An Investigation into

the Value of Unit CohesionCin Peacetime

.9

' .FEB t 7 4

DIVISION OF NEUROPSYCHIATRY A

Walter Reed Army Institute of Research

Washington, D.C. 20307

u0193 a m o 0-pproved-This documeft hasbeen,- r ,d

S o, public release and sale; its

distribution i5

mhe contents of this reoort are the authors' personal. opin-

ions and not those of the U.S. Army or the Walter Reed ArmyInstitute of Research. Approved for public release; dis-tribution unlimited.

C.F. Tyner, .D.COL, MCDirectorDivision of Neuropsychiatrv

-.i7Ae - / ," .Franklin c ,Jr. , M. r."1COL, MCe- T

Di':ectorWalter Reed Army Institute of Research

..............-".- ...... ,--- ..---..-.-.°-- .-....

i ow.- W .. -d -O C; . 6 t - 4 a

Report WRAIR NP-83-5

AN INVESTIGATION INTO THE VALUE OFUNIT COHESION IN PEACETIME

Frederick J. Manning, Ph.D.

Larry H. Ingraham, Ph.D.

Division of Neuropsychiatry202-576-3006; AVN 291-3006

November, 1913

Scientific Manuscript

-4.•

Walter Reed Army Institute of ResearchWashington, DC 20307.

U.S. Army Medical Research and Development CommandFt. Detrick, MD 21701

I This documpnt 11. ------

Sf,! publi.- rece - Qd a, ri, Ji. ' ;'" ""i i t. C3q'l t'."

ax

-, ad *~ - - L s .. . . - . - . -..

PREFACE

This paper is one of a series of occasional, informalaccounts of work in the Division of Neuropsychiatry at theWalter Reed Army Institute of Research. The reportsgenerally address topics in Army preventive medicine forwhich implementation responsibility lies significantlyoutside the Medical Department. Although their contentsmay overlap partly with our publications in the scientificliterature, most papers are based on trip reports, brief-ings, and consultations involving specific Army audi-ences. Comments to the senior author are welcome.

This work was supported by Research Area III -- HealthHazards of Military Systems and Combat Operations -- of theU.S. Army Medical Research and Development Command; MGGarrison Rapmund, Commanding.

The paper will appear as a chapter in the forthcomingbook Contemporary Studies in Combat Psychiatry, G.L.Belenky and F.D. Jones, eds; Kluwer-Nijof, Boston; 1984.

,.9

1* V

[3

*•. - .,

CONTENTS

INTRODUCTION 1

MEASURE MEN OF COHESION 4

RELATIONS BETWEEN SURVEY SCORES AND 7TRADITIONAL MEASURES OF BATTALION PERFORMANCE

FIGURES 12

GLOSSARY •22

- REFERENCES 24

.

M UFr P .- [- -Ca - -7 - li k-Tk -p 77 7 r W 7

INTRODUCTION

The importance of welding soldiers into small groups ofcomrades has been recognized by military leaders for cen-turies. Richardson (1978) takes us back to the firstcentury BC to Asclepipdotus' book Tactics, which tells usthat in the Greek phalanx a front rank man, protostates,together with his rear rank man epistates, were togethercalled parastatai, comrades in arms. ,Richardson goes on tocite Onasander, in the "first century AD, as the first realgeneral to mention the importance of comradeship. Rerecommended that commanders should station "brothers inrank beside brothers; friends beside friends; and loversbeside their favorites." A strikingly similar passageoccurs in the British Regulations for the Rifle Corpsprepared in 1800:

Having found his company... he (the captain) willthen arrange comrades. Every corporal, private andbugler will select a comrade of the rank differentfrom his own, i.e., front and rear rank, and isnever to change him without the permission of hisCaptain. Comrades are always to have the sameberth in quarters; and that they may be as littleseparated as possible, in *either barracks or thefield, will join the same file on parade and go onthe same duties with arms when It is with thebaggage also.

The French officer and military theorist Ardant DuPicq(1958) extended this advice in attempting to persuade hisChief-of-Staff more than a hundred years ago of the import-ance of unit cohesion. *Four brave men", he said, "who donot know each other will not dare to attack a lion. Fourless brave, but knowing each other well, sure of theirreliability and consequently of mutual aid, will attackresolutely."

L4.

Many first hand reports on battle stress casualtiesmade it clear that by the end of World War II, Americanpsychiatrists also recognized the overwhelming importanceof interpersonal relationships in sustaining soldiers inbattle. Historian S.L.A. Marshall (1966) said it besthowever, in writing of his observations in World War II:

I hold it to be one of the simplest truths of war

that the thing which enables an infantry soldier tokeep going with his weapons is the near presence orpresumed presence of a comrade. .

