymorphic warning in a future computing session, it3
may contradict a weaker mental model and beperceived as novel (i.e., cause an orienting reflex). [15;52]. In summary, we hypothesize:H1b: Users habituate less to polymorphic warningsthan to static warnings in computing sessions overconsecutive days.3.2 RecoveryAlthough users will habituate to warnings, we predictthat they will partially recover from the habituationafter a day’s rest period without seeing warnings.Decay theory [6] explains that memory becomesweaker due to the mere passage of time. When awarning is withheld for a day, the mental model of thewarning will become weaker. Therefore, when userssee this warning in the future, it will be less likely tomatch the mental model and will appear novel. Inresponse to this novelty, the response strength willrecover and the sensitization process will increase aperson’s attention to the warning, thus counteractinghabituation [11].Although the mental model diminishes with time, it isunlikely to fade completely within a single day. Thebrain will still inhibit the behavioral response to thestimulus and habituation will occur. However, thisresponse inhibition or habituation is likely to be weakerwhen users see a warning after it has been withheld fora day as compared to when they see it repeatedly withina single computing session [37]. In summary, wehypothesize:H2a: If warnings are withheld after habituation occurs,the response recovers at least partially the next day.We predict that the amount of recovery from day to daywill be greater for polymorphic warnings than for staticwarnings. As previously discussed, the mental modelsof polymorphic warnings are weaker and less stablethan the models of static warnings. Less stable mentalmodels (i.e., mental models that have not received asmuch reinforcement) fade more quickly than stablemodels [37]. Thus, after users do not see a warning fora day, they are more likely to perceive the polymorphicwarning as novel. As a result, user response topolymorphic warnings will recover to a greater degreethan the user response to static warnings.Furthermore, if the polymorphic warning continues tochange its appearance from one day to the next, it iseven more likely to differ from the existing mentalmodel, thus weakening behavioral inhibition, increasingsensitization, and enhancing response recovery [37].Conversely, with static warnings, response recoverywill be weaker because the mental model is morerobust, reinforced by repetitive exposures to the samewarning on previous days [26; 28]. The behavioralresponse will be inhibited to a greater degree, andhabituation will be more pronounced [37]. In summary:H2b: If warnings are withheld after habituation occurs,response recovery is stronger for polymorphicwarnings than for static warnings on the next day.4Polymorphic Warning DesignAnderson et al. [4], developed a polymorphic warningartifact based on an extensive review of the warningscience literature. They created 12 graphical variationsof a warning dialog that was expected to sustainattention. Using fMRI data, they tested the differentpolymorphic variations and found that, in terms ofmaintaining attention, four of the variations performedbetter than the rest: (1) including a pictorial symbol, (2)changing the warning’s background color to red, (3)using a “jiggle” animation when the warning appears,and (4) using a zoom animation to make the warningincrease in size. Figure 2 shows each variation for onesample warning with its supporting sources. Given thissupport, we used these four variations for of thepolymorphic warning to test our hypotheses.Neurophysiological tools can be used to evaluate UIdesigns. Riedl et al. explained that neurophysiologicalmeasures are beneficial “to the design of ICT artifacts”[38, p. ii] and that “researchers could use the theory ofcontrolled and automatic brain processes to . . . allowfor a better design of IT artifacts and otherinterventions” [40, p.250]. Further, Dimoka et al. [22]argued that these measures should be used as dependentvariables in evaluating IT-artifact designs:“Rather than relying on perceptual evaluations of ITartifacts, the brain areas associated with the desiredeffects can be used as an objective dependent variablein which the IT artifacts will be designed to affect (p.700).”We use precisely this approach to evaluate thepolymorphic warning design.5MethodsTo test our hypotheses, we conducted a multimethodstudy, simultaneously collecting both fMRI and eyetracking data. This allowed us to capitalize on thestrengths of each method while mitigating theirlimitations [55]. fMRI is useful in measuring neuralactivity by tracking changes in blood-oxygenationlevels (the blood oxygen level–dependent or BOLDresponse) in specific areas of the brain. This allowsresearchers to identify distinct regions of the brainwhere activity is correlated with cognitive processes.4
