DOCUMENT RESUME ED 220 600
Transcript of DOCUMENT RESUME ED 220 600
DOCUMENT RESUME
ED 220 600 CE 033 394
AUTHOR Hughes, R. G.; And OthersTITLE Applications of Simulator Freeze to Carrier
Guideslope Tracking Instruction. Cooperative StudySeries. Final Report, May 1, 1980-August 31, 1981.
INSTITUTION Canyon Research Group, Inc., Westlake Village,Calif.
SPONS AGENCY Air Force Human Resources Lab., Brooks AFB, Texas.;Naval Training Equipment Center, Orlando, Fla.
REPORT NO AFHRI-TR-82-3; NAVTRAEQUIPCEN-78-C-0060-9PUB DATE Jul 82CONTRAC N61339-78-C-0060NOTE 64p.
EDRS PRICE MF01/PC03 Plus Postage.DESCRLPTORS Aircraft Pilots; *Flight Training; *Instructional
Innovation; Military Training; Outcomes of Education;Postsecondary Education; Program Effectiveness;*Simulation; *Skill Development; *Training Methods;*Transfer of Training
IDENTIFIERS Air Force; Simulator Freeze
ABSTRACTTwenty-five experienced F-4 and F-16 Air Force pilots
weie instructed in carrier landings in the Visual Technology ResearchSimulator (VTRS). The training was conducted under threeinstructional conditions, two of which employed the simulator's"freeze" feature. Additionally, two methods of defining errors forcarrier glideslope tracking were examined. These experimentaltraining techniques were compared to a conventional training approachill which no "freezes" were imposed during the training sequence.4While pilots who were trained under the "freeze" condition developedcontrol strategies that distinguished them from pilots trained byconventional measures, no differences were found between these groupson rate or extent of learning. In response to a post-expetimentalquestionnaire, pilots who were trained under "freeze" conditionsindicated that the simulator "freeze" was "frustrating" and added tothe overall difficulty of the task. These pilots further reportedbeing more motivated to avoid the "freeze" than to perform-t/le taskcorrectly during training. A probe technique was used to examinedifferential transfet in lieu of the more traditionaltransfer-of-training technique. Although this experimental use of theprobe technique was a preliminary effort, it does appear to holdpromise for transfer-of-training experiments of this type.(Author/KC)
**************************************************ft********************
Reproductions supplied by EDRS are the best that can be madefrom the original document.
***********************************************************************
nIca 1 geport: '1aVTRA:CTIPCEN
fl)
411111111101111COOPERAT1VE STUDY SERIES1111111111111=mi
DoD Distribution Statement
Approved for public release:
distribution unlimited.
U.S. DEPARTMENT OF EDUCATIONNATIONAL INSTITUTE OF EDUCATION
ETCATIONAL SiESOURCES INFORMATiONCENTER tERIC
ti 1hs Glonio,nt Rh: EeP,/..c.,,o-d ham TTO oeryon O onianijhon
o+ nittr,i 41
MdIfir h1,), n made to frOfOrf
froanitn't nn ,31,ty
Po (IN nf n N14, th ;r1,,f 1r
!pot nt off c 11F
:CAT:(Yls FEE.ZET'PACK:%G
July 192
NAVAL TRAINING EQUIPMENT CENTERORLANDO, FLORIDA 328)3
AIR FORCE HUMAN RESOURCES LABORATORYBROOKS AIR FORCE BASE, TEXAS 78235
V&
UNCLASSIFIEDSECURITY CLASSIFICATION OF THIS PAGE When Data Enittred)
REPORT DOCUMENTATION PAGEREAD INSTRUCTIONS
BEFORE COMPLETING FORM1. REPORTNUMBER, NAVTRAEQUIPCEN 78-C-0060-9/AFHR121*-82-3
2. GOVT ACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER
4. TITLE (and subtitle) .
APPLICATIONS OF SIMULATOR FREEZE TO CARRIERGLIDESLOPE TRACKING INSTRUCTION
S. TYPE OF REPORT & PERIOD COVERED
Final1 May 1980 - 31 August 1981
6. PERFORMING ORG. REPORT HUMBER
TR-81-0237. Au THOR(s)
R.G. Hughes, G. Lintern, D.C. Wightman,R.B. Brooks, and J. Singleton
(Cont'd)
8. CONTRACT OR GRANT NuMBER(s)
N61339-78-C-0060
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Canyon Research Group, Inc.741 Lakefield Road, Suite BWestlake Village, California 91361
10. PROGRAM ELEMENT, PROJECT, TASKAREA & WORK UNIT NUMBERS
4781-6P1A1123-02-34 (AFHRL)
11. CONTROLLING OFFICE NAME AND ADDRESS NAVAL TRAININGEQUIPMENT CENTER, Qrlando, Florida 32813 andUSAF Human Resources Laboratory/0TWilliams AFB. AZ 85224
12. FiEPORT DATE
July 198213. NUMBER OF PAGES
5014. MONITORING AGENCY NAME & ADDRESS(11 di ffefent from Con(rolling Office) 15. SECURITY CLASS. (o/ the toport)
Unclassified
;Se. DECLASSIFICATION, DOWNGRADINGSCHEDULE
16 DISTRIBUTION STATEMENT (of this Report)
Approved for public release; distribution unlimited.
17 DISTRIBUTION STATEMENT (of the sbet:act entered in Block 20, if different (roth Report)
IS SUPPLEMENTARY NOTES
IS. XEY WORDS (ContInu on reyeraa ide II necwary and Identify by block number)
Aircraft Simulation Skill Acquisition Flight Training
Human Factors Computer-Generated Displays Visual Simulation
Transfer of Training Training Dispiays Instructional
Computer-Generated Imagery Techniques
20. ASSTRACT (Continu on reveres side if nitbeeary and identify by block number)
Twenty-live experienced F-4 and F-16 Air Force pilots were instructed'in carrier landings in the Visual Technology Research Simulator (VTRS).The training was conducted underthree instructional conditions, two of whid'employed the simulator's "freeze" feature. Additionally, two methods of defin-
ing errors for carrier glideslope tracking were examined. These experimental
training techniques were compared to a conventional training approach whereno "freezes" were imposed during the training sequence.
DD FJLRM73 1473S N 0102. LF. 0I4.6601EDITION OF 1 NOV 153 IS OUSOLETE UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS PAGE (When Data Entered)
UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS PAGE ()Mon Data Enforod)
20. Abstract (Cont'd)
While pilots who were trained under the "freeze" condition developedcontrol strategies that distinguished them from pilots trained by Con-ventional measures, no differences were found between these groups onrate or extent of learning. In response to a post experimental question-naire, pilots who were trained under "freeze" conditions indicated that thesimulator "freeze" was "frustrating" and added to the overall difficulty of
the task. These pilots further reported being more motivated to avoidthe "freeze" than to perform the task correctly during training.
A probe technique was used to examine differential transfer in lieuof the more traditional'transfer-of-training technique. Although thisexperimental use of the probe technique was a preliminary effort, it doesappear to hold promise for transfer-of-training experiments of this type.
. Author(s) (Cont'd) -
R. G. HughesUSAF Human Resources Laboratory/OTWilliams AFB, Arizona 85224
G. Lintern
Canyon Research Group, Inc.Orlando, Florida 32803
D. C. WightmanNaval Training Equipment CenterOrlando, Florida 32813
R. B. BrooksUSAF Human Resources Laboratory/0TWilliams AFB, Arizona 85224
J. Singleton, LCDRU.S. Navy
S N 0102- LF.014-6601UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS PAGE(Whon De* &tiered)
ii
NAVTRAEQUIPCEN 78C-0060-9/AFHRLTR-82-3
PREFACE
This report is the fourth in a continuing series of cooperativeventures between the Naval Training Equipmect Center and the Air Force HumanResources Laboratory. This experiment was carried out on the Navy's VisualTechology Research Simulator (VTRS) in Orlando; Florida, and is the first inthis cooperative series to employ the VTRS.
A subject of concern to both the Navy and the Air Force was addressedin the experiment reported here. Issues concerning the implementation ofinstructional strategies and employment:6f certain simulator features inline with their instructional value have been addressed in the recent past.The issue addressed in the experiment reported here concerns theinstructional value of one.common simulator feature, the "freeze" feature.The "freeze" feature allows for a suspension of the simulator task so that astudent may be "given instructional feedback while freed of the requirementto perform the task. The experiment reported hene concerns the "freeze"feature commonly found in flight simulators and whether it should beemployed in the training of a complex flight task.
A number of persons contributed to-this research. Walter S. Chambers, .
Stanley C. Collyer, Patricia Daoust and Edward Holler of the Naval TrainingEquipment Center (Code N-732); Brian Nelson, Daniel /Sheppard and DanielWestra of Canyon Research Group, Inc.; and Jack Davis and Karen Thomley ofthe University of Central Florida provided technic l support.
Twentyfive Air Force pilots from the 56th T ctical Fighter Wing,Macbill Air Force Base, Florida served as subjec s in this experiment. Theyare tc:, be commended for their cooperation. The uthors would like tospecifically thank Major C.P. Dockery of the 9t Air Force, Shaw Air ForceBase, South Carolina and Col. D. McCarcer, the ft1aval Training EquipmentCenter's Air Force Liaison officer for their v luable assistance in securingthe pilots who participated.
1/2
NAVTRAEQUIPCEN 78C-0060-9/AFHRLTR-82-3
TABLE OF CONTENTS
Section Page
INTRODUCTION \..\ 7
SUMMARY OF PRIMARY OBJECTIVES \ 11
II METHOD 1\3'
APPARATUS 13
VISUAL SYSTEM 13
FRESNEL LENS OPTICAL LANDING SYSTEM 18
SIMULATOR CONFIGURATION 18PERFORMANCE MEASUREMENT 18
FREEZE AND RESET FEATURES 20
TRAINING CONDITIONS 20DISPLACEMENT ERROR CRITERION 20
DISPLACEMENT AND DESCENT RATE CRITERION 21
PROCEDURE 21
PERFORMANCE MEASUREMENT AND DATA ANALYSIS 22
III RESULTS 23
TRAINING TRIALS 23
LEARNING EFFECTS 23
EXPERIMENTAL TREATMENT EFFECTS 23
PROBE TRIALS 23
LEARNING EFFECTS 23
EXPERIMENTAL TREATMENT EFFECTS 23
PROBE METHODOLOGY 25
QUESTIONNAIRE RESPONSES . . OOOOOOOOO 28ON THE GENERAL ROLE OF ERROR IN TRAINING 28ON THE INSTRUCTIONAL USE OF THE FREEZE FEATURE 28
IV DISCUSSION 29
RECOMMENDATIONS 31
REFERENCES 33
APPENDIX A RESULTS/DATA SUMMARY TABLES 35
APPENDIX B MEANS AND STANDARD DEVIATIONS OF INDIVIDUALITEMS-OF THE QUESTIONNAIRE 47
3
NAVTRAEQUIPCEN 78C-0060-9/AFHRLTR-82-3
LIST OF ILLUSTRATIONS
Figure
ComputerGenerated Image of the Day Carrier
Page
1
with FLOLS and Portion of Wake 16
2 ComputerGenerated Image of the Night Carrier with FLOLS . 17
3 Configuration of FLOLS Simulation ShowingDatum Bars and Meatball 19
4 Freezes Per Trial Averaged Across Freeze Conditionsand Across 4Trial Blocks of Training Trials 24
5 Means of RMS Glideslope Error 26
6 Means of Aileron Control Activity 27
4
NAVTRAEQUIPCEN 78C-0060-9/AFHRLTR-82-3
LIST OF TABLES
Table Page
1 Biographical Data on Pilot Subjects 14
Al Glideslope RMS Error: Training Trials 35
,A2 Angle of Attack RMS Error: Traitng Trials 36
A3 Average Elevator Stick Activity: Training Trials 37
A4 Average Throttle Activity: Training Trials 38
A5 Average Aileron Stick Activity: Training Trials 39
A6 Average Rudder Pedal Activity: Training Trials 40
A7 Glideslope RMS Error: Probe Trials 41
A8 Angle of Attack RMS Error: Probe Trials 42
A9 Average Elevator Stick Activity: Probe Trials 43
A10 Average Throttle Activity: Probe Trials 44
All Average Aileron Stick Activity: Probe Trials 45
Al2 Average Rudder Pedal Activity: Probe Trials 46
5/6
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
SECTION I
INTRODUCTION
The history of simulation has closely followed the history ofaviation. Early trainers were crude. unsophisticated boxes designed to givepilots some semblance of the flying experience and to prepare them toperform procedures important in flight. With the passage of time, aircraftand flight simulators have undergone evolutionary changes. Modern flightsimulators are capable of reproducing many of the same conditions that areattainable in the aircraft (Caro, 1977). This refinement in simulatorcapability is an outgrowth of advances in technology and is due to anincreased desire on the part of the simulator design engineer to give thepilot an environment that "feels" as much as possible like the real .
aircraft. Pilot self-report and subjective assessment were the rulingcriteria that drove the development of flight simulators.
