ArticlePDF Available

Personality Factors in Flight Operations: Volume I. Leader Characteristics and Crew Performance in a Full-mission Air Transport Simulation

Authors:

Abstract and Figures

Crew effectiveness is a joint product of the piloting skills, attitudes, and personality characteristics of team members. As obvious as this point might seem, both traditional approaches to optimizing crew performance and more recent training development highlighting crew coordination have emphasized only the skill and attitudinal dimensions. This volume is the first in a series of papers on this simulation. A subsequent volume will focus on patterns of communication within crews. The results of a full-mission simulation research study assessing the impact of individual personality on crew performance is reported. Using a selection algorithm described in previous research, captains were classified as fitting one of three profiles along a battery of personality assessment scales. The performances of 23 crews led by captains fitting each profile were contrasted over a one-and-one-half-day simulated trip. Crews led by captains fitting a positive Instrumental-Expressive profile (high achievement motivation and interpersonal skill) were consistently effective and made fewer errors. Crews led by captains fitting a Negative Expressive profile (below average achievement motivation, negative expressive style, such as complaining) were consistently less effective and made more errors. Crews led by captains fitting a Negative Instrumental profile (high levels of competitiveness, verbal aggressiveness, and impatience and irritability) were less effective on the first day but equal to the best on the second day. These results underscore the importance of stable personality variables as predictors of team coordination and performance.
Content may be subject to copyright.
NASA Technical Memorandum 102259
Personality Factors in Flight
Operations: Volume I. Leader
Characteristics and Crew
Performance in a Full-Mission
Air Transport Simulation
Thomas R. Chidester and Barbara G. Kanki
Ames Research Center, Moffett Field, California
H. Clayton Foushee
Federal Aviation Administration, Washington, D.C.
Cortlandt L. Dickinson
Menlo Park, California
Stephen V. Bowles
California Professional School for Psychology, Berkeley, California
April 1990
NationalAeronautics anti
Space Administration
Ames Research Center
Moffett Field, California 94035-1000
https://ntrs.nasa.gov/search.jsp?R=19900014054 2018-12-20T17:24:26+00:00Z
TABLE OF CONTENTS
SUMMARY
INTRODUCTION ............................................................................................................................. 2
5
METHOD .......................................................................................................................................... 5
2.1 Study Desig_ Overview ............................................................................................
5
2.2 Subjectsand Recruitment..........................................................................................
6
2.2.1 Prete,,;ting......................................................................................................
2.2.2 ProfileClassification.....................................................................................6
7
2.3 Confidentiality...........................................................................................................7
2.4 Expcrimentai_Equipment ...........................................................................7
2.5 Personnel...................................................................................................................
8
2.6 Experimental Procedure ............................................................................................
8
2.7 Simulation Scenarios ................................................................................................. 9
2.7.1 Day One Scenarios ........................................................................................ 10
2.7.2 Day Two Scenarios ....................................................................................... 11
2.8 Measurement .............................................................................................................
2.8.1 Flight Experience and Personality Data ........................................................ 12
2.8.2 Subjective Workload ..................................................................................... 12
12
2.8.3 Crew Performance .........................................................................................
2.8.3.1 Observer Ratings ........................................................................... 12
13
2.8.3.2 Crew Errors ................................................................................... 14
2.8.3.3 Aircraft Handling ..........................................................................
14
RESULTS ........................................................................................................................................... 14
3.1 Pretest Data ............................................................................................................... 14
3.1.1 Flight Experience ..........................................................................................
3.1.2 Personality Data ............................................................................................ 14
16
3.2 Subjective Workload ................................................................................................. 17
. ........ .....
3.3 Observer R_fings ......................................................................................... 19
3.4 Error Analyses ...................................................................................................... 20
3.5 Aircraft Handling ......................................................................................................
20
DISCUSSION ................................................................................................................................... 22
4.1 Crew Familiarity .......................................................................................................
4.2 Implicatiom; for Designing Operationally-valid Research ........................................ 24
REFERENCES .................................................................................................................................. 25
111
PRECEDING PAGE BLANK NOT FILMED
APPENDICES
A
B
C
D
E
F
Definitions of Personality Scales ..............................................................................
Flight Experience Items ............................................................................................
Personality Battery ....................................................................................................
Task Load Index Items and Definitions ....................................................................
Items Comprising the Observer's Rating of Crew Performance ...............................
Observer Ratings of Phase Performance ...................................................................
28
29
30
39
40
41
iv
SUMMARY
Crew effectiveness is a joint product of the piloting skills, attitudes, and personality characteris-
tics of team members. As obuious as this point might seem, both traditional approaches to optimizing
crew performance and more recent training development highlighting crew coordination have empha-
sized only the skill and attitudinal dimensions. This volume is the fu'st in a series of papers on this simu-
lation. Asubsequent volume will focus on patterns of communication within crews. This paper reports
the results of a full-mission simulation research study assessing the impact of individual personality on
crew performance. Using a selection algorithm described in previous research, captains were classified
as fitting one of three prof'des along a battery of personality assessment scales. The performances of 23
crews led by captains fitting each profile were contrasted over a one-and-one-half-day simulated trip.
Crews led by captains fitting a positive Instrumental-Expressive profile (high achievement motivation
and interpersonal skill) were consistently effective and made fewer errors. Crews led by captains fitting
a Negative Expressive profile (below average achievement motivation, negative expressive style, such as
complaining) were consistently less effective and made more errors. Crews led by captains fitting a
Negative Instrumental prof'de (high levels of competitiveness, verbal aggressiveness, and impatience and
irritability) were less effective on the first day but equal to the best on the second day. These results
underscore the importance of stable personality variables as predictors of team coordination and
performance.
INTRODUCTION
Theeffectivenessof anaircrewisajoint productof thetechnicalskills,attitudes,andpersonality
characteristicsof itsindividualmembers(Sells,1955)andtheprocessby whichtheyplan,execute,and
solveproblems.As obviousasthismight seem,traditionalapproachesto theoptimizationof crewper-
formancein air transportoperationshaveemphasizedtheskills dimensionsalmostexclusively.As pilots
areselected,trained,andevaluated,theprimaryemphasisisonensuringthateachhasthetechnical
skills necessaryto performhis orherrole in thecockpit.Morerecently,manytrainingprogramshave
beenexpandedto includethe influenceof interpersonalcharacteristicsassociatedwith crewcoordina-
tion, in theform of CockpitResourceManagement(CRM) programs(seeOrladyandFoushee,1987).
Whileit is certainlyencouragingthateffortsarenowunderwayto improveboth technical and interper-
sonal skills, the impact of other factors, particularly stable personality characteristics, may provide an
additional contribution to the crew performance process.
The search for personality predictors of performance in aviation operations has been plagued by
an historic failure to validate links between those dimensions and performance criteria (Melton, 1947,
1957; Ellis and Conrad, 1948; R. S. Melton, 1954). The reasons for these failures are complex, and
include the strategies of performance evaluation or assessment employed and problems with conceptual-
izations of the role of personality. For example, the vast majority of personality research in aviation set-
tings has examined performance during initial pilot training and has employed criteria such as complet-
ing training or obtaining a pilot rating (Dolgin and Gibb, 1988). These types of criteria probably do not
capture the range of performance occurring in day-to-day flight. A focus on training-completion misses
the complex setting in which crews perform.
Judging the operational significance of personality solely on the basis of training-completion
research implicitly assumes that the relationship between personality and performance remains constant
from training through initial experience to routine performance. Helmreich, Sawin, and Carsrud (1986)
have presented evidence that performance during and shortly after training is much less sensitive to per-
sonality effects than performance after the effects of training have begun to subside. As a result, it is
necessary to assess the relationship at the level of routine performance, or in critical situations, to answer
questions of operational relevance. Unfortunately, criteria available from routine flight settings in the
form of regular performance checks typically consist of pass-fail evaluations, with failures being
extremely rare, and tend to focus on very standard tasks like control-manipulation smoothness or com-
pletion of specific maneuvers. Using flight checks as criteria would mask a broader range of perfor-
mance variability and minimize the power of any predictor of performance (Hackman and Helmreich,
1987).
In any event, the research conducted to date has not fully explored the range of valid research
questions beating on the link between personality and performance. This range includes the following
issues: First, one can ask whether those attracted to aviation or space settings differ on average from the
general population. Second, among those sufficiently attracted to present themselves as candidates, does
personality predict successful completion of training? Third, do crewmembers who are more successful
or proficient in their duties over the course of a career differ in personality from those who are less suc-
cessful? Fourth, does personality predict individual performance or do particular combinations of crew
personalities predict crew performance in critical flight situations? Fifth, how does personality interact
with training,taskdesign,cr othervariablesin thepredictionof performance?Thoughsomeresearch
hassuggestedpilotpopulationdifferencesin personalitycharacteristics(e.g.,Fry andReinhardt,1969),
only thetraining-completioaquestionhasbeenadequatelyexplored.Thepresentstudyfocuseson the
impactof personalityin criticalflight situations.
Ourstrategyfor evaluatingthepotentialimpactof personalitycharacteristicsoncreweffective-
nesswas(1) todefinecritical elementsof performance,(2) toidentifydimensionsof personalitytheoret-
ically linkedto theseelements,(3) to identifyaselectionalgorithmto classifyor differentiateindividual
subjectsalongthesedimen,;ions,and(4) toconductahigh-fidelityvalidationstudyto determinewhether
thesetheoreticalrelationshipstranslateintopracticalperformanceconsequences.Asastartingpointfor
research,wechosetofocuson theimpactof crew leadersandthecharacteristicsthat contributeto lead-
ership.Eventhoughourunderstandingof groupphenomenais notwhatit shouldbe,therearea number
of possibleelementsof asuccessfulcrewthathavebeensuggestedbypreviousresearch.First,high
levelsof individual technicalskill, proficiency,andthemotivationtowork hardarethefoundationupon
whicheffectivecrewcoordinationisbuilt. Second,pastresearch(e.g.,Kanki, Lozito, andFoushee,
1988)hasdemonstratedthefteffectivecrewsarecharacterizedbycommunicationspatternsthattendto
bebothpredictableandresponsivelylinked.In addition,we wouldsuggestthateffectiveleadershipis a
joint functionof: 1)maintainingeffectivetaskdelegationanddefinition,2) encouragingcross-checking
andfeedback,and3)creatinganatmospherewheresubordinatesfeelfreeto offer suggestionsand
counter-proposalstoleadel-prescribedcoursesof action(e.g.,Ginnett,1987).Ourtaskwasto seekper-
sonalitymeasuresthatcapturedtheseelements.
A greatdealof eml:hasishasbeenplaceduponthefirst element,individual technicalskillsand
themotivationtoachieve.]::orresearchpurposes,wechosetofocusondimensionsunderlyingthemoti-
vationalcomponentof individual performance,one'soveralllevelof "instrumentality,"whichwedefine
operationallyasaperson'slevelof goalorientationandindependence.We alsochoseto emphasize
"achievementstriving" asanadditionalmeasureof anindividual's dispositionalorientationtowardtask
performancesituations.A .';econd dimension, oriented toward communication and interpersonal
exchange, is commonly defined as "expressivity," or interpersonal warmth and sensitivity. Communica-
tion would be expected to be facilitated in groups led by or composed of individuals characterized by
high levels of expressivity and inhibited in groups composed of individuals characterized by both nega-
tive expressive (e.g., frequent complaining) and negative instrumental traits such as verbal aggressive-
ness, competitiveness, or impatience and irritability. In summary, we theorized that effective leaders are
more often characterized by relatively high levels of both positive instrumentality and positive expres-
sivity (high levels of both concern for people and concern for performance), and that this type of leader-
ship style would facilitate crew performance. Lower levels of positive expressivity and higher levels of
negative expressive and negative instrumental traits were expected to lead to less effective crew com-
munication, coordination, and performance overall. Definitions of each of these dimensions and the
instruments measuring each are displayed in appendix A.
These dimensions were chosen for several reasons. First, a number of personality theorists and
researchers have focused cn dimensions reflecting instrumentality and expressivity as one central set of
personality characteristics. Influential theorists (e.g., Spence and Helmreich, 1978; Fiedler, 1967) have
all in one way or another identified these characteristics as core components of human personality with
strong behavioral relationships. Moreover, many popular management theories have espoused concern
for people balanced with concern for performance as the key to leader success. Blake and Mouton's
(1978)"managerialgrid" is perhapsthemostwidely appliedexampleof thesenotions,andit hasbeen
incorporatedinto manytrainingprograms,includinga numberin aviation.Second,thesedimensions
capturetheoreticallyrelevanttraitscorrespondingto coreelementsof performancein aviation.Third, a
greatdealof real-worldperformancedatahasbeencollectedfor theinstrumentsassessingthesedimen-
sions(SpenceandHelmreich,1983;Helmreich,Spence,andPred,1988).
Sofar,wehavesuggestedpossiblerelationshipsbetweendimensionsof personalityandelements
of crewperformance.But if onewishestoapplypersonalityresearchoperationally,eitherby selecting
individualswithdesirablecharacteristicsor tailoringtrainingto individualpersonalityprofiles,onemust
havesomemeansfor classifyingindividuals.Typically in research,oneconcentratesononeor two
dimensionsof personality.A frequentapproachis todetermine,throughmultipleregressiontechniques,
theuniqueportionof behavioralvariancecontributedby aparticulartrait.But in appliedsettings,a
researchermayneedtoconsiderthedistributionof combinationsof manydifferent traitswithin individ-
ual subjects,or in otherwords,to lookat theconstellationof traitsthatexistin peopleworking in the
appliedsetting.This isoftenmadenecessaryby thesmallnumbersof subjectsavailablefor study.Pro-
viding aresearchdesigncombiningpossiblelevels(evensimplemediansplits)of eachtrait understudy
quicklybecomesimpractical.
