UNCLASSIFIED

AD NUMBER

AD055062

NEW LIMITATION CHANGE

TOApproved for public release, distributionunlimited

FROMDistribution authorized to U.S. Gov't.agencies and their contractors;Administrative/Operational Use; FEB 1955.Other requests shall be referred to ArmyMedical Research Laboratory, Fort Knox,KY.

AUTHORITY

USAMRL ltr, 26 Feb 1970

THIS PAGE IS UNCLASSIFIED

a0 4 11

liMad Sei~vices Technical lfofrmation AgencyReprockced 6y

DOCUMENT SERVICE CENTERKNOTT BUILDING, DAYTON, 21, OHIO0

Because of our limited supy ouae requested toRETURN THIS COPY WHEN IT HAS &0RUVfD YOUR PURPOSE

so that it may be made available to other requesters.Your cooperation will be appreciated.

NOTICE: WHEN GOV];RNMENT OR OTHER DRAWDIOSp SPECIMIATIONS ORt OTHER DATAARE UMR FOR ANY PURPOSE OTHER THAN IN CONNECTIONV/l A DEFDV1LY RELATEDGOVERNMENT PROC UREMENT OPERATION, THE V.S. OOVRRNDMEN THEREBY INCURSNO RESPONSIBILiTY, NOR ANY OBLIGATION WTOVRAND THE FACT THAT TurFGOVERNMENT MAY HAVE FORMULATED, r0 1 FUWUED 080R1I ANY WAY SUPPLIED TRESAID DRAWINGS, SPECIFICATIONS, OR OTHER DATA IS NOT TO 118EAIE BYIMPLICATION OR OTHERWISZ AS IN ANY MANNER LICE -NSO THE HOLDER OR ANlY OTHRPERsON QR CORPORATION, OR CONVYN ANY RIGEIOR PRMUWINKTO MANUFACTURE,USE OR SELL ANY PATENTSD INVENTION THAT MAY IN ZNYWAY 15E RELATED THERETO

JEE!

A.MY MEDICAL RESEARCH LABORATORY__ FORT KNOXI KENTUCKV

U.REPORtT NO. 176

8 February 1955

MRFORMANC2E WITH UIGHT-~WEtGHT GRENADESAS A FUNCTION~ 0? WEIGHT AND D)ISTANCE*

"3motask u4a"v PSYCHOPHYSIOLOGICAL $TUDIF 46 AAL. Prjec tNio. 6-4340-001o Subtaski, Control Coordinatiou Swais-.

=: WIUN fe VLPUN Im"M WI; AV.1Wm OF 1W ARMY

BestAvai~lable

Copy

s

Report No., 176

4Pm Project No. 6-95-20-001

PEPOMIMW10 .3 WITH LIGHT-QEIGT GRNADEB AS AMOiNCTION OF WEIGHT AND DISTANCE

0D. 0. Hartman a.t R.e. Page

Psyohology Depuf t,

Submitted

30 Deeo,!Aber 1954

20 pp & ii 10 illus.

Abstraotl

Granades vvime• i 2, 4, 6, 8 and 10 once were throeo attagets located 10s 15, 20, 25, 30 and 15 yards fm tbh subjeet.Sach increase in distance resulted in a decrement in the accuracyand oomasistenaV of thrwlmdzge Weight signiticant27 altered bothaccawra and omsistmoy. i'ith the e:e'pt.ie ckf the ,rn .igre•aee, there ms a trend for decreasing aoeur&V with 2ncreasing,lit.

~"- .4

v .. 0

C;C41~II4

A 14Im o J b*

II) to .

~~40

I .. to

- . - - - - - -

DISTRIBUTION LIST FOR AMRL REPORTS

Category (A)

Copies CopsmieThe Surgeon General 10 Suprem Headquarters of the Allied 2Department of the Army Powers in EuropeMain Navy Building Medical SectionWashington 25, D.C. Roequencourt. FrancsATINTN Chief. Research and Develop-

ment Division Offioe. Chief of Army Field Forces IFort Monroe, Virginia

DirectorArmed Forces Institute of Pathology 1 Walter H. Moursund. Jr.. Colonel GS 2Wal',er Reed Army Medical Center Office of the Army Attache, Box 36Washington 12, '. C. U. So Navy 100

Fleet Post OfficeArmy Library I New York. New YorkRoam IA 522. The PentagonWashin ton 25. D, C. Commanding GeneralAIIN: Tiational Defense Review Headquarters. QM R&D Command

Q4 R&D Center. U. S. ArmyHeadquarters I Natick. MassachusettsArmy Medical Service Graduate SchoolWalter Reed Army Medical Center Commanding Officer IWashington 1276. C. Medical Nutrition Laboratory