Thus the flight simulator has been regarded as an artificial airplanewherein flight tasks could be performed in a safer, cheaper environment.Economy and safety we'e the specific reasons to substitute flight simulatorsfor aircraft. Howeve recent advances in training technology have increasedawareness that the a craft may not be the best place to begin the processof learning to fly. Th simulator, in contrast to the airplane, has thepotential to be structured as a learning environment that can facilitate theacquisition of the perceptual and motor skills necessary for aircraftcontrol (Caro, 1976).
Reliance upon an "in-flight" model to dictate the limits_of design anduse of simulators for pilot training can unnecessarily restrict thepotential of simulation to facilitate skill acquisition. Recent researchhas begun to address the issue of how to enhance the training value ofsimulation by employing principles of learning and transfer rather thansheer physical fidelity as the guiding criteria (Hughes, 1979): The
employment of certain techniques that follow from these principles maysleadto simulator conditions that, although objectively unrealistic, can enhancethe efficiency of the simulator as a training device.
For example, Hughes, Hannan, and Jones (1979) have explored the use ofa record-playback feature for instruction of a complex flight task whileBailey, Hughes,,and Jones (1980) have tested the instructional value ofbackward chaining with a dive bombing task. Lintern (1980) and Hennessy,Lintern, and Collyer (1981) have examined learning with novel and alteredvisual displays.'
Simulators not only offer the promise of permitting learning to proceedmore rapidly, but also may permit a higher level of skill attainment. Withsome specific flight tasks, the benign and predictable simulationenvironment may permit pilots to quickly perceive critical relationships
NAVTRAEQUIPCEN 78-C -0060 -9/AFHRL -TR -82 -3
that can aid their performance in the aircraft but that would never clearlyemerge if they practiced only in the aircraft.
These studies represent initial attempts to explore the potential offlight simulators as training devices through the alteration of the displaypresentation or through the application of principles of learning astraining features. This point of view diverges from the artificial aircraftmodel and requires that relevant learning principles be addressed as thedriving force behind the conduct of training in flight simulators.
Of all the instructional options that have been, or might be, builtinto an aircraft simulator, "freeze" appears to have gained the widestacceptance. "Freeze," in the simulatar training context, refers to thesuspension of any part or all of the simulated task for instructionalpurposes. Two uses may be made of this "freeze" feature. First, "freeze"can be employed to suspend some aspect of the flight dynamics (e.g.,aircraft roll may be "frozen") so that the trainee may focus _his attentionupon some other, more critical aspect of the task, prior to attempting thewhole task. Another potential use of "freeze" that has high apparentvalidity is for the instructor to stop the action while explaining errorsand strategies to the student. Nevertheless, there must.be some concernthat the level of assistance provided by the use of "freeze," especially inthe latter case, could disrupt retention of the skill (c.f., Snow, 1980) orthat the interruption could disrupt its acquisition.
The present study is concerned with the use of the latter, total tasksuspension "freeze" technique, and how it relates to the specific role of'errors in skill acquisition; in particular, whether a "freeze" should beused as an opportunity to instruct the student in the cause and correctionof errors or whether it should be used to minimize them. While it isapparent that information about the direction and magnitude of errors canfacilitate skill learning (AdAms, 1981), there is same concern that ,.
frequently committed errors could become embedded in a student's responserepertoire (Holding, 1970). Thus approaches that enhance the student'sawareness of the nature of the error and those that minimize errorsrepresent extreme positions in the treatment of errors during learning.
The possibility that repetition will embed errors in a student'sbehavior appears possible in the case of the carrier landing task. Forexample, a marginal approach, while being severely criticized by the LandingSignal Officer (LSO), can result in a safe landing. Although Navy pilotsare consciously and explicitly concerned with technique in making carrierapproaches, successful completion of a dangerous task by any means is, 'in atechnical sense, reinforced. This positive reinforcement may partiallynegate the effects of negative reinforcement frowthe LSO and from thepilot's own cognitive judgments and could be one of the more potentinfluences on learning. The natural consequences of errors apOtar to have apowerful self-correcting influence in perceptual-motor learning, but thenatural consequences of errors in carrier approaches may not be sufficientlynegative with the required frequency to give full force to this effect.
8
1, U
NAVTRAEQUIPCEN 78C-0060-9/AFNETR-82-3
Given the expanded range of options available for simulator instructionversus inflight instruction, the questions arise as to whether errorsshould be permitted, how they should be treated if they are permitted, andhow an instructor might intervene to enhance the learning process. Thecarrier landing task was considered ideal for examining these issues.Control behavior is strictly constrained by the demands of the task, anderrors can be specified clearly. Navy pilots and instructors tend to agreethat error detection and correction are fundamental to safe and consistentcarrier landings.
The present study addressed the issue of how errors should be treatedby examining three techniques for teaching the carrier approach task.Specifically, the study addressed the use of the simulator's "freeze"feature to interrupt an otherwise continuous performance whenever an errorwas detected. One possible advantage of freezing the task in this manner isthat it would allow students to attend to instructional feedback without thesimultaneous need for them to perform the task. Effectiveness of the
-feedback might thus be enhanced and so lead to faster learning.Alternatively, interruption of the continuous task might be disruptive andthereby impede learning. The following instructional conditions werecreated to explore these issues.
"Freeze/Reset." Under the Freeze/Reset condition, the simulator was"frozen" whenever an error was detected. Feedback was given while the"freeze" was in effect. During the "freeze," inside and outsidecockpitreferences and cues were maintaineo as they were when the "freeze"occurred. Before continuing the task, however, the simulator was returnedto the appropriate position (vertically) on glideslope with appropriatelyconfigured angle of attack and airspeed. At the termination of the"freeze," the student continued the task from the "corrected" position.
"Freeze/Flyout." Under the Freeze/Flyout condition, the actions duringthe "freeze" were the sameNas those for the Freeze/Reset condition exceptthat at termination of the "freeze," the student cbntinued the task from theexact point at which it was "frozen." That is, no correction was made.
"Conventional." Under the Conventional condition, students learned thetask without use of the "freeze." Instructional feeftack was given at thecompletion of each approach.
In addition to examlning these three alternative instructionalconditions, the study aeiressed the manner in which errors were defined.One error criterion was based on displacement from the glideslope while theother was based on both displacement from glideslope and deviation from theoptimum rate of descent. The latter criterion is potentially moreinformative in that a descentrate error car give advance warning of a
displacement error so that earlier corrective action can be taken. Theaddition of rate information to displacement information has been shown tominimize glideslope tracking errors of experienced Navy pilots (Kaul,Collyer, and Lintern, 1980).
9
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
One important methodological issue in transfer-of-training (TOT)research is related to selection of an appropriate training period.'Training to a proficiency criterion has often been used, but in a study ofdifferential transfer, the average time for various groups to attainproficiency almost alwos differs. Thus, training time tends to beconfounded with the exNrimental effects of interest. Fixed training times
resolve that problem but selection of an appropriate period can be critical,and necessarily relies heavily on the judgment and experience of theexperimenter. Training times could be too short to allow differences toemerge. Alternaively, they could be so long that worthwhile trainingdifferences are washed out by subjects attaining a high level of proficiencywith even the poorest training conditions. Thus, training times should beextended into, but not beyond, that period in training which showsworthwhile learning differences between instructional methods.
Pre-experimental work could ascertain the most appropriate trainingperiod, but it would require expenditure of a large portion of theexperimental resources to obtain a'reliable answer. Furthermore, in a studyof more than two training conditions, the selected training time may beappropriate for only some of the comparisons. A range of times could beused but would,reduce theipower of the experiment (i.e., its capability toreveal differences between conditions) to the extent that some of theselected training periods were inappropriate. Training time is a specialissue in a study of novel training techniques, such as those considered herewhere an experimenter has limited experience and meager data to provideguidance.
A probe technique in which learning trials or the experimentalconditions are interspersed with test trials on the control condition couldavoid these problems. This technique, which appears to have been used onlyonce in applied transfer-of-training research (Smith, Pence, Queen andWulfetk, 1974) might effectively map the course of learning and thus allowan estimate of the optimum training period for each instructional method.Smith et al. (1974) used a single-trial probe strategy in which training andprobe trill's alternated. Their strategy was probably not optimum.
Presumably an experimental session should be weighted heavily with trainingversus probe trials to limit dilution of the training effects. Nevertheless
probes should be frequent enough to ensure that critical differences are notmissed, and sufficient data would be required at each probe to achieve
worthwhile stability. Probe methodology would seem to offer distinctadvantages for the initial investigation of a novel training method.
However, for evaluating savings in relation to'a standard instructionalparadigm, the traditional transfer-of-training paradigm would still bepreferred. In this experiment, the probe technique was employed in anattempt to solve some.of the dilemmas faced by transfer of trainingexperimenters. Here probe trials, consisting of performance of the transfertask, were interspersed throughout the training sessions in order to assessthe level of proficiency of the trainees as training progressed. Twocritical issues surface as a consequence of the use of a probe technique.First, are the probe performances sensitive to the learning that is taking
14:10
NAVTRAEQUIPCEN 78C-0060-9/AFHRLTR-82-3
place during the training trials? Secondly, are the probe trials.in someway disruptive of the training process? Both of these issues wereaddressed in the experiment.
SUMMARY OF PRIMARY OBJECTIVES
1. To assess the relative effectiveness of three instructionalmethods differing in the degree to which each alt5rs'theinstructional environment following an error.
2. To explore the total task suspension use of "freeze" forinstruction of a continuous tracking skill.
3. To examine the effects on leaening of two error criteria:displacement only versus Aisplacement and rate.
4. To assess.the extent to which the transfer task performancesampled periodicaTly in probe trials is sensitive to what islearned in training trials and what effect, if any, probe trialsmight have upon learning.
ll/12
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
SECTION II
METHOD
Five groups of five experienced Air Force pilots were taught carrierlandings in a flight simulator at the Naval Training Equipment Center undera control or one of four experimental training conditions.
Experienced Air Force pilots were sought to avoid the necessity ofteaching basic aircraft control skills. These pilots were, however,inexperienced with reference_to the specific carrier approach skills thatwere to be learned.in this experiment. Table 1 summarizes the flightexperience of the pilots.
APPARATUS
The Visual Technology Research Simulator (VTRS) consists of a fullyinstrumented T-2C navy jet trainer cockpit, a six-degree-of-freedomsynergistic motion platform, a 32-element G-seat, a wide-angle visual systemthat can project both computer-generated and model-board images, and anExperimenter/Operator Control Station (Collyer and Chambers, 1978). The.motion system, G-seat, and model board were not used in this experiment.
