PROCEEDINGS of the HUMAN FACTORS and ERGONOMICS SOCIETY 56th ANNUAL MEETING - 2012

1639

Missing Critical Auditory Alarms in Aeronautics: Evidence for Inattentional Deafness? Frédéric Dehais1, Mickaël Causse1, Nicolas Régis1, Eric Menant1, Patrice Labedan1, François Vachon2, & Sébastien Tremblay2 1 ISAE, Toulouse, France; 2Université Laval, Québec, Canada The inability of pilots to detect unexpected changes in the environment (e.g., auditory alarms) is a critical problem in aeronautics. The lack of response to alarms is not thought to be a perceptual/attentional issue, but rather that pilots choose to ignore such warnings due to cognitive biases. In the current paper we consider an alternative explanation, by extending the phenomenon of inattentional deafness to aeronautics. Fourteen pilots equipped with an eye tracker and an electrocardiogram performed landings in a flight simulator. During the critical landing, an auditory landing gear alarm was triggered while the volunteers also faced a windshear. Eight out of 14 pilots did not report the occurrence of the critical alarm during the debriefing. Interestingly, all but one of these ‘deaf’ pilots failed to perform the adequate go-around behavior. These findings establish inattentional deafness as a cognitive phenomenon that is critical for air safety.

Copyright 2012 by Human Factors and Ergonomics Society, Inc. All rights reserved. DOI 10.1177/1071181312561328

INTRODUCTION Auditory alerts are known to present various advantages in emergency situations compared to visual stimuli. In aeronautics, they provide information for pilots without requiring head/gaze movements (Edworthy, Loxley, & Dennis, 1991) and provoke faster reaction times (Wheale, 1981). Yet, the analysis of air safety reports reveals that a consequent number of accidents are due to a lack of reaction to auditory alarms (Bliss, 2003). Two reasons related to cognitive biases are usually proposed to account for such a lack of response. A first and major explanation is to consider that the unreliability of these alerting systems, that are likely to provoke the so-called ‘Cry-wolf effect’ (Breznitz, 1984; Wickens et al., 2009), also leads the pilots to mistrust the alarms (Shapiro, 1994; Song & Kuchar, 2001; Sorkin, 1988) especially under high workload conditions (Bliss & Dunn, 2000). A second explanation is that the aggressive, distracting, and disturbing nature of auditory alarms (Doll, Folds, & Leiker, 1984; Edworthy et al., 1991) can considerably increase pilot stress levels during warning events (Peryer, Noyes, Pleydell-Pearce, & Lieven, 2005). In fact, the immediate reaction for many pilots is to find a way to consciously silence the noise, rather than analyze the meaning of the alert. Nevertheless, these considerations are not sufficient to fully account for the misperception of critical auditory warnings— that is, the absence of reaction to auditory alarms—as often reported in accident analyses (Bea, 1993) and observed in flight simulators (Dehais, Tessier, Christophe, & Reuzeau, 2009). A complementary explanation is to consider the role of perceptual and attentional processes engaged in the cockpit. There is evidence that tasks involving high perceptual load consume most of attentional capacity, leaving little or none remaining for processing any task-irrelevant information (see Lavie, 1995). Yet, high-load contexts tend to prevent the perceptual processing of task-irrelevant information and thus promote various forms of inattentional blindness (Mack & Rock, 1998). This propensity to remain unaware of unexpected, though fully perceptible stimuli is not limited to vision, however. Indeed, there is evidence that unexpected

salient sounds can remain unnoticed under attentiondemanding settings (e.g., Koreimann, Strauß, & Vitouch, 2009; Spence & Read, 2003; Vachon, Tremblay, Nicholls, & Jones, 2011; Wood & Cowan, 1995). Although less wellknown than its visual counterpart, this inattentional deafness phenomenon is likely to have important consequences for safety-critical situations such as military or commercial flights as pilots are often overloaded with visual information. In fact, the capacity of acknowledging an unexpected auditory stimulus has been shown to diminish when engaged in a visual task of high perceptual load (e.g., Macdonald & Lavie, 2011). Hence, since visual activity is a key element of flying, it is possible that during critical flight phases, visual information processing could interfere with concurrent appraisal of auditory alarms and then induce inattentional deafness. The objective of the present study is therefore to assess whether inattentional deafness is likely to occur in the context of flying and, if so, the potential impact of such phenomenon on the pilot’s behavior. An experiment was conducted using a 3-axis flight simulator and 14 pilots equipped with an eye tracker and an electrocardiogram (ECG) had to perform landings in various conditions. During the scenario, a landing gear failure was triggered together with an associated triple chime auditory alarm. On the detection of this alarm, the pilot should abort the landing and perform a go-around maneuver. In order to increase processing load, the landing gear failure occurred while pilots simultaneously faced a buffetinginducing windshear. We expected the pilots to be so focused on handling the windshear that they would become particularly susceptible to fail to notice the landing gear auditory alarm. Inattentional deafness and subsequent maneuvers were assessed by combining behavioral metrics (e.g., pilot’s actions, self-report, ocular behavior) to psychophysiological measures (e.g., heart rate). METHOD Participants Fourteen healthy pilots (all men; mean age = 29.9 years, SD = 9.5; flight experience: 419 hours, range = 30-3500 hours), all

