J. Kozey, J. McCabe and J. Jenkins School of Health and Human Performance Dalhousie University, Halifax, Nova Scotia The effect of different training methods.

J. Kozey, J. McCabe and J. JenkinsJ. Kozey, J. McCabe and J. JenkinsSchool of Health and Human PerformanceSchool of Health and Human PerformanceDalhousie University, Halifax, Nova ScotiaDalhousie University, Halifax, Nova Scotia

The effect of different training The effect of different training methods on egress performance methods on egress performance from the Modular Egress from the Modular Egress Training SimulatorTraining Simulator

IntroductionIntroduction

Helicopters provide an important mode Helicopters provide an important mode of transportation in the civilian, industrial of transportation in the civilian, industrial and military sectorsand military sectors

They are especially important for rapid They are especially important for rapid transport to locations with limited spacetransport to locations with limited space

Unfortunately like all transportation Unfortunately like all transportation systems adverse incidents happen – systems adverse incidents happen – example ditchingsexample ditchings

DitchingsDitchings Most ditchings result in the helicopter inverting Most ditchings result in the helicopter inverting

(77%, Taber & McCabe, SAFE – 2005).(77%, Taber & McCabe, SAFE – 2005). Survivors must content with immersion, Survivors must content with immersion,

disorientation and debrisdisorientation and debris Statistical evidence that training will improve the Statistical evidence that training will improve the

chance of surviving a ditching (Cunningham, 1978)chance of surviving a ditching (Cunningham, 1978) 234 Helicopter Mishaps/1093 Occupants234 Helicopter Mishaps/1093 Occupants Survival Rate: Survival Rate:

66%66% without dunker training without dunker training 91.5%91.5% with dunker trainingwith dunker training

Egress Training Egress Training

While the evidence indicates that training While the evidence indicates that training is beneficial:is beneficial: Little is written about skilled performance in Little is written about skilled performance in

egress training egress training Little is known about practice, training fidelity Little is known about practice, training fidelity

and retentionand retention

The Development of Skilled The Development of Skilled Behaviour Behaviour

(Wickens, Lee, Liu, & Becker, 2004)(Wickens, Lee, Liu, & Becker, 2004)

PracticePractice More practice = better performanceMore practice = better performance

www.croquetworld.com/images/Bamfordquote1.gif

Simulations/Training fidelitySimulations/Training fidelity

Lower fidelity Higher fidelity

Success Success “Trainees are required to demonstrate the ability, underwater in an inverted HUET, to operate an escape exit mechanism, release a representative seat restraint, and effect an escape unaided.”

(Mills and Muir, 1998)

Research QuestionsResearch Questions Training Methods:

Is there a difference in HUET performance as a result of different training protocols?

Fidelity: Is escape success affected by prior training of

underwater window ejection skills? Practice:

Is escape success affected by the number of prior training trials?

HypothesesHypotheses There is no difference between the There is no difference between the

Pass/Fail rates of the different groups Pass/Fail rates of the different groups during Session 1: Training.during Session 1: Training.

There is a difference in Pass/Fail rates There is a difference in Pass/Fail rates when a window must be ejected.when a window must be ejected.

There is a difference in HUET egress There is a difference in HUET egress Pass/Fail rates among the different Pass/Fail rates among the different training groups at Session 2: Re-test. training groups at Session 2: Re-test.

Overall PlanOverall Plan

Train three groups of people in HUET Train three groups of people in HUET escape (Session 1)escape (Session 1) Each group will receive a different training Each group will receive a different training

experience experience Bring the groups back to perform 1 HUET Bring the groups back to perform 1 HUET

trial (Session 2) and measure their trial (Session 2) and measure their performanceperformance

MethodsMethods Study was conducted over a 6 month period Study was conducted over a 6 month period

Session 1: Training session (Oct. 2005)Session 1: Training session (Oct. 2005) Session 2: Testing session (Apr. 2006)Session 2: Testing session (Apr. 2006)

SubjectsSubjects Recruited via posters, flyers and ads through Recruited via posters, flyers and ads through

newspapers, fire stations and TVnewspapers, fire stations and TV Inclusion/Exclusion:Inclusion/Exclusion:

No prior HUET experienceNo prior HUET experience Successfully complete a medical examSuccessfully complete a medical exam Indicate a willingness to attend both sessionsIndicate a willingness to attend both sessions Sign a document of informed consent Sign a document of informed consent