VISUAL SYSTEM. The background subtended 50° above to 300 below the pilot'seye level, and 800 to either side of.the cockpit. The aircraft carrierimage, which was a representation of the Forrestal (CVA 59) was generated bycomputer and projected onto the background through a 1025-line videosystem. A carrier wake and Fresnel Lens Optical Landing System (FLOLS) werealso generated by this method. Both daytime and nighttime carrier images
could be displayed (Figures 1 and 2).
Average delay between control inputs and generation of thecorresponding visual scene was approximately 116 msec (calculation of newaircraft coordinates required 50 msec while calculation of the visual scenecorresponding to the viewpoint from the.new aircraft coordinates requiredApproximately 50 msec and'generation of the new scene required 17 msec). An
updated aircraft position was computed every 33 msec, but the pictureposition was mpdated every 17 msec by extrapolating aircraft position inbetween each computed aircraft position.
The sky brightness for the day scene was 0.85 fL (foot-lambert) and theseascape brightness was 0.6fL. The brightest area of the day carrier was
4.0 fL. Except for the horizon, no features were represented in either the
sky oy sea. The night background luminance was 0,04 fL and the horizon andseascape were not visible. The night carrier appeared as lights of 0.8 fL ,1brightness outlining the landing deck and other features.
13
TABLE 1. BIOGRAPHICAL DATA ON PILOT SUBJECTS
Subject-
GroupAssignment
Age
Flight Hrs.
Aircraft
Simulator
Last 30 Days
Hrs. Aircraft
1
(3)
35
2030
10
15/
F-16
2
(3)
35
2363
470
20/
F-16
3
(1)
27
1233
550
22/
F-4D
4
(4)
32
2050
360
20/
F-16
5
(2)
33
1900
300
20/
F-4
6
(5)
32
1680
350
20/
F-4D
7
(4)
29
1250
300
17/
F-4D
8 9 10 11 12 13
(1) (5) (2) (3) (4) (3)
37 32 31 33 34 32
3040 1500 1825 1500 1956 2700
450 400 194 210 90 300
20/ 22/ 12/ 20/ 24/ 20/
F-4 F-4D F-16 F-4 F-16 F-16
continued
TABLE 1. BIOGRAPHICAL DATA ON PILOT SUBJECTS (Cont'd)
Subject
GroupAssignment
Age
Flight Hrs.Aircraft
Simulator
Cast 30 DaysHrs. Aircraft
14
(5)
37
2500
215
20/F-4
15
(2)
,38
3040
170
25/
F-4
16
(1)
45
3100
150
0
17
(1)
47
5500
500
15/
F-4
18
(4)
43
3500
295
5/
F-4
19
(2)
31
2175
200
20/
F-4
20
(5)
32
2080
210
18/
F-16
21
(4)
36
2035
300
0
22
(2)
42
5300
350
15/
F-4D
23
(5)
35
1600
390
25/
F-4D
24
(1)
39
2300
260
25/F-4
25
(3)
30
1480
250
14/F-4D
Key to Group Assignment
1) Freeze/Flyout Display2) Freeze Reset/Displacement3) Conventional Display4) Freeze/Flyout Displacement and Rate5) Freeze Reset/Displacement and Rate
Figure 1, Computer-Generated 1mage,of the Day Carrier, with FLOLS and Portion of Wake.
J2/I
20
a ,...;
a
/S.
a .
:
I
0 0
Figure 2. Computer-Generated Image af the Night Carrier, with FLOLS.
2.i.
NAVTRAEQURCEN 78-C-0060-9/AFHRL-TR-82-3
FRESNEL LENS OPTICAL LANDING SYSTEM. The configuration of the FLOLS isshown in Figure 3. In contrast to a carrier,FLOLS, which is generated byincandescent lights and can derefore be much brighter than other parts ofthe carrier, the simulated FLOLS was generated by the same sYstem as thecarrier image. It was therefore only as bright as the brightest areas ofthe ship (e.g., the white lines on the landing deck). To compensate for itslower relative brightness, the FLOLS was enlarged by a factor of 4.5 whenthe distance behind the ramp was greater than 2250 ft. From 2250 ft., thesize of the FLOLS was linearly reduced until it a.ttained 1.5X its normalsize at 750 ft. It remained that size throughout the remainder of theapproach. The FLOLS was centered 414 ft. down the landing deck and 61 ft.to the left of the centerline. It was set at a nominal 3.5° glideslope andwith a lateral viewing wedge of 52°.
SIMULATOR CONFIGURATION. The simulator was initialized with the aircraft at9000 ft. from the ramp, on the glideslope and centerline, and in theapproach attitude and configuration (hook and wheels down, speed brake out,15 units angle-of-attack (A0A), and power at 83%). The T-2C is normallylanded with full flaps, but flaps were at half extension for this experimentto more closely simulate approach speeds of typical fleet aircraft. Fuel
was set at 1320 lbs.to give 10,000 lbs.gross weight. A landing trial wasflown from the initial condition to wire arrestment or, in the case of abolter, to 1000 ft. past the carrier.
The carrier was pt on a heading of 360° at 5 knots. Environmentalw.nd was set at 349.5 with a velocity of 20.1 knots. This combination ofcarrier speed and environmental wind produced a relative component of 25knots down the landing deck.
Turbulence was used to increase the difficulty. The turbulence modelbuffeted the simulator computed aircraft model with a random forcingfunction.
PERFORMANCE MEASUREMENT. VTRS has the capacity to assess performance at30 Hz for three classes of measure. First, aircraft position can bedetermined throughout the approach. Secondly, aircraft control surfaceactivity can be measured throughout the approach for elevator, aileron andrudder displacement. Finally, aircraft control position data can also becollected in order to determine throttle, control stick and rudder pedalactivity throughout the approach.
For each of these categories of measure, specific measures such as RootMean Square (RMS) error, absolute error, and percent time within apredefined tolerance band may be collected throughout the course of anapproach. In addition, these measuees can also be collected for anypredefined segment (or segments) of the approach (e.g., 1500 ft. totouchdown).
18
'
,- DATUM BAR/
-
r---11 1
I -II I
MEATBALL(SHOWINGRANGE OFMOVEMENT) I I
iL.---J1-------1
I
I
I I
\
DATUM BAR
... l' 'I*'
1 -I8.5' 4.125' 2.75 4.125'
,
8.5'
Figure 3. Configuration of FLOLS Simulation, Showing Datum Bars and Meatball.(Dimensions Shown are in Ft.).
2 4
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
Aside from continuous measures of glideslope performance, up to twoseparate snapshot measures may be collected at any point on the approach.This snapshot can contain any of the classes of measures deemed important.
[lie capacity also exists for collecting touchdown performance data fromwhich a measure of success of the landing may be calculated.
FREEZE AND RESET FEATURES. The simulator has the capacity to suspend theongoing simulation either manually, by the use of the "freeze" buttonlocated on the console, or the "freeze" feature may be instigatedautomatically based upon a specified error being committed by the trainee.Regardless of how the "freeze" is initiated, thiree actions are possiblefollowing the suspension of the simulation. First, the simulator may be ,
xeset to a corrected initialization point and the trainee can begin to flyfrom that point. Secondly, the current simulator state may be restored andthe trainee can continue the task from that point. Finally, the simulatormay,be set to some other, entirely different, initial condition and thetrainee can be required to perform some other task. The first two of thesepost "freeze" actions were employed here as experimental trainingconditions. These training conditions and the criteria for defining an
error are described below.
TRAINING CONDITIONS. Two experimental training procedures were used in theexperiment. For both procedures the simulator was frozen during theapproach if the pilot's vertical deviations from the glideslope exceededspecific criteria. Under one procedure, known as Freeze/Reset, when thesimulator was frozen, the pilots were advised on how they had incurred theirvertical error and were then reset to the glideslope with the simulator inits optimum approach attitude. Longitudinal distance from the carrier andlateral distance from the extended centerline of the landing deck were notchanged. Pilots continued their approach from the reset position. Underthe other procedure known as Freeze/Flyout, pilots were advised on how thvhad incurred their vertical error and how to correct it once they werereleased. They then continued their approach from the position and attitudein which the simulator had been frozen.
Two experimental training conditions were derived for each Freezeprocedure by applying two different criteria.
DISPLACEMENT ERROR CRITERION. This criterion "froze" the system if
Gi > Gc (1)
mhere ei = angular displacement of the aircraft from the3.5° glideslope,
ec . 0.5625 -r(0.3125 x 10-4), 0 < r < 6000and
r . range in feet from the carrier ramp.
This algorithm linearly increased the criterion in meatball units from 1.0at 6000 feet from the ramp to 1.5 at the ramp. "Freezes" did not occurbeyond 6000 feet from the carrier or past the ramp.
20
NAVTRAEQUIPCEN 78C-0060-9/AFHRLTR-82-3
DISPLACEMENT AND DESCENT RATE ERROR CRITERION. The second error criterionwould result in a "freeze" if vertical deviation from the 3.5° glideslope,descent rate error, or some combination of the two was excessive. "Freezes"would occur:
if Mi > oc (2)
. 1for M = oi + 0.5625 oc ,
. angular rate of displacement in degrees/secondfrom the glideslope,
oc . 0.405 (0.49 x 10-4) (r + rK),
and rK . 524 feet, the distince from the carrier ramp tothe FLOLS origin.
This algorithm established a criterion that was a'weighted sum of thepreviously described displacement criterion and a descent rate error limitthat decreased linearly from 600 fpm at 6000 feet from the ramp to 200 fpmat the ramp.
"Freezes" were not permitted within 10 seconds of restarting the.approach after a previous "freeze." In addition, a "freeze" would not occurif, at the end of this 10second period, the subject was outside of theperformance criterion but was decreasing the error.
In the fifth training condition, designated the Conventional, thesimulator was not "frozen" during the approach but the subjects were givenfeedback as to their error (equivalent to that given the Freeze/Flywt.group) at the end of each trial.
PROCEDURE
Two subjects arrived at the simulation facility each day, Mondaythrough Thursday, during the experiment. They viewed a vi6e9 tape oncarrier landings which described the FLOLS and carrier landings. They werethen given detailed instructionstly *a Navy LSO on carrier landingtechniques. This instructional period lasted approximately 45 minutes.When convenient,.subjects were given this preliminary instruction in pairs,but the remaining experimental work was undertaken with only one subject inattendance 'except that subjects were occasionally permitted to monitor theperfonPance of others from outside the simulator if they had entirelycompleted their experimental work. Subjects were assigned to trainingconditions as they arrived at the simulator facility in accordance with apredetermined sequence that ensured the number of subjects having beentrained with each condipon remained approximately equal throughout theexperiment.
21
<r
NAVTRAEQUIPCEN 78C-0060-9/AFHRLTR-82-3
After preliminary instruction, subjects were familiarized with the
cantrols of the simulator. They were then given a brief flight of
approximately two minutes before they commenced their carrier landing
training. The training sequence consisted of 24 approaches to the day
carrier,on the afternoon of their first day at the simulator facility, and
24 approaches to the night carrier on the morning of the second day. The
two 24trial blocks were diVided into sixtrial subblocks, the first four
trials of weth were flown under the appropriate training condition. The
last two trials of each subblock were used as probe trials to assess theprogress of learning, and were flown under the control condition. The LSO
gave no instructions during or following probe trials. Subjects were given
a ten minute rest after the twelfth trial of each 24trial block.
PERFORMANCE MEASUREMENT AND DATA ANALYSIS
Position, attitude, and state of the simulated aircraft and elevator,throttle, aileron and rudder control positions were sampled at 30 Hz. The
simulator position, attitude, and state variables were used to deriveperformance measures for glideslope and AOA tracking. The control position
variables were used to derive control activity measures: The continuous
measures were derived for nonaverlapping 1500foot segments of the
approach.
Repeated measures analyses of variance were applied to root mean square(RMS) glideslope and RMS AOA errors and to flight control activity
measures. In addition, multiple discriminant analyses were applied to the
probe data to find linear combinations of scores that maximallydiscriminated the instructional treatments.