Chidester(1987)andGregorichetal.(1989)haveemployedthetechniqueof clusteranalysis
within severalsamplesof pilotstodeterminethedistributionsof differingcombinationsof positiveand
negativepersonalattributesalongthepersonalitydimensionsdescribedabove.Clusteranalysisis asta-
tisticaltechniquewhichcombinessubjectsinto groupsor clustersbaseduponeachsubject'ssimilarityto
othersubjectsalonganyspecifiedsetof dimensions(Anderberg,1973).Thosesubjectsthatfell into
eachclusteraresimilar tooneanotheralongthedimensionsanalyzedanddifferentfrom individualsin
otherclusters.Thesubgroupsthatwereidentifiedthroughthisprocessreflectmeaningfulconstellations
of traitsastheyaredistributedacrossindividuals.Threedistinctclustershavebeenfoundacrosssamples
of pilots, onewith highlevelsof positivetraitsandtwootherswith differentconstellationsof negative
traits.Pilotsin thepositiveclusterarecharacterizedbyhighlevelsof instrumentality,expressivity,
achievementstriving,work, andmasteryandaredesignatedthepositiveinstrumental-expressiveor
"IE+" cluster.Oneof thenegativeclustersis definedbyhighlevelsof negativeexpressivityandlow
levelsof instrumentalityandachievementstriving.This clusteris characterizedby traitsassociatedwith
tendenciesto expressoneselfin anegativefashion(e.g.,complaining)andlower thanaveragegoalori-
entation.It hasbeenlabeledthenegativeexpressiveor"EC-" cluster.Thesecondnegativeclusteris
characterizedby higherthanaveragelevelsof verbalaggressiveness,negativeinstrumentality,andcom-
petitiveness.Thisclustercomprisesamore"authoritarian"orientationandmaywell beassociatedwith
elementsof a profilepopularlyknownas"the fight stuff." It hasbeenlabeledthenegativeinstrumental
or "I-" cluster.Chidester(1987)foundsomeevidencethattheseclustersmayberelatedtodeterminants
of crew performance.Pilotsrespondeddifferentially totrainingin crew coordinationasafunctionof
theseprofiles.IE+ pilotsbenefittedthemostfrom trainingasassessedbychangesin attitudesconcern-
ingcockpitmanagement.
Havingspecifiedasetof elementscriticaltocrewperformance,a setof personalitycharacteris-
ticslinkedto theseelements,andameansfor classifyingindividualsalongmultipledimensions,ourtask
wasto assesswhetherthesecharacteristicswereoperationallyrelevant.Thatis, dothesetheoreticalrela-
tionshipstranslateinto real-worldperformancedifferences?Weput thesecharacteristicstothetestin a
full-missionsimulationresearchexperiment.
4
METHOD
2.1 Study Design Overview
The current study wa,; designed to (1) evaluate whether the personality characteristics of aircraft
commanders significantly impact the crew performance process and (2) evaluate the experimental clas-
sification algorithm as a possible countermeasure for the prevention of crew coordination problems. A
two-day full-mission simulation study was designed, in which crews flew five flight segments under
varying conditions of workload and in which crews were chosen according to personality criteria. Three
different types of crews were composed. The crew types contrasted were based upon cluster membership
as described by Chidester (1987). Crew types represent selection for leadership; that is, only the cap-
tain's personality characteristics were considered when crews were chosen. The first crew type was
composed of a randomly assi gned first officer and flight engineer flying with a captain from the IE+
cluster. We hypothesized that these captains would be good leaders and their leadership would translate
into effective crew performaace. The second type was composed of a randomly assigned first officer and
flight engineer flying with a captain from the I- cluster. The third type was composed of a randomly
assigned first officer and flight engineer flying with a captain from the EC- cluster. We hypothesized
that these two negative leader crews would be less effective at crew coordination in the high workload
flight segments. These hypolheses were tested using data collected from four sources: self-reports,
expert observation, video-based coding of errors, and aircraft handling parameters.
2.2 Subjects and Recruitment
Twenty-three, three-person crews (69 pilots) completed a one and one-half day full-mission
simulation of airline operations in the Ames Man-Vehicle Systems Research Facility (MVSRF) Boe-
ing 727 simulator. All crews were employed by the same major U.S. air carder, all crewmembers were
currently operating the B-727 exclusively in passenger operations, and all crewmembers were at the time
qualified in the B-727 crew position (captain, first officer, second officer, or flight engineer) they occu-
pied in the simulation.
All crewmembers had completed an initial course in crew coordination or CRM and all had par-
ticipated in at least one line-oriented flight training (LOFT) session during recurrent training. LOFT
(Lauber and Foushee, 1981) is a form of training in which a crew completes a full-mission simulation
very similar to that in this research project. Unlike traditional simulator training, the focus is not on the
completion of a specified set of maneuvers or procedures, but on training the crew to deal with problems
in the manner required in thz line environment; as a team working and exchanging information with
each other, with air traffic c,:mtrol, and with company dispatch and maintenance services.
Subjects were recruited through an announcement letter delivered to their company mailboxes at
their local domicile. This lel:ter described the study as a major simulation examining the factors influenc-
ing crew coordination in routine line operations and described the degree of participation requested.
Subjects were asked to notify a local union council member if they did not wish to be contacted by
NASA investigators by tele phone. Twelve of the 394 pilots and flight engineers in the domicile declined
to participate at this point. Telephone numbers for the remaining 382 pilots were released to the research
teamby thecompanythroughthelocal unionexecutivecommittee.Onememberof theresearchteam
thenattemptedto contactall of thesepilotsby telephoneandmailedacopyof thepretesttothosepilots
contactedandwilling to participate.A totalof 161questionnairesweremailed; 121pilotsreturned
usablepretests.Of thatnumber,69subjects(18%of thosein thedomicile)weresubsequentlyscheduled
for thesimulation.While thismayappearto bea lowpercentageof participation,thedegreeof partici-
pationrequired(travelingtotheresearchcenterontwoconsecutivedaysoff duty),thegeographicdis-
persionof pilotsassignedto thedomicile (manypilotscommutegreatdistancesto begineachduty
cycle),andthelimitedamountof availablesimulatortimemadeahighercomplete-participationrate
unlikely. Wewereableto fill virtually all of thesimulatortimeslotsavailabletousdespitethreelast-
minutecancellationsoverthecourseof thestudy.Giventheseconstraints,therateofretumof pretest
questionnaires(31%of thedomicile)is amoreaccuratereflectionof willingnessof subjectsto
participate.
2.2.1 Pretesting- Priortoschedulingfor thesimulation,candidatesubjectpilots completeda
batteryof personalityinstrumentscomposedof theExpandedPersonalAttributesQuestionnaire(EPAQ;
Spence,Helmreich,andHolahan,1979),theWork andFamilyOrientationQuestionnaire(WOFO;
SpenceandHelmreich,1978),andtheAchievementStrivingandImpatience/Irritabilityscales(A/S, IA;
Pred,Spence,andHelmreich,1986)derivedfromtheJenkinsActivity Survey(JAS)measureof the
TypeA BehaviorPattern(Jenkins,Zyzanski,andRosenman,1971)alongwith a numberof itemsfocus-
ingonflight experience.Scoringof theseinstrumentsbytheirpublishedinstructionsresultsin 10scale-
scores:instrumentality,expressivity,negativeinstrumentality,verbalaggressiveness,negativecommu-
nion,work,mastery,competitiveness,achievementstriving,andimpatience/irritability(seeappendixA).
2.2.2 Profile Classification- Thepersonalitybatterywasscoredsothatclusterprofilecouldbe
determinedfor eachsubject.A scoringroutinedevelopedby Chidester(1987)wasutilizedin lieu of
conductinga newclusteranalysisin this smallsample.This routinecomparesasubject'sscoreoneach
dimensiontonorms(samplemedian)basedonasampleof over400airline pilots(Chidester,1990).
Eachsubjectisthenconsideredfor inclusionintoaclusterbaseduponhis/herrelativestanding(aboveor
belowthemedian)oneachdimensionascomparedto thepatternof medianscoresfoundfor thatcluster
in Chidester'ssample.Individualswereassignedto (1) theIE+ clusterif theyscoredabovethemedian
onthreeof thefollowing dimensions:instrumentality,expressivity,mastery,andwork, (2)theI-cluster
if theyscoredabovethemedianonnegativeinstrumentalityandverbalaggressivenessandbelowthe
medianonexpressivity,or (3) theEC- clusterif theyscoredabovethemedianon NegativeCommunion
andbelowthemedianon threeof thefollowing dimensions:instrumentality,achievementstriving,
mastery,work, andimpatience/irritability.If anindividualmetnoneof thesecriteria,he/shewaslisted
asunclassifiable.Unclassifiablecaptainswerenotpursuedfurtherfor recruitment,butunclassifiable
first andsecondofficerswererecruitedsincetheirassignmentto crewswasintendedto berandom.
Thecarefulreaderwill recognizea numberof alternativewaysof assigningnewindividualsto
previously-definedclusters.Forexample,thescoresofeachindividualmaybecomparedtoclustercen-
troids(seeNorusis,1988)andincludedin theclusterto whichtheindividualis closest.Chidester's
(1987)scoringroutinewaschosentoemphasizedistinctionbetweenclusters.Sincewewouldcontrasta
smallsample,largepersonalitydifferencesbetweenclustersweredesirableandborderline-caseleaders
(theunclassifiable)werenotrecruited.Mostalternativeclassificationschemeswouldresultin lessdis-
tinct clusters.
2.3 Confidentiality
Because of the sensitivity of pilot performance data in general and the focus upon operational
significance in this investigation, it was necessary to guarantee all participating pilots complete confi-
dentiality. All data are identified by a five-digit code number. Thus, it is not possible for anyone, includ-
ing the NASA investigators, to identify any participating pilot by name. The code number merely pro-
vides a link between each subject's pretest data and his or her participation in the simulator.
2.4 Experimental Equipment
A Boeing 727-200 simulator operated by the MVSRF at Ames Research Center was utilized.
This simulator has a six de_;ree-of-freedom motion platform and four-window visual system. It was
manufactured by Singer-Link Corporation, and is equipped with special effects and programmed with
the aircraft performance data required to meet Federal Aviation Administration (FAA) Phase II certifi-
cation. The MVSRF was constructed specifically for research purposes and is configured for detailed
data collection. Simulator computers record aircraft configuration and handling information, and multi-
camera videotape and multichannel audiotape recording systems are installed for capturing crew com-
munication and action. A remotely-located Air Traffic Control (ATC) facility with flight-progress moni-
toring displays and voice-disguising equipment provides for highly realistic ATC support. The facility
also supports equipment to provide Automated Terminal Information System (ATIS) information to the
crew over VHF radio and background recordings of ATC communication provided by the FAA with air-
craft operating in airspace controlled by Air Route Traffic Control Centers (ARTCC). Both of these
capabilities greatly enhanc_;d the realism of flights conducted in the simulator. We gratefully acknowl-
edge the assistance of the FAA in obtaining recordings of Oakland and Los Angeles ARTCC
communications.
2.5 Personnel
Simulator operatiors were accomplished with a basic staff of six people. An expert observer
(described in section 2.8.3.1) rode in the simulator cab along with the subject crew. A simulator operator
and the experimenter occupied an Experimenter Operations Station (EOS) located remotely from the
simulator cab. The EOS rogm included a data-entry terminal through which all simulator functions were
initiated, modified, and terminated. (Aircraft setup and events were pre-programmed and the operator
needed only to initiate the program and make any changes necessitated by crew decisions, such as
adding fuel). Video and au:lio monitors and recording equipment were also located in the EOS, allowing
the experimenter and operztor to monitor crew communications and to detect simulator problems. The
experimenter started and stopped audio-video recording, communicated with the observer in event of
simulator problems, servec as ground-crewman and company dispatcher, and supervised the operation of
the simulator and support tacilities. The remote ATC facility housed an air traffic controller, a pseudo-
pilot, and an ATC facility manager. The controller was a retired FAA-certified traffic controller who
provided all clearances and ATC services to the subject crew. The pseudopilot initiated calls to ATC
over the radio frequency ir use by the subject crew. These calls were scripted to simulate "traffic" as it
would normally occur in the operating environment. The pseudopilot also operated "target aircraft"
which were visible to the crew though the four-window visual system. A number of aircraft operating
7
alongapproachpathsandairwaysalongtherouteof flight weresimulatedinanattemptto increasereal-
ism.An ATC facility managerworkedwith thecontrollerandpseudopilot,operatingtheATC facility
computersandmonitoringthecourseof thesimulation.TheATC manager also served as the interface to
facility maintenance and support personnel.
2.6 Experimental Procedure
Crews were met upon arrival at the simulator facility on the first day by the experimenter. In an
initial briefing, the importance of operational realism was emphasized and pilots were urged to treat the
simulation just as if it were an actual two-day trip. Crews were informed that they would have access to
all resources that they would normally have in flight, including complete ATC services, dispatch, access
to maintenance, ATIS, and so on. Crewmembers then received a detailed briefing on the differences
between the simulator's configuration and that of their company's aircraft. This briefing was conducted
by the expert observer. Following the briefing, crews were given a schedule of operations, weather infor-
mation and flight plans for two familiarization flight segments, and an opportunity to hold an initial crew
briefing. Crews were then escorted to the simulator cab, where they completed a round trip between
San Francisco and Stockton, California. After completing the familiarization segments, crewmembers
had lunch, then returned to initiate the experimental flights.