9937 TU U. S. ArmyChief. Human Relations & Research 2 Fitaismos ArayHospital

Branch Denver 7. ColoradoMilitary Personnel Management DivisionOffice of Assistant Chief of Staff Headquarters 1DeparLuent of the Army 06th Medical General LaboratoryRoam 2C724. The Pentagon AMO 500. Co70 PostmasterWashington 25. D. C. San Francwiso. California

Office Assistant Chief of Staff, G4 I HeadquartersDepartment of the Army Camp DetrickThe Pentagon Frederick. MarylandWashington 25. D.. C, ATrIN: Technical LibraryATZN: Research & Development Sblications

Commanding Officer 1 and Reports OfficeChemical Corps Medical Laboratories QM Food and Container InstituteAray ChemcalCenter. Maryland for the Armed ForcesATIM: Chief, Technical Information 1819 West Pershin9 Road

Branch Chicago 9. Illinois

Army Ehvrezmental Health Laboratory I Commanding OfficerBuilding No. 1235 Arctic Test Branch. OCAFCArmy Cheiscal Center. Maryland (AAU 8576)

APO 733. c/o PostmasterCommanda t 1 Seattle. WashingtonMarine Corps Schools

jtantico. Virginia LibrarianATIN: Library and Record Section. N= Veterans Administration Hospital

Little Rock. ArkansasMedical Field Service School 2Fort Sam Houston. Teas

PRVY NAVY

U. S. Naval School of Aviation I Commanding Officer 1U. S. Naval Air Development Center

U. S. dkNa Air Station Johnsville. PennsylvaniaPenaoola. Florida ATM?: Aviation Medical Acceleration Lab.

Commanding Officer I Officer in Charge 1Naval Meical Research Institute U. S. Naval Medical Research LaboratoryNational Naval Medical Center U. S. Naval Submarine BaseBethesda 14, Maryland New London. ConnecticutATIN: Librarian

Category (A) Continued

NAc eV

Chief. Bureau of Medicine & Surgery 1 Special Assistant for Blo Sciences IDepartment of the Navy Office of Naval ResearchWashington 25. D. C, Department of the Navy

Commanding Officer I Washington 25. D. C.

Naval Medical Field Research Lab. SuperintendentMa-ine Barracks U. S. Naval Postgraduate SchoolCump Lejeune, N. C. MoAterey, CaliforniaAT: Librarian

AIR FC AIR FORC

Comnandant I Department of the Air Force IUSAF School of Aviation Medicine Headquarters USAFRandolph Air Force Base Director, Research & Development DCS/DRandolph Field. Texas Washington 25, D. C.ATTN: Research Secretariat ATIN: AFDRD-HF

Commander I Arctic Aeromedical LaboratoryHQ Wright Air Development Center APO 731, c/o PostmasterW~iqht-Patterson Air Force Base. Ohio Seattle. WashingtonAM:Aero Medical Laboratory (WCD) ATIN:Librarian. AAL

Directorate of ResearchCommander

Director 1 Air Research and Development CommandAir University Library ATM: FOHMaxwell Air Force Base. Alabama P, 0. 0. 35

Baltimore 3. Maryland

OM1 A( IES OM AGENCIES

Exchange and Gift Division 1 Armed Services Technical Information 5Library of Congress AgencyWashington 25. D. C. Dooment Service Center

Knott BuildingDirector. Armed Forces Medical Library I Dayton 2, Ohio7th Street & Indepedence Ave. SW ATN.: DC-SD32Washington 25. D. C.ATIW: Aaquisitions Division National Scionce Foundation

2101 Constitution AveOerations Research Office 1 Washington, D. C.The ohms HoFkins University7100 Connecticut Ave Dertment of PhysiologyChvy Chase. Maryland Unversity of RochestorATI: Lbrarian School of Modicito and Dentistry

250 Crittedn BlvdNational Institutes of Health 2 Rochester. New YorkDivision of Research GrantsBethesda 14. Maryland National Research Council

Division of Medical Science2101 Constitution AveWashington, D. C.

JCREIG AMFESSEES FOWGNIQ iADOSDefense Research Member 2AAC Liaison Officer 2Canadian Joint Staff (W) Surgeon General's Offioe2001 Connecticut Ave.. NgW Department of the ArmyWashington 8. D. C. R=am 2342

Main Navy BuildingWashington 25. D. C.

Tropical Research Medical Laboratory I Dr. 0. G. FdholmSchool of Tropical Medicine Head. Division of Human PhysiologyUniversity of Puerto Rico Medloal Reearch Commoil LaboratorySan Juas. Puerto Rico Holly Hill. Hampstead

Londn. NW 3qglcad

Category (A) Continued

EMEIGNCopiesTIHRU: The Foreign Service of theUnited States of AmericaUnited States Army Liacison OfficeAmerican Consulate GeneralSingapore

(for clearance and forwardinq tot)The Dputy Director of Army HealthFar East Lad ForcesSingapore

British Naval & Army Medical 2Liaison Officer

Bureau of Medicine & SurgeryBuildinq 4. Room 60A23rd and E StreetsWashington. D. C.