2, 2 2
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
SECTION III
RESULTS
TRAINING TRIALS
LEARNING EFFECTS. Several learning trends are apparent and show up asreductions in RMS error for glideslope (See Appendix A, Table Al) and AOA(Table A2), and as reductions in activity of elevator (Table A3) andailerons (Table A5). There was comparable reduction in throttle activity(Table A4), while average y'uddee pedal activity (Table A6) showed noreliable effects. Figure 4 shows a consistent trend towards reduction innumber of "freezes" (dith a reversal in transitioning from Day to Nightapproaches) in the experimental conditions. This also indicates a steadyimprovement in glideslope control throughout the experiment.
EXPERIMENTAL TREATMENT EFFECTS. Only RMS glideslope error scores showed anystatistically reliable effect of experimental training condition's (TableAl). Glideslope errors of the Freeze/Reset groups were lower than for othergroups, .an unsurprising result in view of the fact that errors wereconstrained and frequently zeroed by the experimental manipulation. Thissuggests.that "freezes'" occurred frequently enough to reduce error andthereby validates the experimental manipulation to some extent.
PROBE TRrALS
LEARNING EFFECTS. While error scores and control activity tended todecrease throughout probe trials, these trends were generally reliable onlyfor the final,segment (last.1500 feet) of the Day approaches (session 1)with measures of RMS glideslope error (Table A7), RMS AOA error (Table A8),and elevator and aileron control activity (Tables A9 and All). In contrastto other measures, the reductions in aileron control activity werestatistically reliable over all four 1500-foot approach segments for bothDay,and Night conditions. RMS glideslope error and aileron control activityshowed a sharp increment in the transition from Day to Night approaches(which coincided with the break between sessions I and 2).
EXPERIMENTAL TREATMENT EFFECTS. 4.1nivariate analyses showed no reliableinstructional treatment effects between probe-trial measures of RMSglideslope error, RMS AOA error and control activity. Tables A7 to Al2summarize the tests of statistical reliability for these measures. Thusneither instructional technique nor the method of determining error appearedto have any clear effect on learning.
In a further ,effort to seek a relationship between instructionaltechniques and learning, stepwise discriminant analyses were applied tothe*six measures of Tables A7 to Al2. The analyses were restricted to thedata of the final approach segment (1500 to 0 feet) because they showed themost consistent learning trends. Data from the first pair of probe trials
23
1.40
1.35
1.30
1.25
1.20
1.15
.,
1.10
1 05
1.00
.95
.90
.85
.80
SESSION 2STARTS HERE
(1-4) (7-10) (13-i 6) (19-22) (25-28) (31-34) (37-40) (43-46)
TRIALSSESSION 1 (DAY) SESSION 2 (NIGHT)
Figure--4. -Freezes per Trial--Averaged--AcrossF-reeze Conditrions
and Across 4-Trial Blocks of Training Trials.
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
were not included because they did not appear to follow the pattern of laterprobe trials, and because it was considered that differences due to learningmight not have become established at that point. Pairs of probe trials wereused to define dependent measures so that the numbers_of dependent measuresentered into the discriminant analyses were 18 (three probe pairs by sixmeasures) for the day analysis and 24 (four probe pairs by six measures) forthe night analysis. Day and night trials were analyzed separately toimprove the ratio of subjects to dependent measures. Nevertheless, theanalyses did not conform to the normal constraints (the size of the smallestgroup should exceed the number of variables, the total number of.subjectsshould be two to three times the number of dependent measures (Tatsuoka,1970)) and should be considered as purely exploratory.
The cumulative proportion of total dispersion accounted for by thediscriminant function of the Day data was 86% and the average correctclassification of subjects into groups (using the jackknifed procedure) was36% (expected chance value of 20%). The overall approximate F ratio wasstatistically reliable at p<.01. Aileron control activity for the third andfourth probe trials contributed most to the separation of the groups on thefirst canonical variable. The trend was towards smoother aileron controlinputs for pilots in the Conventional (no-"freeze") condition.
The cumulative proportion of the total dispersion accounted for by thediscriminant function of the Night data was 94% and the average correct.classification of subjects into groups was 68%. The overall approximate Fratio was statistically reliable at p<.01. Pedal control activity at thesixth and seventh probe pairs, throttle control activity at the fifth probepair and aileron,control activity at the sixth probe pair contributed mostto separation of\the groups on the first canonical variable. The trend wastowards smoother throttle and aileron control inputs for pilots in theCOnventional group, and for higher thnottle activity for both Freeze.groupsthat used the Displa ement-only criterion.
PROBE METHODOLOGY
Means of RMS glideslope errors and of aileron control activity in thefinal 1500 feet of the approach were examined for disruptive effects oftransitions between training and probe trials (Figures 5 and 6). Thissegment was chosen for analysis because it showed the strongest and mostconsistent learning trends in both training and probe trials. Theglideslope measure was examined because the experimental manipulations wereintended to:impact glideslope tracking. Aileron control activity was alsoexamined because of the strong and consistent learning effects that weredemonstrated with this measure.
There were no consistent differences between probe trials for theexperimental groups and the trials for the Conventional groups that were inthe probe locations. Nor were there consistent differences between firstand second probe trials, between first and second training trials followinga probe, or between training trials prior to and subsequent to a probe, thatwould suggest any disruption resulting from transitions between training andprobe trials.
2S
co PO cs)
Co
1 kw.
L?.
CO
NT
RO
L A
CT
IVIT
Y (
UN
ITS
/SE
C A
SD
EF
INE
D F
OR
TA
BLE
6)
e, ...C...
....
.....
/0"'
4,
......
......
.....
'NO...
.
......
...
.....
.....
.... .
......
KW
,
ar.
... S
.
.....
......
. ....
.....
......
.
0C
P-4
CO
CO
0
<'m
.1m
.M...
......
......
....
. ...
-4
C-Z8-211-121HdV/6-0900-0-8L N30dInInnlAVN
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
QUESTIONNAIRE RESPONSES
Descriptive data based on pilot responses to the paper-and-pencilquestionnaire are given by individual item in Appendix B. The results aresummarized below.
ON THE GENERAL ROLE.OF ERRORS IN TRAINING. Pilots generally.disagreed that"errors served little purpose" and also disagreed with the notion that"students may actually learn the errors they commit" (Item 12). Pilots alsodisagreed with the contention that "instructional methods that allow errorsto occurare inefficient". (Item 14). Instead, pilots in-the study pointedto error recognition as a basis for the development of correct performance(Item 18). Errors were seen as helping the student to focus on the criticalelements of task performance (Item 13), as well as exposing the student toout-of-tolerance situations which may, under later conditions, result fromfactors such as adverse weather, visibility/ceiling limitations, etc. (Item15). On the issue of whether correct performance is best thought of asresulting from a process of eliminating errors or from a process of shapingdesired resp,onses, pilots were undecided (Item 17).
ON THE'INSTRUCTIONAL USE OF THE FREEZE FEATURE. Pilots agreed that it waseasier to attend to the LSO's feedback while the simulator was "frozen" thanwhile trying to listen and fly the aircraft at the same time (Item 3).Pilots also agreed that use of the "freeze" aided development of errorrecognition (Item 5) but were undecided as to whether it might significantlydecrease the overall time required to learn the landing task. On thenegative side, pilots indtcated that the occurrence of the "freeze" early intraining was,"frustratinr (Item 8). In fact, pilots in the Freeze/Resetcondition indicated that they were more motivated by "trying to avoid the"freeze" than by trying to fly the task correctly" (Item 7). Pilots in theFreeze/Flyout condition tended to respond in somewhat the opposite manner.Regardless of the "freeie" condition to which subjects were assigned, allindicated that regaining control of the simulator following a "freeze"significantly added to the difficulty of the flying task (Item 2) and thatthe difficulty increased the closer the occurrence of the "freeze" to theterminal portion of the task (Item 2).
In general, the questionnaire data indicated that pilots perceivederrors as contributing positively to training, that the present use of the"freeze" feature was in some instances aversive, and that it served to addto the difficulty of learning the task in the simulator despite the factthat the "freeze" made it easier to attend to feedback from the LSO. So faras being -able to potentially reduce the time needed to learn the task,pilots perceived the present application of the "freeze" to have littlevalue.
28
NAVTRAEQUIPCEN 78C-0060-9/AFHRLTR-82-3
SECTION IV
DISCUSSION
Learning appeared to be limited to the final 1500 feet of the approach;an observation that suggests that the experienced Air Forde pilots used inthe experiment were able to perform the early part of the task withstrategies similar to those used in normal landings. Only in the last partof the approach, where the altitude tolerances became very small-, did thepilots appear to develop new techniques. This probably reduced theeffectiveness of the expermental manipulation since it appears to haveoffered limited opportunity to "freeze," and to process errors subsequentlyduring the "freeze," in that part of the approach that required newlearning. Thus, the results did not clearly bear on the hypotheses relatingto the treatment of errors when they occur and the procedure for definingthe error criterion.
Nevertheless, the Freeze/Reset condition did result in smaller errorsin the final approach so that a modified hypothesis, that the magnitude of.errors permitted in training will affect learning, can be examined bycontrasting the probe performances of the Freeze/Reset groups with the probeperformances of the Conventional and Freeze/Flyout groups. This hypothesisvaries sli-ghtly from the original'one that postulated effects ofdifferential treatment of errors when they occurred.
The notion thatlearning will proCeed more quickly if students canpractice with fewer or smaller errors hasintuitive appeal. As notedpreviously'in this report, Holding (1970) has suggested that frequentlycommitted errors could become embedded in a student's response repertoireand thereby impede progression to asymptotic behavior. Lintern and Roscoe(1980) have suggested that training might benefit from manipulations thatconverge quickly on desirable control behavior. In addition, a criterionsetting process might be postulated where students who become used toperforming with few errors endeavor to acheive similar standards even undermore difficult conditions. Given the popularity of these notions, it seemsnoteworthy that there is do suggestion in the data that the reduced errorsduring Freeze/Reset training resulted in any differential rate of learningor terminal level of probe performance. While some instructional strategiesthat promote more accurate learning behavior do assist skill acquisition(e.g., Lintern, 1980)* it would appear that they do so by means other thantheir errorlimiting function.
There is some evidence, albeit insubstantial, that "freeze" had someadverse effects on performance during probe trials; in particular, thatsubjects training under the "freeze" conditions exhibited higher controlactivity. Pilot comments support the vlew that the "freeze" may be aversiveand that some applications of the "freeze" (e.g., the Freeze/Reset) maychange the nature of students' motivation in performing the task. That one
29
NAVTRAEQUIPCEN 78C-0060-9/AFHRLTR-82-3
or more of these processes can affect perceptualmotor acquisition isindicated by the work of Payne and his associates (Payne, 1970; Payne andArtley, 1972; Payne and Dunman, 1974; Payne and Richardson, 1972). Thus
while there is no evidence that the use of "freeze" can facilitate learning,there is some suggestion that it can disrupt the integrity of the task andthereby impede learning on some aspects of the skill. In surinary, the
implications of the data for instructional procedures are that attempts toenhance performance during learning will not necessarily facilitate skillacquisition, while injudicious use of the "freeze" function may disrupt it.
From a methodological standpoint, the study is significant in that itsupports the use of the probe technique as an alternative to the moretraditional transferoftraining methodology in the preliminaryinvestigation of instructional treatment effects. In the present case, theprobe technique proved to be sensitive to learning effects as well as tosubtle performance differences which transferred from training trials to thesubsequent probe (criterion) trials.
The main methodological issues explored by examining the probe datawere in relation to possible disruptive effects from frequent transferbetween training and control conditions. There was no evidence of anydisruptive effects, although this conclusion must be tempered with theobservation that learning effects were minimal. The probe methodology couldbe valuable in a learning experiment, and its further examination with.datathat show a strong learning trend would be useful. Stability of the probe
trials was similar to stability of similarly located control trials. The
question of how many probe trials are necessary might best be answered by apower analysis (Hays, 1963) and would require data that show some worthwhiledifferences between groups. Nevertheless this preliminary analysis of theprobe technique suggests that it could be useful in future studies of skillacquisition.