PROCEEDINGS of the HUMAN FACTORS and ERGONOMICS SOCIETY 56th ANNUAL MEETING - 2012

Frencch defense stafff from Institut Supérieur de l’Aéronautique l e et dee l’Espace (ISAE) campuss, were recru uited by locall adverrtisement. Fligh ht Simulator

1640

d Figurre 2) and a tripple chime auditory alarm. Thhis event should lead pilots to aboort the landingg and perform m a go-around proceedure. The sceenario ended w when participannts reached the landiing ground touuchdown area, whatever theeir altitude and the ddisplays were sw No crashes weree simulated. witched off. N

A 3-axis motio on flight simulaator was used to conduct thee experriment (see Fig gure 1). It sim mulates a twin-engine aircraft ft flightt model, and itt is able to rep produce aerody ynamic effectss such as buffeting (ii.e., aircraft vib bration during stall). Its userr interfface is compossed of a Primary Flight Dissplay (PFD), a simpllified Head Up U Display (H HUD) compriising a speed d vectoor, a Navigatio on Display (ND D), and the up pper Electronicc Centrral Aircraft Mo onitoring Display (ECAM) page. p The pilott has a stick to contro ol the flight, a rudder, r and two o thrust levers.

Figure re 2: The landing gear indicator w was situated in thee lower part of th he PFD. In the scenario,, the undercarriaage sequence failed as the “nose” ked at 900 feet (““one red” and “tw wo green” instead wheell was still not lock of “th hree green”).

Proccedure

Figuree 1: The ISAE flig ght simulator used for the experim mentation.

T Two stereophon nic speakers, located l under the t displays on n each side of the cabin, c were ussed to broadcaast continuouss radio communicatio on and engine sound (77dB) and to triggerr ms (single chiime, triple chiime, repetitivee four types of alarm chimee and pull up [86.3dB] 8.5 times t louder than the globall ambieent cockpit sound). A dedicated software s wass impleemented to au utomatically manage m the diifferent eventss (e.g., failure, gusts of wind) that occurred durin ng the landingss (see nnext section). “Crittical landing” scenario T The scenario consisted c of a manual landin ng on the 14R R runwaay at Blagnacc airport (Tou ulouse, Francee). The initiall condiitions were deefined as follo ows: 2500 feett, heading 142 2 degreees, 130 knots, visibility 8100 0 m, slight rain n, landing flapss configguration, the landing l gear was w in transit (“three red”). The landing gear sequence wass supposed to o be completee (“threee green”) beffore the aircrafft reached an altitude a of 900 0 feet. At 900 feet, participants p faaced a slight windshear w thatt criticaally dropped the t speed of the t aircraft. Simultaneously,, the ffailure of thee undercarriag ge sequence occurred and d particcipants were warned w through h the landing gear indicatorr (“twoo green” and “one red”, in nstead of “threee green”; seee