DISTRIBUTION LIST FOR AMRL REPORTS

Category (D)

coies

&=ol 1General to Army Field ForcesAM,1t: Human Engnern,?mertr Fort Bliss. Teuas

Commanding Of ficer IChief. Anmy Field ForcesArY Field Forces Human Research Nunmn Research limit 02

Uni #1 P.O. Box 446The Armored Center Fort Ord . Cali forniaFort Kuo=. Kentucky Drco

Qirtermas ter Board IHuman Rlesuces Research Of f.'CFoart Lee. Virginia P.O. BOX ý590

Washington 7. D. C.President IATTN4: LibraryBoard 92. OCAMTFort Knox. Kentucky Humnan Research Unit 93

Office, Chief of Army Field ForcesPresident I Fort Boozing. GeorgiaBoard 93. 0CAFFFort Beozing. Georgia operations Research Off ice

B410 Connecticut AvenueCommanding Of ficer I Co C e.o Marylandf~aakford ArsemalAT Tbe Library

ATN:Heiam Eninern Miso 1C Office Chief of OrdnanceAflN:HumanDagineringMissinTN: &OD3.PS, The Pen tagon

Ordnance Corps I Washington 25. D. C.Detroit ArosenalCaonter Lime. Mickloan Operations Personnel Research DivATfl4: Mr. J. D. WIlembardo VWE of fice of Chief. Research & Dowelopment

Do ty Chief of Staffso for Pious&Ce at. Pdl yant Generalso School I osearch

Ofieof; &howl Saea" Room 2C 734. The PentagonFofficenofatin Hns Washington 25. D. C.Indiamaj-Iols 16. Sodaa Office of (1iasf Signal Of ficer1

Adjutant Generails Office 1Room x 25, The PentagonPersomel Research Branch Washington 25. D. C.Department of the ArmyWashington 25. D. C. Commanding OfficerI

Trowspertation Research &President. Army Field Forcesm Deselogsnt Station'Board 01 Fort Emstis. VirginiaFort Bragg. North Ozrolina

Director IDirector. U. S. Nayml Air Ebhperi- INoval Research LAboratory mostal StationWaNShington, 25. D. C. ATIN: gvp.. Aere Medical Equipment

LbratoryOseadigOfficer IPhiladerlpi 12. Pennsylvmani

U S Naswall Medical Research Unit 91"Bid T-19. University of California CommanderIBeratele 4. Callfornia U. S. Naval Air Test Canter

Patwuot Riser. MarylandOf ficer-in-Morge AM01 Physiological Toot Section ofU. S. Nasal Prmemmel Research Field Slervice Test

ActivityTo *T U. S. Electronics Laboratory '2

Wahngoe2,7 .. on NDiego 52. California Di o

Category (D) Continued

Dieto aOerations Evaluation 1 Of floe of Naval Research51o 374e PsiophysiolLou Pscoogy

Washington 25. 1). C. Department of the Navy

Ps~chological Research & Applications 1 ahntu 5 .C

* Special Devices Center* Port Washington. L. I.. Now York

AIR FCRCEAIR IMCCommander I DirectorAir Fore Oprtionai Test Center Armament Systems, Personnel ResearchATTN: TcialReports Branch LiaboratoryEglin Air Force Base. Florida Lowry Air Force Base. Color"d

Commander 1 DiroectorLackiand Air Force Base Basic PilotSan Antonio. Texas Good fellow Air Force Base. Texma

Assistant for Ground Safety, 1 )Itmom Factors Operations 4IrS/Pers * Headquarters USA Research LaboratoryWashington 25. -D. C. Air Research and Development

Comand. USAFAFPT Research Center I Boiling Air Force BaseLiaison Officer Washington 25, D. C.Deputy Chief of Staffrrations, Head:Furters 7he Rand CorporationrrTraining Commnd 1700 NO= Street

Scott Air Force, Base. Illinois Samta Monica. CaliforniaATHI: Mr. Victor M. Hunt

Dr. Frank A, Geldard I Dr. William A. HuntIPeabody Hall Departuest of PsychologyUniversity of Virginia ~o t omutr UniversityCharlottesville. Virginia SE ta-mstr. Illinois

Dr. William E. Kappuf. Jr. 1 Dr. Antheny Loan IDepartment of Pschlogy Institute of Industrial MedicineUniversity of IliesNew Yolk Universit

Urbana Illiois BolovesMedical bso477 First As' oDr. Lorrin A. Rigge 1 New York. Now Yo*rk

Department of PsychologyD.Hrl SiebrBrow Universit Dr.ihode!eaorProvidence 12. RosIsland P

Drobam waoratorDr. Eliot Stellar I Providemies 12tIbede IslandJobes Hopkins University

* Psycholoigy Dq'stemet Dr. Willard MfthleBeltimore, le. Maryland iee. matem. Florid.