2.7 Simulation Scenarios
Past full-mission simulation research (Ruffell-Smith, 1979; Lauber and Foushee, 1981; Foushee
et al. (1986) has shown that successful simulation scenarios have at least five essential elements. First,
they are designed to be completely representative of the actual operational environment, and all details
are faithfully represented. Second, they are complicated enough to require the coordinated action of all
crewmembers for successful completion, but not to the extent that they induce complete crew failure
such as a "crash." Third, problems presented to crews have ongoing consequences which must be dealt
with in flight, but cannot be fixed in flight. Fourth, the problems involved axe very ambiguous, and there
is usually no simple corrective "by the book" solution. And fifth, the original problem is usually com-
pounded by other events such as weather-induced complications (e.g., landing on a rain-slick runway
with partial brake failure). It is also interesting to note that these characteristics have been seen in past
incidents and accidents.
In the process of scenario design, outlines of potential events were developed by the principal
investigators using accident case studies and incident reports. These were reviewed by investigators with
checking and training experience in the particular aircraft type being simulated and by simulator opera-
tional personnel from the MVSRF. Typical environmental conditions for the proposed area of flight
(November-February weather patterns for coastal and central California) were considered in great detail,
so that weather patterns and scenario events would seem realistic to the experimental flight crews. Air-
craft documentation and airline dispatch procedures were assembled for each flight segment in coopera-
tion with airline management and members of the local pilot labor union executive committee. Follow-
ing this development process, selected scenario outlines were programmed into the simulator computer
and eight pretest runs were conducted using qualified flight crews to refine procedures, train facility per-
sonnel and the experiment staff, and test scenario events. These pretest crews were carefully debriefed to
8
assesstherealismof thesceaariosandproceduresusedbytheexperimentstaff.This feedbackallowed
continualrefinementuntil tl',escenarioswerefinalized.
Crewsflew five experimentalflight segments(legs).Eachsegmentwasplannedandflown as
closelyaspossibletorealoperations.Crewswereprovidedwith all of theirnormalflight documenta-
tion, completedall normalflight andcockpitpreparations,andcommunicatedwith all groundsupport
personnelnormallyavailabletothem.Flight routingscorrespondedto typical clearancesalongroutesin
centralandsouthernCaliforniaandwereadaptedfrom flight planscreatedfor anearlierNASAinvesti-
gation.Airfields availableto thecrewincludedSanFrancisco(SFO),Stockton(SCK),Sacramento
(SMF),andLosAngeles(La,X). Segments1through 3 were flown on the first day, segments 4 and 5
occurred on the second day. Routine levels of workload were designed into segments 1, 2, and 4, but
segments 3 and 5 were far more demanding than normal and involved continuing abnormal conditions
that could not be resolved completely in flight. Crews were led to believe, however, that they would fly
six legs. The last leg (SMF-SFO) was not intended to occur, except as a part of preflight planning,
because of scenario events. This deception was intended to counteract suspicions that might be associ-
ated with the last segment of the study, particularly since the last segment of day one involved an
abnormality.
2.7.1 Day One Scenarios- On the first day, all legs except the third were relatively routine.
However, one irregular item was included in the dispatch paperwork for the full-mission segments. The
deferred item list and accompanying minimum equipment list (MEL), indicated that the plane was to be
dispatched with the no. 3 generator inoperative. This was a legal procedure, but given the weather con-
ditions (night, fog, and low cloud ceilings), prudent crews should have considered delaying departure
until either the weather improved or the generator was repaired. If the crew elected to request the repair,
the generator was "repaired," but the dispatcher warned the crew that they might expect it to malfunction
again requiring it to be reset. If the crew did not request repair, the generator was "repaired" later when
the crew reached the maintenance base at SFO. In either case, each time the no. 3 engine was started fol-
lowing the "repair," the generator field light illuminated, and at least one field reset was required to
bring the generator on line. This manipulation was intended to complicate the decision-making process
on a later high workload se;,_ment.
On segment 3, crews flew from Sacramento to Los Angeles. Following a normal takeoff and
climb, a combination of system failures were activated automatically. First, the vertical stabilizer trim
system began running uncommanded and jammed in a nose-down condition at a predetermined point.
Second, and shortly thereafter, when the aircraft crossed a specified navigational point, the no. 2 engine
low-oil-pressure warning light illuminated and the indicated pressure fell to the cautionary range. This
combination represented a relatively high-workload situation, but was compounded by neither weather,
traffic, nor ATC problems. No diversion from the flight plan was necessary, except for actions required
to stabilize the aircraft and land safely. Segment 3 is displayed graphically in figure 1.
The scenario prese_ ted the crew with two independent failures which could have impacted flight
safety. The jammed stabilizer trim system was a serious control problem, disabling the autopilot, and
requiring (due to the descent configuration) constant nose-up control inputs and considerable back pres-
sure from the flying pilot. "]_is made the approach and landing much more difficult and was physically
fatiguing (given the need to hold constant, firm back pressure). The procedure for dealing with the
9
Takeoff:
Smog forecast for
LA. Basin
SFO
SJC
Descent and approach:
Stabilizer runs sway and
JamsIn position
Oil filter bypass light
Illuminates 5 min. later
Completion of checklists
Very high workload
SMF RNO
Climb:
Steep crossing
restriction
lLow celllnga and fog
I Cruise: /
=1 Uneventful 1
I/Landing:
I/High speed, reduced
Jcontrol landing
_J_ Physicaland procedural
_coordination required
Figure 1.- Scenario for segment 3.
jammed stabilizer trim, once it was identified, was to limit flap setting to 15° for landing, increase
approach speed by 15 knots, and to establish the landing configuration as early as possible so as to get a
feel for the control forces necessary for landing. This procedure is difficult but not unreasonable since it
is required for all type ratings in the B-727 aircraft. The low oil pressure light and corresponding cau-
tionary gauge reading were also covered by a checklist procedure, but the outcome of the checklist left
the crew a choice. At the captain's discretion, the crew could either shut the engine down or reduce
thrust on the engine to idle. Shutting the engine down required the completion of more checklists
(greatly increasing workload and time requirements), may have required the dumping of fuel (again
increasing workload), and removed one generator from operation. Recall that the crew had already
encountered minor problems with the no. 3 generator, so the aircraft could have easily been down to one
should conditions have deteriorated further. The most prudent course would appear to be to keep the
engine operating as a reserve while continually monitoring its operating parameters. However, a pilot
could reasonably argue that the possibility of the engine failing catastrophically was sufficient to warrant
the workload consequences of shutting it down. Once this decision was made, the crew had to land the
aircraft in an abnormal configuration.
2.7.2 Day Two Scenarios-- On the second day, the weather continued to be characterized by
fog, overcast, and minimal visibilities in the central California valley. The In'st leg (SCK-LAX) went
without programmed incident except for fog in the Los Angeles area (ceiling 400 ft, visibility one mile).
By the time of the second leg (LAX-SMF), weather had deteriorated further in the central valley, result-
ing in poor visibility in the Sacramento area (ceiling 300 ft, visibility one-half mile; approaching the
legal minimums for ILS approaches). Weather in Los Angeles remained foggy but above legal mini-
mums. Following a normal departure and climb out of the Los Angeles area and normal cruise, the crew
received clearance for the Wraps Four Arrival to Sacramento. As the crew entered the Sacramento ter-
minal area, runway visual range (RVR) was reported by the approach controller as 2000 ft, just above
10
theCategory1approachmir imumsof 1800ft. Prudentcrewsshould,at thispoint, haveconsideredand
briefedfor apossibleCategcry2approachandlanding.Theminimumsfor thisapproachwereRVR
1200with adecisionheightof 126ft. However,aftertheaircraftcrossedtheoutermarker,RVR was
reportedbythe toweraslessthanI000 ft, which wasbelowtheminimumfor theinitiation of an
approach.However,sincetheaircraftwasinsidethemarker,it waslegaltocontinueuntil thepublished
decisionheightandtolandi:_thecaptainhadanadequateviewof therunwayenvironment.If thecrew
attemptedthelanding,therunwaywasnotvisible atthedecisionheight,requiringtheapproachto be
aborted.Duringthemissed_pproach,asthecrewretractedtheflaps,ahydraulicsystemA failure
occurred,causedby aleakthatdepletedall of thehydraulicfluid. At this point,it wasimmediately
apparentthata diversiontoan alternate airport would be necessary. Weather conditions at various
nearby alternates were poor (all ceilings less than 800 ft and visibilities less than one mile). Reno
(RNO), San Francisco, Oakland, and San Jose (SJC) were the best alternates. Reno was unacceptable
because the prevailing wind direction made its shortest runway the active runway and the use of this
runway for B-727 operation:_ was prohibited by company policy. SFO had the best weather of the
remaining alternates, with clearing conditions, a 2000-ft ceiling, and 2 miles visibility with light rain.
SFO also provided a very loag runway at just under 12,000 ft. However, SJC was listed as the flight's
legal alternate, because SFO weather conditions were below alternate minimums at the time of dispatch.
Segment five is displayed graphically in figure 2.
This scenario confrot_ted the crew with a number of hazards and limitations. First, the hydraulic
failure disabled a number of aircraft systems. The landing gear had to be extended by hand crank, and
once extended could not be retracted. Flaps also had to be extended by alternate means, and this system
does not allow leading edge flaps to be retracted once extended. In addition, the trailing edge flaps were
not protected against asymmetric extension using the alternate system. Alternate extension required
more time than extension by normal means, and was limited to 15 ° in case of a missed approach (flap
retraction from the normal 20 ° setting by the alternate system following a missed approach would be too
slow to reduce drag sufficie atly to allow the aircraft to climb to obstacle clearance altitude). The
hydraulic failure also disabled nosewheel steering, ground spoilers, and outboard flight spoilers. All of
these limitations caused a combination of higher than normal approach speeds and reduced stopping
ability. Finally, the crew had to select an airport (SFO was suggested by the circumstances) and execute
an approach and landing under adverse circumstances. When the crew extended the flaps on approach to
SFO, they received an outboard trailing-edge flap asymmetry indication resulting from the lack of pro-
tection discussed above. This condition changed the handling characteristics of the aircraft and required
that crews discuss and estimate a landing speed, because none is given in the flight manuals for the
condition of split inboard/outboard flaps combined with a hydraulic system failure.
2.8 Measurement
A variety of workload and performance variables were measured. Measurement instruments and
procedures are described in this section.
11
Landing: /Jb_ RNO
Flapasymmetry problem
High landing speed //_SMF_ Hlyo=raUlIC.Vailure _Low ceilings and fog
No nose w_l stesr_/__axe _ Descont and
approach:
Fast, noisy
descent required
by ATC for
crossing traffic
Climb and cruise:
Light turbulence
Takeoff:
Poorforecast
High takeoff weight
LAX
Figure 2.- Scenario for segment 5.
2.8.1 Flight Experience and Personality Data- Flight experience and personality data were
collected in the pretest described earlier. Crewmembers were asked to report their age, gender, total of
number of years they had been with their airline; total number of flight hours completed with their air-
line, in the military, in private aircraft, and any other setting; total flight hours in the B-727; and flight
hours in their present crew position (see appendix B). Personality characteristics were assessed using the
EPAQ, WOFO, and the A/S and I/I scales. These questions from these scales are listed in appendix C.
The scales are described in appendix A.
2.8.2 Subjective Workload-- Following each flight segment, all crewmembers completed the
NASA Task Load Index (TLX; Hart and Staveland, 1988). This rating form consists of 6 workload-
related dimensions: mental demand, physical demand, temporal demand, performance, effort, and frus-
tration. Subjects reported their responses along a seven-point Likert scales with high numbers indicating
greater amounts of the dimension in question. Hart and Staveland (1988) also advocate the completion
of a scale-weighting measure wherein subjects evaluate the importance of each dimension to their per-
ception of workload in the performance setting. However, completion of the TLX by pretest crews used
in the scenario development process (24 subjects) indicated that a simple summed composite of the six
items correlated 0.92 with the weighted scale scores. As a result, the forced-choice procedure was elimi-
nated and only the summed rating will be reported. Items forming the TLX are listed in appendix D.
2.8.3 Crew Performance-- Crew performance data were collected from three sources: expert
observation, video coding of crew errors, and computer recording of aircraft handling parameters.
2.8.3,1 Observer ratings: A recently-retired, highly experienced (39 yr in Part 121 operations,
30,750 total flight hours) airline captain served as an expert observer and was present in the simulator
cab with every flight crew. He was blind to the experimental condition, and evaluated crew performance
12
following everyflight segment,andindividualperformanceduring specificphasesof thehigh-workload
segments.
The observer completed an overall rating form immediately after every segment. This form was
adapted directly from Helmreich and Wilhelm's (1987) Line/LOFT Checklist, which was, in turn,
derived from previous NASA simulation studies. The observer was instructed to complete the form
evaluating the crew as a unit. Ratings were made on five-point Likert scales and were intended to assess
the expert observer's overa_.l impression of performance on each dimension for each segment. These
ratings were summed and averaged to form a crew-level composite for each segment. Items forming this
scale are listed in appendix E, and will be referred to as observer ratings of crew performance.
The observer completed a second rating form for the high workload segments, three and five.
This form was adapted frora the Foushee et al. (1986) study and was organized by phase of flight and
abnormal events during segments three and five. Within each section of the form (e.g., preflight,
taxi/takeoff, climb, and cruise), the observer rated the performance of each individual crewmember.
Ratings were completed in real time as the flight progressed using a five-point Likert scale and were
summed and averaged to form a composite for each crewmember during each phase of flight. Items
forming this scale are listecl in appendix F, and will be referred to as observer ratings of individual phase
performance.