P Participants were told that thhe purpose of the experimen nt was to analyze carrdiac responsees and visual ppatterns during g landiings. A 20-m min PowerPoinnt presentationn detailed the functtioning of the simulator (usser interface, im mportant fligh ht param meters). In pparticular, theyy were explaained that five differrent events w were likely too occur duringg landings: an n antiskkid failure (sim mulated by an auditory “singgle chime”), an n enginne failure (sim mulated by ann auditory “reppetitive chime” and a red warningg on the corressponding enginne indicator), a decission height isssue (poor exxternal visibiliity); a ground proxiimity issue (siimulated by ann auditory “Puull Up” alarm)); and eventually a landing geaar failure (sim mulated by an n audittory “triple chiime” warning and “one red aand two green” on thhe landing geaar indicator iff the undercarrriage sequence was nnot completed at 900 feet). T The associate pprocedures were explaained (respectiively: antiskidd failure: “do nnot exceed 130 knotss at touchdow wn”; engine faiilure: “set thee corresponding g throtttle lever on iidle and use tthe rudder”; ddecision heigh ht issuee: “perform a ggo-around if the runway is noot visible at 200 feet””; ground proxximity issue: “perform an immediate go oarounnd”; and a landding gear failuure: “proceed too an immediate go-arround to furthher recycle thee landing gearr”). Participantts sat inn the flight sim mulator and thhe sensors (ECG, eye trackerr) weree set before sttarting a 5-minn resting periood without any y n, stimuulation. They then completeed a 1-hour trraining session withoout simulator motion, in whhich they perfformed manuaal landiings, in particuular supervisingg the automaticc undercarriage sequeence that was supposed to ennd before 900 feet (every 250 feet, a wheel wass locked one after the othher). Each new w landiing was progreessively more ddifficult due too slight changes in lannding conditionns (i.e., stronger crosswind, llower visibility y, etc.).. During the trraining, the diffferent alarms were presented threee times and thhe participantss were asked to identify the eventts and recall thhe associated pprocedures. Paarticipants were also trained to fill out a questionnnaire and a sself-report afteer

PROCEEDINGS of the HUMAN FACTORS and ERGONOMICS SOCIETY 56th ANNUAL MEETING - 2012

each landing (see next section). After the 9th practice landing, the simulator motion was engaged to reproduce realistic flight feelings and a continuous radio communication was also broadcasted to reproduce more realistic flight conditions. The “landing-gear failure” scenario was then initiated (see preceding section).

1641

Wilcoxon Signed-rank test was used for every within-subject comparison while the Mann-Whitney test was employed for between-subjects comparisons. RESULTS Alarm Detection and Expected Behavior

Measurements Subjective measurements: Participants were asked to fill out a 4-item questionnaire directly after the end of the scenario. The questions were: “Describe the weather and wind conditions”, “Describe the status of the aircraft”, “Describe the special events you have faced”, and “Describe your actions and decisions”. From this questionnaire, we extracted information to check the perception of the auditory alert (“Auditory alert perception”: Yes/No), to check the perception of the failure via the landing gear indicator (“Visual alert perception”: Yes/No) and to analyze the decision that led to a go-around when one had been performed (“Origin of the goaround”: Failure/Unstabilized approach; see Table 1). Moreover, a self-report of psychological stress and mental workload was collected using a visual analog scale (1 for very low, 7 for very high). Ocular measurement: A Pertech® head-mounted eye tracker was used to analyze participants’ ocular behavior. This 80-g non-intrusive device has 0.25° of accuracy and a 50-Hz sampling rate. A dedicated software (EyeTechLab) provided data such as timestamps and the x,y coordinates of the participants’ eye gaze on the visual scene. Eye-tracking data were used to check whether participants glanced at the landing gear indicator during the scenario and the associated timing (“Timing to glance at landing gear indicator”; see Table 1). Moreover, we calculated and compared the proportion of the time spent performing saccades during the two segments of the flight using dedicated algorithm (Regis, Dehais, Tessier & Gagnon, 2012): S1 (beginning of the scenario until the simultaneous occurrence of the failure and the windshear) and S2 (just after the occurrence of the failure until the aircraft had reached the landing ground threshold). Heart rate measurement: An ECG was used to collect participants’ cardiac activity at a sampling rate of 2048 Hz with the Biopac® system. Three electrodes connected to an extender cable were applied to participants’ chest using UniGel to enhance the quality of the signal. The Biopac Acqknowledge software was used to export and filter the heart rate (HR) derived from the inter-beat-interval. Due to a commonly observed difference in HR baseline values among participants, HR values were then standardized to provide an inter-participant comparison. The mean HR of the resting period was subtracted from the mean HR during the scenario (Bonner & Wilson, 2002). We calculated and compared the mean HR during the two segments S1 and S2. Statistical Analysis All behavioral data were analyzed with Statistica® 7.1 (StatSoft). A Kolmogorov-Smirnov goodness-of-fit test indicated that our variable distributions were not normal, so a

