Transcript

JOBNAME: Kluwer Journals − GR PAGE: 1 SESS: 14 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

Group Decision and Negotiation, 6:3 231–254 (1997)© 1997 Kluwer Academic Publishers

Group Development (II): Implications for GSSResearch and Practice

LAKU CHIDAMBARAMUniversity of Hawaii, College of Business Administration, 2404 Maile Way, Honolulu, Hawaii 96822, andDelft University of Technology, SEPA, Postbus 5015, 2600 GA Delft, The Netherlands

ROBERT P. BOSTROMUniversity of Georgia, College of Business Administration, Brooks Hall, Athens, Georgia 30602

Abstract

This article is the second of a two-part series that examines the issue of group development and its impact onthe study and design of group support systems. It builds on the ideas developed in the preceding article anddiscusses the implications of group development for group support systems (GSS) research. Specifically, thepaper examines the impact of computer support systems on group development processes.

Our previous paper reviewed models of development—based on two competing paradigms, sequential andnonsequential—proposed by various researchers. Despite differences between the two paradigms, individualmodels share some common aspects, from the types of processes experienced to the nature of outputs produced,that jointly help describe group development. Common processes displayed in the evolution of a group, distilledfrom the various models, include cohesiveness, conflict management, balance between socioemotional and taskneeds, effective communication, and involvement in group activities. Additionally, the various models alsoindicate that groups have critical periods in their developmental path.

This paper develops the idea that GSS structures can support groups with their developmental processes andhelp them deal with critical periods. Most GSS provide global structures like anonymity, simultaneity, and thecapacity for enhanced information processing. These structures, if appropriated effectively by groups, caninfluence their developmental paths over time. In many cases these structures can help the development ofgroups by improving their ability to manage conflict, increasing their cohesiveness, developing a sense of groupidentity, and enhancing open communication. This paper develops a set of testable propositions that can guideresearchers of group behavior, organizational users of GSS, and developers of these systems.

1. Introduction

In the previous paper, we expanded on the framework developed by Gibbard, Hartmann,and Mann (1974) to classify the models of group development into two broad categories:sequential and nonsequential. Sequential models of group development posit unitarysequences of development that groups pass through during the course of their lives. Theprimary focus of these models is to describe the actual sequence of behaviors exhibited bygroups over time. Nonsequential models propose no predetermined sequence of events;instead, they focus on explaining the underlying factors that cause shifts in group devel-opment. In contrast to the sequential models, these models shift their focus from merelydescribing the developmental pattern of groups to explaining the relationships among the

Kluwer Journal@ats-ss10/data11/kluwer/journals/group/v6n3art2 COMPOSED: 04/15/97 12:21 am. PG.POS. 1 SESSION: 7

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 2 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

various causal factors underlying group development. For more details of the variousmodels, please refer to the preceding paper in this journal.

As the previous paper illustrates, both models viewed development differently. Sequen-tial models rarely accommodated contextual variables (which often were absent in thesterile lab environments) or contingency factors (which were strictly controlled). In con-trast, however, nonsequential models explicitly recognized the impact of such factors ongroup development. Despite differences, the sequential and the nonsequential models canbe viewed as somewhat complementary. While primary intent of sequential models was todescribe the patterns of development, the intent of the nonsequential models was tounderstand the processes underlying such development. Thus, these models can be viewedas explaining similar phenomena from different perspectives.

Moreover, both models agreed about the existence of critical periods in the life of agroup. These include the initial meetings of the group, crisis points (some at the midwaypoint of the group’s life), and the final meetings of a group. The models also suggest thatgroup development can be characterized by the exhibition or existence of certain behav-iors. These include being cohesive, managing conflict effectively, balancing task andsocioemotional needs, communicating effectively, and being involved actively in groupactivities.

In this paper we suggest that a GSS, if used appropriately, may be able to assist groupdevelopment by helping members deal with critical periods and improving their interac-tion processes. We offer a set of propositions that can be tested empirically. The intent ofthis paper is to apply the lessons learned from four decades of research on group devel-opment to our more recent examination of computer supported group behavior and per-formance.

A GSS provides computerized structures for organizing group and individual thoughtprocesses and actions. GSS tools, in most situations, automate particular manual1 proce-dures for carrying out specific activities, such as a brainwriting procedure to generateideas or electronic voting to evaluate alternatives. Moreover, given the nature of computertechnology, certain structural components are “embedded” in the technology, providingaccess to additional structuring capabilities (Applegate, Konsynski, and Nunamaker1987). Examples of these embedded GSS components include structures for simultaneouscommunication, anonymous idea input, electronic recording and display, and enhancedinformation processing.