Message Content: Pictorial symbols (e.g., an exclamationpoint) [32; 48]Warning Appearance: Color [7; 42]Animation: Jiggle, scale/zoom [9; 24; 33]Figure 2. Symbol, background color, zoom and jigglevariations.fMRI identifies regions in terms of voxels or small 3mm cubes, which makes it ideal when high spatialresolution is required [20]. A neural manifestation ofhabituation to visual stimuli in the brain is calledrepetition suppression (RS): the reduction of neuralresponses to stimuli that are repeatedly viewed [26]. Inour case, high spatial resolution was important becauseit allowed us to disentangle RS effects from sensoryadaptation or fatigue effects [37].We used fMRI to capture evidence of the RS effect,which is a reduction in the degree of fMRI activation(as measured by the BOLD response) that occurs as aparticipant is exposed to multiple repetitions of astimulus—a robust indicator of habituation [26]. Weutilized the differential RS effect in various brainregions to map sensitivity to repetitive security warningstimuli.Concurrent with the fMRI scan, we used an eye trackerto measure the eye-movement memory (EMM) effect—another robust indicator of habituation [43]. The EMMeffect manifests in fewer eye-gaze fixations and lessvisual sampling of the regions of interest within thevisual stimulus. Memory researchers have discoveredthat the EMM effect is a pervasive phenomenon inwhich people unconsciously pay less attention toimages they have viewed before. With repeatedexposure, the memories become increasingly available,thus requiring less visual sampling of an image [30].One strength of eye tracking is its temporal resolution,which allows researchers to measure with millisecondprecision the attentional process of participants’responses to repeated stimuli. Thus, fMRI (with highspatial resolution) and eye tracking (with high temporalresolution) complement each other, measuring both abehavioral manifestation of attention (i.e., eyemovements) as well as the neural activity that drivesattention.5.1ParticipantsWe recruited 16 participants from a large US university(eight male, eight female). This number of participantsis consistent with other fMRI studies [21]. Participantswere between 19 and 29 years of age (the mean age was23.3 years), right-handed, native English speakers, hadnormal or corrected-normal visual acuity, and wereprimarily PC users. One subject was excluded from thestudy due to scanner malfunction, resulting in 15 totalparticipants (eight male, seven female).1 Eachparticipant engaged in five fMRI scans: one at the sametime each day for five consecutive days. Upon arrival,participants were screened to ensure MRI compatibility.They were then given instructions about the task andplaced in the scanner. Each scan lasted 30 minutes,beginning with a structural scan and followed by twofunctional scans that displayed the warnings andimages.5.2EthicsThe university Institutional Review Board (IRB)approved the protocols used. Upon arrival at thefacility, participants completed a screening form toensure MRI compatibility. Participants were verballybriefed about MRI procedures as well as the task andpurpose of the experiment before entering the scanner.5.3Experiment DesignOur experimental design (Figure 3) consisted of fivesteps. In Step 1, computer-security warning images1We conducted a pilot study that revealed a large estimated effectsize for the repetition effect (partial eta2 .7). Using this estimatedeffect size, an a priori power analysis indicated that we would needfour subjects to achieve power greater than .8, indicating that asample size of 15 is more than adequate.5
6Figure 3. fMRI repetition-suppression-effect (RSE) longitudinal protocol.
were randomly split into two pools: one for the staticcondition and the other for the polymorphic condition.In Step 2, warnings in the polymorphic pool wererandomly assigned to one of the four variationsdepicted in Figure 2, with the order of polymorphicvariations also randomized. In Step 3, general softwareimages were randomly split into two sets of images. InStep 4, 20 of the images in the first set were shown fourtimes each, whereas the other 20 images were displayedonly once. These unique general software images wereused to create a baseline of unique presentationsthroughout the task. By comparing the responses foreach repeated image to the unique baseline images, wewere able to distinguish the habituation effect fromattention decay attributable to participants’ fatigue overtime.Overall, there were 260 images, randomized for eachparticipant, across two blocks of 7.7 minutes each, witha two-minute break in between blocks. Images weredisplayed for 3 seconds each, with a 0.5-secondinterstimulus interval. The technical details of the fMRIscans and procedures are documented in the appendix.6Figure 4. Activity in the right inferior temporal gyrus inresponse to each presentation of static and polymorphicwarnings. Beta values were extracted from a whole-brainanalysis for each subject and then averaged acrosssubjects according to stimulus condition.AnalysisWe analyzed each hypothesis separately for the fMRIand eye-tracking data. Our analyses are describedbelow, followed