Eight participants (57.1%) reported that they did not hear any alarms during the scenario (see Table 1). We will refer to them as the “deafness group”. Among this group, four pilots (50%) landed despite the landing gear failure and three (37.5%) performed a go-around in the vicinity of the landing ground because they felt too unstabilized by the windshear. In this group, only one participant (see COPJE in Table 1) declared that he detected the failure visually and that this led him to go around. This was confirmed by eye-movement data showing that the participant glanced at the landing gear indicator 4.0 s after the failure occurred and before initiating the go-around maneuver. No other member of the deafness group glanced at the visual landing gear indicator after the failure. The remaining six participants (42.9%) declared they have heard the triple chime warning during the scenario. They are the “alarm group”. Among this group, five participants (83.3%) reported that they also visually detected the failure and the combination of these visual and auditory warnings led them to correctly perform a go-around. This was confirmed by eye-tracking data showing that four of these five participants glanced at the landing gear indicator just before initiating the maneuver. The remaining participant (see CELOL in Table 1) declared that the combination of the auditory alarm and the windshear directly led him to a reflex action to go around. He did not report being aware of the landing gear failure. Table 1. Behavioral performances of the participants.

NIZJE HUORE COPJE GAKAR MINPH CROCH VIGYA PAPDA BRELU GORYA CELOL BREGR MONTH COUBE

Auditory alert perception

Visual alert perception

No No No No No No No No Yes Yes Yes Yes Yes Yes

No No Yes No No No No No Yes Yes No Yes Yes No

Timing to glance at landing gear indicator (s) 4.00 8.24 00.5 0.49 0.50 -

Origin of the go-around Failure Unstabilized Unstabilized Unstabilized Failure Failure Unstabilized Failure Failure Failure

To assess whether the conscious perception of the auditory alarm was associated to the appropriate go-around behavior, we performed a chi-square test on the frequencies displayed in Table 2. With a phi coefficient of .708, the test revealed a strong significant relation between the subjective detection of the alert and the subsequent application of the

PROCEEDINGS of the HUMAN FACTORS and ERGONOMICS SOCIETY 56th ANNUAL MEETING - 2012

1642

expeccted maneuver, χ2(1) = 7.02, p = .008. In faact, odds ratioss reveaaled that a pilott had 35 times more chances to perform thee approopriate go-arou und maneuver following thee landing gearr failurre if he was aw ware of the aud ditory alarm thaan if the alarm m remaiined unnoticed d. Table 2. Number of pilots who performed the exp pected go-around d uver or not accorrding to whether they reported ha aving detected thee maneu auditoory alarm or not.

Auditory alert detected

Yes No

Expected go-arround due to gear fa ailure Yes No 5 1 1 7 F Figure 4: Mean H HR response (+SE E) for the two fligght segments.

Saccaadic Behaviorr T The analysis revealed r that the percentag ge of saccadess durinng S2 (20.44%,, SD = 15.87) was w significantly higher than n durinng S1 (12.25%, SD = 9.79), Z = 3.23, p = .00 01. bjective Stress Subjeective and Ob A As shown in Figure F 3, the subjective s stresss score of thee alarm m group was sttatistically high her than the deafness d group,, U = 44.50, Z = -2.02 2, p = .043, sug ggesting that pilots who havee detected the auditory alert felt more m stressed th han those who o failedd to consciou usly perceived d the alarm. However, no o differrence in subjective workload d was found between b the 2 groupps (p = .807).

F Figure 3: Mean subjective stress score s (+SE) for th he deafness group p (N = 8)) and the alarm group g (N = 6).

W With regards to t the objectivee stress level, the analysis off the caardiovascular response r reveaaled a significaantly faster HR R durinng S2 compared d to S1, Z = 2.91, 2 p = .003 (see ( Figure 4). This rresult indicates that the levell of stress incrreased after thee simulltaneous occurrrence of the landing l gear failure f and thee windsshear. Howeveer, no differencce in HR respo onse was found d betweeen the deafneess group and the t alarm grou up, both during g S1 (pp = .651) and S2 (p = .477). Thus, T although pilots from thee alarm m group reporteed a higher streess level than those from thee deafnness group, such difference did not transllate to a moree objecctive measure of o stress.