Subject recruitment and assignmentSubject recruitment and assignment

211 initial volunteers from whom 191 entered the 211 initial volunteers from whom 191 entered the Training session after medical screening and Training session after medical screening and informed consentinformed consent

Subjects assigned to 1 of 3 Groups balanced for Subjects assigned to 1 of 3 Groups balanced for sex (males and females) and self-reported “water sex (males and females) and self-reported “water experience/comfort”experience/comfort”

Water experience:Water experience: Level 1: no formal swimming trainingLevel 1: no formal swimming training Level 2: swimming trainingLevel 2: swimming training Level 3: competitive swimmer, SCUBA diving Level 3: competitive swimmer, SCUBA diving

experienceexperience

Training sessionTraining session

Fidelity

Experimental daysExperimental days Session 1:Session 1:

40 minute classroom 40 minute classroom sessionsession

Change into coverallsChange into coveralls Tested in pairsTested in pairs Pre-test saliva sample Pre-test saliva sample

& questionnaire& questionnaire HUET trials (subjects HUET trials (subjects

and evaluator blind to and evaluator blind to group assignment)group assignment)

Post-test saliva & Post-test saliva & questionnairequestionnaire

Session 2:Session 2: 10 minute classroom 10 minute classroom

sessionsession Change into coverallsChange into coveralls Pre-test saliva sample & Pre-test saliva sample &

questionnairequestionnaire Watched a “pre-flight” Watched a “pre-flight”

video (45 seconds)video (45 seconds) 1 HUET trial (evaluator 1 HUET trial (evaluator

blind to previous group blind to previous group assignment)assignment)

Tested in alone or pairs Tested in alone or pairs (same side of METS)(same side of METS)

Post-test saliva & Post-test saliva & questionnairequestionnaire

Training sessionTraining sessionGroup Conditions   Description   # of trials

1 1 Immersion Straight In No window 1

2 Immersion 180° inversion No window 1

2 1 Immersion Straight In No window 1

2 Immersion 180° inversion No window 1

  3 Immersion 180° inversion Window in 1

3 1 Immersion Straight In No Window 1

2 Immersion 180° inversion No Window 1

  3 Immersion 180° inversion Window in 4

Dependent measuresDependent measures

Egress time (seconds): The time recorded Egress time (seconds): The time recorded from when the METS touches the water to from when the METS touches the water to when the subject breaks the water when the subject breaks the water surface. surface.

Pass/Fail: The performance of the subject Pass/Fail: The performance of the subject in executing the egress according to a in executing the egress according to a predefined rating scale.predefined rating scale.

Two subjectsTwo instructorsTwo Evaluators

Timers and other research assistants

Two safety divers

Performance Evaluation Performance Evaluation Categories (METS™ Underwater Categories (METS™ Underwater

Escape)Escape) A4: Unaided escape, correct sequence, window A4: Unaided escape, correct sequence, window

released on first attempt.released on first attempt. A3: Unaided escape, correct sequence, window not A3: Unaided escape, correct sequence, window not

released on first attempt.released on first attempt. A2: Unaided escape, incorrect sequence, window A2: Unaided escape, incorrect sequence, window

released on first attempt.released on first attempt. A1: Unaided escape, incorrect sequence, window not A1: Unaided escape, incorrect sequence, window not

released on first attempt.released on first attempt. B: Instructor assistance required with window release.B: Instructor assistance required with window release. C: Instructor assistance with seatbelt required.C: Instructor assistance with seatbelt required. D: Instructor assisted evacuation from the METS™.D: Instructor assisted evacuation from the METS™. E: Unwilling to proceed, voluntarily leaves the pool E: Unwilling to proceed, voluntarily leaves the pool

deck.deck.

Subject Allocation forSubject Allocation for training training sessionsession

(group by sex by water levels) (group by sex by water levels)Group Level 1 Level 2 Level 3 Grand

Female Male Female Male Female Male totals

1 1 17 7 17 6 14 62

2 3 15 6 17 5 13 59

3 4 13 8 16 4 21 66

All 8 45 21 50 15 48 187

4 people withdrew and did not complete their trials

Session 1 : Egress timesSession 1 : Egress times

Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6

Mean 12.7 13.2 17.0 16.4 16.6 16.2Max 24.3 21.6 25.6 24.0 23.5 25.0Min 8.4 8.5 10.9 10.0 11.7 11.5SD 2.7 2.4 2.8 2.6 2.4 2.9

Training Results: Percent success versus Training Results: Percent success versus trialstrials