Later on, he answers his own question of what induces aman to face death bravely: I

...largely the same things which induce him to facelife bravely -- friendship, loyalty to responsi-bility, and the knowledge that he is a repositoryof the faith and confidence of others.

The importance of unit cohesion in time of peace, itseems to us, is much less well accepted. Indeed, one couldargue, with Marshall, that:

It is from the acquiring of the habit of workingwith the group and of feeling responsible to thegroup that his (the soldier's) thoughts are apt toturn ultimately to the welfare of the group whentactical disintegration occurs in battle.

One could argue as DuPicq, that while esprit-de-corpsmay improve with experience in war, wars are becomingshorter and shorter, demanding, therefore that we createesprit in advance. However, it has been our experiencethat these arguments are often ineffective with command-ers. Their posture may be summarized by the answer wereceived on one occasion: "The enemy will take care of ourcohesion building. Right now, my job is training, notmaking the troops feel good." The project we will describebelow was our attempt to find an answer to the basicquestion implicit in that response: How-does the presenceor absence of unit cohesion affect the peacetime perform-ance of basic individual and unit 6kills?

Our investigation is of course not the first in thisarea. There exists an extensive literature devoted to therelationship between interpersonal attraction and produc-tivity. Results, however, in studies of the peacetimemilitary, athletic teams, and industries have all provedequivocal. Goodacre (1951) found a high positive correla-tion between sociometric measures of cohesiveness and theproblem-solving scores of combat units engaged in field

-2-

** ~ j~a% .. %'* % V" v. J1; ., V* " . '.. . ". X.v . , "

exercises. Hemphill and Sechrest (1952) studied bombercrews in combat .over Korea. Sociometric scores of crewcohesiveness were positively correlated with bombingaccuracy scores. French (1951) on the other hand, wasunable to show a significant relationship between hissociometric index of cohesiveness within military recruitcompanies and a variety of measures of performance, andPalmer and Myers (1968), observing radar crews of fortyanti-aircraft batteries for a period of three months, foundsociometric measures of group cohesiveness negativelyrelated to productivity.

Results are no less diverse in the area of teamathletic competition. Klein and Christiansen (1969),VanderVelden (1971), and Wydmeyer and Martens (1978), forexample, all found highly cohesive bas-ketball teams weremore successful than less cohesive teams. Fiedler (1954)and Grace (1954), however, found a negative relationshipbetween cohesion and performance, and Melnick and Chemers(1974) found that cohesiveness had neither a positive nornegative relationship to team success in basketball.

Stoqdill's (1972) review of the experimental and civi-* lian work force literature produced the same diversity of

results: twelve studies showed a positive relationshipbetween, productivity and cohesiveness, eleven showed anegative relationship, and eleven showed no relationshipwhatever. -In the analysis of these results, Stogdill(1972), points out that cohesiveness and produtivity tendto be positively related under conditions ofl high groupmotivation and negatively related under condiiions of lowmotivation. An even more elementary explanation, however,is the wide variation in the measurement, indeed even thedefinition, of cohesiveness. Cartwright (1968) has pointedout three rather different uses of the term: (a) attrac-tion to the group, including resistence to leaving it;

. (b) the motivation of members to participate in groupactivities; and (c) coordination of the efforts of mem-

" bers. Although he felt that sociologists and socialpsychologists had more or less come to a de facto agreementlimiting their use to the first of these three (cf. Lottand Lott, 1965), Zander's (1979) view was that " .... in theabsence of a reliable method for measuring cohesiveness ina natural setting, or a reliable procedure for creating itin the laboratory, one cannot be sure to what phenomenainvestigators are attending when they examine its originsor effects."

-3-

€, €4'.; €€,5 4€ ' ". .'';-. . ,"-.'. .- ,S. ggg 2 ."" -"." . "." *X* 5."~ . .. .2 " :. *.

Military writers, at any rate, tend to use a workingdefinition which includes group motivation or direction aswell as group attractiveness. The Chief-of-Staff of the USArmy thus defined unit cohesion as follows: "The bondingtogether of soldiers in such a way as to sustain their willand commitment to each other, the unit, and mission

. accomplishment despite combat or mission stress (ARCOSTAction Team, 1980; see also Hauser, 1979). Thisdefinition, which incorporates the added concept of group

9drive, implies that the group member's identification withL-: leaders of his unit and his group often results in commit-

ment to the norms of the formal organization which theseleaders represent. It also emphasizes the critical role ofthe small unit leader, who is in fact a member of at leasttwo groups simultaneously. The "link-pin" concept ofLikert (1961) is helpful in this regard. For Likert, lead-ers occupy positions in a hierarchy between levels; theyare simultaneously members of their small face-to-face workgroups and members of the next higher managerial echelon.It is thus possible for cohesion to be transmitted and