The ratings of crew and individual performance presented are limited to evaluations by a single
expert observer. The psych:_metfic ambiguities and potential biases resulting from use of a single
observer or rater are a problem for this study. However, as in previous NASA investigations, we have
elected to collect what mat be very important data in the only manner in which it could be collected,
rather than simply not collect it, and to bolster that data with information from other sources. For safety
reasons, only one observer can be in the MVSRF simulator cab at a time, so inter-rater reliability could
not be established for these ratings. We have chosen to err on the side of maximizing operational credi-
bility over psychometric sc.phistication in these performance ratings, in a situation where maximizing
both was not possible. The observer is a highly experienced pilot in terms of both operational experience
and aircraft accident investigation, and as such, his evaluations would be considered highly relevant by
the aviation community. Much of the information required to make these ratings is available only in real
time and inside the simulator cab. The situational awareness of an experienced pilot sitting directly
behind the action of a flightcrew solving a problem simply cannot be matched or replaced by the eval-
uations using multichannel audio and multicamera video presentations. The use of multiple observers to
evaluate crews using these media would allow assessment of rater consistency, but at the expense of lost
information and reduced operational credibility. We have chosen to collect and analyze this opera-
tionally significant information, recognizing its psychometric limitations. At the same time, we have
attempted to compensate for these limitations by collecting data from other sources which reflect on the
same domain, crew effectiveness, but capture only smaller portions.
2.8.3.2 Crew errors: Error analyses were undertaken using two independent sources of data to
assure the reliability of performance assessment. First, during test runs, the expert observer kept a record
of all errors he observed. The second source of error data required a complete review of the videotape
records. Using these records, two condition-blind observers reviewed each flight for operational errors.
The first observer was a retired pilot, the second a NASA researcher. Both observers studied the B-727
training and flight operaticns manuals provided by the airline prior to initiating error identification.
13
Whenanerrorwasrecognizedbyoneor bothobservers,thetapewasstoppedandthesegmentcontain-
ing anallegederrorreviewed.Theerrorwascountedin theanalysisonlyif bothobserversagreedthat
theerroroccurredandagreedonadescriptionof theerror.All errorsidentifiedby thevideotape
observerswerethenpresentedto theexpertobserver,whohadtheoptionof eliminatinganerroronthe
basisof hisnotestakenduringthesimulationsorhisoperationalexperience.This wasaconservative
errortabulationprocessandassuredthateveryerrordatapoint wasreviewedatleasttwice.Sincesome
performanceerrorsweremoreoperationallysignificantthanothers,errorswerecategorizedaccordingto
levelof severity.Thisprocesswasaccomplishedbytheexpertobserverandbybothof theobservers
involvedin thevideotapeerroranalysis.A three-levelclassificationwasutilized.Type 1errorswere
definedasminor,with a low probabilityof seriousflight safety consequences. Type 2 errors were
defined as moderately severe, with a stronger potential for flight safety consequences. Type 3 errors
were defined as major, operationally significant errors having a direct negative impact upon flight safety.
2.8.3.3 Aircraft handling: Data collected and stored by the simulator computer allowed the
assessment of aircraft handling. Deviations from prescribed paths during specific flight phases
(instrument approaches) were calculated for the high workload segments. Deviations from glide slope,
localizer, and target approach airspeed were collected by the simulator computer once per second from
the outer marker to the runway threshold on final approach during segments 3 and 5. The absolute value
of the deviations per second was summed for each dimension during each approach.
RESULTS
3.1 Pretest Data
Crewmembers who participated in the study are described by their years and hours of flight
experience and by their mean scores on personality scales in comparison to other samples of pilots.
3.1.1 Flight Experience-- Flight experience differed, predictably, by crew position. Captains
had been with their airline an average of 22.5 yr and had completed an average of 12,493 flight hours in
airline operations (not necessarily all with the same carrier), with 3,259 of those hours as Captain. Cap-
tains averaged 51 yr of age. First Officers had been with their airline an average of 13.8 yr and had
completed 4,011 airline flight hours, with 2,145 of those hours as First Officer. First Officers averaged
43 yr of age. Second Officers had been with their airline an average of 2 yr and had completed
2,680 airline hours, with 1,505 of those hours as Second Officer. It was not possible to determine with
certainty where the Second Officer's had completed airline hours other than as B-727 Second Officers,
because the background questionnaire did not request these details. However, because the B-727 was a
relatively junior (lower seniority) aircraft at the airline, these hours most likely consisted of Pilot-in-
Command or Copilot hours completed with another airline, perhaps a commuter or regional airline. Sec-
ond Officers averaged 33 yr of age.
3.1.2 Personality Data- Average scores on each of the personality characteristics assessed are
displayed in table 1. No differences among crew positions were found in the personality data. These
sample means can be compared with means among pilots sampled by Chidester (1990) and norms
reported by the authors of each instrument (male college students appear to be the closest comparison
14
groupreported).Comparisonsof thesemeanssuggeststhatthis samplefalls somewherebetweenthe
normsfor malecollegestudertsandnormsforpilots reportedby Chidester(1990).
Table I. Sample means and norms for personality scales
I
Personality c h_tracteristics by
test administered Sample Pilots a College males
Characteristics PAQ
Instrumentality
Expressivity
Negative instramentality
Verbal aggres:_iveness
Negative comnunion
40.47 43.87 38.92
22.55 24.24 22.08
11.89 10.30 13.69
5.26 4.14 5.55
5.74 5.84 6.36
Charzcteristics WOFO
Mastery 20.32 22.87 19.26
Work 20.48 22.00 19.80
. Competitiven_._ss 12.31 13.18 13.63
Char_cteristics JAS (derived scales)
Achievement striving 23.73 26.17 22.89
Impatience/ir_ itability 14.01 14.91 16.48
|aBased on a sample of 469 pilots in domestic short-haul operations
(Chidester, 1990).
Table 2 displays the mean personality scale scores of captains classified into each cluster profile
for the purposes of this experiment. As would be expected given the use of an algorithm to assign pilots
on the basis of previously-de_Sned clusters, each cluster closely resembled its profile as defined by
Chidester (1987) and Gregorich et al. (1989). IE+ captains were distinguished by elevated scores on
instrumentality, expressivity, mastery, work, achievement striving, and unexpectedly, competitiveness.
In simple terms, IE+ captains can be described as being both instrumental and expressive relative to
other pilots. I- captains were distinguished by elevated scores on negative instrumentality, verbal
aggressiveness, impatience/irritability, and mastery. Perhaps most importantly, I- captains scored very
low on expressivity. I- capta: ns can be described as being instrumental, but not at all expressive. EC-
captains were distinguished by elevated scores on negative communion and expressivity and by very low
scores on instrumentality, mzstery, work, and achievement striving. EC- captains can be described as at
least moderately expressive, but not at all instrumental, relative to other pilots.
15
Table2. Personalityscalesmeansfor eachprofileamongcaptains
Personalitydimensionsbytest
administered IE+ I- EC-
Characteristics PAQ
Instrumentality
Expressivity
Negativeinstrumentality
Verbalaggressiveness
Negativecommunion
43.09 40.30 36.11
24.18 19.20 23.72
10.18 14.10 11.50
4.40 6.40 5.28
4.00 6.00 6.50
Characteristics WOFO
Mastery 24.09 21.30 18.39
Work 22.18 19.60 18.89
Competitiveness 14.55 12.60 11.89
Characteristics JAS(derivedscales)
Achievementstriving 24.27 24.00 22.17
Impatience/irritability 13.09 15.40 14.39
3.2 Subjective Workload
Analysis of TLX workload ratings revealed flight segment as a significant main effect
(F (4,80) = 54.26, p < 0.01). Post hoc comparisons revealed that segments 3 and 5 were rated as signifi-
cantly higher in workload than segments 1, 2, and 4. This served as a manipulation check, indicating that
the test segments required greater effort than the normal segments. Mean workload ratings for each seg-
ment are presented in table 3. While this analysis suggested no impact of captain personality on average
crew workload ratings, it has been argued that agreement or disagreement among crewmembers in work-
load assessment may be sensitive to differences in crew coordination (R. D. Blomberg, personal com-
munication). However, disagreement contributes only to error variance in a repeated measures analysis.
An index of disagreement among crewmembers was calculated for each flight segment. Analysis of dis-
agreement revealed a significant flight segment by captain personality interaction (F (8,80) = 2.27,
p < 0.05). Simple-effects tests suggested that crews led by IE+ captains disagreed more concerning
workload in both segments 3 and 5 than did crews led by I- or EC- captains. Follow up analyses using
analysis of covariance revealed that this was attributable to disagreement between IE+ captains and their
subordinate crewmembers. Specifically, IE+ captains rated segments 3 and 5 as less demanding than did
their subordinates, but their subordinates rated workload the same as subordinates or captains of I- or
EC- crews.
16
3.3 Observer Ratings
Analyses of the observer's ratings of crew performance during the full-mission segments
revealed a significant interaction between leader personality and flight segment (F (8,80) = 2.80,
p < 0.01). Means for each group of captains during each segment are presented in Table 4 and these
results are presented graphi,:ally in Figure 3. Examination of these means revealed that crews led by IE+
captains were rated as consistently effective, and these ratings were higher than the other crew types for
the segments overall (though not every comparison for every segment was statistically significant).
Crews led by EC- captains were rated as consistently less effective over all segments than those led by
IE+ captains (though not al_ comparisons were statistically significant). Crews led by I- captains
received ratings that varied considerably across segments. For segments one, two, and three, I- led
crews were similar to EC- crews; they were rated as less effective than IE+ crews. However, on segment
five, I- led crews were rateA as performing as well as IE+ led crews, and significantly more effectively
than EC- led crews.
Table 3. Average ratings of workload
Segment Task load index score
Segment 1 2.97
Segment 2 3.27
Segment 3 4.51 a
Segment 4 2.69
Segment 5 4.56 a
aMeans for segments 3 and 5 differ from
all others at the 0.05 level.
Table 4. Observer ratings of crew performance
Captain Personality profile
Segment IE+ I--a EC-
Segment 1 3.85 3.40 3.23
Segment 2 3.88 3.55 3.59
Segment 3 3.95 b 2.97 2.90
Segment 4 4.37 3.79 3.12
Segment 5 4.22 3.98 2.90 b
aThe following within-crew comparisons are signifi-
cant among I- led crews: 5 vs. 1, 2, and 3; 3 vs. 4.
bIndicates significant between-crew differences
17
IE÷
4.5 F _ Ec-
I
4.0
3.5
3.0
O2.5
2.0 1 2 3 4 5
Flight Segment
Figure 3.- Observer ratings over five flight segments
Observer ratings of individual Phase Performance during the high workload segments were con-
sistent with the crew-level ratings. That is, crews led by IE+ captains were rated as outperforming I- and
EC- led crews throughout segment three (F (2,20) = 3.69, p <. 0.05, post-hoe comparisons via Tukey's
Honestly Significant Difference (HSD)), and EC- crews were rated as performing more poorly than IE+
or I- led crews on segment five (F (2,20) = 4.27, p <. 0.05). There were mean differences in ratings
between phases, but no reliable differences among crew positions nor interactions between phase or
position and captain personality profile. Means for each phase are displayed for segment three in table 5
and for segment five in table 6.
Table 5. Observer ratings of phase performance--segment 3
Captain personality profile
Phase of flight IE+ I- EC-
Prestart 3.93 3.62 3.51
Taxi/takeoff 3.83 3.59 3.42
Climb 3.69 3.38 3.40
Cruise 3.69 3.24 3.04
Stabilizer 3.46 2.72 3.00
Oil pressure 3.63 2.99 2.51
Approach 4.15 3.55 3.27
Mean 3.76a 3.29
L i
aMean for IE+ differs from I- and EC- at the 0.05 level.
3.16
18
Table,5.Observerratingsof phaseperformance--segment5
Captainpersonalityprofile
Phaseof flight IE+ I- EC-
Prestart 4.42 3.79 3.40
Taxi/takeoff 4.30 3.64 3.41
Climb 4.39 3.94 3.41
Cruise 4.23 3.59 3.44
Missedappro_Lch 4.26 4.13 3.25
Hydraulicfail'are 4.14 3.91 3.24
CruisetoAltitude 3.96 3.89 3.12
Approach 3.63 3.99 3.06
Mean 4.16 3.86
I
aMean for EC- differs from IE+ and I- at the 0.05 level.
3.29 a
3.4 Error Analyses
Atotal of 998 error,; were identified by the video observers across the 23 crews. Of that number,
85 (8.5%) were eliminated by the expert observer as logical choices made by the crew in response to
available information. Of those errors that were eliminated, 60 (71%) were initially rated as type 1 errors
by the video observers; the remainder were initially rated as type 2 errors. None of the eliminated errors
was initially rated as a type 3.
Aconsensus severity classification was reached by the video and expert observers for the
remaining 913 errors. Tabh: 7 shows the relative frequency of each type of error. Examples of type 1
errors included: failure of the pilot not flying to make and announce a crosscheck dictated by company
policy, incorrect briefing of a missed approach procedure which was immediately corrected by the pilot
not flying, and failure to notice an "expect" provision on a standard arrival procedure. Examples of
type 2 errors included: delayed discovery of a system failure (hydraulic loss) following aural and visual
indications of the failure, flying at airspeeds above or below those required for the aircraft configuration,
and failure to use a proper ,:limb airspeed to make an ATC crossing restriction imposed due to intersect-
ing traffic. Examples of type 3 errors included: airspeed deviations resulting in stick-shaker (stall) or
mach-overspeed warnings, identification and initiation of an incorrect checklist or failure to run a
required emergency checklist (jammed stabilizer) resulting in an improper landing configuration, and
failure to consider an alternate airport other than that filed (SJC) or to determine alternate weather fol-
lowing a weather-induced missed approach.