Dr., Merl. La= Rese ac 1~g Dr. Norten No~esoS. B. rear ol eead Imstitute of bmdintrial Medcine

Un v r i y of M c i a N ow York U mive.sI tAniverebyof. Michigan w' mme Imica RaterIan Abor.Michian ~First Aveame

New York. New York.

Di~eotor BibliograpbLoaL Derv~os I V We mt of PayoboAmeIPOect Unversity, of foetor

Zastiutoe for Applied 1hopeimesatal lchestor 3, New York

NomrahwlIai!Tafts CollegeBoston 14. Massachusetts

REPORT NO. 176

PERFORMANCE WITH LIGIT-WEIGHT GRENADESAS A FUNCTION OF WEIGHT AND DISTANCE*

by

Bryce 0. Hartman, Capt, MSC, andRobert E. Page, SFC

from

PSYCHOLOGY DEPARTMENTARMY MEDICAL RESEARCH LABORATORY

For Ktiox, Kent?1kyFO zerua~ry 191

*Subtask under PSYCHOPHYSIOLOGICAL STUDIES, AMRL ProjectNo. 6-95-20-001, Subtask, Control Coordination Studies.

Report No. 176Project No. 6-95-20-001Subtask AMRL S-1MEDEA

ABSTRACT

PERFORMANCE WITH LIGHT-WEIGHT GRENADESAS A FUNCTION OF WEIGHT AND DISTANCE

OBJECT

In this study, the effect of weight and distance on the accuracyand consistency of throwing was investigated. Five weights wereselected from a range which was considerably below the standard.Six distances were used.

RESULTS AND CONCLUSIONS

Each increase in distance resulted in a decrement in the accu-racy and consistency with which the subjects threw. With the ex-ception of the 2-ounce grenade, there was a trend for a decrease inthe accuracy and consistency of throwing with increasing weight.Accuracy with the 2-ounce grenade was significantly different fromthat with all other weights. The graphic analyses and observationsduring the experiment suggested that throwing the 2-ounce grenadewas a qualitatively different task than throwing the other weights.It was not possible to relate performance to effort in any systematicfashion.

RECOMMENDATIONS

Light-weight grenades should not be designed to weigh less than6 ounces until the effect of changes in the characteristic trajectoryand throwing technique with light objects is evaluated.

The relationship between accuracy and effort should be moreprecisely determined. A reactive force platform appears to be apromising instrument for this determination.

Submitted 30 December 1954 by:Bryce 0. Hartman, Capt, MSCRobert E. Page, SFC

APPROVED:4 lor ofRe earch

APPROVED: '__ 160_1_______,_____LWILLIAM W. COXLt Colonel, MCCommanding

PERFORMANCE WITH LIGHT-WEIGHT GRENADESAS A FUNCTION OF WEIGHT AND DISTANCE

I. INTRODUCTION

The hand grenade is one of the weapons used by a soldier inclose combat. Its effectiveness is determined by many factors,including the skill and training of the thrower, the conditions underwhich it is thrown, the environment, and the design of the grenadeitself. This laboratory has been conducting a series of studies inwhich the design of the grenade and the conditions under which it isthrown have been investigated. The basic approach has been thatgrenade throwing is a gross psychomotor task, involving a largeamount of force and a lesser amount of fine control (coordination)and skill.

In two previous reports, the variables under study were theshape and weight of the grenade and the distance it was thrown.None of the five shapes which were studied significantly improvedperformance (1, 2). Both weight and distance significantly alteredperformance. In AMRL Report No. 153 (2), weight and distancewere combined into an effort measure and a curve was drawn show-ing the relationship between effort and accuracy.

This study is an extension of the attempt to combine weight anddistance into an effort measure as well as an evaluation of perform -

ance with light-weight grenades. Five weights and six throwingdistances were used. The distances were 10, 15, 20, 25, 30, and35 yards. The weights were 2, 4, 6, 8, and 10 ounces. While theseweights are considerably below the weight of the standard fragmen-tation grenade, the results can be generalized to a variety of motortasks in which the gross effort of throwing is a component.

Because of the intention to combine weight and distance into aneffort measure, a factorial design was used in this study; both varia-bles were studied simultaneously. Both accuracy (radial distancefrom the target center) and consistency (dispersion, or the scatterof the throw) were used as measures of performance.