4 distributed throughout a sizeable collection of groups that4' are not coextensive in their memberships but are linked to

one another by members who occupy positions in more thanone group. We generally speak of this larger collective ashaving esprit-de-corps or esprit when this process issuccessful. In any case, we began our inquiry into the

-9 value of. cohesion in peace time with a clear realizationthat it would need a measure of cohesion that included notonly attraction to peers, but also identification withleaders and/or organizational goals. Our survey of theliterature, and that of Stogdill (1972), made us confidentthat if we could devise such a measure the importance ofunit cohesion to peacetime military performance wouldbecome apparent to commanders.

9 MEASUREMENT OF COHESION

Conversations, interviews, and test runs with soldiersand experienced leaders, as well as close inspection of theliterature cited above, led to a battery of questions whichwere put to a sample of each of the 20 battalions visitedby the 7th U.S. Corps Inspector General (IG) in the courseof a 9 month period in late 1979 to early 1980. A total of37 people in each battalion were questioned by IG teammembers: the battalion persbnnel officer, the Company

I -4-

S.-J -. V. .- S 1

Commanders of Headquarters (HHC) and Alpha Companies; theFirst Sergeants of Charlie and Service Companies; platoonleaders from 1st platoon Bravo Company, 2nd platoon CharlieCompany, and the scout platoon from Service Company;platoon sergeants were quizzed from the communications

.4 platoon of HHC, 3rd platoon, Alpha Company, and 4thplatoon, Bravo Company; Alpha, Bravo and Charlie Companieseach contributed two squad leaders, and HHC and ServiceCompany one each. Fifteen junior enlisted soldiers were-selected at random from the unit manning roster, as well as3 soldiers below the rank of E-4 who had arrived onlywithin the prior month. The sets of questions were ofcourse tailored to fit the position of the person withinthe unit, and tapped both subjective feeling ("How do youlike being in this unit?") and objective information ("Whodo you spend time with after duty hours?"). The juniorenlisted men were questioned about their squad, squadleaders about their squads and their platoons. The PlatoonSergeant was questioned about his platoon and the company,and so on up the line so that although we ultimately derivea battalion score, this is merely a compilation of theattachment the surveyed members feel to their immediategroup (including the leaders). The leftmost column ofTable 1A is a list of the questions asked the juniorenlisted soldiers. The central three columns (headed byplus, 0, and minus) are sample high-cohesion, low-cohe-sion, and zero-cohesion answers. The inspection teammember asking the questions compared the answers receivedwith the sample answers and simply circled the sampleanswer most similar to that given by the subject. We thenawarded one point for each plus answer and subtracted apoint for each minus answer. )The individual's score was

-.S simply the algebraic sum, and the battalion score the sumtotal accumulated across all ranks, positions, andquestions. The three columns on the far right of Table 1Aare in fact the percentage of subjects giving high, low andzero cohesion answers to these questions. These data arebased on the answers of 300 junior enlisted soldiers in 20battalions. Table 1B is a similar display of the questionsput to company commanders.

A question that arises immediately in the developmentof any new measuring instrument, of course, is that ofreliability. In the present case, a skeptic might ask ifwe were actually learning something about the battalionsinvolved or about the IG team members asking the ques-tions. In fact, that doubt has 'been almost entirelyresolved by the 35 interviews that were scored independ-ently and simultaneously by a team member and one of the

-5-

authors. The correlation between the two sets of scores soderived is .98, so whatever the questions may be measuring,.they almost certainly involve differences among battalionsrather than differences among our questioners.

The question of validity, however, is somewhat moredifficult to answer. In the words of more than a few ofthe battalion commanders whose units scored on the lowside, "Are we really measuring unit cohesion?" Thequestion itself, of course, assumes that there is some

•. standard out there against which we car hold our newmeasuring instrument to assess its adequac much like thestandard yard, foot and inch measures the GreenwichObservatory. In fact, if such a standard sts at all, itis in the minds of people like the cru 'I old Infantrycolonel who was the Inspector General fo . Corps. Ouronly goal was to make a handy instrumen . one doesn'tneed 30 years' experience to tell whetheL a unit has areasonable level of cohesion. Viewed in this way, themeasurement appears to have a fairly high degree ofvalidity. There were only two instances out of 20 batta-lions where the Inspector General saw unit cohesion asmarkedly different than our scores indicated. In one case,he felt the scores'.were too high, and in another case hefelt they' were too low. Further; evidence for "face"validity came from the scores of the two armored cavalrysquadrons we assessed. These two units, the closest wehave to elite troops in US Army, Europe, gathered 82 and 79percent of all possible points on our cohesion measure. Therest of the units tested scored between 65 and 74 percent.