Table 7. Frequency of errors by severity classification
ii
Flight segment Type 1 Type 2 Type 3
I I
Segmer t 3 168 137 81
Segmer,t 5 242 161 124
19
Erroranalysesrevealedapatternof f'mdingssimilarto thatseenfor observerratings.A repeated
measuresanalysisincludingtype 1,2,and3 errorcountsfor eachday,with severitylevelenteredasa
designfactor,revealedthreemaineffects.First,EC- led crewstendedto makemoreerrorsthanIE+ or
I- ledcrews(F (2,20)= 4.03,p <0.05;means= 23.2,17.7,and18.3perday,respectively).Second,asis
apparentin table7,crewstendedto makemoretype1(minor)thantype2 (moderate)errors,andmore
type2thantype3(major)errors(F(2,40)= 22.60,p< 0.01;means=9, 6.6,and4.5perday,respec-
tively). Third, crews tended to make more errors on day two than on day 1(F (1,20)=20.64, p < 0.01;
means= 16.5 and 22.9, respectively). However, this difference between the two days appeared to be
entirely attributable to a difference in the length of segments three and five (F (1,20) = 151.75; p < 0.01;
means 66.5 and 100.1 minutes, respectively). An analysis of covariance, that controlled for this time dif-
ference eliminated the day effect, but left the captain personality (F = 4.05) and error severity (F = 22.6)
main effects intact. This indicates that the error rates were statistically equivalent on both days; given
equal flight times, equivalent numbers of errors would occur.
While error analyses were consistent with ratings by the expert observer in discriminating per-
formance of EC- led crews from IE+ or I- led crews, they did not reveal a change in I- led crew per-
formance over the course of the simulation. That is, the captain personality by flight segment interaction,
which was significant for expert observer ratings, was not significant for crew errors. One possible
explanation for this discrepancy is that in the process of systematically rating crew performance,
observers may (and probably should) take more than errors into account. This might be expected, since
the observer has access not only to errors, but also to the crew communications process which is signifi-
cantly related to crew performance (Kanki, Lozito, and Foushee, 1989).
3.5 Aircraft Handling
Deviations from glide slope, localizer, and target approach airspeed were collected by the sim-
ulator computer once per second from the outer marker to the runway threshold on final approach during
segments 3 and 5. The absolute value of the deviation per second was summed for each dimension dur-
ing each approach. These sums were submitted to a multivariate analysis of variance, which revealed a
main effect for flight segment (multivariate approximate F (4,17) = 7.55, p < 0.01). Univariate analyses
revealed that this difference was due to greater glide slope deviation on segment three than segment five
(F (1,20) = 22.25, p < 0.01). Follow-up analyses indicated that this effect was accompanied by greater
airspeed variability on segment three. These deviations could be expected to result from operation of an
aircraft with a jammed stabilizer, the problem presented to the crew on segment three. Aircraft handling
appeared to be unrelated to captain personality.
DISCUSSION
These results present a relatively consistent picture of the impact of leader personality on crew
performance, particularly in critical high-workload situations. As might be expected, consistently effec-
tive performance was found among crews led by IE+ captains and less effective performance among
crews led by EC- captains. Also, IE+ captains rated their workload during segments three and five as
lower than did their own crews or I- or EC- leaders and crews. This finding may be a clue to
20
understandingall theotherresults,butit requiresthetesting of hypotheses based upon process data.
Specifically, IE+ captains may evaluate the abnormal segments as less demanding due to specific
strategies they apply in task delegation, or they may simply adopt differing standards of evaluating
workload. Since we predicted that IE+ captains would better organize both people and tasks, a finding of
differing strategies would strongly validate this theoretical conceptualization. If IE+ captains do indeed
delegate tasks more effectively, one would expect their workload to be lower, but not necessarily that of
their subordinates. These jurdor officers may have a number of tasks to accomplish equivalent to those
led by less effective captains, yet the crew may have a greater capacity to deal with that workload. Dis-
criminating between these two competing explanations will require further research into the process by
which these crews accompli:_hed their tasks.
Somewhat surprisinldy, I- led crews performed comparably with IE+ crews on the second day
and apparently improved ov,_r time. From previous research (Foushee et al., 1986), one might expect
that increasing crew familim'ity would result in better crew performance in the later flight segments,
regardless of crew type. Instead, familiarity apparently facilitated performance only among I- led crews.
These f'mdings are it_triguing because they support logical hypotheses about crew effectiveness
and are consistent with recent research examining individual pilot performance. Foushee and Helmreich
(1988) have hypothesized that relatively high levels of both instrumental and expressive traits might
facilitate crew performance. IE+ individuals have elevated levels of both of these traits, and as noted
earlier, Chidester (1987) found that IE+ pilots appeared to benefit most from CRM training. This group
showed the most positive and enduring attitude change up to six months after participating in the train-
ing program. Results from the present study suggest that personality factors, in general, may contribute
significantly to crew effectiveness, and provide further support for the notion that both instrumentality
and expressivity are import_ nt predictors of team performance in aerospace environments.
Somewhat surprising was the lack of relationship between captain personality and approach
deviations. In the Foushee et al. (1986) simulation study, approach deviations were sensitive to differ-
ences in crew familiarity, just as were ratings and errors. This would appear to argue for a reliable corre-
spondence between approach smoothness and crew errors or observer ratings, and lead to an expectation
of similar patterns of person ality effects in the handling data. One could argue that at the least, familiar-
ity influenced approach smoothness, since segment five approaches were smoother than segment three
approaches. But this explanation is confounded in the experiment with the jammed stabilizer manipula-
tion, and the handling differences identified would appear _to be attributable to that manipulation. That
approaches flown in this study were not influenced by experimental variables may be due to a number of
factors, ranging from the aircraft type to the meaning of personality variables. This study utilized a
B-727 simulator, while Foushee et al. (1986) used a B-737 simulator. The variety of differences between
these two aircraft types might explain differences in findings concerning aircraft handling. Additionally,
personality characteristics should not be expected to influence all aspects of performance. Aircraft han-
dling examines only the most individual physical and technical portions of the flying task, while the per-
sonality battery emphasizes differences in motivation towards organizing and performing tasks. Predic-
tions of influence should be carefully defined to correspond to these motivational and organizational
portions of the task.
Perhaps this is the primary lesson to be drawn from decades of personality research in aviation
(cf. Dolgin and Gibb, 1988): that the variety of tasks comprising crew performance may be differentially
21
influencedbyanumberof factors.Theseincludecharacteristicsof individuals(i.e.,skill, personality,
andexperience),their organizations(military or airline),theirmission(tactical,commuter,or long-haul),
training(technicalproficiencyandcrewcoordination)andtaskdesign(oldvs.newtechnology).For
example,while theliteratureonpersonalityeffectsonpilotperformanceis overwhelminglydominated
by failuresto demonstratereliablelinkages,mostof thisresearchexaminesonly whetherpersonality
predictscompletionof initial pilot training.Thevarietyofotherpossiblelinks betweenpersonalityand
performance-relevantfactorshavebeenleftrelativelyunexplored.Oneexceptionisthatpersonality
characteristicsappearto reliably differentiatethosewhoareattractedtoaviationfromthegeneralpopu-
lation(Fry andReinhardt,1969;Novello andYoussef,1974a,1974b).Thepresentstudysuggeststhat
personalitymayplay acritical rolein day-to-dayand,especiallyhigh-workloadtaskperformance.Char-
acteristicsof thecaptainandprobablythecrewimpacttheeffectivenesswith whichthe crewplansand
respondsto inflight problems.Perhapsaviationpsychologistsinterestedin personalitycannowbeginto
distinguishmorepreciselywhereandwhenthesecharacteristicsinfluenceperformanceoutcomes.
At leasttwo otherissuesareraisedbytheresults:definingthelimits of crew familiarity effects,
and devising strategies for conducting operationally valid research. The remainder of this discussion
highlights those issues.
4.1 Crew Familiarity
The familiarity effect among I- led crews raises a number of important questions. Why was the
pattern of lower performance levels seen in the observer ratings not reflected in the errors committed by
I- led crews on the first day? The expert observer saw these crews as relatively ineffective, giving them
the lowest average ratings for segment three. However, these crews made no more errors than IE+ led
crews. One reason may be that process observers are, by definition, integrating more than errors into
their observations in any group task situation. In previous studies (Ruffell-Smith, 1979; Foushee and
Manos, 1981; Foushee et al., 1986; Kanki, Lozito, and Foushee, 1989), patterns of flight crew commu-
nications were significantly related to crew performance. As a result, we have incorporated communica-
tions dimensions into our observer rating scales. These scales require the observer to use his experience
and professional judgment to evaluate how crews make decisions, handle inter- and intracrew communi-
cations, prioritize problems, deal with distractions, and distribute workload. In short, these ratings seek
to evaluate the process by which crewmembers coordinate their activities. Since problems of crew
coordination do not always produce observable errors, we should not expect ratings reflecting process to
be perfectly correlated with performance outcomes. They are related, but they are not the same. Prob-
lems reflected in observer ratings are important in their own fight because they may raise the probability
that errors will be committed or not corrected quickly. We would argue that our observer ratings reflect
the fact that there were significant process problems within I- crews.
This argument is also consistent with the conceptual framework proposed by McGrath (1964,
1984). In this model, the link between input variables (such as personality profiles) and group outcomes
(such as errors of communication or action) is mediated by the process of group activities (such as pat-
terns of communication). While it is possible to identify links between input and process or process and
outcome variables, these relationships may diverge somewhat from input-outcome relationships. The
observer ratings may be viewed as integrating both process and outcome information.
22
Theideathatfamiliarity mayhaveaffectedtheprocessof crewinteractionin I- crewsraises
anotherquestion.Whatbehaviorswerechangingoverthecourseof thesimulation?Thereareatleast
threepossibilities.(1)Thecaptainsmayhavealteredtheirbehavior,(2) thejunior crewmembersmay
haveadaptedto thecaptain'sstyle,or (3) theperformancesituationsoneachdaymayhaveemphasized
differing elementsof leadershipor performance.Adaptationby subordinatesmaybethemostplausible
explanation.It is notunusualfor acrewto becomposedof individualswhohavenevermetpriorto the
beginningof a trip.Accordingly,anadjustmentperiodis likely duringthe first few flight segments,and
it is probablethatsubordinatecrewmembersoftenattempttotailortheirbehaviorsto thecaptain's
expectations.Sincethecaptain'sroleismorecentralto cockpitorganizationthanotherpositionsand
becauseit carriesfinal decisionauthority,we suspectthat acaptainis lesslikely to changehis behavior
toadapttothecrew(althoughthismayoccurtoalesserdegreeaswell).
Fousheeetal.(1986)demonstratedsignificantprocessdifferencesbetweencrewsthathad
recentlyflown togethervs.Ihosethathadnot.Themostlikely explanationis that,all subordinatecrew
membersareinitially tentativein theirbehaviorbecausetheyareawaitingsignalsfrom theleaderabout
howheor sheexpectsthecockpitcrewtooperate.Ingeneral,IE+ leaderswouldbeexpectedto very
quicklycreateanatmospherein whichopencommunicationis encouraged.Theoretically,I- leaders
wouldnot beaslikely todo soandmightby naturetendto discouragequestioningby subordinates.
After theinitial adjustmentprocess,subordinatesin I- crewsmayhavebeenableto work moreeffec-
tively becausetheykneww trottoexpect.On theotherhand,EC- ledcrewsneverseemedto makeany
adjustmentsoverthecourseof thesimulation.ThatI- andEC- leadersdiffer substantiallyin taskmoti-
vationor instrumentalitymaybesufficientto explainthedifferencesin theeffectsof familiarity, but
tangibleevidenceof bothchangeamongtheI- crewsanddifferencesbetweenEC-andI- crewswill
havetoawaittheprocessanalysesof datageneratedby this study.
Anotherimportantideais relatedto thegeneralityof thisfamiliarity effectandits application
overtime.ThecurrentstudyandtheFousheeetal.(1986)studycomparedcrewswho hadworked
togetherfor only twoor threedays,andfamiliarity seemedto provideaperformancebenefitin a sub-
stantialnumberof thesecases.However,weknowlittle aboutteamperformanceoverlongerdurations,
andit is quitepossiblethatincreasingfamiliarity couldultimately resultin worseperformance.In this
report,wehavesuggestedthatanumberof characteristicsof I- leadersmight beviewedasaversive
undercertaincircumstances.I- leadersarecharacterizedby highlevelsof impatienceandirritability,
competitiveness,andverbalaggressivenesscombinedwith low levelsof expressivity.Sofar, wehave
shownthatleaderspossessiJ_gsuchcharacteristicsarecapableof operatingin crewsthatperformatrela-
tively high levelsafteraninitial adjustmentperiod,but it maynotbe possibleto maintaintheselevels
overlongperiodsof time.Over time,we wouldpredictthatindividualspossessingtheseattributes
wouldhavedifficulty maintainingeffectivecrewperformance.If thisis thecase,wewouldbeparticu-
larly concernedaboutcrewmembersfitting this profileandparticipatingin long-durationoperationssuch
asship,submarine,or spacestationoperations.Thetwo-daytimeperiodof thissimulationstudywasnot
sufficientto exploretheselimits, andit is importantthatresearchonlongerdurationflights be
accomplished.