By its very design, a GSS tool establishes a particular structure (e.g., an anonymousround-robin data entry procedure for brainwriting) to which additional structure may beadded (e.g., no verbal discussion during electronic brainstorming). A group responds tothis combined collection of structures when using GSS technology. Several studies (e.g.,Gopal, Bostrom, and Chin 1992–93) have suggested that groups using a GSS can makehigher quality decisions, communicate more effectively, examine more alternatives, andbecome more cohesive, if they learn to effectively appropriate these structural components(Poole and DeSanctis 1990).

However, if the appropriation is ironic, that is, contrary to the spirit in which the GSSwas designed to be used, it can result in group regression. For instance if an embedded

232 CHIDAMBARAM AND BOSTROM

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 2 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 3 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

structure of the GSS like anonymity is ironically appropriated, “flaming” behaviors canerupt, causing greater social loafing or increased hostility. Another example is whenelectronic voting is used to cut off a discussion of alternatives instead of facilitating theirexploration. Over time the training of such a pattern can slow the group’s developmentprocesses and perhaps even cause it to regress. Occasionally, use of a GSS itself maycause structural imbalances; for instance, authors have posited that a GSS tends to in-crease task focus (DeSanctis and Gallupe, 1987), often at the expense of socioemotionalfactors. Such behavior, as suggested by the landmark work of Bales (1951), can causegroup development to falter. Thus, how a GSS is used—not that it is used—will determinethe nature and extent of group development.

Over time, as groups learn to use the GSS technology and grow more comfortable withit, they may become increasingly capable of effectively appropriating2 some of the struc-tures offered by the GSS, like its ability to provide anonymous input and its capacity forsimultaneous communication. Effective appropriation of these structures can help groupsdevelop faster into mature, high-performing teams (Chidambaram, Bostrom, and Wynne1990–91). In addition to speeding up the developmental process, GSS structures also canhelp groups with the developmental attributes identified earlier: becoming cohesive, man-aging conflict effectively, balancing task and socioemotional needs, communicating ef-fectively, and being involved actively in group tasks.

The next section discusses those GSS structures most likely to influence the develop-ment of groups; it also examines the applicability of adaptive structuration theory tostudying group development. Section 3 presents research propositions derived from theevidence in the previous section and the arguments in the first paper on group develop-ment. Finally, the paper concludes with a discussion of what group development means toGSS researchers, designers, and users.

2. Group development and GSS structures

While each GSS has different tools—and so, different structures—to support group in-teraction, they provide certain generic structures that are comparable (e.g., GroupSystemsand Visionquest). Of these structures, some can have a more significant impact on groupdevelopment than others. Among the most important GSS structures (Dennis and Gallupe1993; Nunamaker et al. 1993) that, if appropriated faithfully, can help groups develop areanonymous input and evaluation, simultaneous data entry, electronic recording and dis-play, structured interaction processes, and enhanced information processing. Table 1 iden-tifies the impact these features can have on group development. For example, simultaneityand anonymity support active and equal member participation while anonymity and elec-tronic display (especially, the public screen) promote members being critical of ideas notpersons.

GROUP DEVELOPMENT (II) 233

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 3 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 4 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

Tabl

e1.

Impa

ctof

grou

psu

ppor

tsy

stem

stru

ctur

eson

grou

pde

velo

pmen

t

Gro

upde

velo

pmen

t

fact

ors

GS

Sst

ruct

ures

Ano

nym

ity

Sim

ulta

neit

yE

lect

roni

cre

cord

ing

and

disp

lay

Pro

cess

stru

ctur

ing

Enh

ance

din

form

atio

npr

oces

sing

Tim

e: Init

ial

mee

ting

(s)

Can

help

brea

kth

eic

e;m

ayle

adto

“flam

ing”

Can

prov

ide

equa

lity

ofpa

rtic

ipat

ion

and

open

com

mu-

nica

tion

Sep

arat

esin

divi

dual

wor

ksp

ace

and

grou

psp

ace

Pro

vide

sad

diti

onal

stru

ctur

eto

init

ial

mee

ting

s

Off

ers

inte

grat

edse

tof

tool

sfo

rgr

oup

task

s;no

vice

user

sm

aybe

inti

mid

ated

Cri

sis

poin

tsC

anre

duce

“us

vs.