.4,

Another approach to the topic of validity forsakes thesearch for an outside standard altogether, and simply askswhether the measure helps organize our experience at all.Does it show any orderly relationships to other availabledata? If so, are they the relationships one would expectif the measure reallx measured cohesion? The data inTables Two through Six provide an affirmative answer toboth of these questions.

A varimax factor analysis conducted on the averagescores of the 8 subgroups (personnel officer, C.O., FirstSergeant, etc.) of each battalion, usinq the StatisticalPackage for the Social Sciences, yielded a primary factormost strongly represented by the scores of junior enlisted

'soldiers, followed closely by those of the commandingofficer and first sergeant. Two further factors, strongly-dominated by the scores of the personnel officers (S-l) and

-6-

-" r t acnd .. .the remainder o the

'

I! "newbies" respectively, accounted for the remainder of the

variance. Therefore, for the sake of clarity and simpli-city, the..data in Tables 2 through 6 are limited to thescores of the junior enlisted. Each of these tablesinvolves dividing the total enlisted sample into sub-samples based on type of battalion, type of job, months onthe job, rank, or race, and then noting what percentage ofthe scores in each of these subgroups fell into the lowthird, the middle third and the high third of all juniorenlisted scores. Simply put, the important number in eachof these tables is 33: if there are no differences amongthe sub-samples, then all of the entries in the table

should read 33. In Table 2, howeVer, we see that only 23%of junior enlisted soldiers in the armor battalions weinvestigated had scores which placed them among the lowone-third of all junior enlisted. Thirty-two percentscored in the middle third, and 45% scored in the high one-third. Further inspection of the column labelled "HIGHTHIRD" reveals that Armor and Cavalry, both units organizedaround small groups of soldiers in a fighting vehicle, showdisproportionately high cohesion scores. If we look atTable 3, which shows as its sub-sample career managementfields (type of job), we see that 46% of Armor crewmenscore among the high one-third of enlisted soldiers.Tables 4 through 6 also show reasonable results for apurported measure of cohesion. Scores increase with rank,and with time.on the job, and, as we might expect, minori-ties tehd to identify less with their battalion tbanCaucasians. We could perhaps continue this analysissomewhat further, but it is clear that the findings are atleast consistent with the hypothesis that we're measuring

.." "the bonding together of soldiers in such a way as tosustain their will and commitment to each other and theunit". We will now turn to the subject of whether ourmeasures have anything to do with mission accomplishment.

-'.

RELATIONS BETWEEN SURVEY SCORES- ANDTRADITIONAL MEASURES OF BATTALION PERFORMANCE

Table 7 shows the intercorrelations among nine mea-. sures of battalion performance. A glossary of acronyms is

included at the rear of the paper, so we will not go intogreat detail at the moment on this table. We started out

*with a much larger list --- 23, in fact --- which constitu-ted just about all the quantifiable information we could

-7-

,.-~ .. ., . . . . . .. - . . .. - -...- . .. - - . .- - .**. ', . ' " . -. . . . .- - - . ,

obtain on the units. Many of them, however, were closelyrelated to one of these nine or, in a couple of instances,showed the same score for nearly all battalions. Thesenine are not very closely related, and our combat armsbrethren assure us that taken as a set, they provide apretty fair picture of battalion functioning. Figure 1contrasts the performance of the 5 battalions with thehighest cohesion scores with the 5 lowest scoring, andTable 8 shows the correlation of cohesion scores with thevarious measures of performance across all 20 battalions.The bottom line of the table shows the rank-order correla-tions between the battalion cohesion scores (i.e, summingover all 37 interviews in each of the 20 battalions) andeach of the nine performance measures. The lines abovethis one show the statistically significant correlationsbetween these measures and various subsamples of thebattalion. Ignoring the far right hand column for themoment, the table shows very strong relations betweencohesion scores and the results of the annual generalinspection (AGI), with physical fitness testing (PT),.operational readiness testing (ORT), and with the number ofbattalion members arrested in the previous 12 months(CRIME). Considerably less impressive relations existedbetween unit cohesion scores and the. percentage of batta-lion members passing the. LG-administered skill qualifi-cation tests CSQT), the battalion's reenlistment (REUP6),disciplinary (UCMJ6),. afid administrative discharge (AD6)rates. No relationship whatever was seen in the case ofyearly battalion level tactical testing (ARTEP). We wereinitially quite disappointed that all our cohesion measuresdid not correlate strongly with all our performance mea-sures, and spent considerable time and effort evaluating