23
4.2 Implications for Designing Operationally-Valid Research
Foushee (1984) argued that flight simulation provides an ideal environment in which to conduct
research meeting both basic and applied criteria. Since aircraft simulators provide high levels of both
realism and experimental control, they make ideal laboratories for experimentation, and there are clearly
fewer problems in generalizing to real-world behavior. On the negative side, high-fidelity simulation
demands a large investment, and sound use of these expensive facilities demands that proposed theories
or models be conceptualized with real-world applications in mind. This suggests in turn that researchers
should collect evidence for the ecological validity of their theories prior to testing in a high-fidelity envi-
ronment. The tradeoff identified here argues for a program of research moving from lower-fidelity to
increasingly high-fidelity environments.
We believe that high-fidelity studies represent an important direction for psychological research.
Helmreich (1983) has argued that researchers have overly restricted their work to the extent that it does
not apply to real-world phenomena. The failure of personality researchers to demonstrate strong links
with important behavioral dimensions may be in large part the result of the "sterile" laboratory tradition
so predominant in psychological research. The structuring of artificial tasks for laboratory experimenta-
tion, a process viewed as necessary for both control and assessment, may also tend to create artificial
conditions that account for far more behavioral variation than the experimental variables themselves.
We chose the personality battery utilized in this study in large part because of the substantial
body of real-world performance-relevant data collected by Spence, Helmreich, and their colleagues
(Spence and Helmreich, 1983; Helmreich, 1982, 1986; Chidester, 1987). For example, Spence and
Helmreich's (1983) measures of achievement motivation were shown to predict performance among
academic scientists and engineers. Moreover, Helmreich (1986) found instrumentality and expressivity
to be significantly correlated with check airman evaluations of individual pilot performance. This study
has provided further evidence for the validity of instrumentality and expressivity as meaningful and
important components of individual personality. As we consider the development of selection criteria for
future aerospace operations, these dimensions appear to be strong candidates for representation. In the
absence of such preliminary validation work, the application of personality models or measures to pre-
dicting pilot performance amounts to what Ellis and Conrad (1948) referred to as a criterion shift in the
absence of empirical justification. Given the cost of high-fidelity research, applying untested measures
or models becomes highly risky. This phenomenon may have been a factor in the problem-plagued
search for personality predictors of performance in past research. Thus, high fidelity research environ-
ments may put us in a better position to resume our search, but only if researchers accomplish prelimi-
nary validation work. Personality theory and research should move towards a closer association with
real-word performance.
24
REFERENCES
Anderberg, M. R.: Cluster Analysis for Applications. New York: Academic, 1973.
Blake, R. R.; and Mouton, J. S: The New Managerial Grid. Houston: Gulf, 1978.
Bluth, B. J.: Soviet space su'ess. Science, 81, 30-35,1981.
Chidester, T. R.: Trends and individual differences in response to short-haul flight operations. Aviation,
Space, and Environraental Medicine, vol. 61, Feb. 1990, pp. 132-138.
Chidester, T. R.: Selection tor optimal crew performance: Relative impact of selection and training. In
R. S. Jensen, ed., Proceedings of the Fourth International Symposium on Aviation Psychology,
Columbus, OH, 473-479, 1987.
Dolgin, D. L.; and Gibb, G. D.: A review of personality measurement in aircrew selection. NAMRL
Monograph no. 36, Naval Aerospace Medical Research Laboratory, Pensacola, FL, 1988.
Ellis, A.; and Conrad, H. S.: The validity of personality inventories in military practice. Psychological
Bulletin, vol. 5,385-426, 1948,
Fiedler, F. E.: A theory of l,_adership effectiveness. New York: McGraw-Hill, 1967.
Foushee, H. C.: Dyads and triads at 35,000 feet: Factors affecting group process and aircrew perfor-
mance. American P:_ychologist, vol. 39, 885-893, 1984.
Foushee, H. C.; and Helmreich, R. L.: Group interaction and flightcrew performance. In E. L. Wiener
and D. C. Nagel (Fxls.), Human Factors in Aviation. San Diego: Academic, 189-227, 1988.
Foushee, H. C.; Lauber, J. K.; Baetge, M. M.; and Acomb, D. B.: Crew factors in flight operations III:
The operational significance of exposure to short-haul air transport operations.
NASA TM-88322, 1986.
Foushee, H. C.; and Manos, K. L.: Information transfer within the cockpit: Problems in intracockpit
communications. In C. E. Billings and E. S. Cheaney, eds., Information transfer in the aviation
system. NASA TP-1875, 1981.
Fry, G. E.; and Reinhardt, R. F.: Personality characteristics of jet pilots as measured by the Edwards Per-
sonal Preference Schedule. Aerospace Medicine, vol. 40, 484-486, 1969.
Ginnett, R.: The formation process of airline flight crews. In R. S. Jensen, ed., Proceedings of the Fourth
International Symp:_sium on Aviation Psychology, Columbus, OH, 399-405, 1987.
25
Gregorich,S.;Helmreich,R.L.; Wilhelm,J.A.; andChidester,T. R.:Personalitybasedclustersaspre-
dictorsof aviatorattitudesandperformance.In R.S.Jensen,ed.,Proceedingsof theFifth Inter-
nationalSymposiumonAviationPsychology,Columbus,OH, 1989.
Hackman,J.R.;andHelmreich,R. L.: Assessingthe behaviorandperformanceof teams in organiza-
tions: The case of air transport crews. In D. R. Peterson and D. B. Fishman, eds., Assessment for
decision. New Brunswick, NJ.: Rutgers University Press, I987.
Hart, S. G.; and Staveland, L. E.: Development of a multi-dimensional workload rating scale: Results of
empirical and theoretical research. In P. A. Hancock and N. Meshkati, eds., Human Mental
Workload. Elsevier, Amsterdam, The Netherlands, 1988.
Helmreich, R. L.: Pilot selection and training. Paper presented at the annual meeting of the American
Psychological Association, Washington, DC, 1982.
Helmreich, R. L.: Applying psychology in outer space: Unfulfilled promises revisited. American Psy-
chologist, vol. 38, 1983, pp. 445-450.
Helmreich, R. L.: Studying flight crew behavior: A social psychologist encounters the real world.
Invited address to the annual meeting of the American Psychological Association, Washington,
DC, 1986.
Helmreich, R. L.; Sawin, L. L.; and Carsrud, A. L.: The honeymoon effect in job performance: Delayed
predictive power of achievement motivation. Journal of Applied Psychology, vol. 71, 1986,
pp. 1085-1088.
Helmreich, R. L.; Spence, J. T.; and Pred, R. S.: Making it without losing it: type A, achievement moti-
vation, and scientific attainment revisited. Personality and Social Psychology Bulletin, vol. 14,
1988, pp. 495-504.
Helmreich, R. L.; and Wilhelm, J. A.: Evaluating cockpit resource management training. In R. S. Jensen,
ed., Proceedings of the Fourth International Symposium on Aviation Psychology, Columbus,
OH, 1987, pp. 440-446.
Jenkins, C. D.; Zyzanski, S. J.; and Rosenman, R. H.: Progress toward validation of a computer scored
test for the Type A coronary prone behavior pattern. Psychometric Medicine, vol. 33, 1971
pp. 193-202.
Kanki, B. G.; Lozito, S. C.; and Foushee, H. C.: Communication Indexes of Crew Coordination. Avia-
tion, Space, and Environmental Medicine, vol. 60, 1989, pp. 56-60.
Lauber, J. K.; and Foushee, H. C.: Guidelines for line-oriented flight training. NASA CP-2184, 1981.
McGrath, J. E.: Social psychology: A brief introduction. Holt (New York), 1964.
McGrath, J. E.: Groups: Interaction and performance. Prentice-Hall (Englewood Cliffs, NJ), 1984.
26
Melton,A. W., ed.:Apparatu:;Tests,Air ForcesAviation PsychologyProgramResearchReports,No.4.
Washington:U. S.GovernmentPrintingOffice, 1947.
Melton,A. W.: (1957).Military psychologyin theUnitedStatesof America.AmericanPsychologist,
vol. 12,1957,pp. 740-746.
Melton,R. S.:Studiesin theevaluationof thepersonalitycharacteristicsof successfulnavalaviators.
Aviation Medicine,vol.25,1954,pp.600-604,.
Norusis,M. J.:Chapter3: ClusterAnalysis.InM. J.Norusis,ed.,.SPSS/PC+AdvancedStatisticsV2.0.
Chicago:SPSSInc., 1988.
Novello,J. R.; andYoussef,Z. I.: Psycho-socialstudiesin generalaviation:I. Personalityprofile of
malepilots.AerospaceMedicine,vol. 45, 1974a,pp.185-188.
Novello,J. R.; andYoussef,7.I.: Psycho-social studies in general aviation: II. Personality profile of
female pilots. Aerospace Medicine, vol. 45, 1974b, pp. 630-633.
Orlady, H. W.; and Foushee, H. C., eds.: Proceedings of the NASA/MAC workshop on cockpit resource
management. NASA (;P-2455, 1987.
Pred, R. S., Spence, J. T.; anc[ Helmreich, R. L.: The development of new scales for the Jenkins Activity
Survey measure of the Type A construct. Social and Behavioral Science Documents, 1986.
Ruffell-Smith, H. P.: A simulator study of the interaction of pilot workload with errors, vigilance, and
decisions. NASA TM-78482, 1979.
Sells, S. B.: Development of a personality test battery for psychiatric screening of flying personnel. Avi-
ation Medicine, vol. 25, 1955, pp. 35-45.
Spence, J. T.; and Helmreich, R. L.: Masculinity and femininity: Their psychological dimensions, corre-
lates, and antecedents. University of Texas Press (Austin), 1978.
Spence, J. T.; and Helmreich. R. L.: Achievement-related motives and behavior. In J. T. Spence, ed.,
Achievement and achievement motives: Psychological and sociological approaches. W. H.
Freeman and Co. (New York), 1983.
Spence, J. T.; Helmreich, R. L.; and Holahan, C. K.: Negative and positive components of psychological
masculinity and femininity and their relationships to self-reports of neurotic and acting out
behaviors. J. Personality and Social Psychology, vol. 37, 1979, pp. 1673-1682.
27
APPENDIX A
DEFINITIONS OF PERSONALITY SCALES
Extended Personal Attributes Questionnaire (EPAQ; Spence, Helmreich, and Holahan, 1979)
Instrumentality- a cluster of positive attributes reflecting goal-orientation and independence (active,
self-confident, can stand up to pressure)
Expressivity- a cluster of positive attributes reflecting interpersonal warmth and sensitivity (gentle,
kind, aware of the feelings of others)
Negative Instrumentality- negative characteristics reflecting arrogance, hostility, and interpersonal
invulnerability (boastful, egotistical, dictatorial)
Negative Communion- self-subordinating, subservient, or unassertive characteristics (gullible,
spineless, subordinates self to others)
Verbal Aggressiveness- verbal passive-aggressive characteristics (complaining, nagging, fussy)
Negative Instrumentality- cluster of negative attributes reflecting emotional invulnerability,
cynicism, and hostility (arrogant, boastful, egotistical, dictatorial)
Work and Family Orientation Questionnaire (WOFO; Spence and Helmreich, 1978)
Mastery- a preference for challenging tasks and striving for excellence ("If I am not good at some-
thing, I would rather keep struggling to master it than move on to something I may be good
at")
Work- a desire to work hard and do a good job ("I find satisfaction in working as well as I can")
Competitiveness- a preference for tasks with clear winners and losers and a desire to outperform
others ("It annoys me when other people perform better than I do")
Achievement and Impatience Scales (Pred, Helmreich, and Spence, 1986)
Achievement Striving- a cluster of characteristics related to hard work, activity, and seriousness in
approaching work tasks ("How much does your job 'stir you into action? .... Compared to
others, how much effort do you put forth?")
Impatience/Irritability- ("How easily do you get irritated? .... When a person is talking and takes too
long to come to a point, how often do you feel like hurrying the person along?")
28
APPENDIX B
FLIGHT EXPERIENCE ITEMS
Crew Position (circle one) CAPT FO SO
How long have you been employed by your present airline? __.yrs. __
How many hours have you logged as a pilot in the following categories?
Military
General Aviation
Airline
Other
Age Sex
How many hours have you lo_;ged in your present crew position (Captain,
First Officer)?
mnths.
How many hours have you lo_r,ged in the B-727?
29
APPENDIX C
PERSONALITY BATTERY
Extended Personal Attributes Questionnaire (EPAQ)
The items below inquire about what kind of aperson you think you are. Each item consists of apair of
characteristics, with the letters A - E in between. For example:
Not at all artistic Very artistic
A ........ B ........ C ........ D ........ E
Each pair describes contradictory characteristics -- that is, you cannot be both at the same time, such as
very artistic and not at all artistic.
The letters form a scale between the two extremes. You are to choose a letter which describes where you
fall on the scale. For example, if you think you have no artistic ability, you would choose A. If you think
you are pretty good, you might choose D. If you are only medium, you might choose C, and so forth.
Be sure to answer every question, even if you're not sure.