them

”sy

ndro

me;

redu

ces

taki

ngan

dgi

ving

cred

it

Faci

lita

tes

quic

kan

def

fici

ent

dial

ogue

onis

sues

;ca

nle

adto

info

rmat

ion

over

load

Focu

ses

mem

bers

ongr

oup

goal

san

dta

sks;

may

redu

cein

terp

erso

nal

inte

r-ac

tion

and

com

mu-

nica

tion

Pro

vide

sad

equa

test

ruct

ure;

exce

ssiv

est

ruct

ure

can

bede

trim

enta

l;us

eof

faci

lita

tor

may

bene

eded

can

acce

ssre

cord

sof

past

mee

ting

s;pr

o-vi

des

elec

tron

icre

min

ders

Fina

lm

eeti

ng(s

)M

ayre

duce

sepa

ra-

tion

anxi

ety

Can

help

deal

wit

hta

skpr

eocc

upat

ion

seen

inla

stst

ages

ofa

grou

p’s

life

(in

som

ese

quen

tial

mod

els)

May

incr

ease

task

focu

sfu

rthe

r;ca

nli

mit

form

atio

nof

soci

oem

otio

nal

ties

Pro

vide

sad

diti

onal

stru

ctur

eto

fina

lm

eeti

ngs;

exce

ssiv

est

ruct

ure

may

dis-

cour

age

free

and

open

dial

ogue

Enh

ance

sgr

oup

mem

ory

and

ca-

paci

tyto

tran

smit

grou

pno

rms

tosu

cces

sors

Beh

avio

r:C

ohes

iven

ess

Faci

lita

tes

open

ness

;m

ayre

duce

“per

-so

nal

touc

h”

Impr

oves

flow

ofco

mm

unic

atio

n;ta

skfo

cus

can

re-

duce

soci

oem

o-ti

onal

inte

ract

ion

Enh

ance

sgr

oup

own-

ersh

ipof

idea

san

dth

ereb

yse

nse

of“g

roup

ness

Red

uces

topi

cw

an-

deri

ngbu

tm

ayal

sore

duce

grou

pm

aint

enan

cefu

nc-

tion

Offl

oads

tedi

ous

task

sto

syst

eman

dal

-lo

ws

grou

pto

forg

ea

com

mon

bond

234 CHIDAMBARAM AND BOSTROM

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 4 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 5 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

Tabl

e1.

(Con

tinu

ed)

Gro

upde

velo

pmen

t

fact

ors

GS

Sst

ruct

ures

Ano

nym

ity

Sim

ulta

neit

yE

lect

roni

cre

cord

ing

and

disp

lay

Pro

cess

stru

ctur

ing

Enh

ance

din

form

atio

npr

oces

sing

Beh

avio

rC

onfl

ict

man

agem

ent

Red

uces

form

atio

nof

cliq

ues

and

sub-

grou

ps;

mak

esit

diffi

cult

tode

alw

ith

indi

vidu

alde

trac

tors

Faci

lita

tes

equa

lity

ofpa

rtic

ipat

ion

and

open

com

mun

ica-

tion

Can

keep

disc

ussi

ons

focu

sed

Pro

vide

sm

embe

rspr

oced

ural

prot

o-co

lsfo

rad

dres

sing

issu

es

Pro

vide

sm

echa

nism

sfo

rin

stan

tlyju

dg-

ing

moo

dof

grou

p;ca

nm

ini-

miz

ero

leof

per-

sona

liti

esTa

sk-e

mot

ion

bala

nce

Red

uces

pers

onal

inhi

biti

on;

enco

ur-

ages

part

icip

atio

n;su

ppor

tsso

cioe

mo-

tion

alin

tera

ctio

n

Impr

oves

task

-rel

ated

crea

tivit

y;m

ayim

pede

soci

oem

o-ti

onal

aspe

cts

Com

pute

rm

ediu

mca

nbe

used

for

both

task

-rel

ated

and

soci

oem

otio

nal

com

mun

icat

ion

(e.g

.,us

eof

“em

o-ti

ons”

)

Can

beus

ed,

wit

had

equa

tefa

cili

ta-

tion

and

trai

ning

,fo

rso

cioe

mot

iona

lac

tivit

ies

(in

addi

-ti

onto

task

-rel

ated

ones

)