-- hypotheses explaining this particular spectrum of find-* ings. Platoon leaders, platoon sergeants, and section or

squad leaders show very low correlations with battalionperformance. Why this should be so is not clear. Thesimplest explanation is that the questions asked of thesegroups were simply not good ways of measuring the extent towhich these men felt themselves a part of a cohesiveunit. It is of course possible that the questions are finemeasures, but that some unique characteristics of thesegroups or their positions (e.g., first level supervisors,former junior enlisted, etc) make their cohesion scoresunrelated to unit performance. We cannot presently deter-mine which of these alternatives is correct, and would thusopt for the simpler.

On the performance side, we finally recalled conver-sation with commanders in which they talked about juggling

-8-

ipriorities, even selective disobedience, in the face of too

i little time for too many tasks. Indeed, if everyone picked~his priorities slightly differently, we would be doomed toi exactly the kind of results we see in the table. Under. these circumstances perhaps the most useful description of

our results would be that performance is a function of bothknowledge (itself a functin of such things as training

.. .. time, instructor ability and diligence and training aids aswell as native ability) and motivation (a very direct

~function of unit cohesion and esprit as well as traditionalcreature comforts): Performance - f (knowledge x motiva-

Stion). Factors like the battalion'sproiesthlvlof technical sk 11 required for a given task, and available

' resources will determine which specific aspects of a givenbattalion's performance are affected most strongly by level

~of unit cohesion.

The right hand column of Table 8 might be seen as atest of this notion of cohesion as a non-specific "multi-plier". It shows the correlation of cohesion scores andthe average ranking of the battalion on the nine perform-ance measures in the table. As expected, the correlationof battalion cohesion with this measure of overall per-formance is quite high (.81). Scores of the junior enlist-

; ed soldiers (.72) and the company' commanders (.68) alsoSshowed exceptionally high correlations, It seems likelyJ'. then that unit cohesion, "bonding together of unit mem-" bers.. to sustain their will and comnfitment to each other,- the unit, and mission accomplishment," is indeed not only a-- "force multiplier" in combat, but a powerful "training

multiplier" in time of peace.

.. Some might argue (and have) that this discovery by nomeans .implies that esprit or cohesion causes high per-

-'.i formance, but that in fact it is more likely to be otherway around -- that high performance produces high

Sesprit. There is certainly nothing in our data that would" , allow us to choose between these two positions (if indeed- . we must choose rather than accept the seemingly obvious

middle ground of a reciprocal interaction) . Our Army isjust initiating a substantial number of changes aimed atdrastically increasing 'unit cohesion, including introduc-tion of a regimental system of some sort and unit rotations

-<Wto overseas assignments. Perhaps we will soon know theS.answer to the question of primacy. In the meantime,

however, we can ask where the high scoring units in the- present study step away from the pack. Even more. precisely, which questions on our survey differentiated the< five most cohesi ve units from the five least cohesive

.- 9-

. .

units? Nine of the junior enlisted questions so qualified,but the best of these were:

(1) How often, aside from meetings, does the CO talkwith you personally?

(2) Is your squad (section) leader ever included inafter-duty activities?

(3) If we went to war tomorrow, would you feel confi-dent going with this unit, or would you rather go withanother?

(4) How often, aside from meetings, does your platoonleader talk with you personally?

(5) Who would you go to first if you had a personalproblem, like being in debt?

Question number three, on confidence in going to war,was intended as a broad sort of summary question, and. itdoes not provide much guidance for creating cohesion,

4however well it may measure it. *The other four questions,however, seem to us to have profound implications forleadership.

"Solidarity and confidence cannot be improvised. Theyare born only of mutual acquaintanceship ... pride existsonly among people who know each other well". This adviceof DuPicq (1946) is apparently nowhere more applicable thanin the relations of leader to led. Not only does the groupmember's commitment to the norms of the formal organizationdepend upon identification with the leaders, in the "link-pin" fashion described above (George, 1971), but personswho are made to feel like valued members of a group willfeel far more attraction to the group than those who do nothave much social worth. We would argue from this thatbuilding cohesion requires interaction beyond the worksetting,. where rank and duties so clearly delimit"worth." Unit athletic teams provide excellent examples ofsettings where a private' might outperform superiors, mighteven teach them a thing or two, and in the process, come tobe known by them as other than first rank, fourth file inthe heavy weapons platoon.

Which activities are not so important as who partici-

-10-JI.