II. EXPEIRIMENTAL PROCEDURE

A. Apparatus

The grenades consisted of facsimiles of the standard frag-mentation grenade, except that there was no arming mechanism and

/

the body was not serrated. They were made of wood, drilled out orweighted with lead, as necessary to obtain 2, 4, 6, 8, and 10 ounces.Care was taken in adding lead to assure that the grenades wereproperly balanced. Five grenades of each weight were used. Theweights obtained were within one-half ounce of the desired weight.

The target consisted of a horizontal, bull's-eye type of targetarea, with a two-by-four post one foot tall at the center. A seriesof reference pins (concentric to the center) were laid out in a rirclewhich had a radius of 11 feet. The pins were set at 50 intervals,running clockwise, with zero being on the line between the centerpost and the subject. Tags, giving the angular values, were placedupon the pins. A movable barricade, measuring 4 feet high and 4feet wide, was used to insure that the thrower was at the specifieddistance from the target and all throws were made overhand. Dis-tance pegs were placed in a line at 5-yard intervals, starting 10yards from the target and going out to 35 yards. Since this studywas run in the desert, care was taken to select an experimental areawhich had sufficient visual structure to provide effective depth cuesand reduce the contribution of the perceptual component to the totalerror.

B. Subjects

Twelve men from the laboratory were employed as subjects.They were in good health, had no physical deformities of the throw-ing arm or hand, and had served as subjects in previous grenadestudies.

C. Scoring

"All throws were recorded in terms of distance to the nearestinch from the bull's-eye, and in terms of direction to the nearest S0.A hit on the center post was recorded as zero in distance and direc -tion. The score was taken at the initial point of-contact, with theroll of the grenade disregarded.

D. Procedure

Subjects made five throws of each weight of grenade atevery distance. To minimize fatigue, subjects threw only thirtytimes in each day. All twelve subjects threw one weight in a singleday. All thirty throws by a subject were made in succession. Sub -jects threw from the six distances in a predetermined random order.Throwing order was systematically balanced between subjects and

2

no two subjects threw in the same order. Each throwing order wasused five times (once for each weight) for the subject assigned to it,with the orders reversed in the second and fourth days of the series.

For each session the men were organized into teams of three.These teams included a thrower, a scorer, and a tape man. Sub-jects rotated at all three positions. In addition to the scoring teams,there were 3 supervisors on the field during the session.

The procedure for each throw was as follows: The movablebarricade was set up at the designated distance. The subject crouchedbehind the barricade with a grenade in his hand. On the command,"Throvw," the subject stood up, threw, and returned to the crouchedposition.

At the target area, the scorer pushed a pencil into the groundat the initial point of contact for the throw. The tape man movedaround the outside ring of the target so that the tape stretched tightagainst the side of the pencil. The throw was recorded in distancefrom the center of the target to the nearest inch and in direction tothe nearest 50 marked on the reference pin closest to the tape as itcrossed the outer circle.

III. RESULTS

A total of 1, 800 scores were obtained, five for each of the 12subjects, on every combination of 5 weights and 6 distances. Amean wvas calculated for each group of 5 throws. These individualmeans were then summarised into 30 group means, one for eachcombination of weight and distance. Two kinds of means can becomputed from these types of scores. The first, which will be calledthe radial mean, can be obtained from the sum of the error (distancefrom the center of the target) disregarding the angular values whichwere recorded. The second, which will be called the trigonometricmean, is obtained by converting distance and direction scores torectangular coordinates, summing the two coordinates separately,computing coordinate means, and then combining them trigonometric-ally into a radial mean by the formula

MR = M.2 + My2 *

The second type of mean was chosen for analysis because it defines

*The procedure and an example are presented in (1).

2

a specific point in space, gives a better representation of a distribu-tion of throws, and permits better predictions.

In addition to the means, which are a measure of accuracy foreach of the grenades, standard deviations were obtained trigonometric-ally. The procedure used for this computation was similar to thatused for computing the mean error scores. Standard deviations forthe X axis scores and the Y axis scores were computed separatelyand then combined in the formula

= + r2*

The resulting 30 standard deviations, which are measures of con-sistency, are an important tool for evaluating performance. Agrenade with a large scatter but whose mean score is exactly on thetarget may not be as good a grenade as one with a very smallscatter and a mean close to, but not on, the target. With this lattergrenade, the thrower has a better chance that any single throw willbe close enough to the enemy to cause damage.