by tests of our hypotheses.6.1fMRI AnalysisMRI data was analyzed using the Analysis ofFunctional NeuroImages (AFNI) suite of programs [16](see appendix for details). Whole-brain, multivariatemodel analyses were conducted on the fMRI data toidentify significant clusters of activation, or regions ofinterest (ROIs), consistent with the hypothesizedpattern. All of our hypothesis tests utilized the sameROIs. Graphs of brain activity in response topolymorphic and static warnings over consecutive daysare presented for two brain regions in Figures 4 and 5.6.2Figure 5. Activity in the right ventral visual pathway inresponse to each presentation of static and polymorphicwarnings. Beta values were extracted from a whole-brainanalysis for each subject and then averaged acrosssubjects according to stimulus condition.Eye-Tracking AnalysisEye-tracking data was collected using an MRIcompatible SR Research EyeLink 1000 Plus (see Figure6). Fixations were defined as periods of time betweeneye movements that were not also part of blinks.Fixation count was used as the dependent variable ineach analysis.22We chose fixation count as a more appropriate measure ofhabituation than fixation duration because the warning stimuli weredisplayed to subjects for the same duration. However, we replicatedall analyses using fixation duration as the dependent variable and theresults were the same as those obtained using fixation count as thedependent variable.Figure 6. EyeLink 1000 Plus long-range eye tracker,mounted under the MRI viewing monitor.7
The number of fixations for polymorphic and staticwarnings per warning repetition per day is shown inFigure 7. The mean and standard deviations of fixationcount and fixation duration per day are shown in Table1. Some of the polymorphic warnings were animated,which prevented participants from fixating upon thewarning during the animation. To control for this, wenormalized all intercepts to zero and controlled forwarning type in the analysis, allowing for individualwarning intercepts. This control allowed us to focus onand accurately analyze how fixations changed over timeas an indicator of habituation.6.3Hypotheses Results6.3.1H1a Analysis: Users habituate to warningsover consecutive days.fMRI Analysis: We conducted a whole-brain,multivariate model analysis [12] on the fMRI dataholding gender,3 day, repetition number, and stimulustype (static warning and polymorphic warning) fixed, tofind areas that responded to a linear trend on daynumber, collapsing across repetitions and stimulustypes. In this analysis, two main ROIs were identified:the right and left insula. To quantify the extent of thedecrease in these ROIs, beta values were extracted forthese regions and tested using a within-subjects,repeated measures ANOVA. Both the right [F (1, 597) 67.87, p .001] and left insula [F (1, 597) 86.19, p .001] exhibited a significant habituation effect acrossdays (Table 2). Thus, the fMRI analysis supported H1a.Eye-Tracking Analysis: In a linear mixed-effectsmodel, we included fixation count as the dependentvariable and the subject ID and warning ID as randomfactors. The presentation number (across days) wastreated as a fixed factor, and visual complexity4 wasincluded as a covariate. The eye-tracking analysissupported H1a; the beta of presentation number acrossdays was significantly negative [χ2 (1, N 11,976) 212.89, p .001, β -0.1031], indicating habituation.Visual complexity was also significant [χ2 (1, N 11,976) 34.85, p .001, β 0.3815]. The R2 of themodel was 0.13.ROIs for Main Effect of 1FixationdurationSD (ms)450325384441444 .001 .001ROIs for Day by Stimulus-Type InteractionRegionp Value2.4867.8786.19p Value2.2730F Value2.18F Value2.65-16-16Peak zFixationcount SDPeak z7.35-4340Peak yDay 57.71Peak yDay 48.09160158Peak xDay 38.08Peak xDay 29.1R. insulaL. insula# VoxelsDay 1Fixationcountmean# VoxelsFigure 7. Change in eye gaze fixations across viewingsRegionL. middlefrontalgyrus19049-31185.19.02L. middleoccipitalgyrus1182576394.70.03Table 2. ROIs for habituation across days.Table 1. Absolute fixation count and fixation duration byday.3We controlled for sex because it has been shown to have asignificant effect on behavior relating to technology [e.g., 25; 39; 54].4A MATLAB script was used to calculate visual complexity [41].8