* o 5.*. ** *,?, * , :,'-:-,. ,;.P.' ... ', '. , w5 , ,.'..* 2 ."" "". ,.. . .• " ' C .... . - .". , , ".

-4p.

pates in how many different settings. Company leadersusually acknowledge the necessity of "command presence* inthe barracks after duty hours, but all too often find theyhave nothing to say once they get there. They find their

4only shared experiences are the formal interactions of theworkday. Hence, their presence after work is oftenresented. The more people, the more varied the settings,and the more time the group maintains stable membership,the more the members have in common and the higher theresultant cohesion. S.L.A. Marshall (1966) provides asuccinct and appropriate closing which is consistent withthis view:

The good company has no place for an officer whowould rather be right than be loved, for the timewill quickly come when he walks alone, and inbattle no man may succeed in solitude.

A -

we.

4%

'U, ;::" 4, % ' , . ,?;,)"?; €" '- -22 :" '?'""''"2."",""" " €'

ifS * 0 fl 0 P4

44 *. q * VS - VIn

.P54

P4 on4 ev 0% 'a 'a VI

ro f -

ni

z = U Z' i

ZO at£

0z 0 c

*E CA A 0L

fi aU 2 i

uZ2 0U .0ue.2 a" OE34.4:~X 4, -

.... a z- z~ , ~Z 0.o 0 zo 0

0 zz

ZU -6

a 69 a,00

No - JA' 30 0-#~A a C e

z zA

).. Y. J- - ~xo<a

-C-

~.. U UD

4A03. '

0 c sq -

I "- a.c @a cr a, C A

SM~~ M II "0 3a . AUCUc l

4 .8 .1 11 - -- .26

44 A , -

4-4 1% P, in K 0% In

min 0 0 ' O N C

.4 Uc 1-o

t mlot

-C .

AA A =

va 4A ** -

A "m W.* ~ V

P as

x. .. w- .A 2

C! Z0 133 C 0z Z.2

-AZ--

2IA z

tA 4c

~0 0.2 E I 1 -

x#A m 'e A C A * A 0

Z C c

53 c

0,= 0 j!50 0'~~weoo -u -= V

Ay A0 V.S

g 5 0*

E- 'Go

UCL 0 i

-0-. a OU

IA 30 a. t '1

K~~~ r cs to ' *

a C - *j ICUs 2

44~- A: 4 1 l i 0

13.-f

4>

6-44

Iii0

00

0-

PU4 ono

--

0

00

o z

coM 0D

I-a

ga

in,ca

Lo=

z

C4 M4 r4.C122

0E UzI-g

U3g

22

4r0

"IWI

en r.

4 I4Az

>4-

0 -4F., I-S C43

0-4

00

.- r4 - .* *-.- q. * -- UGo

-J 2d

0

u

u

'S I-. ca

C.C4

* z

-' g

all

i dc

-S C.,

0~ m 0N

zz

Kf.A

zz

z

0 i

40 do=) N

U3

00

U3z

6.4 .

C1 Q) 00- 0*-4 0

I-* n r= .I- ) -

Un &Q C

0 cq M

r- P a)

4r 0 (1) ;

,-4 > 0-0 '-1 (1 C.

Z ~0

4) ' 0 0CO)_* * 4-~ ) . 4

0

z c m~ r.

=4 0 (d

-3 CI2

9 4z 0~ q-a0 d 0-. k 0

0o i S q iI- gg ~0CJ4ad

V c

ga0 u

00 C.)

'0 C. C.

d 0 V.

M3 a -4 N. .v-4

(n C6~ Ci

01.

E-

'0 0

C-)

'0 0

C6-

4c. aa4..:

iC. C-D

"a' 0. 0

We % Co

C4

z co C) 00-

o PC

I-' 3-' P--4-

-200

- _. _S ,5 -S a , ... , . ,- .- . . . .. . .- .- ,-4.*, *..-.4

,

4

TABLE 9: RANK ORDER OF 20 BATTALIONS ON COHESION

AND PERFORMANCE (Correlation = 0.81).

'-a UNIT COHESION AVERAGE PERFORMANCE

2 2%96 3 6.

4 4

5 8

i 6 10

7 7

8 5

.9 3

.10 15

11 .14

12 11

13 6

14 13

15 16

16 17.4,

17 12

* .18 9

19 20

~4420 17

I -21-

'4,.

"'W'.%'',," " "" "" "" %" "" ." ." ." ," -" -; ". ",,L'. " ,,. " %L . "_% '. ,..,,.% .,, ,% ,' ". ," % * ,.'.., p ,,4."

0V 0 0

0l 0

.CL

0 0 0 0 0 00(D N~ 0 (0 N

.I.