30
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
SECTION V
RECOMMENDATIONS
Caution should be exercised when using the flight simulator's "freeze"feature d4ring the performance of a continuous control task such as thatinvolved(ip the approach to landing task. Other tasks to which this advicemight alsó apply are aerial refueling training and weapons delivery training.
The/fact that an instructional technique may result in better traineeperformahce during training should not be the sole criterion for itsimplementation. While an effective instructional technique can be expected\to aid and improve learning performances,,better performance on aninstructional task does not necessarily lead to improved performance on thecriterion task.
The probe methodology is recommehded as an alternative to thetraditional transfer-of-Veining paradigm, especially for exploratorystudies where training effectiveness may vary not only as a function ofinstructional approach but also as a function.of amount of training.
31/32
NAVTRAEQUIPCEN 78-CJ0060-9/AFHRL-TR-82-3
REFERENCES
Adams, J.A. A closed-loop theory of motor behavior. Journal of MotorBehavior, 1971, 3, 111-149.
Bailey, J.S., Hughes, R.G., and Jones, W.E. Application of backward chaininto air-to-surface weapons delivery training. AFHRL-TR-7963, WilliamsAF , AZ: Operations Training Division, Air Force Human ResourcesLaboratory, April 1980. AD-A085610.
Caro, P.W. Some factors influencing transfer of simulator training. Profes-sional Paper HUMRRO4P-1-76. Alexandria, VA: Human Resources ResearchOrganization, August 1976.
Caro, P.W. Some current problems in simulator design testing and use. Pro-fessional Paper HUMRRO-PP-2-77. Alexandria, VA: Human ResourcesResearch Organization, March 1977.
Collyer, S.C. and Chambers, W.S. AWAVS, a research facility for definingflight trainer visual requirements. Proceedings of the Human FactorsSociety, 22 Annual Meeting, Detroit, 1978.
Hays, W.L. Statistics. 'New York: flolt, Rinehart and Winston, 1963.
Hennessy, R.T., Lintern, G., and Collyer, S.C. Unconventional visualdisplays for fli9ht trainina. NAVTRAEQUIPCEN 78-C-0060-5, Orlando,FL: Naval Training Equipment Center, November 1981.
Holding, D.H. Learning without errors. In L. Smith (Ed.), Psychology of' Motor Learning.. Chicago, IL: The Athletic Institute, 1970.
Hughes, R.G. Enabling features versus instructional features in flyingtraining simulation. Proceedings of the 1st Inter-service/IndustryTraining Equipment Conference. Orlando, FL: NAVTRAEQUIPCEN IH-316,
Hughes, R.G., Hannan, S., and Jones, W. Application of flight simulatorrecord/playback feature. AFHRL-TR-79-52, Williams AFB, AZ: OperationsTraining division, Air Force Human Resources Laboratory, AD-A081 752,December 1979.
Kaul, C.E., Collyer, S.C. and Lintern, G. Glideslope descent-rate cuingto aid carrier landings. NAVTRAEQUIPCEN IH-322. Orlando, FL: NavalTraining Equipment Center, October 1980.
Linterd, G. Transfer of landing skill after training with supplementaryvisual cues. 'Human Factors, 1980, 22, 81-88.
33
NAVTRAEQUIPCEN 78C-0060-9/AFHRLTR-82-3
Linterri",-,G., and Roscoe, S.N. Visual cue augmentation in contact flightsimulation. In Roscoe, S.N. Aviation Psychology, Ames, Iowa: Iowa
State University'Press, 1980.
Payne, R.B. Functional properties of supplementary feedback stimuli.Journal o, Motor Behavior, 1970, 2, 37-43.
Payne, and Artley, C.W. Facilitation of psychoffiotor learning byclassically differentiated feedback,cges. Journal of Motor Behavior,
1972, 4, 47-55.- 1
Payne, R.B. and Dunman, L.S. Effects of classical differentiation on thefunctional properties of supplementary feedback cues. Journal of Motor
Behavior, 1974, 6, 47-52. /
Payne, R.B. and Richardson, E.T. Effects of classically differentiatedsupplementary feedback cues on tracking skill Journal of Motor
Behavior, 1972, 4, 257-261.
Smith; R.L., Pence, G:G., Queen, J.E., and Wulfeck, J.W. Effect of apredictor instrument on learning to land a simulated jet trainer.Inglewood, CA: ,OunTap and Associates, Inc., 1974.
Snow, R.E. Aptitudes and instructional methods: Research bn individual
differences in learning--related processes. Stanford, CA: Stanford
University School of Education, 1980.
Tatsuoka, M.M. Dis&iminant analysis: The study of 9roup differences.Champaign, IL: Institute for Personality and Ability Testing, 1970.
7 4
34
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL=TR
APPENDIX A.
RESULTS/DATA SUMMARY T 8LES
TABLE Al. GLIDESLOPE RMS ERROR (IN FEET):MEANS, STATISTICAL RELIABILITIES (*:p<.05, **:p<.01),
AND VALUES OF ETA SQUARED (n2): TRAINING TRIALS
S.
Distance From6000 - 4500 4500 - 3000 3000 - 1500 1500 - 0the Ramp (Ft)
Means Day Nighi ';'Day Night Day Night Day Night
CONDITIONS
Freeze/Flyout 15.3 14.2 13.2 13.5 11.2 10.7 8%9 9.3Di§placement
Freeze/FlyoutDisp & Rate
13.0 13.5 14.4 11.5 10.2 9.1 8.1 8.0
Conventional 16.4 14.7 14.2 15:4 10.5 12.9 7.1 11.9
Freeze/Reset.Displacement 13.1 '9.7 11.0 11.3 8.6 7.2 4.9 4.9
Freeze/ResetDisp & Rate 8.6 9.1 7.7 8.8 6.7 6.6 4.1 4'2
TRAINING TRIALS
1-4 14.5 14.3 14,5 14.1 12.4 11.4 8.5 11.0
7-10 14.1 12.1 10.8 10.8 8.4 8.8 6.3 7.1
13-16 13.2 11.4 13.3 11.8 9.2 8.4 5.9 6.3
19-22 12.1, 11.1 10.6 11.5/1 8.4 8.9 6.4 6.5
Reliabilitiesand n2 p n
2
CONDITION
Day
Night
TRAINING
Day
Night
CT
Day
Night
* * .078
.053 .077 * *
**
Wel=
ANIP NO
n2
n2
n2
.079
.023
.032
Nadia
**
.055
.055
.023
we we
**
.118
.030
35 ,
,4
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
TABLE A2. ANGLE OF ATTACK RMS ERROR (IN AOA UNITS):MEANS, STATISTICAL RELIABILITIES (*:p<.05, **:p<.01),
AND VALUES OF ETA SQUARED (n2): TRAINING TRIALS
Distance Fromthe Ramp (Ft)
6000
Means Day
CONDITIONS
Freeze/Flyout.576
Displacement
Freeze/FlyoUt.415
Disp.& Rate
Conventfonal .370
Freeze/ResetDisplacement .597
Freeze/Reset.508
Disp & Rate
TRAINING TRIALS
1-4 . :562
7-10. .491
13-16 .463
19-22 .436
Reliabilitiesand n2 p
CONDITION
Day ....
Night......
TRAINING
Day , **
Night **
CT
Day....Night ....
- 4500 4500 - 3000 3000 - 1500 1500 - 0
Night' Day Night Day Night Day Night
:425 :778 .580 133 .573 1.067 .845
.358 .653 .661 .590 .579 .795 .706
.355 .482 %437 .423 .407 .695 .701
.420 .697 .617 :661 .511 .736. .675
374/ .587 .494 .572 .493 .673 .554
.476 .727 .634 .671 .591 .905 .829
'389 .623 .565 .570 .561 .780 ..693
.363 .583 .520 .574 .467 .727 .646
.623 .473 .560 .427 ' .792 .622
n2 n2
n2 n2
...... .....
...... ...... ......
\1022 * .023 * .019 .053 .021
.033 ** .035 .** .040 ** .035
...... ..... .....
** .o7p ** 1,
.076 ......
36
NAVtRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
TABLE A3. AVERAGE ELEVATOR STICK ACTIVITY (IN UNITS/SECWHERE,RANGE OF CONTROL DISPLACEMENT IS FROM -1 TO +1 UNITS):
MEANS, STATISTICAL RELIABILITIES (*:p<.05, **:15<.01),AND VALUES OF ETA SQUARED (n2): TRAINING TRIALS
Distance Fromthe Ramp (Ft)
6000 - 4500 4500 - 3000 3000 - 1500 1500 - 0
Means Day Night Day Night Day Night Day Night
CONDITIONS
Freeze/Flyout.108 .068 .110 .070 .128 .088 .202 .132Displacement
Freeze/Flyout.097 .085 .098 .085 .107 .090 .153. .110
Disp & Rate
Conventional .071 .086 .067 .083 ,075 .089 .127 .141
Freeze/Reset.154 .063 .142 .063, .210 .070 .299 .089
Displacement
Freeze/Reset.129 .087 .139 .083 .184 .089, . .324 .104
Disp & Rate
TRAINING TRIALS
.079 .134 .078 .169 .090 . .253 .127
.078 .106 .079 .135 .087 .220 .120
.077 .098 .076 .123 .085 .191 413
.079 .096 .076 .117 .083 .188 .106
1-4 .143
7-10 .103
13-16, .100
19-22 .091
Reliabilitiesand n2 p n
2n2
n2 n2
CONDITION
Day
Night
TRAINING
Day
Night
CT
Day-
Night
06 .101.
.101.
.044 .029 * * .023 * * .015
37
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
TABLE A4: AVERAGE THROTTLE ACTIVITY (IN UNIfS/SECWHERE RANGE OF THROTTLE DISPLACEMENT IS FROM 0 TO +1 UNITS)
MEANS, STATISTICAL RELIABILITIES (*:p<.05, **:p<.01),AND VALUES OF ETA SQUARED (n2): TRAINING TRIALS
Distance Fromthe Ramp .(Ft)
6000 - 4500 4500 - 3000 3000 - 1500 1500 - 0
Means pay Night Day Night: Day Night Day Night
CONDITIONS
'.081
.083
.081
'.080
.0B1
.081
.081
.080
.082
.041
.009
.008
.024
.012
.012
.015
.026
.623
.010
.011
.008
.009
.010
.069
.010
.069
.010
.043
.011
.010
.022
:015
.015
.017
.026
.024
.013
.013
.009
.010
.011
.011
.012
:010
.012
.063
.024
.019
.033
.023
, .031
.030
.0.35
.036
.028
.025
.Q22
.025
.016
.022
.024
.022
.023
Freeze/FlyoutDisplacement .112
Freeze/Flyout.081
Disp & Rate
Conventional .081
Freeze/ResetDisplacement
.095
Freeze/Reset.086
Disp &-Rete
TRAINING TRIALS
1-4 .087
7-10 .087
13-16 \.096
. 19-22 '.094
Reliabilitiesand n2 p n
2n
2n2
n2
CONDITION
Day
Night
TRAINING
Day
Night
CT
Day
Ntght .071
_
_
38
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
TABLE A5. AVERAGE.AILERON STICK ACTIVITY (IN UNITS/SECWHERE RANGE OF CONTROL DISPLACEMENT IS FROM -1 TO +1 UNITS):
MEANS, STATISTICAL RELIABILITIES (*p<A5, **:p<.01),AND VALUES OF ETA SQUARED (n2): TRAINING TRIALS
Distance Fromthe Ramp (Ft)
6000 - 4500 4500 - 3000 3000 - 1500 1500 - 0
Mans Diy Night Day Night Day. Night Day Night
CONDITIONS
Freeze/FlyoutDisplacement
Freeze/FlyoutDisp & Rate'
Conventional
Freeze/ResetDisplacement
Freeze/ResetDisp & Rate.