Mean radikl error scores, mean Y-axis error scores, andstandard de'iations for each of the weights and distances are pre-sented in Table 1. The mean Y scores will be used in the figures

TALE I

MEANS (RADIAL ADM '¥, AXIS) AM STANDARD DEVIATIKISFOR EACH CMWNATI3X OF VEIC1T OF ORIAn: AND DISTANCE Or TAOET

'Wleight {In Oon"") |

2, 4 6 1 10

ILI# K my S Ur SD r my) rýVqW DMr. o~W r ED10 1.30 1.50 2.16 1.32 1.2. 1.251.4 1.12 1.01 0.30 1.03 0.63 1.51 1.51 1.06

15 0.29 -0,2 1.96 0.15 0.15 2.03 0.62 -0.42 1.74 0.11 0.0 1.41 1.66 1.66 1.

20 2.73 -1.12 2.72 1.02 1.02 1.56 0.51 -0.10 1.01 0.44 0.4 12. 0.30 043 0.1S25 2.8 -2.07 2,71 1.07 0.2? 2.03 1.10-0.10 2.15 1.44 1.4 2.43 2.15 -2.60 2.46

30 4.35 -4.24 2.94 2.01 -2.01 2.54 1.92 -0.54 2.3 2.34 2.13 LI3 -2.?6 -2.23 2.28

35 36I.631-68.26 3.65 2.32 -2.231 2.5 1.9 -1.58 2.5 4.02 3.71 2.71 5.6 -5.3• 4.23

r a les radia error, obtained trlgonmetrliclly.

My - Mba error a th. Y axis.

* a Stadard deviatlon. abtamLed trlmvsostricalLy.

*The procedure and an example are presented in (1).

because they permit throws short of the target to be plotted separate-ly from throws beyond the target. These scores were subjected tographic analyses and also to analysis of variance, with the appropri-ate Flo and t's being computed.

A. Graphic Analyses

Figures I through 5 present the 360 individual subjectmeans plotted on a schematic view of the target. Each point is themean of 5 throws for a single subject. Each figure represents allthe points for one weight at every distance. The X on each bull's-eye represents the trigonometric mean for the 12 points which havebeen plotted. With minor exceptions, the general results suggestedin these figures can be summarized in the following way. As dis -

tance increases, accuracy changes in a systematic fashion. Thefigures show that subjects in general overthrow on short distancesand progressively underthrow with increasing distance. In general,as distance increases, the scatter of 12 points on each bull's-eyeincreases. When the five figures are examined as a group, a similartrend for weight can be seen. As weight increases mean error in-creases, with the same general overthrow-underthrow relationsobtained with distance. Changes in the scatter as weight increasesare not clear in these figures and will be discussed later. It isinteresting to note that all the means for the 10-yard throws are be -yond the target.

Figures 6 and 7 summarize changes in accuracy (meanerror on the Y axis) and as a function of distance and weight. Eachpoint is the trigonometric mean of all throws for each weight ordistance. The pronounced effect of distance is clear in Figure 6.In Figure 7 a marked deviation from a trend of increasing error withgreater weight can be seen. Mean error for the 2-ounce grenadeis grossly greater than with the other weights. Examination of theplots in Figures I through 5 shows that this deviation increasesprogressively as distance increases, both for individual subjectsand for the means of all twelve subjects.

Figures 8 and 9 summarize changes in consistency as afunction of distance and weight. The marked and progressive effectof distance on consistency is clear In Figure S. An optimum forweight at 6 ounce* is indicated in Figure 9. This may be an artifactdue to the deviation from the general trend when the subjects threwthe lightest grenade, which has previously been discussed.

0Sr

9k

0

S1Va 35

S

*5

Fig I. M S0 H O N E h E E K H i0IIO iLtYWO GETOVE uSAEAOOO TE0 RN MEA %O W 1C

S0 00IAW V''

Ko , . , - VD.-.-- - ..

9,OL25D

XS

006 K6. 00 A TI. 4aý AREm aAEmnTGiopig uvmý "m am po EO*'n l

OV. 0 VD.

(40

0D j2

D

X0

PS a w"w HK*w Me WarhISa POTTW am A ionIinc im W UU AM aT

am~ sw ow no Cm A ud@"DIAm By X

go

30Y0

C 101-C Lr NA CO M O Kf

ON4INOTW GMU ME"O0 STOB

* 0 O.5y.0

Jew

NIA 5. MUM FOR THE IOOINC WdOEM AT EMHThif

..069"C PLOTTED ON. A SC EMATI VIEW. f MM ARE AT

ON" -TRWL. -L AWAN FOR EACH DISTANCE

a NOAMTE ByX.