DISCUSS SION T The objective of this study w was to show thhat inattentionaal deafnness could be oone cause of aiircraft pilots’ innability to reacct to auuditory alarms. The results off this experimeentation seemed d to suupport this hypoothesis as 57.11% of our pilotts reported they y neverr perceived thee alarm. At first blush, one ccould claim thaat this iinability to reccall the presennce of the alarm m ensued from m shortt-term memoryy rather than attentional lim mitations given n that pparticipants weere debriefed 330 s after the aauditory stimulli (insteead of 2 s as in the tyypical inattenttional deafnesss paraddigm; see Maacdonald & Laavie, 2011). H However, taken n togetther, the ressults were rrather consisttent with the inatteentional deafneess hypothesiss as awareness of the auditory y alert was strongly related to thee application oof the expected go-arround maneuvver. More preciisely, of those pilots who did not rreport hearing tthe alarm, all bbut one failed tto looked at the visuaal landing gearr indicator afteer the alarm annd also did no ot perfoorm the approopriate go-arouund behavior; in contrast, of o thosee pilots who ddid consciouslyy perceive thee auditory alertt, all bbut one delibeerately executeed the go-arouund proceduree. Moreeover, the analyysis of the subj bjective stress rresults appeared d to al so support ourr hypothesis: tthe subjective stress score of o the ddeafness group was statisticallly inferior to tthe score of the alarm m group. Thiss could suggeest that ‘deaff’ pilots felt a posteeriori less stresssed as they w were unaware oof the failure in n compparison to pilots who were aw ware of it and had to manage the ffailure correcttly. Taken together, these findings poin nt towaards the exiistence of the inattentioonal deafness phennomenon in thhe cockpit andd suggest that such cognitive limitaation may leaad to inapproppriate decisionn making and d, conseequently, havee dramatic conssequences. Succh a conclusion n in th e auditory dom main is in linee with researchh demonstrating g that failures to vissual awareness in safety-criitical situation ns may be disastrous ((see Varakin, L Levin, & Filderr, 2004). IIn this experriment, we trriggered a criitical event to increease the processsing load in oorder to promoote inattentionaal deafnness. Physiological results sshowed that thhe managemen nt of thhe second flighht segment indduced a higherr HR responsee, sugg esting an incrreased mobilizzation of menntal energy and psychhological stress (Causse, S Sénard, Démonet, & Pastorr, 20100). Moreover, the managem ment of this seecond segmen nt inducced a higher saccadic activvity reflecting a more active inforrmation sourcee scanning (G Goldberg & K Kotval, 1999)),

PROCEEDINGS of the HUMAN FACTORS and ERGONOMICS SOCIETY 56th ANNUAL MEETING - 2012

probably to maintain a correct flight path and avoid stall. Overall, these results tend to confirm that the postfailure/windshear segment increased mental workload compared to the first flight segment. One could consider that the management of the critical windshear saturated the attentional abilities of some of the participants and led to inattentional deafness. However, the current study was unable to provide strong empirical support for this attentional hypothesis given that the small sample did not reach a level of statistical power sufficient to distinguish between the deafness and alarm groups on the basis of their physiological and ocular responses. Conclusions regarding the origin of inattentional deafness in the present setting will certainly benefit from the addition of new participants. Nevertheless, the present study provides a striking indication as to the prevalence and potential outcomes of inattentional deafness in aeronautics. The present study highlights a paradox that user interface designers must face: How can one alert the human operators to potential inattentional deafness, if the alarm systems designed to warn them are neglected? Rather than adding new alarms, an optimal solution could be therefore to use cognitive countermeasures (Dehais, Tessier, & Chaudron, 2003), shown to be effective with light aircraft pilots and commercial pilots (Dehais et al., 2010), and unmanned vehicle operators (Dehais, Causse, & Tremblay, 2011) who failed to take into account critical warnings. Derived from a neuroergonomics approach to cognitive biases (Parasuraman & Rizzo, 2007), cognitive countermeasures are based on a temporary simplification of the user interface in a way to display solely relevant and critical information. REFERENCES BEA. (1993) Bureau Enquête Analyse - Accident investigation report GF072. Technical report a40-ek000823a. Bliss, J. P. (2003). Investigation of alarm-related accidents and incidents in aviation. The International Journal of Aviation Psychology, 13(3), 249-268. Bliss, J. P., & Dunn, M. C. (2000). Behavioural implications of alarm mistrust as a function of task workload. Ergonomics, 43, 1283-1300. Bonner, M. A, et G. F Wilson. 2002. « Heart rate measures of flight test and evaluation ». The International Journal of Aviation Psychology 12 (1): 63–77. Breznitz, S. (1984). Cry wolf: The psychology of false alarms. Lawrence Erlbaum Associates. Causse, M, Sénard, J. M., Démonet, J. F., & Pastor, J. (2010). Monitoring cognitive and emotional processes through pupil and cardiac response during dynamic versus logical task. Applied Psychophysiology and Biofeedback, 35, 115-123. Dehais, F., Causse, M., Tremblay, S. (2011). Mitigation of conflicts with automation. Human Factors, 53, 448-460. Dehais, F, Mercier, S, & Tessier, C. (2009). Conflicts in human operator and unmanned vehicles interactions. In R. Goebel, J. Siekmann & W. Wahlster (Series Eds.) & D. Harris (Vol. Ed.), Engineering psychology and cognitive ergonomics: Lecture Notes in Computer Science, Vol. 5639 (pp. 498-507). Berlin: Springer-Verlag. Dehais, F., Tessier, C., Christophe, L., & Reuzeau, F. The perseveration syndrome in the pilot's activity: Guidelines and cognitive countermeasures. Human Error, Safety and Systems Development, 68-80. Dehais, F, Tessier, C, & Chaudron, L. (2003). Ghost: Experimenting conflicts countermeasures in the pilot's activity. Proceedings of the