TRAINING TRIAL
.112
.107'
.071
.102
.099
.101
.094
.083
.079
.110
.104-
.098
.090
.090
.117
.113
.073
.106
.103
.111
.101
.100
.098
.107
.093
.090
.,082
.117
.108
.104
.092
.095
.129
.122
.079
.118
.118
.132
.110
.104
.105
..134
.099
.103
,102
.130
.128
.119
.107
:112
.168
.136
.100
.145
.147
:163
.133
.126
.132
.146
.123
.112
-.115
..130
.150
.127
.119
.112
1-4'
7-10
13-16
L9-22
.111
.099
.092
.090
Reliabilitiesand n2 p p n
2n2
n2
CONDITION
WO Oar
**
.....
....
.032
WM*
....
.053
*
.009
.050
NIO
**
__
.047 **
**
..,
/NO
.050
.044
Day
Night
TRAINING
Day
' Night
CT
Day
Night
39
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
TABLE A6. AVERAGE RUDDER PEDAL ACTIVITY (IN UNITS/SEC,WHERE RANGE OF-PEDAL DISPLACEMENT IS FROM -1 TO +1 UNITS):
MEANS, STATISTICAL RELIABILITIES (*:p<.05, **:p<.01),AND VALUES OF ETA SQUARED (n2)': TRAINING TRIALS
..4*Dlstance Fromthe Ramp (Ft)
6000 - 4500 4500 - 3000 3000 - 1500 1500 - 0
Means Day Night Day Night Day Night Day Night
CONDITIONS
Freeze/FlyoutDisplacement
Freeze/FlyoutDisp & Rate
Conventional
Freeze/Reset.
Displacement
Freeze/ResetDisp & Rate
TRAINING TRIALS
.063
.059
.046
.060
.058
.055
.057
.060
.056
.049
.058
.062
..040
.061
.054
.056
.053-
.056
.058
.056
.041
.054
.056
.050
.052
.057
.052
.046
.057 .
.058
.034
.056
.049
.052
.050
.053
.061
.057
.041
.053
.057
.054
.053
.054
.053
.051
.060
.060
.036
.057
.052
.053
.054
.055
.074
.059
.045
.059
.060
.061
.059
.059
.058
.064
.061
.063
. .038
.056
.057
.059
.056
.059
1-4
7-10
13-16
19-22
Reliabilities
and n2 p n2
n2
n2
n2
CONDITION
Day
Night
TRAINING
Day
Night
CT
Day
Night
_
_
_
Th
WOE
MI6 410410
.
40
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
TABLE A7. GLIDESLOPE RMS ERROR (IN FEET):MEANS, STATISTICAL RELIABILITIES.(*:p<.05, **:p<.01),
AND VALUES OF ETA SQUARED Cn2): PROBE TRIALS
Distance Fromthe Ramp Ft)
6000 = 4500 4500 - 3000 3000 - 1500 1500 - 0Z
Means Day Night Day Night Day Night Day Night
CONDITIONS
Freeze/FlyoutDisplacement
Freeze/FlyoutDisp & Rate
Conventional
Freeze/ResetDisplacement
'Freeze/ResetDisp &.Rate
TRIALS
14.8
13.0
12.8
12.9
13.3
14.7
14.3
13.1
11.3 .
14.2
12.8
15.0
11.0
11.1
14.9
11.7
12.3
12.3
10.8
11.1
11.3
14.9
11.8
12.3
11.6
13.0
11.1
12.7
11.1
12.5
14.2
9.9
12.9
11.7
11.3
12.4
10.0
8.9
8.3
10.9
8.1
9.8
9.0
, 10.0
8.1
9.3
9.4
10.1
10.7
8.8
10.6
10.1
8.8
9.2
7.6
6.6
6.1
7.8
7.6
9.3
6.3
6.7
'6.2
7.4
9.7
8;4
8.1
7.9
*9.6
8.6
7.3
7.8
,BROBE
' 11-12
17-18
23-24
Reliabilitiesand n2 n
2n2
CONDITION
-_
'"1
WO. OP
.11.
* * .082
Day
Night
PROBE
Day
Night
CP
Day
Night
41
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
TABLE A8. ANGLE OF ATTACK RMS ERRORMEANS, STATISTICAL RELIABILITIES (*:p<.05, **:p<.01),
AND VALUES OF ETA SQUARED (i2): PROBE TRIALS
Distance Fromthe Ramp (Ft)
6000 - 4500 4500 - 3000 3000 - 1500 1500 - 0
Means Day Night Day Night Day Night' Day Night
CONDITIONS
Freeze/FlyoutDisplacement
Freeze/FlyoutDisp & Rate
Conventional'
Freeze/ResetDisplacement
Freeze/ResetOcisp & Rate
PROBE TRIALS
.59
.43
.35
.55
.53
.51
.43
.47
.38
.43
.31
.43
.42
.40
.42
.37.
.37
'.61
.66
.45
.77
.70
.67
.66
.62
.62
.59
.73 ,
.39
.77
.53
.65
.59.
.58
.58
.76
.59
.45
.61
.54
.67
.57
.56
.56
.59
.57
.39
.66
.50
.57
.58
.52
.49
1.07
.70
.60
.87
.90
1.14
.72
..80
.65
.71
.83
.61
.70
.55
.68
.74
.65
.65
5-6
11-12
17-18
23-24
Reliabilitiesand n2 p n
2na n
2n2
CONDITION
Day
Night
PROBE
Day
Night
CP
Day
Night WOO/
* * .073
42 40
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
TABLE A9. AVERAGE ELEVATOR STrCK ACTIVITY (IN UNITS/SECWHERERANGE OF CONTROL DISPLACEMENT IS FROM -1 TO +1 UNITS):
MEANS, STATISTICAL RELIABILITIES (*:p<.05, **:p<.01),AND VALUES OF ETA SQUARED (n2): PROBE TRIALS
Distance Fromthe Ramp (Ft)
6000 - 4500 4500 3000 3000 - 1500 1500 - 0
Means Day Night Day Night Day Night Day Night
CONDITIONS
.099
.101
.066
.105
.139
.117
.098
.101
.093
.071
.089
.073
.066
.089
.074
.083
.077
.077
.108
.104
.065
.136
:147
.143
.104
.105
.096
.070
.090
.074
.o66,
.087
.075
.082
.076
.076
.126
.11
.071
.179
..193
.164
.127
.129
.124
.082
.099
.081
..077
.097
.086
.094
.085
-.084
.195
;169
.132
.282
.348
.269
.225
.205
.2
.121
.147
.130
.115
.123
.127
.14
.116
.126
Freeze/FlyoutDisplacement
Freeze/FlyoutDisp & Rate
Conventional
Freeze/ResetDisplacement
Freeze/ResetDisp A Rate
PROBE TRIALS
5-6
11-12
17-18
23-24
Reliabilitiesand n2 p n
2P n
2p n
2P n
2
CONDITION
Day ....
Night
PROBE
Day .... -- -- * -.01
Night .... .... ....
CP
Day.... _-
Night ....
43
4,9
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
TABLE A10. AVERAGE THROTTLE ACTIVITY (IN UNITS/SECWHERE RANGE OF THROTTLE DISPLACEMENT IS FROM 0 TO +1 UNITS):
MEANS, STATISTICAL RELIABILITIES (*:p<.05, **:p(.01),AND VALUES -OF ETA SQUARED (n2): PROBE TRIALS
Distance From6000 - 4500 4506 - 3000 3000 - 1500 1500 - 0the Ramp (Ft)
Means Day Night Day Night Day Night Day Night
CONDITIONS
Freeze/FlyoutDisplacement
Freeze/FlyoutDisp & Rate
Conventional
Freeze/ResetDisplacement
Freeze/ResetDisp & Rate
PROBE TRIALS
.103
.081
.079
.083
.084
.084
.086
.093
.082
.081
.080
.079
.080
.092
.081
.090
.08
.079
.035
.008
.008
.012
.014
.012
.015
.023
.011
.010
.008
.007
.009
.010
.009
.01
.008
.009
.038
.009
.009
.013
.016
.015
.016
.024
.013
.013
.010
.009
.010
.012
.011
.012
.010
.010
.053
.024
.016
.036
.027
.031
.032
.039
.024
.025
.027
.017
.027
.020
.025
.024
.022
.022
5-6
11-12
17-18
23-24
Reliabilitiesand n2 p n
2n2
n2
n2
CONDITION
Day
Night
PROBE
Day
Night
CP
Day
Night
IMAM
Ma Oak
OROS
.022
MA ORO
1
44 h.
NAVTRAEQUIPaN 78-C-0060-9/AFHRL-TR-82-3
-
TABLE All. AVERAGE AILERON STICK ACTIVITY (IN UNITS/SECWHERE,RANGE OF CONTROL DISPLACEMENT IS FROM -1 TO +1 UNITS):
MEANS, STATISTICAL RELIABILITIES (*:p<.05, **:p<.01),AND VALUES OF ETA SQUARED (n2): PROBE TRIALS
Distance Fromthern Ramp (Ft)
6000 - 4500 4500 - 3000 3000'- 1500 1500 - 0
Means Day Night Day Night Day Wight DaY Night
CONDITIONS
.104
.11
.068
.089
.097
.104
.091
.093
.088
.095
.116 4
.085
.096
.110
.107
.109
.089
:096
.118
.117
.068
.096
.105
.113
.098
.098
.094
.101
.123
.091
.102
.120
.114
.119
.097
.098
.137
.125
.08 ,
.115
.113
.127
:113
.114
.103
.128
.142
.109
.125
.134.
.133
.146
.117
.115
.162
.133
.109
.147
.14
.159
.138
.14
.115
.145
.158
.133
.160
.140
.174
.156
.133
.125
Freeze/FlyoutDitplacement
Freeze/FlyoutDisp & Rate
Conventional
Freeze/ResetDisplacement
Freeze/ResetDisp & Rate
PROBE TRIALS
5-6
11-12
17-18
23-24
Reliabilitiesand n2 P n
2P n2 P n
2P n
2
CONDITION
Day __ __ __
Night
PROBE
Day * .019 ** ..024 * .026 ** .059
Night * .023 ** .031 ** .032 * .044
CP
Day -- * .063
Night __ __
45
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
TABLE Al2. AVERAGE RUDDER PEDAL ACTIVITY (IN UNITS/SECWHERE RANGE OF PEDAL DISPLACEMENT IS FROM -1 TO +1 UNITS):
MEANS, STATISTICAL RELIABILITIES (*:p<.05, **:p<.01),AND VALUES OF ETA SQUARED (n2): PROBE TRIALS
Distance Fromthe Ramp (Ft)
60006
- 4500" 4500 - 3000 3000 - 1500 15Q0 - 0
Means Day Night Day Night Day Night Day Night
CONDITIONS
.057
.058
.042
.051
.058
.053
.053
.056
.051
.051
.062
.054
.042
.062.
.055
.056
.054
.051
.057
.056
.038
.047
.056
.052
.05r
.051
.049
.045
.057
.052
.037
.058
.049
.052
.05
.048
.065
,.057
.038
.048
.058
.058
.053
.053
.049
.051
.062
.052
.040
.058
.053
.056
.054
.049
.073
.056
.046
.055
.062
.066
.059
.057
.052
.064
.067
.059
.044
.060
(1/4/
.067
.059
.056
.051
Freeze/FlyoutDisplacement
Freeze/FlyoutDisp & Rate
Conventional
Freeze/ResetDisplacement
Freeze/ResetDisp & Rate
PROBE TRIALS
5-6
11-12
A 17-18
23-24
Reliabilitiesand n2 p n
2n2 n2 n2
CONDITION
Day
Night
PROBE
Day
Night
CP
Day
Nighi
-
t-
46
4
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
APPENDIX B
MEANS AND STANDARD DEVIATIONS OFiNDIVIDUAL ITEMS OF THE QUESTIONNAIRE:FREEZE/RESET AND FREEZE/FLYOUT CONDITIONS.