50

55

60

65

70

75

80

85

90

95

100

1 2 3 4 5 6

Trial number

Per

cen

t su

cces

sfu

l (%

)

Push-outexit

No Push-out exit

14%

Groups 1,2 & 3 Groups 2 & 3 Group 3

97%

94%

Session 2 – April 2006Session 2 – April 2006

153 of the initial 187 returned for session 2153 of the initial 187 returned for session 2 11 dropped out of Group 111 dropped out of Group 1 13 dropped out of Group 213 dropped out of Group 2 14 dropped out of Group 314 dropped out of Group 3

Reasons:Reasons: 6 not comfortable, 2 attended but did not participate6 not comfortable, 2 attended but did not participate 8 had moved away and 3 had conflicts8 had moved away and 3 had conflicts 8 medical/illness8 medical/illness 13 scheduled but did not attend13 scheduled but did not attend

Session 2: Subject allocation Session 2: Subject allocation (group by sex) (group by sex)

Subjects Group 1 Group 2 Group 3 All

Males 41 38 46 125

Females 11 9 8 28

Combined 52 47 54 153

No difference in group by sex assignments

Session 2: Egress times (s)Session 2: Egress times (s)

Group Mean SD

1 16.9 2.5

2 17.6 2.8

3 17.3 2.7

No difference in the group by time

Session 2: Pass/Fail ratesSession 2: Pass/Fail ratesGroup Total (n) Pass (n) Pass (%)

1 52 28 54%

2 47 38 81%

3 54 52 96%

Total 153 118 77%

(2 = 27.591, df = 2, p< 0.000)

DiscussionDiscussion

Drop outs from Session 1 to 2Drop outs from Session 1 to 2 38 (19.9%) out of the 191 subjects dropped 38 (19.9%) out of the 191 subjects dropped

outout In the planning we had built in a drop out of In the planning we had built in a drop out of

20% therefore we maintained the statistical 20% therefore we maintained the statistical power necessary for the experimentpower necessary for the experiment

We also ran the statistical analysis including We also ran the statistical analysis including all the drop outs as failures. There was no all the drop outs as failures. There was no change in overall Group resultschange in overall Group results

DiscussionDiscussion

Males vs femalesMales vs females In terms of the overall results the percentage In terms of the overall results the percentage

of males who were successful was greater of males who were successful was greater than the femalesthan the females

The window required 80 lbs (36 Kg) of force to The window required 80 lbs (36 Kg) of force to openopen

This would affect the overall success values This would affect the overall success values but it would not change the effect by groups but it would not change the effect by groups

ConclusionsConclusions

From Session 1:From Session 1: Use of a push out window increases the task Use of a push out window increases the task

difficultydifficulty Increases average egress time by 4 secondsIncreases average egress time by 4 seconds Decrease percentage of successful egresses by Decrease percentage of successful egresses by

14%14% With 3 additional trials percentage of With 3 additional trials percentage of

successful egresses rises to successful egresses rises to ~~ 94%94%

Conclusions (continued)Conclusions (continued)

Session 2:Session 2: No effect of previous training on successful No effect of previous training on successful

egress timesegress times Without previous push-out experience the Without previous push-out experience the

percentage of failures is close to 50% (46%)percentage of failures is close to 50% (46%) Just 1 training trial with a window increases Just 1 training trial with a window increases

percentage of successful egresses to 81%percentage of successful egresses to 81% With 4 training trials the percentage of With 4 training trials the percentage of

successful egresses is 96%successful egresses is 96%

Future research issuesFuture research issues

Will changing the period between training Will changing the period between training and testing affect the success rate of and testing affect the success rate of HUET egress?HUET egress?

Does changing the task difficulty alter Does changing the task difficulty alter performance? (Cross cabin, other window performance? (Cross cabin, other window types)types)

Practical implicationsPractical implications

The closer the match of the training to the The closer the match of the training to the “real” situation (fidelity) the better the “real” situation (fidelity) the better the person will perform when needed.person will perform when needed.

Design considerations for push out forces Design considerations for push out forces and mechanism locations is still needed.and mechanism locations is still needed.

Simulations/TrainingSimulations/Training

Lower fidelity Higher fidelity

AcknowledgementAcknowledgement

Thank the subjects for their time and Thank the subjects for their time and participationparticipation

Thank Survival Systems Ltd for support Thank Survival Systems Ltd for support and timeand time

Thank Borge Hognestad and Daniel Thank Borge Hognestad and Daniel McInnis for being the evaluatorsMcInnis for being the evaluators