Pri

mar

ilysu

ppor

tsta

sk-r

elat

edac

tivi-

ties

;hi

ghte

chm

ayre

duce

high

touc

h

Com

mun

icat

ion

Enc

oura

ges

open

and

free

exch

ange

ofid

eas;

may

caus

e“fl

amin

g”an

dso

-ci

allo

afing

Can

incr

ease

amou

ntof

com

mun

icat

ion;

may

caus

eco

gni-

tive

over

load

Pro

vide

sa

vehi

cle

for

sepa

rati

ngpe

rson

alco

mm

unic

atio

nfr

omgr

oup

com

-m

unic

atio

n

Impr

oves

task

-rel

ated

com

mun

icat

ion;

can

beus

edto

sup-

port

othe

rco

mm

u-ni

cati

onac

tivit

ies

Off

ers

ane

wap

-pr

oach

toco

mm

u-ni

cati

on;

can

bea

diffi

cult

stru

ctur

eto

appr

opri

ate

quic

kly

Invo

lvem

ent

Min

imiz

esro

leof

pers

onal

itie

s;di

ffi-

cult

togi

vecr

edit

whe

redu

e—m

aylo

wer

mor

ale

Enc

oura

ges

broa

der

part

icip

atio

n;m

ayex

clud

epo

orty

p-is

tsan

dte

chno

logy

aver

sem

embe

rs

Pro

vide

sa

cons

tant

and

visu

alre

pre-

sent

atio

nof

all

activ

itie

s;en

cour

-ag

espa

rtic

ipat

ion

Perm

its

deve

lopm

ent

ofa

com

mon

agen

daan

dse

lf-

dete

rmin

atio

nof

proc

edur

alde

tail

s

Can

help

tran

smit

grou

pgo

als

easi

ly;

prov

ides

inst

ant

acce

ssto

grou

pre

cord

s

GROUP DEVELOPMENT (II) 235

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 5 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 6 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

2.1. Anonymity

An important GSS structure that has the ability to contribute to the development of groupsis the anonymity offered by the system. It enables members of GSS groups to focus on anidea, independent of who generated it (Connolly, Jessup and Valacich 1990). Members ofsuch computer supported groups also are more likely to think of an idea or comment asthe “group’s” rather than as “his or hers,” which is in contrast to many non-GSS groups.On the other hand, in manual groups, the lack of anonymity (Connolly et al. 1990) and thefear of chastisement (Jablin and Seibold 1978) may prevent members from freely dis-cussing and evaluating ideas, thereby resulting in a lack of consensus.

Two factors may contribute to the gradual regression among traditional groups: in-creased individual inhibition over time and covert evaluation of one member’s ideas byother members (Chidambaram and Bostrom 1993). These tend to occur more in the laterstages of a group’s life, after a “pecking order” has been established. In non-GSS settings,group members frequently use verbal and nonverbal cues—looks, frowns, smiles or head-shakes—to signify their acceptance or disapproval of a certain idea. The impact of suchcontinuous, covert evaluation, which is not uncommon in many groups, is to reducecreativity and communication due to increasing fear of social disapproval. Anonymousinteraction structures can reduce such dysfunctional behavior (Connolly et al. 1990).Additionally, the protocols for blind evaluation, like anonymously ranking alternatives,offered by many GSS can also help the group development process.

However, anonymity of input and evaluation is a double-edged sword. As mentionedearlier, anonymity can result in increased social loafing and interpersonal hostility. It alsoprevents giving and taking credit. As with other GSS structures, this structure has thepotential for “ironic” appropriation (i.e., contrary to the spirit in which the structure wasconceived by the developer). With adequate process support and facilitation, such negativeappropriations may be minimized (Anson 1990).

2.2. Simultaneity

Simultaneity can assist group development in two ways: It can provide an equal platformfor all participants and it can enhance group creativity (Dennis and Gallupe 1993, Nuna-maker et al. 1993). Simultaneous input provides every team member equal access tocommunicate with the group. In traditional meetings, yielding the floor to the mostdominant member(s) often results in introverted or meek members not contributing theirideas. Equal and open access to communication channels can help foster group cohesive-ness, improve conflict management, and increase participation—all important aspects ofgroup development.

In organizations where creativity and innovation are critical for success and time islimited, a GSS can help groups by reducing production blocking and enabling members toperform parallel processing of issues (Fellers, 1989). Previous studies have shown that,despite time pressures, GSS groups generally tend to generate more alternatives andexamine more issues than traditional groups (Chidambaram et al. 1990–91). Hence, a

236 CHIDAMBARAM AND BOSTROM

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 6 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 7 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

GSS can enable groups to examine a wider range of alternatives particularly when time isof the essence.

As in the case of anonymity, simultaneity can also be a double-edged weapon. Ifappropriated ineffectively, it can cause team members to “talk” without listening to theirpeers. Over a period of time, such ineffective appropriations can cause members to loseinterest in group activities and ultimately lead to the group’s regression. Moreover, thisstructure, if left completely unrestricted, can result in information overload. Thus, simul-taneity, like other GSS structures, must be appropriated effectively to accelerate groupdevelopment.