.42

-442

-2,241-g

I -1.36

#0226

0 IAN -O ASE

~ SMTM TO TARUET(K TAPDS

PVL S. THE BWFEOT Of DGUMEC ON THE AOCURDCV OFTHV 9HM RA

-5

.4

-3W

-1O4

.0,3

41

oF -G m ouN UcM

PFI.?. THE FF OF WeMf ON THE AWGLMY Of

TOWKW "M -.*

.5

1.70 150

DISTANC TO TARifT(H VRS)

6.L THE UECE OF 0ISTANC ON THE CONWISENCYOF TMRO WNG HAND GRENACE6

5

4

3

2 4 6 6A -1OWEIGHT OP GRItmAC (14 OUCE)

FSIB 9.HE EFFECT OF WEIGHT ON THE CONSISTENCY orTHROWES "AN GRMNACES.

B. Statistical Analyses

The effect of weight and distance on accuracy was evaluatedby an analysis of variance, using subject means in each cell. Thedesign of this analysis of variance can be described as a doubleclassification analysis, 12 subjects (scores) per cell, the samesubjects in every cell. Table 2 presents the results of the analysisof variance of mean error Lcores. As the table shows, a significantF was obtained for the interaction between weight and distance whenit was tested against the residual. The interaction term was there-fore used in the test of significance for weight and for distance.In both cases significant F's were obtained. It should be noted thatthe variance -X4 divided by the df in Table 2) for distance is largerby a factor of approximately 4 than the variance for weight or theinteraction of weight and distance.

TABLE 2

ANALYSIS. OF VARIANCE OF ACCURACY (MEAN ERROR)

Source of yff2 df FVariation

weight 173.415 4 18.61""

Distance 882.123 5 S.44"

Wt. X Dist. 159.496 20 2.00"

Residual 1155.892 319

SubJects 100.539 11

Total 2251.365

"'Significant at 1% level.

Table 3 presents results of "t" tests of the significanceof the differences between means for weights and for distance.These "t's" were obtained by the difference technique. For distance,all pairs were significant except for the adjacent pairs in the 15,20, and 25 yard portion of the range. For weight, accuracy withthe 2-ounce grenade is significantly poorer than that with any othergrenade. The 6-ounce grenade differed significantly from the 10-ounce grenade. All other combinations were not significant.

TAILE 3

-XDTS OF SIGNIFICA•T E n(W O rTE'E M Mr•. MEAN EM FMR WEIM AND DISTANCE

Mom Error (ft) 0to Mec Error (ft) 0to.t 2 . 4 4.30** tio"15 - 2.12-

2 os - 4.56 t 2. 6 3.42"* 10, yda, a 1.75 t 1 O. 20 a 2.91"4 p4 - 2.80 t2 -8 *•4 15 yds = 2.29 t 1 0. 25 - 3.84"*6 oz - 2.60 t2-10 3.57" 20 yda -2.43 t 1 0 .3 5 . 8.91"

0 oz - 2.89 t4-6 = 0.98 25 ydo 3.03 tlO-35 . 8.57*"

10 oz a 3.36 t 4-8 = 0.00 30 yds = 4.36 tl 5-2D - 0.40t4-10 .lSO 35 yde - 5.68 t 1 5 .2 5 . 2.19"

-7-8 , 1.11 t 1 5 - 30 a 6.62""

(d% a 71) 99t.6 10 - 2.64" (df = 59) t15.33 = 9.64"*

(5% - 1.99) - 1.66 (5% - 2.00) t20.23 - 1.97

(1% 2.65) (1% 2.66) t20- 0.m77-

t20•. 35 . 7.52"*t25.30 . 4.01"t25 - 35 - 5.550*

t 30-35 " 2.97--

Significmnt at :% level., -

o S1•iflcnt at 1% lovel.

The effect of weight and distance on consistency was eval-uated by an analysis of variance, using only one score (the standarddeviation of 12 subject means) in each cell. *This analysis 9f vari-ance can be described as a double classification analysis, one scoreper cell. This design was selected to be used on consistency scoresbecause there were insufficient throws (5 per subject) in each of the30 conditions to permit individual standard deviations to be computed.For the same reason, t's on consistency for pairs of weights anddistances could not be computed. While this design is sufficientfor examining the influence of primary variables upon performance,it does not permit a test of the interaction between the two primaryvariables. The results of the analysis of variance are shown inTable 4. Both weight and distance gave "Fls" which were significantbeyond the 1% level of confidence. The variance ( EX 2 divided bythe df in Table 4) associated with distance is more than half of thetotal variance.

TALE 4ANALYSIS OF VARIANCZ OF CONSUI1IY (STANDARD DEVIATICN)

Source ofVariation 1[2 df Fweight 3.199 4 4.97"Distanc 13.907 5 17.27*Remainder 3.227 20

Total 20.333

" SignLf Lct at 19 level.