6.3.2H1b Analysis: Users habituate less topolymorphic warnings than to staticwarnings over consecutive days.fMRI Analysis: We conducted a whole-brain analysisfor a day by stimulus-type interaction. Two ROIs, theleft middle frontal gyrus [F (1, 595) 5.188, p .05]and left middle occipital gyrus [F (1, 595) 4.697, p .05], displayed a significant habituation interactionacross days and between stimulus types (Table 2).Eye-Tracking Analysis: We specified the same mixedeffects model as in H1a, except that we included aninteraction term between the presentation number(across days) and a polymorphic dummy variable(coded as 1 for polymorphic and 0 for static). The eyetracking analysis supported H1b; the interactionbetween the presentation number and polymorphicdummy was significantly positive [χ2 (1, N 11,976) 10.70, p .001, β 0.024], indicating that participantshabituate less to polymorphic warnings than to staticwarnings over the course of several days. The maineffects for both presentation number [χ2 (1, N 11,976) 493.42, p .001, β -0.115] and polymorphism [χ2(1, N 11,976) 64.71, p .001, β -0.725] were alsosignificant. Visual complexity, however, was notsignificant: χ2 (1, N 11,976) 0.17, p .05, β 0.026. The R2 of the model was .137.6.3.3H2a Analysis: If warnings are withheldafter habituation occurs, user responserecovers at least partially the next day.fMRI Analysis: We first calculated recovery scores bysubtracting the mean beta value of the last display ofeach stimulus type from the first display of thatstimulus type on the following day (i.e., Day 2 Display1 – Day 1 Display 4; etc.). A whole-brain, multivariatemodel analysis was then conducted to test for regionsthat displayed changes from baseline activation, which,collapsing across days, revealed four ROIs where therewas significant recovery. Post hoc analysis comparingspecific days showed significant recovery for Days 2–4in nearly every area, with no significant recovery onDay 5 (Table 3). Thus, H2a was supported by the fMRIdata.Eye-Tracking Analysis: We subtracted the fixationcount for the first viewing of a warning on a given dayfrom the fixation count of the last viewing of thewarning on the previous day. We then tested thishypothesis using a t-test. The test results supportedH2a: participants experienced significantly positiverecovery (m 0.369, sd 3.171) from day to day[t(2377) 5.672, p .001, d 0.233].6.3.4H2b Analysis: If warnings are withheldafter habituation occurs, response recoveryis stronger for polymorphic warnings thanfor static warnings the next day.fMRI Analysis: We analyzed the same ROIs found forH2a, but augmented the model by including stimulustype (polymorphic or static) as a factor. None of theregions displayed a significant recovery by stimulustype interaction (Table 3). Thus H2b was not supported.Eye-
3.1 Response Decay SMCT [49] explains that the brain creates a mental model when exposed to a stimulus (e.g., when seeing a warning). When people see the same stimulus again, they automatically and unconsciously compare the stimulus to this model. If the model and stimulus are similar, an blocking system in the brain inhibits
Eye Drops to Hold 48 Hours Before Allergy Testing Prescription Nasal Sprays Astelin Nasal Spray All Over-the-Counter Eye Drops Visine A Eye Drops Op-Con A Naph-Con A Alomide Eye Drops Prescription Eye Drops Patanol Eye Drops Zaditor Eye Drops Optivar Eye Drops Elestat Eye Drops Medicines That You MAY CONTINUE .
(plural teeth) Short, straight, brown hair and brown eyes Long, wavy, red hair and green eyes Short, curly, brown hair and green eyes Medium length, straight, black hair and dark eyes Long, straight, black hair and dark eyes Short, straight, blonde hair and blue eyes Long, wavy, brown hair and brown eyes Medium length, straight, dark blonde .
Tobii Eye Tracker 5, that could be simply transferred to other eye tracking devices, within the created Toolkit. The capacity of the Tobii Eye Tracker 5 to offer an assessment of the eye gaze independently for the left and right eyes, which is not afforded by the eye-tribe gadget, is an intriguing capability.
Eye Consultants of Atlanta Eye Doctors of Washington Eye Physicians and Surgeons of Atlanta Glaucoma Care Center Great Lakes Ophthalmology Henry Ford Hospitals Johns Hopkins University Jules Stein Eye Institute, UCLA Kellogg Eye Center Kresge Eye Institute Krieger Eye Institute Maryland Center for Eye Care
The term eye gaze tracking includes also some works that track the whole eye, focus on the shape of the eye (even when it is closed) and contains the eyebrows in the eye tracking. Consequently, the term eye gaze tracking represents a lot of different types of eye tracking [6]. 2.1.2 Eye Gaze and Communication Gaze is one of the main keys to .
Dry eye syndrome or dry eye disease, is a common condition that occurs when the eyes do not make enough tears, or the tears evaporate too quickly. Most cases of sore tired eyes resolve themselves. Advice to patients Patients should be encouraged to manage both dry eyes and sore eyes by implementing some self care measures such as
My views on Hong Kong Eye Care Professionals They are caring and patient to teach us about eye care A professional on eye care to assess our eye conditions Help us to protect our eyes Check my eyes for short-sightedness and long-sightedness Examine my eyesight and help me to choose my spectacles Provide eye care and information on improving our .
the non-dominant (recessive) form. Returning to the example of eye colour: brown eyes are more common than blue eyes because the gene for brown eyes is dominant over the gene for blue. This means that whenever the brown eye gene is present, it will overrule the recessive blue eye gene. The same applies to dominant and recessive mito genes.