L

I I 4. ~ . * .

o L o

.I.

0UV = -I

C.05~

CM, P.-

S~~~a HujI I-22

a..

GLOSSARY OF ABBREVIATIONS/ACRONYMS

AD6 Number of "AdministrativeDischarges" (i.e., not medical orpunitive, but prior to scheduledtermination of the term ofenlistment) in the previous sixmonths.

AGI Annual General Inspection. Theweek-long check of unitadministration and maintenanceperformed by the inspector qeneraland his team. The actual scoresused were the percentages of sub-areas passed, weighted by the teamby importance of sub-areas.

ARTEP Army Testing and Evaluation Program,a standardized, unit level, gradedfield exercise testing the unit'sability to perform its wartimemission. Scores are % of missionspassed.. Testing performed, sometimein previous 12 months.

• CO Commanding Officer.

CRIME The number of apprehensions, bylocal military police, of battalionmembers, for all crimes, during theprevious 12 months.

EM Enlisted Member. Soldiers in thelowest four pay grades.

l G/FSG First Sergeant, the highest rankingnon-commissioned officer in thecompany.

ORT Operational Readiness Test, a full.scale "alert", in which thebattalion is required to deploy toits wartime position with allequipment. A standard NATO ratinqsystem provides the scores (4 =best, 16 is worst).

PFC Private First Class.

-23-I" z-

PLD Platoon Leader, generally a Second*lieutenant.

PSG Platoon Sergeant, generally aSergeant First Class (E-7).

PT Physical Training. Score used is %of unit members passing the standardphysical fitness test administeredduring the week-long AGI.

PVT Private. Either of the two lowestpay grades.

REUP6 Reenlistments by battalion members

* in the previous six months, as a %of the battalion's assigned quota.

SLD Squad or Section Leader, generally aSergeant (E-5) or Staff Sergeant (E-6), in charge of 5 to 15 men,depending on the type of unit.

SP4 Specialist Fourth Class, a soldierin pay grade E-4, in a positiondemanding technical but notsupervisory skills.

SOT Skill Qualification Test, astandardized test of individual jobskills. All members of one companywere tested during the AGI on map

Sreading, disassembling and_* reassembling the M-16 rifle, first

aid for a leg wound, and use of the- protective mask. The battalion's

score was the % passing.

UCMJ Uniform Code of Military Justice.The law governing militarymembers. Score for battalion wasnumber of non-judicial and court-imposed punishments in the previoussix months.

-24-

V5

REFERENCES

ARCOST Action Team. Improving stability and enhancing* cohesion in the US Army; Strategy and compendium of

initiatives. Washington, D.C.: Office of theChief of Staff, Army, 1980.

Cartwright, D. The nature of group cohesiveneks. InD. Cartwright and A. Zander (Eds.), GroupDynamics. New York: Harper & Row, 1968.

DuPicq, A. Battle Studies. Harrisburg, PA: TheMilitary Service Pub. Co., 1958.

Fiedler, F. Assumed similarity measures as predictionsof team effectiveness. J. Abnormal and SocialPsychology, 49, 381-388, 1954.

French, R.L. Sociometric status and individualadjustment among naval recruits. J. Abnormal andSocial Psychology, 46, 64-71, 1951.

George, A,L. Primary groups, organization, andmilitary performance. In R.W. Little (Ed.),Handbook of Military - Institutions. BeverlyHills: Sage Publication Inc., 1971.

Goodacre, D.M. The use of a sociometric test as apredictor of combat unit effectiveness.Sociometry, 14, 148-152, 1951.

Grace, H. Conformance and performance. J. of SocialPsychology, 40, 233-237, 1954.

Hauser, W.L. A smaller Army? Adapting to the all-volunteer situation. Parameters, 9(3), 2-7, 1979

Hemphill, J.K. and Sechrest, L. A comparison of threecriteria of air crew effectiveness in combat overKorea. J. of Applied Psychology, 36, 323-327,:i 1952. -

Klein, M. and Christiansen, G. Group composition,group structure, and group effectiveness ofbasketball teams, in Loy, J.W. and Kenyon, G.S.(Eds) Sport, Culture, and Society. London:MacMillan Co., 1969.

Likert, R. New Patterns of Management. New York:McGraw-Hill, 1961.

-25-

S'.

a7F

.9

Lott, A.J. and Lott, B.E. Group cohesiveness asinterpersonal attraction. Psychological Bulletin,64, 259-309, 1965.

Marshall, S.L.A. Men Against Fire. New York: WilliamMorrow, 1966.

Melnick, M. and Chemers, M. Effects of group socialstructure on the success of basketball teams.Research Quarterly, 45, 1-8, 1974.