SYMBOLS: FREEZE/FLYOUT = 0; FREEZE/RESET = A
1) Use of the freeze ifeature may be used to significantly decrease the.overall training time required to learn the landing task.
1
StronglyDisagree
2 .L.CL3
Neut
.8 T. 2.9c = 1.32 a = 1.10
4 5_
StronglyAgree
2) -Regaining control of the simulator following a freeze significantlyadded to the difficulty of the flying task in the simulator (whenresponding, consider each of the following phases of the maneuver
separately): ,
(a) ."IN THE MIDDLE" (first 1/5)
1
Stronglypisagree
\ral Strongly41e. Agree
7 = 2.2 'T. 2.6a = .79 a = 1.07
4 5
(b) "IN THE GROOVE" (second 1/3)
01 5
Strongly utral Strongly
2
DiSagree, Agree
a = 1.1 a = 1.1.
(c) "IN CLOSE" 5-10 seconds from,the ramp)
1
itronglyDisagree
2
Neu
3
T. 3.7 = 4.2a = 1.49 a = .92
47
5
4- 11 5
StronglyAgree .
3
4
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
(d) "AT THE.RAMO"
/ 1 2 3
Strongly NeutraDisagree
= 4 = ;5
a = 1.49 a = .97
5
StronglyAgree
3) It was significantly easier to.attend to the LSO's feedback while'frozen than while trying to listen and fly the Aircraft at the same time.
1
StronglyDisagree
2
Neu
3 -4 5
StronglyAgree
= 3.6 7 .8
a = .70 o 1.14
4) Improvements in performance were highly correlated with a decrease in
the number of freezes.
,1
StronglyDisagree
2
Neu
3 4 5
StronglyAgree
= 3.5 7= 3.5,a .85 a = .85
5) Using the freeze feature to explicitly identify pilot errors duringthe "training" trials made it easier to detect errors on "test" trialswhtn no feedback was given and when no freezes were in effect.
1
StronglyDisagree
2
A
-)-(= 3.4
a = 1.35
3
al
= 3.5a = 1.08
4 5
StronglyAgree
6) Compared.with the usUal practice of givii,g detailed feedback at the
conclusion of a task, providing feedback immediately following an rror is
more effective.
1 2 a-
StronglyDisagree
7' 3.3
a = 1,16
ral
5
StronglyAgree
7) In learning the task, I was more motivated by trying to avoid a freeze
than by trying to fly the task correctly.
1
StronglyDisagree
3
Neutral
a . 1.42 a48
5
StronglyAgree
NAVTRAEQUIPCEN 78-C-0060-9/AFHK-TR-82-3
8) The occurrence of the freeze was "frustrating" early in training.
1 2 3
StronglyDisa9ree
= 3.5a = 1.39
Ne
4-
= 4.3a = 1.06
StronglyAgree
9) A helpful feature would be to present a "warning" signal (such asan auditory.tone) prior to freezing the visual system.
1 2 3 -4 5
Strongly Neu StronglyDisagree ,Agree
r.-Y = 3.5
. ' a = 1.58 a = .82
10) Night approaches were more dqficult to learn than th*e Day approaches.
1
1-
5
Neutra
4
StronglyDisagree Pe Agree
L 2
7 = 2.6ay1.35 a . 1.05
11) Errors were more difficult to detect during the Night approaches thanduring the Day approaches. A
StronglyDisagree
2
Neutra
7 = 3a = 1.05
4 5
StronglyAgree
12) "'Errors serve little purpose, since students may actually learn theerrors that they commit."
DisagreeStrongly
2
tral
3- .-.. 4 5
StronglyAgree1
-
7 = 2.3 7= 2.4a = .95 a = 1.07
13) "Errors help th student to focus on the critical elements bf task
/performance."
1
StronglyDisagree
. 2
49
5
Strongly.Agree
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
14) "Instructional methods that allow errors to freely occur are ineffi-cient, since students spend valuable training time practicing incorrectresponses."
1 2- -4- -3 1 4 5
Strongly Neu al StronglyDisagree Agree
7 = 2.3 7 = 2.8a . 1.06 a = 1.03
15) "In committing errors, students learn how to recover from situationswhich at some later time may be caused not be task-specific errors but byconditions beyond their control (for example, by adverse weather, visi-bility, turbulence, etc.)."
1
StronglyDisagree
2 3
Neutr
= .74 a = .52
16) "Pointing out 'errors' frustrates students, whereas pointing out whata student is doing, 'right' is reinforcing."
5
StronglyAgree
1
StronglyDisagree
2
StronglyAgree
17) "Correct perfprmance resulfs from a process of eliminating errors andnot from a process of shaping desired performance."
02 3- 4 5
StronglyAgree
1
StronglyDisagree
-X- = 3.0
a = .94
Neutral
X = 3.2a = 1.03
18) "A student's recognition of what is consfdered correct is dependentupon his being able to recognize what is incorrect (that is, an error)."
1
StronglyDisagree
2
= 3.8a = .42
Neu
3 4 5
StronglyAgree
= 4.3a = .67
t-
50 t)
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
DISTRIBUTION LIST
Naval Training Equipment CenterOrlando, Florida 32813
Document Processing DivisionDefense Documentation CenterCameron StationAlexandria, VA 22314
JSAS Manuscript Office1200 Seventeenth Street, NWWashington, D.C. 20036
American Psychological AssocPsyc INFO Document Control Unit1200 Seventeenth Street, NWWashington, D.C. 20036
Human Factors SocietyAttn: Bulletin EditorP. O. Box 1369Santa Monica, CA 90406
Acquisitions LibrarianERIC Processing & Reference
FacilityRugby Avenue
Wthesda, MD 20014
65 LibraryDivision of Public DocumentsGovernment Printing Office
12 Washington, D.C. 20402
Technical LibraryNaval Training Equipment Center
,Orlando, FL 328133
Technical Library (DRXAM-TL)U.S. Army Material Development& Readiness Command
1 5001 Eisenhower AvenueAlexandria, VA 22333
The Van Evera LibraryHuman Resources Research
2 Organization300 North Washington StreetAlexandria, VA 22314
Unit X2 Documents Expediting Project
Library of CongressWashington, D.C. 20504
Literature Resources Department 1
BioScience Information Service2100 Arch StreetPhiladelphia, PA 19103
Acadeinic Library 2
ATZI-AG-ALBBldg 400, Room 205Ft Benjamin Harrison, IN 46216
AULSE-61-296Maxwell AFB, AL 36112
Center LibraryNavy Personnel Research and
Development CenterSan Diego, CA 92152
,
'CONRAD Technical LibraryU.S. Army gbhal Schoolht,Gordon4 GA 30905
USAF DFSLBD 1
USAF Academy, CO 80840
Milne Library, Document Section 1
State University Collegeof Arts & Sciences
Geneseo, NY 14454
U.S. Army Transportation School 1
Attn: LibraryFt Eustis,,VA 23604
1 Aerospace Systems, Inc.Research Library121 Middlesex Turnpike
3 Burlington, MA 01803
ARI Field Unit - USAREURc/o DCSPERAttn: Library
1 APO, NY 09403
1 of 8
Chief of Naval Education& Training Support
Research LibraryPensacola, FL 32509
r; tf
1
1
2
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
Old Domilion University 1
Performance Assessment Lab1425 W..49th StreetNorfolk, VA 23508
Prof. Ralph E. FlexmanUniversity of Illinois'Institute of KviationTerminal Bldg - Willard AirportSavoy, IL 61874
ATC/XPTD 10
Randolph AFB, TX 78148
HQ USAF/DPXHMMWashington, D.C. 20330
HQ USAF/MPCDWBolling AFB, Bldg 626
2 Washington, D.C. 20332
Deputy Chief of Naval Operations(MPT)
uP-01TWashington, D.C. 20370
Headquarters, US Marine Corps 1
(Code MPI-20)
Washington, D.C. 20380
HQDA (DAPE-MBR)Washington, D.C. 20310
US Coast Guard (G-P-1/62)400 Seventh Street, SWWashington, D.C. 20590
OAD (E & LS)ODDR&E
Room 3D129, PentagonWashington, D.C. 20301
DASD (MRA & 0/TrainingRoom 3B922,,PentagonWashington, D.C. 20301
ODASD(PM)
Attn: Major W.S. SellmanRoom 3B930; PentagOnWashington, D.C. 20301
HQ AFSC/DLSAndrews AFBWashington, D.C. 20334
AMD/RDUBrooks AFB, TX 78235
1
1
1
HQ USAF/MPPTWashington, D.C. 20330
AQ USAF/SAMI,Room BD936, The PentagonWashington, D.C. 20330
ATC/XPTIARandolph AFB, TX 78148
DCT/TTGHStop 62Chanute AFB, IL 61868
.Keesler/TTGHXKeesler AFB, MS 39534
LTTC/TTSLowry AFB, CO 80230
.STTC/TTGHR
Stop 32Sheppard AFB, TX 76311
SAF/ALRWashington, D.C. -20330
AFMPC/MPCYPR2 Randolph AFB, TX 78148
AFOSR/NL (Dr. A. R. Fregley)Bolling AFBWashington, D.C. 203321
1
2 of 8
Center for Naval AnalysesAttn: Dr. R. F. Lockman'2000 N. Beauregard StreetAlexandria,'VA 22311
CommanderHQ, TRADOCAttn: ATTNG-PAFt Monroe, VA 23651
NAYTRAEQUIPCEN 78-C-0060-9/AFHRL-TR-82-3
DR, Naval Recruiting,CoMmand(Code N22)
401\5 Wilson Blvd.