View publication stats

1643

International Joint Conference on Artificial Intelligence, 18, 163168. Doll, T. J., Folds, D., & Leiker, L. A. (1984). Auditory information systems in military aircraft: Current configurations versus the state of the art. Final Report, 1 May-30 Sep. 1983 Georgia Inst. of Tech., Atlanta. Systems Engineering Lab. Edworthy, J., Loxley, S., & Dennis, I. (1991). Improving auditory warning design: Relationship between warning sound para-meters and perceived urgency. Human Factors, 33, 205-231. Goldberg, J. H. & Kotval, X. P. (1999). Computer interface evaluation using eye movements: Methods and constructs. International Journal of Industrial Ergonomics, 24, 631-645. Koreimann, S., Strauß, S., & Vitouch, O. (2009). Inattentional deafness under dynamic musical conditions. Proceedings of the 7th Triennial Conference of European Society for the Cognitive Sciences of Music (pp. 246-249). Lavie, N. (1995). Perceptual load as a necessary condition for selective attention. Journal of Experimental Psychology: Human Perception and Performance, 21, 451-468. Macdonald, J. S. P., & Lavie, N. (2011). Visual perceptual load induces inattentional deafness. Attention, Perception, & Psychophysics, 73, 1780-1789. Mack, A., & Rock, I. (1998). Inattentional blindness. Cambridge: The MIT Press. Parasuraman, R, & Rizzo, M. (2007). Neuroergonomics: The brain at work. New York: Oxford University Press. Peryer, G., Noyes, J., Pleydell-Pearce, K., & Lieven, N. (2005). Auditory alert characteristics: A survey of pilot views. International Journal of Aviation Psychology, 15, 233-250. Regis, N., Dehais, F., Tessier, C., & Gagnon, J.F. (2012), Ocular metrics for detecting attentional tunnelling, Human Factors and Ergonomics Society – Chapter Europe, Toulouse, France. Shapiro, N. (1994). NBC Dateline. New York: National Broadcasting Corporation. Song, L., & Kuchar, J. K. (2001). Describing, predicting, and mitigating dissonance between alerting systems. Proceedings of the 4th International Workshop on Human Error, Safety, and System Development. Linkoping, Sweden. Sorkin, R. D. (1988). Why are people turning off our alarms. Journal of the Acoustical Society of America, 84, 1107-1108. Spence, C., & Read, L. (2003). Speech shadowing while driving: On the difficulty of splitting attentions between eye and ear. Psychological Science, 14, 251-256. Vachon, F., Tremblay, S., Nicholls, A. P., & Jones, D. M. (2011). Exploiting the auditory modality in decision support: Beneficial “warning” effects and unavoidable costs. Proceedings of the Human Factors and Ergonomics Society 55th Annual Meeting (pp. 14021406). Santa Monica, CA: Human Factors and Ergonomics Society. Varakin, D., Levin, D. T., & Fidler, R. (2004). Unseen and unaware: Implications of recent research on failures of visual awareness for human-computer interface design. Human-Computer Interaction, 19, 389-422. Wheale, J. L. (1981). The speed of response to synthesized voice messages. British Journal of Audiology, 15(3), 205-212. Wickens. C. D.. Rice. S.. Keller. D.. Hutchins. S.. Hughes. J.. & Clayton. K. (2009). False Alerts in Air Traffic Control Conflict Alerting System: Is There a “Cry Wolf” Effect? Human Factors: The Journal of the Human Factors and Ergonomics Society. 51(4). 446. Wood, N., & Cowan, N. (1995). The cocktail party phenomenon revisited: Attention and memory in the classic selective listening procedure of Cherry (1953). Journal of Experimental Psychology: General, 124, 243-262.