Palmer, F. H. and Meyers, T.I. Sociometric choices andgroup productivity among radar crews. InBibliography of Publications. Washington, D.C.:Human Resources Research Office, 1968.

Richardson, F.M. 'Fighting Spirit. London: Leo CooperLtd, 1978.

Stogdill, R. Group productivity, drive, andcohesiveness. Organizational Behavior and Human

.. Performance, 8, 26-43, 1972.

VanderVelden, L. Relationships among members, team andsituational variables and basketball success: asocial-psychoLogfcal inquiry. Unpublished doctoral

* dissbrtation, Univ. of Wisconsin, .1971.

Widmeyer, W.N. and Martens, R. When cohesiondetermines outcome in sport. The ResearchQuarterly, 49,372-380, 1978.

Zander, A. The psychology of group process. AnnualReview of Psychology, 30, 417-451, 1979.

2'.

~-26-

Division of Neuropsychiatry Report Series

Title Date

THE FUTURE BATTLEFIELD: HUMAN DIMEN- December .1982

SIONS AND IMPLICATIONS FOR DOCTRINEAND RESEARCH

DEPLOYMENT THREATS TO RAPID DEPLOY- December 1982MENT FORCES

A RESEARCH PERSPECTIVE ON DRUG AND August 1983ALCOHOL USE IN THE ARMY

ISRAELI BATTLE SHOCK CASUALTIES: August 19831973 AND 1982

AN INVESTIGATION INTO THE VALUE November 1983OF UNIT COHESION IN PEACETIME

-

,wbThese reports have been filed with the Defense Technical' Information Center. Copies are also available from:~Director, Division of Neuropsychiatry, Walter Reed Army

Institute of Research; Washington DC 20307. AVN 291-3556;., 202-576-3556.

'p.

'.."-2

tRM~ASSW** %* ~~~SECURITY CLASSIFICATION OF THIS PAGE (When, Dol emd _________________

REs" ISTRUCTIONSREPORT DOCUMENTATION PAGE SINRE COMPLETMG oEM1. NI PBEFO

V A C SI N O - RE P H .C LO i G NUOIE

4% TITLE (O ubdde) S. TYPE OF REPORT & PERIOD COVERED

An Inmestigation into the Value of Scientific Manuscriptnit Cohesio in P~eet ~~ ~S. PERFORMING ORG. REPORT NUMBER

7. AUTHOIiq) 4. CONTRACT ON GRANT NUINSE111(m)

",- Frederick J. Haming, Ph.D..Larry R. Ingraham, Ph.D.

. PEIRPORMINO ORANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT PROJECT. TASK

,alter Reed Army Institute of Research AREA & WORK UNIT HUNGERS

Walter Reed Army Medical Center 3E162777A879Washington,_D.C. 20307It. CONTROLLING O NICE NAME AND ADDRES 12. REPORT OATS

Noember 1983

!VO US AMCY IEICAL RESSUM a E~4~ CL. 12. NUMBER OF PAGES

14. MONITORING AGENCY NAMU • AODRESS(It difernl hom CeonhUlh Office) IL SECURITY CLASS. (41 hie Mr)

4IS. OECASIICATION/OOWNGRAOING

* 16. ISTR I UTIOM STATEMENT (of1 he oreplt)

" Approved for public release; distribution unlimited

17. DISTRIGUTION STATIIUNT (W the abetote atered In olek 210, It 11lEum hm Reawt)

m. q Il. SUPPLEMENTARY NOTES

It. KEY WORDS (CSt...,.mo r.vee, ..ld of a.....naad Id.t.,.b bock anmber)

cohesion, motivation, morale, military performance; military effectiveness;soldiers; interpersonal relations; leadership

. ~ IM AS1SRACT (XWd0W emooso oft Wnneeamp iWde@altr6 bpeek mumbe)-:A sample of soldiers and their leaders in 20 combat-ready US Ary battalions

asked a series of questions about interpersonal relations in their immediawork group. Their answers were scored on a scale intended to reflect grwsolidarity oc "unit cohesion" summing across individuals yielded a battaliscore, which proved to be highly correlated with a measure of overall battali

v. * performance

D * FOM"7 U3 EDITION, or Novs is oSSOLTa UrtASWIIDO AoSECURITY CLASSIFICATION OF TrHtS PAGE (When. Dateal

" "'- ,".'"""1 _: ' "Lh >." " ":".: "",'- "-, **""~ .' S%:;. 9, *," ,'" "% v, ... " - " ' ... S ''.. , '*-. .S. .-.. ,.... .. ,." "" " • - .. ,-... .,."•" -