Arltngton, VA 22203
Det 4,\ HQ MAC
Wright4atters-o-n-AF8, OH 45433
Director\Training Analysis & Evaluation
Group \
Department 6f the NavyOrlando, FL 32813
DMDC (DPMYL)550'Camino El EsteroMontero, CA 93940
MEPCOM/MEPCTFt Sheridan, IL 60037
MIISA Branch OfficeBldg S-237Naval Air Station MemphisMillington, TN 38054
PERI-OUUS Army Institute for 'the
Behavioral & Social Sciences5001 Eisenhower AvenueAlexandria, VA 22333
Personnel & Training ResearchPrograms
Office of Naval Research (Code 458)Psychological Sciences Division800,N. Quincy StreetArlington, VA 22217
1 AFHRL Technology OfficeAttn: MAJ Duncan L. DieterlyNASA-Ames Research CenterMS 239-2Moffett Field, CA 94035
1
7-AFSC-L-taison Office
NASA-Ames Research Center1 Moffett Field, CA 94035-
1
Commanding OfficerAttn: ESD/ESS
Marine Corps Communication-Electronics School
Marine Corps BaseTwentynine Palms, CA 92278
National Defense University2 Research Directorate
Ft McNair, D.C. 20319
1
1
1
1
1 Dr. Michael Letsky (OP 102B/WEM) 1
R&D Studies Branch (Mppower,Personnel and thining)
Navy DepartmentWashington, D.C. 20350
1 H.P. VanCottNational Bureau of StandardsA-375 Metrology BldgWashington, D.C. 20234
3 Naval Civilian Personnel CommandAttn: James S. HerndonSouthern Field DivisionNorfolk, VA 23511
uS Army Institute of Administration 1
Attn: ATZI-DCO-EAft Benjamin Harrison, IN 46216
US Army Research Institute forthe Behavioral and Social
/SciencesAttn: PERI-OBP. O. Box 2086Ft Benning, GA 31905
1550 ATTW (MAC) (DOTEE) Stop #64 1
Kirtland AFB, NM 87117
3 of 8
COL A. J. AffleckDirector of Psychology - NavyDepartment of Defense (Navy Office)Camberra ACT 2600AUSTRALIA
Mr. J. AndersonHead of Personnel PsychologicalDtvision
C/O RAE, APREFarnborough, Hants, ENGLAND
LT COL Joseph BirtAerospace Medical Research
Laboratories/HEWright-Patterson AFB, OH 45433
5d
1
1
1
1
2
NAVTRAEQUIPCEN
Dr. Ralph R. CanterUS Army Research InstituteField UnitP. O. Box 16117Fort Harrison, IN 46216
Mr. K. CorkindaleAD/SAG(A)3Main BuildingWhitehall, London SWIA 2HBENGLAND
MAJ Charles A. U. Cotton'Canadian Forces PersonnelApplied Research Unit
1107 Avenue RoadToronto, Ontario M5N 2E4CANADA
Dr. Ralph DusekUS Army Research Institute
for the Behavior0 and SocialSciences
5001 Eisenhower AvenueAlexandria, VA 22333
Mr: E. ElliottSP (N)Archway South, Old Admiralty BldgSpring GardensLondonSWLA-2BE, ENGLAND-
Dr. John FordNaval Personnel Research andDevelopment Center
San Diego, CA 92152
Dr. K. GardnerHead of APUAdmiralty Marine Technology
EstablishmentQueens RoadTeddington, Middlesex, UK
USAFSAM/VN (Dr. Bilfce Hartman)
Brooks AFB, TX 78235
Dr. D. HopkinRAF Institute of Aviation and
MedicineFarnborough, HampshireENGLAND
78-C-0060-9/AFHRL-82-3
1 Mr. Les InnesBehavioral Sciences Divis.ionDefense and, Civil Institute
of Environmental MedicineP. O. Box 2000Downsview, Ontario M3M
1 CANADA
1
Mr. J. JamesResearch BranchHQ Support CommandRAF UpwoodHuntingdon, Cambridgeshire PE17 1PJENGLAND
Dr. Ed JohnsOnUS Army Research Institute for the
Behavioral and Social Sciences
5001 Eisenhower Avenue2 Alexandria, VA 22333
1
1
1
MAJ E. J. LewisProductivity Development DivisionDepartment of ProductivityCanberra ACT 2600AUSTRALIA
LibrarianRAN Research LaboratoryP. O. Box 706Darlinghurst, NSW 2010AUSTRALIA
Dr. John E.'MayhoodDirector General, PersonnelResearch & Development
National Defense HeadquartersOttawa, Ontario KlA 0K2CANADA
COL A. G. OwensD. Psych-ADepartment of DefenseCanberra ACT 2600AUSTRALIA
4 of 8
Dr. D. G. PearceBehavioral Science Division, DCIEMP. O. Box 2000Downsview, Ontario M3M 3B9CANADA
1
1
1
1
1
1
1
1
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-82-3
Dr. M. RadomskiAssociate Chief, Defense and
Civil Institute of EnvironmentalMedicine
P. O. Box 2000Downsview, Ontario M3M 3B9CANADA
Dr. L. A. RyderDirectorate of Psychology - Air
ForceDEFAIR - Russell Offices
Canberra A.C.T. 2600AUSTRALIA
Dr. H. Wallace SinaikoSmithsonian InstitutionManpower Research801 N. Pitt StreetAlexandria, VA 22314
Dr. K. N. AcklesDefense Research and DevelopmentCanadian Defense Liaison Staff2450 Massachusetts Avenue, NWWashington, D.C. 20008
Mr. K. W. TilleyACS(P) (RAF)
Main BuildingWhitehal1,4London SWIA 2HBENGLAND
Dr. Martin TolcottOffice of Naval Research800 N. Qdincy StreetDepartment of the NavyArlington, VA 22217
Dr. Martin WiskoffNavy Personnel Research andDevelopment Center
San Diego, CA 92151
Dr. Russell BrownDefense and Civil Institute ofEnvironmental Medicine
P. O. Box 2000Downsview, Ontario M3M 3B9CANADA
2 IPC Science and Technology PressLimited
Westbury HouseBury StreetGuildford, Surrey, GU2, 5AWENGLAND
1.
1
4
1
1
1
1
Commanding OfficerCanadian Forces Personnel AppliedResearch Unit
Suite 6004900 Yonge StreetWillowdale, Ontario MZN 6B7CANADA
LT COL Dennis J. Armstrong1 Psych Research UnitDepartment of Defense (Army Office)Campbell Park OfficesCANBERRA ACT 2600AUSTRALIA
The Air Attache (S38)Embassy of Australia1601 Massachusetts Avenue, NWWashington, D.C. 20036
1
1
2
1
Dr. J. Huddleston 1
Head of Personnel PsychologyArmy Personnel Research Establishmentc/o RAE, FarnboroughHants, ENGLAND
.The Naval AttacheEmbassy of Australia1601 Massachusetts Avenue, NWWashington, D.C. 20036
5 of 8
Dr. Jesse OrlanskyScience and Technology DivisionInstitute for Defense Analyses400 Army-Navy DriveArljngton, VA 22202
HumRRO/Western Division/CarmelOffice
27857 Berwick DriveCarmel, CA 93923
Sev.ille Research Corporation
Suite 400Plaza BuildingPensacola, FL 32505
'( 4
1
1
1
1
NAVTRAEQUIPCEN 78-C-00601-9/AFHRL-82-3
4444th OPS SQDN (OTD)Luke AFB, AZ 85309
TAC/DOOSLangley AFB, VA 23665
LTATG/TTDILittle Rock AFB, AR 72076
National Aviation FacilitiesExperimental CenterLibrary
Atlantic City, NJ 08405
Dr. Donald W. ConnollyResearch PsychologistFederal Aviation AdministrationFAA NAFEC ANA-230 Bldg 3Atlantib City, NJ 08405
CoMmanding OfficerAir Force Office of ScientiflicResearch
Technical Library'Washington, D.C. 20025
Technical LibraryOUSDR&ERoom 30122Washington, D.C. 20301
OUSDR&E (R&AT) (E&LS)CAPT Paul R. ChatelierWashington, D.C. 20301
CommanderNaval Air Development CenterAttn: Technical LibraryWarminster, PA 18974
Chief of Naval Operations (MPT)OP-115Attn: M. K. MalehornArlington AnnexWashington, D.C. 20350
Assistant Secretary of the NavyResearch, Engineering & SystemsWashington, D.C. 20350
2 Office of Deputy Chief of Naval. 1
OperationsManpower, Personnel and Training
1 (0P-01)
Washington, D.C. 20350
1 Chief of Naval OperationsOP-987HAttn: Dr. R. G. Smith
1 Washington, D.C. 20350
1
Chief of Naval OperationsOP-596
Washington, D.C. 20350
Chief of Naval OperationsOP-5938Washington, D.C. 20350
Chief of Naval Material1 MAT 08D2
CP5, Room 678Attn: Arnold I. RubinsteinWashington, D.C. 20360
1
Cgmmander
Naval Air Systems CommandTechnical LibraryAIR-950DWashington, D.C. 20361
1 Commander
Naval Air Systems CommandAIR 340FWashington, D.C. 20361
1
1
CommanderNaval Air Systems CommandAIR 413Attn: LCDR J. H. AshburnWasqngton, D.C. 20361
Naval Research LaboratoryAttn: LibraryWashington, D.C. 20375
1 HQ Marine CorpsCode APCAttn: LTC J. W. BiermasWashington, D.C. 20380
6 of 8
1
1
1
1
1
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-82-3
Scientific AdvisorHeadquarters US Marine CorpsWashington, D.C. 20380
1
Scientific Technical Information 1
OfficeNASA
Washington, D.C. 20546
Federal Aviation Administration 1
Tednical LibraryBureau Research and DevelopmentWashington, D.C. 20590
CommanderNaval Air Test CenterCT 176Patuxent River, MD 20670
Dr. J. D. FletcherDefense Adv. Research ProjectsAgency (CTO)
1400 Wilson BoulevardArlington, VA 22209
Comanding OfficerNaval Education Training Programand Development Center
Attn: Technical LibraryPensacola, FL 32509
1
1
1
Chief of Naval Education and 1
TrainingCode 01APensacola, FL 32509
Commanding Officer 1
Naval Aerospace Medical ResearchLaboratory
Code L5Department of PsychologyPensacola, FL 32512
ChiefARI Field UnitP. O. Box 476Ft Rucker, AL 36362
Chief of Naval Air Training 1
Attn: Code N2NAS
Corpus Christi, TX 78419
7 of 8
Chief of Naval Education andTr4ining Liaison Office
Human Resource LaboratoryFlying Training DivisionWilliams AFB, AZ 85224
CommanderEacificAissile Test CenterPoint Mugu, CA 93042
Commander,Naval Weapons CenterHuman Factors Branch (Code 3194)*Attn: Mr. Ronald A. EricksonChina Lake, CA 93555
Officer in ChargeNaval Aerospace Medical ResearchLaboratory
Box 29407New Orleans, LA 70189
CAPT James GoodsonCode L-32Naval Aerospace Medical Research,Laboratory
Pensacola, FL 32512
Mr. James BasingerASD/YWEWright-Patterson AFB, OH 45433
Mr. Don GumAFHRL/OTTWright-Patterson AFB, OH 45433
Mr. Eric MonroeAFHRL/OTRWilliams AFB, AZ 85224
Dr. Thomas LongridgeAFHRL/OTRWilliams AFB, AZ 85224
Dr. Kenneth BoffAFHRL/LRWright-Patterson AFB, OH 45433
Mr. Brian GoldiezPM TRADEAttn: DRCPM-PND-RENaval Training CenterOrlando, FL 32813
1
1
1
1
1
1
1
1
1
1
1
NAVTRAEQUIPCEN 78-C-0060-9/AFHRL-82-3
Mr. Robert WrightAeromechanics Lab (USAAVRADCOM)Ames Research Center, MS 239-2Moffett Field, CA 94035
Dr. David C. NagelLM-239-3NASA Ames Research CenterMoffett Field, CA 94035
Mr. Will BickleyUSARI Field UnitP. O. Box 476Fort Rucker, AL 36362
Mr. James L, CopelandNASA Langley Research CenterMS 125-BHampton, VA 23365
AFHRL/OTLNWilliams AFB, AZ 85224
AFHRL/TSZBrooks AFB, TX 78235
Dr. Genevieve HaddadAFOSR/NLBolling AFBWashington, D.C. 20332
Lt. Col. Jefferson KoonceUSAFA/DFBLUSAF Academy, CO 80840
CDR Joseph FunaroCode 602.Human Factors EngineeringDivision
Naval Air Development CenterWarminster, PA 18974
Lt. Col. David L. Hosley,OASD/MRA&L, Room 2D261
Cv/ PentagonWashington, D.C. 20330
15 AF/DOTPMarch AFB, CA 92518
93 BMW/D05Castle AFR, CA 95342
1 USAFTAWC/TNEglin AFB., FL 32542
U.S. Army Research Institute 2
Field Unit1 Attn: Mr. C.A. Gainer
P. O. Box 476Fort Rucker, AL 36362
1
Study Center1 National Maritime Research Center
Kings Point,.NY 11024
Officer CommandingRAF School of Education
1 RAF NewtonNottingham NG13 8HLENGLAND
AD/HF(A)20 Ministry ef Defense
Main BuildingWhitehall, London SW1A 2HB
38 ENGLAND
Dr. William E. Collins1 FAA, CAMI (AAC 118)
P. O. Box 25082Oklahoma City, OK 73125
ASD/YW-AZS (Capt. L. D. Puckett) 1
1 Wright-Patterson AFB, OH 45433
1
1
1
1
1
8 of 8
HQ USAFA/RREUSAF Academy, CO 80840
Chief of Naval Education andTraining
TDT&E BranchPensacola, FL 32508
Army Training Support Center 1
ATIC-SMDAttn: James S. HerndonFort Eustis, VA 23604
3400 TCHTW/TTS (Lt. Eckert) 2
Lowry AFB, CO 80230
4