Not at all aggressive Very aggressive
A ........ B ........ C ........ D ........ E
Very whiny Not at all whiny
A ........ B ........ C ........ D ........ E
Not at all independent Very independent
A ........ B ........ C ........ D ........ E
Not at all arrogant Very arrogant
A ........ B ........ C ........ D ........ E
Not at all emotional Very emotional
A........ B ........ C ........ D ........ E
Very submissive Very dominant
A ........ B........ C ........ D ........ E
Very boastful Not at all boastful
A........ B........ C ........ D ........ E
Not at all excitable Very excitable
in a major crisis in a major crisis
A ........ B ........ C ........ D ........ E
30
Verypassive Very active
A........B........C........D........E
Notatallegotistical Very egotistical
A........B........C........D.......E
Not atall abletodevote Able todevote
selfcompletelytoothers selfcompletelytoothers
A........B........C........D........E
Notatall spineless Very spineless
A........B........C........D........E
Veryrough Very gentle
A........B........C........D........E
Notatallcomplaining Very complaining
A........B........C........D........E
Not atall helpfultoothers Very helpfulto others
A........B........C........D........E
Notatallcompetitive Verycompetitive
A........B........C........D........E
Subordinatesoneselftoothers Neversubordinatesselftoothers
A........B........C........D........E
Veryhomeoriented Veryworldly
A........B........C........D........E
Very greedy Not atall greedy
A........B........C........D........E
Not atall kind Verykind
A........B........C........D........E
IndifferentIoothers' approval Highly needful of others' approval
A ........ B ........ C ........ D ........ E
Very dictatorial Not at all dictatorial
A ........ B ........ C ........ D ........ E
ii:eelings not easily hurt Feelings easily hurt
A ........ B ........ C ........ D ........ E
31
Doesn'tnag Nagsalot
A........B........C........D........E
Notatallawareof Very aware of
feelings of others feelings of others
A ........ B ........ C ........ D ........ E
Can make decisions easily Has difficulty making decisions
A ........ B ........ C ........ D ........ E
Very fussy Not at all fussy
A ........ B ........ C ........ D ........ E
Give up very easily Never gives up easily
A ........ B ........ C ........ D ........ E
Very cynical Not at all cynical
A ........ B ........ C ........ D ........ E
Never cries Cries very easily
A ........ B ........ C ........ D ........ E
Not at all self-confident Very self-confident
A ........ B ........ C ........ D ........ E
Does not look out only Looks out only
for self for self
A ........ B ........ C ........ D ........ E
Feels very inferior Feels very superior
A ........ B ........ C ........ D ........ E
Not at all hostile Very hostile
A ........ B ........ C ........ D ........ E
Not at all understanding Very understanding
of others of others
A ........ B ........ C ........ D ........ E
Very cold in relations Very warm in relations
with others with others
A ........ B ....... C ........ D ........ E
Very servile Not at all servile
A ........ B ........ C ........ D ........ E
32
Very little needfor security Very strongneedfor security
A........B........C........D........E
Not atall gullible Very gullible
A........B........C........D........E
Goestopiecesunderpressure Standsupwellunderpressure
A........B........C........D........E
Achievement and Impatience Scales
For each question below, pie ase select the alternative that best describes yourself or your opinion. Indi-
cate the alternative you chocse by circling the appropriate letter on the scale, A, B, C, D, or E.
How much does your job "stir you into action"?
---A ............... B ............. C ............... D ............. E ....
Much less About average Much more often
than others than others
When a person is talking anti takes too long to come to the point, how often do you feel like hurrying the
person along?
---A .............. B ............. C ............... D ............. E ....
Very frequently Occasionally Almost never
Nowadays, do you consider yourself to be:
---A ............... B ............. C ............... D ............. E ....
Very hard- Slightly Very relaxed
driving hard-driving and easy-going
How would your best friend or others who know you well rate your general level of activity?
---A ............... B ............. C ............... D ............. E ....
Too slow, About Very active,
should be more active average should slow down
Typically, how easily do you get irritated?
---A ............... B ............. C ............... D ............. E ....
Extremely Somewhat Not at all
easily easily easily
33
How seriouslydo youtakeyourwork?
---A.............. B............. C ............... D............. E ....
Much less About Much more
seriously average seriously
than most than most
How often do you set deadlines or quotas for yourself at work or other activities?
---A .............. B ............. C ............... D ............. E ....
Very often Sometimes Almost never
Do you tend to do most things in a hurry?
---A .............. B............. C ............... D ............. E ....
Not at all More true Definitely
true than not true
Compared with others in my occupation, the amount of effort I put forth is:
---A .............. B ............. C ............... D ............. E ....
Much more About average Much less
How is your "temper" these days?
---A .............. B ............. C ............... D ............. E ....
Very hard Sometimes get I seldom get
to control angry but easy angry
to control
Compared with others in my occupation, I approach life in general:
---A .............. B ............. C ............... D ............. E ....
Much more About Much less
seriously average seriously
When you have to wait in line such as at a restaurant, the movies, or the post office, how do you usually
feel?
---A .............. B ............. C ............... D ............. E ....
Accept it Feel very impatient
calmly and refuse to stay long
34
Work and Family Orientation Questionnaire
The following statements describe reactions to conditions of work and challenging situations.
For each item, indicate how much you agree or disagree with the statement, as it refers to yourself, by
choosing the appropriate letter on the scale, A, B, C, D, or E. When you have decided on your answer,
circle the letter that best describes your attitude. There are no right or wrong answers.
I would rather do something at which I feel confident and relaxed than something which is challenging
and difficult.
---A .............. B ............. C ............... D ............. E ....
Strongl3, Slightly Neither agree Slightly Strongly
agree agree nor disagree disagree disagree
It is important for me to do my work as well as I can even if it isn't popular with my co-workers.
---A .............. B ............. C ............... D ............. E ....
Strongly Slightly Neither agree Slightly Strongly
agree agree nor disagree disagree disagree
I enjoy working in situations involving competition with others.
---A ............... B ............. C ............... D ............. E ....
Strongl3, Slightly Neither agree Slightly Strongly
agree agree nor disagree disagree disagree
When a group I belong to phms an activity, I would rather direct it myself than just help out and have
someone else organize it.
---A ................ B ............. C ............... D ............. E ....
Strongl) Slightly Neither agree Slightly Strongly
agree agree nor disagree disagree disagree
I would rather learn easy fun games than difficult thought games.
---A .............. B ............. C ............... D ............. E ....
Strongly Slightly Neither agree Slightly Strongly
agree agree nor disagree disagree disagree
It is important to me to perform better than others on a task.
---A .............. B ............. C ............... D ............. E ....
Strongly Slightly Neither agree Slightly Strongly
agree agree nor disagree disagree disagree
35
I find satisfactionin working aswellasI can.
---A.............. B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
If I amnotgoodatsomethingI wouldratherkeepstrugglingtomasterit thanmoveonto somethingI
maybegoodat.
---A.............. B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
OnceI undertakeatask,I persist.
---A.............. B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly
agree agree nordisagree disagree
I preferto work in situationsthatrequireahighlevelof skill.
---A.............. B............. C............... D............. E....
Strongly
disagree
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
Thereis asatisfactionin ajob well done.
---A.............. B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
I feelthatwinning isimportantin bothwork andgames.
---A.............. B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
I moreoftenattempttasksthatI amnot sureI candothantasksthatI believeI cando.
---A.............. B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
36
I find satisfactionin exceedingmypreviousperformanceevenif I don't outperformothers.
---A.............. B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
I like to work hard.
---A.............. B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
Partof my enjoymentin doiagthingsisimprovingmypastperformance.
---A.............. B............. C............... D............. E....
Strongb._ Slightly Neitheragree Slightly
agree agree nordisagree disagree
It annoysmewhenotherpeopleperformbetterthanI do.
Strongly
disagree
---A............... B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
I like to bebusyall thetime
---A............... B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
I try harderwhenI'm in competitionwith otherpeople.
---A.............. B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
It is importanttomethatmyjob offersopportunityfor promotionandadvancement.
---A............... B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
37
It is importantto myfuturesatisfactionthatmyjob payswell.
---A.............. B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
It is importanttomethatmy job bringsmeprestigeandrecognitionfromothers.
---A.............. B............. C............... D............. E....
Strongly Slightly Neitheragree Slightly Strongly
agree agree nordisagree disagree disagree
38
APPENDIX D
TASK LOAD INDEX ITEMS AND DEFINITIONS
Mental Demand Low/High How much mental and perceptual activity was required
(e.g., thinking, deciding, calculating, remembering, look-
ing, searching, etc)? Was the task easy or demanding,
simple or complex, exacting or forgiving?
Physical Demand Low/High How much physical activity was required (e.g., pushing,
pulling turning, controlling, activating, etc.)?
Temporal Demand Low/High How much time pressure did you feel due to the rate or
pace at which the tasks or task elements occurred? Was
the pace slow and leisurely or rapid and frantic?
Performance Perfect/Failure How successful do you think you were in accomplishing
the goals of the task set by the experimenter (or yourself)?
How satisfied were you with your performance in
accomplishing these goals?
Effort Low/High How hard did you have to work (mentally and physically)
to accomplish your level of performance?
Frustration Level Low/High How insecure, discouraged, irritated, stressed and
annoyed versus secure, gratified, content, relaxed and
complacent did you feel during the task?
All items were rated on a seven-point scale.
39
APPENDIX E
ITEMS COMPRISING THE OBSERVER'S RATING OF CREW PERFORMANCE
Items scored "not at all" to "very much" on afive-point scale:
Communications were thorough, addressing coordination, planning, and problems anticipated.
Open communications were established among crewmembers.
Timing of communications was proper.
Active participation in decision making process was encouraged and practiced.
Alternatives were weighed before decisions were made final.
Crewmembers showed concern with accomplishment of tasks at hand.
Crewmembers showed concern for the quality of interpersonal relationships in the cockpit.
Work overloads were reported and work prioritized or redistributed.
Crewmembers planned ahead for high workload situations.
Appropriate resources were used in planning.
Items scored on a five-point bipolar scale:
Overall vigilance Inattentive
Interpersonal climate Hostile
Preparation and planning Late
Distractions avoided or prioritized Poor -
Workload distributed and communicated Poor
Overall Crew Effectiveness Poor -
Alert
Friendly
Well in Advance
Excellent
Excellent
Excellent
40
APPENDIX F
OBSERVER RATINGS OF PHASE PERFORMANCE
Items (scored on a 5-point scale ranging from below average to above average):
Procedures, Checkli:;ts, and Callouts
ATC and Company Communications
Planning and Situation Awareness
Crew Coordination _Lnd Communication
Overall Performance and Execution
Phases of flight in which ratings were completed:
_egment 3
Prestart
Taxi/Fakeoff
Climb
Cruise
Runaway/Jammed Stabilizer
Oil Filter Bypass
Approach/Landing
Segment 5
Prestart
Taxi/Takeoff
Climb
Cruise
Approach and Missed Approach
System A Hydraulic Failure
Cruise to Alternate
Approach/Landing
41
RI/ A
NaNon_ Aeronmu_c$ and
SpaCe Administzs_on
I. Report No.
NASA TM- 102259
Report Documentation Page
2. Government Accession No. 3. Recipient's Catalog No.
4. Title and Subtitle
Personality Factors in Flight Operations: Volume I. Leader
Characteristics and Crew Performance in a Full-Mission Air
Transport Simulation
7. Author(s)
Thomas R. Chidester, Barbara G. Kanki, H. Clayton Foushee (Federal
Aviation Administration, Washington, DC), Cortlandt L. Dickinson
(Menlo Park, CA), and Stephen V. Bowles (California Professional
School for Psychology, Berkeley, CA)
9. Performing Organization Name and Address
Ames Research Center
Moffett Field, CA 94035-1000
5. Report Date
April 1990
6. Performing Organization Code
8. Performing Organization Report No,
A-90018
12. Sponsoring Agency Name and Address
National Aeronautics and Space Administration
Washington, DC 20546-0001
10. Work Unit No.
199-06-12
11. Contract or Grant No.
13. Type of Report and Period Covered
Technical Memorandum
14. Sponsoring Agency Code
15. Supplementary Notes
Point of Contact: Thomas R. Chidester, Ames Research Center, MS 262-5
Moffett Field, CA 94035-1000
(415) 604-5785 or FTS 464-5785
16. Abstract
Crew effectiveness is a joint product of the piloting skills, attitudes, and personality characteristics of team
members. As obvious asthis point might seem, both traditional approaches to optimizing crew performance and
more recent training development highlighting crew coordination have emphasized only the skill and attitudinal
dimensions. This volume is the first in a series of papers on this simulation. A subsequent volume will focus on
pattems of communication within crews. This paper reports the results of a full-mission simulation research study
assessing the impact of individual personality on crew performance. Using a selection algorithm described in
previous research, captains were classified as fitting one of three profiles along a battery of personality assessment
scales. The performances of 23 crews led by captains fitting each profile were contrasted over a one-and-one-half-
day simulated trip. Crews led by captains fitting a positive Instrumental-Expressive profile (high achievement
motivation and interpersonal skill) were consistently effective and made fewer errors. Crews led by captains fitting
a Negative Expressive profile (below average achievement motivation, negative expressive style, such as complain-
ing) were consistently less effective and made more errors. Crews led by captains fitting a Negative Instrumental
profile (high levels of competitiveness, verbal aggressiveness, and impatience and irritability) were less effective
on the first day but equal to the best on the second day. These results underscore the importance of stable personality
variables as predictors of team coordination and performance.
17. Key Words (Suggested by Author(s))
Crew coordination
Personality factors
Pilot performance
lB. Distribution Statement
Unclassified- Unlimited
Subject Category - 03
19. Security Classif. (of this report)
Unclassified
20. Security Classif. (of this page)
Unclassified 21. No. of Pages
42 22. PriceA03
NASA FORM 1626 OCT_ For sale by the National Technical Information Service, Springfield, Virginia 22161
... İkinci grup uçuş ekibi tüm uçuş safhalarında, birinci gruba oranla daha düşük bir performans; üçüncü grup uçuş ekibi ise diğer iki gruptan daha düşük performans gösterdikleri, ancak uçuşun başlangıç safhasında daha kötü bir performans sergilemelerine rağmen son safhada göreceli olarak bir performans artışı gözlemlenmiştir. Üçüncü grup uçuş ekiplerinin kaptan ile birlikte geçirdikleri zaman içerisinde, bu tarz bir kaptan profiline nasıl ayak uyduracaklarını zamanla öğrenmiş oldukları değerlendirilmiştir [11,27,43,45]. ...