00

.7 9

22

0 'a • i+2 2I

EF.R (DAM x + W)N MO

re. 1. A LOT F DO -- REAO TO WOI F L

NISI

NQ

" d 2 0 •o

-4 II

In Figure 10, an attempt has been made to relate the accur -

acy of throwing to effort, by plotting error against a combination of

weight and distance. The effort scale (which is, at best, a rough

approximation of how hard the grenade was thrown) has units desig-nated yard-ounces, obtained by multiplying the distance to the targetby the weight of the grenade. In the previous study (2), this pro-cedure was used and a systematic trend was evident. A curve wasfitted by visual approximation. The distribution of the points inFigure 10 is such that no attempt has been made to approximate. acurve plot.

Since an extension of the effort curve from the previousstudy is one objective of this study, it was necessary to have somemethod of evaluating the comparability of scores in the two studies.For the instructional period in this study, every subject made 10throws at each of two distances (20 and 35 yards) with the 18-ouncegrenade. Target order was reversed for half of the subjects. All20 throws were made in succession. Eight of the 12 subjects usedin this study had served in the previous study. Their scores forthese combinations of weight and distance in the two studies werecompared. Differences were tested for significance by computingt's. A "t" of 1. 63 was obtained for throws on the 20-yard target. A"It" of 1. 01 was obtained for throws on the 35-yard target. Neitherwas siglnifidant ?t the 5% level of confidence.

IV. DISCUSSION AND CONCLUSIONS

Throughout the graphic and statistical analyses the marked effectof distance on both the accuracy and consistency of throwing grenadesis evident. As distance increased both accuracy and consistencydecreas..d.

The results with weight were not so clear-cut. kn the analysesof variance, weight yielded significant "F's" for both accuracy andconsistency. Examination of the "t's" for accuracy in Table 3 showsthat this significance is primarily the result of the difference betweenthe 2-ounce grenade and all other grenades. Figure 7 confirms thisgeneral conclusion. Figure 9, which shows the effect of weight onconsistency, suggests the same general though less positive con-clusion. Examination of the mean Y scores for the 2-ounce grenade.given in Table 1, shows a marked increase in error at the 30-and35-yard distances. This is probably the reason for the significantdifference of the 2-ounce grenade. It appears, from an inspectionof the individual means with the 2-ounce grenade on the 30-and 35-yard targets in Figure 1, that these two distances are too far for aman to throw a very light-weight o0ject effectively. A similar butless marked effect can be noted in Table I for the 4-ounce grenadesat 30- and 35-yards. Observation during the experiment revealed aninteresting difference in the way the light grenades were handled.

At the shorter distances, although the throw was overhand, the menmade throws which had almost no trajectory at all. Instead, theymade a hard, flat throw, directly at the target post. In general,it must be concluded that throwing this extremely light object was aqualitatively different task than, that required by the remainder ofthe grenades.

The attempt to relate performance of throwing grenades with acrude approximation of effort was fruitless. It was pointed out abovethat the throwing task with the lighter grenade3 at the shorter dis-tances was qualitatively different from that of other combinations.This same general effect can be noted in the effort plot, which sug-gests that some combinations of light weights and short distances,although plotted at the low end of the effort scale, might more ap-propriately be at the high end of the scale. It is possible that changesin the way a man throws a light object a short distance while strivingfor maximum accuracy markedly alters the relationship betweenaccuracy and effort found in moderate and high parts of the range.

More adequate measurements of this effort could be obtained bya different technique. When the human throws, he exerts a reactiveforce with his body which can be directly related to the force impartedto the object thrown. If he threw while standing on a platform instru-mented to measure the reactive force, a direct measure of effortmight be obtained. This technique is now under consideration.

Three general conclusions come out of this study. First, asdistance increases, accuracy and consistency decrease in a systematicfashion. Second, weight significantly alters accuracy and consistency.Third, changes in performance as a function of weight may be relatedto qualitative differences between throwing light and heavy objects.

The data suggest that these differences occur around 6 ounces,both in terms of accuracy and consistency.

V. RECOMMENDATIONS

Light-weight grenades should not weigh less than 6 ounces.

The relationship between effort and performance at the grosspsychomotor task of throwing should be investigated with more preci-sion. Varying amounts of effort could be obtained by having subjectsthrow objects of several weights at targets located at severaldistances. A reactive -force platform might provide a more directand precise measure of the effort expended.

VI. BIBLIOGRAPHY

1. Hartman, B.O., J. T. Burke, and R. Y. Walker. Theaccuracy of throwing hand grenades as a function of theirweight, shape, and size. Army Medical Research Labora-tory, Report No. 117, 18 June 1953, Fort Knox, Kentucky.

2. Hartman, B. 0. The accuracy of throwing hand grenadesas a function of their weight, shape, and distance. Army

Medical Research Laboratory, Report No. 153, 2 Novemn-ber 1954, Fort Knox, Kentucky.