... Uçuş ekibi kaynak yönetimi ve kişilik ilişkisi başlığı altında ifade edilen araştırma sonuçları [3,11,21,27,43,45,47] bağlamında belirlenen araştırma sorusu ve hipotezler aşağıda sunulmuştur: ...
Article
Full-text available
zet Alan yazındaki pek çok araştırma sonucu elde edilen verilere göre, uçak kazalarının en önemli kök nedenleri arasında ilk sırada insan faktörleri ve insan faktörleri ile ilişkili diğer nedenler yer almaktadır. Bu bağlamda, Uçuş Ekibi Kaynak Yönetimi yaklaşımı; "emniyetli, etkin ve verimli bir uçuş operasyonu için elde mevcut bilgi, donanım ve insan gibi tüm kaynakların kullanımı" olarak tanımlanmaktadır. Bu çalışmanın amacı, öncelikle pilotların UEKY tutumlarını, bu tutumlar ile kişilik yapıları arasındaki ilişkileri saptamaktır. Araştırma kapsamında, UEKY'ne ilişkin pilotların tutumları ile meslekî kökenleri, statüleri, yaşları, uçuş tecrübeleri ve eğitim düzeyleri gibi demografik değişkenler arasındaki ilişkileri incelenmiştir. Araştırma, İstanbul'da ulusal bir havayolu işletmesindeki 340 pilotun katılımıyla gerçekleştirilmiştir. Araştırmada Uçuş Ekibi Kaynak Yönetimi Tutumları Ölçeği ve Meslekî Kişilik Envanteri kullanılmıştır. Verilerin analizinde tanımlayıcı istatistikler, korelâsyon, regresyon, ANOVA ve Scheffe testleri kullanılmıştır. Yapılan istatistikî analizlere göre pilotların UEKY tutumları ve kişilik yapıları arasında istatistikî olarak anlamlı ilişkiler olduğu, demografik değişkenlere göre istatistikî olarak anlamlı farklılıklar olduğu saptanmıştır. Anahtar Sözcükler: Havacılıkta İnsan Faktörleri, Uçuş Ekibi Kaynak Yönetimi, Kişilik Crew resource management: The relationships between personality types and crew resource management attitudes of pilots Abstract In the literature there are many results about the human factors and the other human related factors as the most important root causes of aircraft accidents. In this regard, Crew Resource Management approach is defined as using all of the resources such as knowledge, equipment and people for safe, effective and efficient flight operations. The main purpose of this study is to determine the pilots' attitudes towards CRM and to determine the relationship between CRM attitudes and personality types. Within this study, the relationship between the pilots' CRM attitudes and the demographic variables such as the status, ages, levels of education, flight experiences, and professional origins of pilots researched. 340 pilots of a national airline in Istanbul were participated in the research. Crew Resource Management Attitudes Questionnaire and Vocational Personality Inventory were used. The data is analyzed by descriptive statistics, correlation analysis, regression analysis, ANOVA and Scheffe tests. According to the statistical analysis, there are significant correlations between the personality types and CRM attitudes.
... [1] Personality factors, in general, have been found to contribute to crew effectiveness and are important predictors of team performance in aerospace environments. [2] Numerous studies have been conducted to link cognitive functioning with flight performance, and different measurements of cognitive efficiency have been identified as crucial to the piloting ability, including time-sharing, [3] speed of processing, [4] attention, [5] and problem-solving. [6] Executive functions (EFs) underlie goal-directed behavior and adaptation to novel and complex situations. ...
... [10,11] Agreeableness and its relation to performance in aircrew have also advocated that there is a universal pilot personality, irrespective of experience, or position. [2,11] Another set of studies investigating associations between cognition and flight performance aimed to measure the predictive validity of pilot candidate selection tests, as determined by the relationship between students' performance on the selection tests for entry into the training program and the training outcome (indexed by flight performance). [12][13][14][15] Another study was able to explain 45% of the variance of the flight simulator performance with four Cogscreen-Aeromedical Edition predictors (speed/working memory, visual associative memory, motor coordination, and tracking) in a cohort of 100 aviators aged 50-69 years. ...
Article
Introduction Military flying entails sound psychological health along with a high level of skill and aptitude. Pilots are selected on the basis of their psychological profile, psychomotor ability, and cognitive ability. Hence, the performance of pilots can be construed as a product of skill, attitude, and personality factors. Personality factors and cognitive ability have a vast impact on the flying performance and flight safety. This acquires a whole new dimension for flying instructors, as teaching the nuances of flying to a young flight cadet is a very challenging task. Materials and Methods To study the personality and cognitive profile of the special group of flight instructors, 45 instructors and trainee instructors from flying instructors’ school, Tambaram, were administered NEO five- factor inventory and four cognitive tests comprising Stroop test, symbol-digit coding and digit span-backward and forward. The data were analyzed using descriptive statistics, t-test, and analysis of variance. Results and Discussion Trainees and instructors constituted the sample wherein they were considered as a common group; hence, comparison between them was not carried out. The aircrew manifested emotional stability, high extraversion, willingness to experience novel experiences, low to average agreeableness, and average to very high conscientiousness. Their cognitive performance was compared with available norms for general population and was found to be above average. No significant difference was found in the personality dimensions or cognitive performance on the basis of mode of entry, stream, marital status, etc. Conclusion Instructors came out to be more emotionally balanced, extraverted, and keen to have new experiences and had high conscientiousness compared to non-instructor pilots. The instructors showed above- average performance when compared to normative data provided by the test authors. It has implications for future studies with large samples and similar protocols.
... flight segment and crew-member characteristics) and makes detailed analyses of crew processes possible (cf. Chidester, Kanki, Foushee, Dickinson, & Bowles, 1990;Foushee & Manos, 1981;Kanki & Foushee, 1989;Kanki, Lozito. & Foushee, 1988). ...
Article
Full-text available
A revised version of the Cockpit Management Attitudes Questionnaire (CMAQ) is introduced. Factor analyses of responses from 3 different samples reveal comparable factor structure (previous attempts to factor analyze this measure had produced equivocal results). Implications for the measurement of attitudes and the assessment of attitude change are discussed. It is argued that the CMAQ will benefit both special training programs and efforts to explore attitude–performance linkages in air-transport operations.
... 1. Engage in contingency planning that helps set up shared mental models for emergencies. 2. Have leaders who establish a democratic environment that supports better sharing of information, and who explicitly state more plans, strategies, and intentions, consider more options, provide more explanations and give more warnings or predictions (Chidester, Kanki, Foushee, Dickinson, & Bowles, 1990;Orasanu, 1990). 3. Develop a shared understanding of the problem prior to looking for solutions, thus avoiding getting bogged down (Hirokawa, 1983), which is particularly important the more diverse the team members' backgrounds are (Citera et al., 1995). ...
Chapter
Situation awareness (SA), a person's understanding of what is happening in the current situation, is critical for performance in a wide variety of domains, including aviation, air traffic control, driving, military operations, emergency management, health care, and power grid operations. This chapter discusses SA in teams, and the factors that underlie the development of team SA which significantly affects performance in many settings. Alternating between "goal driven" and "data driven" is characteristic of human information processing that underpins much of SA development in complex worlds. Interactive SA trainer attempts to bootstrap the natural process by exposing the trainee to many, many situations in a very short period of time using computer-based training tools. The chapter concludes with a discussion of validated techniques for measuring situation awareness in individuals and teams that can be used to evaluate new system designs and training programs, and to conduct further research on this important construct.
... Considerable research has been conducted on the types of processes that teams use and their impact on team coordination and TSA. Orasanu and Salas (1993) summarize a number of studies involving aircrew and military teams to show that effective teams (1) engage in contingency planning that helps setup shared mental models for emergencies, (2) have leaders who establish a democratic environment that supports better sharing of information, and who explicitly state more plans, strategies, and intentions, consider more options, provide more explanations and give more warnings or predictions (Chidester, Kanki, Foushee, Dickinson, & Bowles, 1990;Orasanu, 1990), and 3) develop a shared understanding of the problem prior to looking for solutions, thus avoiding getting bogged down (Hirokawa, 1983), which is particularly important the more diverse the team members backgrounds (Citera et al., 1995). Klein, Zsambok and Thordsen (1993) additionally point to the importance of (1) clear delineation and understanding of tasks, goals, and roles, (2) avoidance of micro-managing but willingness to help other teammates , (3) avoidance of fixation and willingness to examine various factors of the situation, (4) encouragement of different opinions and a process to come to convergence, and (5) the ability to manage time and make changes as needed. ...
Chapter
Situation awareness (SA) has been studied extensively in individuals and teams over the past three decades. SA, an understanding of what is happening in the current situation, has been shown to be critical for performance in a wide variety of domains, including aviation, air traffic control, military operations, emergency management, healthcare, and power grid operations (Endsley, 2015b; Parasuraman, Sheridan, & Wickens, 2008; Wickens, 2008). Both overall team SA and the development of shared SA on relevant information have been found to be crucial for effective team performance in many of these domains. A model of team SA is discussed that highlights the importance of team SA requirements, team SA devices, team SA mechanisms, and team SA processes for achieving SA in teams. Common challenges for team SA and shared SA in team environments are presented, and the effectiveness of different measurement approaches for assessing both team SA and shared SA are reviewed for supporting important research in this area.
... Orasanu and Fischer, 1997), or low-®delity, such as several process control tasks (Roth et al., 1992) or medical decision tasks (Gaba, in press). Realistic features can be built in, such as temporal parameters, distractions, and workload, and subjects' behavior can be analyzed as a function of relevant factors, such as differences in levels of experience or personality (Cohen and Freeman, 1986;Chidester et al., 1990), or availability of tools or aids (Roth et al., 1987;Woods, 1993). Salas et al. (1995) argued that NDM both can and should be studied in the lab as well as`in the wild', although doing so means giving up some of the contextual features that de®ne the phenomena in the real world. ...
... Some of these findings include characteristics such as stable, tough minded, extroverted (Bartram, 1995), logical, outgoing, and attention to detail (Picano, 1991). Chidester, Kanki, Foushee, Dickinson, & Bowles (1990) found that personality factors contributed to crew effectiveness and that goal orientation, independence and intrapersonal personality characteristics were predictors of team performance in aerospace environments (Fitzgibbons, Davis & Schutte 2004). Most of these studies have focused on the military and commercial pilots, but there has been less research done concerning personality profiles of student pilots, specifically student pilots enrolled in an academic aviation flight program. ...
... A breakthrough study, conducted by Ruffel Smith [13] suggested a correlation between the leadership and communication style of the captain with overall crew performance. Further research reinforced this hypothesis [14,15], and during the 1980's, industry and government came together to form what was at the time, a new discipline of aviation team training called Cockpit Resource Management, or CRM. As it evolved and became recognized as applicable to the larger aviation community, it was renamed Crew Resource Management. ...
... Orasanu and Fischer, 1997), or low-®delity, such as several process control tasks (Roth et al., 1992) or medical decision tasks (Gaba, in press). Realistic features can be built in, such as temporal parameters, distractions, and workload, and subjects' behavior can be analyzed as a function of relevant factors, such as differences in levels of experience or personality (Cohen and Freeman, 1986; Chidester et al., 1990), or availability of tools or aids (Roth et al., 1987; Woods, 1993). Salas et al. (1995) argued that NDM both can and should be studied in the lab as well asìn the wild', although doing so means giving up some of the contextual features that de®ne the phenomena in the real world. ...
Article
Full-text available
We review the progress of naturalistic decision making (NDM) in the decade since the first conference on the subject in 1989. After setting out a brief history of NDM we identify its essential characteristics and consider five of its main contributions: recognition-primed decisions, coping with uncertainty, team decision making, decision errors, and methodology. NDM helped identify important areas of inquiry previously neglected (e.g. the use of expertise in sizing up situations and generating options), it introduced new models, conceptualizations, and methods, and recruited applied investigators into the field. Above all, NDM contributed a new perspective on how decisions (broadly defined as committing oneself to a certain course of action) are made. NDM still faces significant challenges, including improvement of the quantity and rigor of its empirical research, and confirming the validity of its prescriptive models. Copyright © 2001 John Wiley & Sons, Ltd.
Chapter
The selection of space crewmembers can be a tricky activity. Not only is it important to find people without obvious psychological problems, but it also is important to select people capable of achieving the mission goals. Issues to be considered include professional background and training that are congruent with these goals, along with dedication and persistence to get the job done. In addition, it is important to pick crewmembers who can get along with each other, form a compatible team, and enjoy each other’s company during meals, celebrations, and other social events. Hence, the two pillars of crewmember selection: the ability to work diligently in one’s area of expertise to accomplish the mission goals, and the ability to interact collegially and socially with one’s crewmates during work and leisure time activities. In this chapter, we will consider a number of important intrapsychic and interpersonal qualities that have been found to be necessary for mission success.
Article
Cluster analysis is a collective term covering a wide variety of techniques for delineating natural groups or clusters in data sets. This book integrates the necessary elements of data analysis, cluster analysis, and computer implementation to cover the complete sequence of steps from raw data to the finished analysis. The author develops a conceptual and philosophical basis for using cluster analysis as a tool of discovery and applies it systematically throughout the book. He provides a comprehensive discussion of variables, scales, and measures of association that establishes a sound basis for constructing an operational definition of similarity tailored to the needs of any particular operational definition of similarity tailored to the needs of any particular problem, and devotes special attention to the problems of analyzing data sets containing mixtures of nominal, ordinal, and interval variables. (Author)