2.3. Electronic recording and display

The difficulty of taking, keeping, and distributing the minutes of a meeting among tradi-tional groups serves to underscore the importance of this electronic structure for organi-zations. The ability of the system to capture and store data generated during meetings canhelp build a comprehensive organizational database. Researchers have suggested thatrepeated use of this GSS structure can enhance group and organizational memory (Hofferand Valacich 1993). Moreover, this structure, which provides a writing-intensive channelof communication (versus a verbal channel in traditional meetings), ensures the accuracyof the meeting record. When arguments erupt over what happened or what was discussedin any meeting, a ready on-line reference is available to all group members.

The related ability of the GSS in separating individual work space from group workspace is an important structural element that can also enhance group development. Forinstance, in many GSS implementations, the public screen supports group discussion andfocuses attention on group goals. Research shows that members’ attention during discus-sion is focused on the comments and ideas displayed on the public screen (Chidambaramand Bostrom 1993). Such a task-oriented focus, verified in other studies (e.g., Jessup andTansik, 1991), helps participants perceive electronic meetings as effective. Members alsotend to be more objective in such meetings, because entering ideas into the system andviewing them on the public screen allows them to disassociate themselves from theirideas. As groups begin to incorporate this dynamic into their regular interactions, acommon group identity starts to develop. Members, via the public screen, are continuallyreminded of their “groupness,” an essential ingredient of a well-developed group.

2.4. Process structuring

Process structures provide groups a vehicle for regulating interaction (Delbecq and Van deVen 1971). Effectively regulating group interaction is a key ingredient in managing in-tragroup conflict. Putnam (1986), in her review of group conflict research, states that oneof the key issues in effective conflict management is related to handling procedural details(e.g., what ideas to discuss first, how to set an agenda, when to follow an agenda). A GSScan offer enhanced procedural support, on an ongoing basis and especially during periods

GROUP DEVELOPMENT (II) 237

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 7 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 8 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

of intense socioemotional discussion (Miranda and Bostrom 1993). For instance, in GSSsettings, the technology often provides mechanisms for enforcing an agenda (agreed to bythe group), while in traditional meetings such enforcement is difficult and can lead tofurther conflict. Moreover, in these settings—due to the nature of the public screen and theseparation of individual and group work spaces—the attention of the group remainsfocused and usually does not wander (Dennis and Gallupe 1993).

Previous research suggests that a key reason for differences between the conflict man-agement profiles of GSS and manual groups is the formation of cliques and subgroups(Chidambaram 1989). Coalition formation is an integral part of group behavior (Putnam1986); however, such behavior, if left unchecked, can have a detrimental effect on groupdevelopment. A likely scenario as groups meet more frequently is for coalitions to becomestronger and conflict to increase. In manual groups, coalition formation may occur fasterthan in GSS groups because other members’ attitudes and personalities are easier toidentify.

In GSS groups, however, the process of coalition formation tends to be more difficultbecause of the task-oriented nature of the discussion and equality of participation(McLeod 1992). Also, the system’s anonymity helps reduce the emergence of strong,dominant personalities early in the development of the group. Hence, over time, thecoalitions in manual groups may grow stronger, making conflict management and cohe-sion building more difficult, while in GSS groups, this may not occur as easily.

Users must be cautioned that procedural structure, if appropriated ironically, can retardgroup development. Excessive structure can reduce creativity and involvement (Van deVen and Delbecq 1974, 1971). Separation of individual and group work spaces, availabil-ity of the public screen, and enforcement of a schedule are all structures that, withincreasing experience, can be exploited by dominant members to enforce their own agen-das. Hence, as with other GSS structures, care must be taken to effectively appropriateprocess structures.

2.5. Enhanced information processing

Many GSS implementations provide enhanced information processing capabilities such asinstant tabulation of voting results, group editing of complex documents, conversion ofmeeting data from one format into another, and key word searches of data (Nunamaker etal. 1993). These capabilities can enhance the development of groups. For instance, “moodmeters,” using some form of voting procedure exist in many GSS. These mood meters canbe used to judge how the group feels about certain issues; results then are tabulated anddisplayed instantly. In another example, use of group editing tools may provide all mem-bers equal opportunity to contribute to a collaborative writing task (Burke and Chidam-baram 1994). Over time, the use of such mechanisms can improve openness and com-munication among group members—essential ingredients of group development.

This structure, like other GSS structures, has the potential of being misappropriated.The complexity of tools can overwhelm users, instant voting access can restrict discus-sions, and group editing tools can force confrontations. Over time, the training in these

238 CHIDAMBARAM AND BOSTROM

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 8 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 9 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

negative patterns can slow group development or even reverse it. Adequate training andproper facilitation can be used to achieve effective appropriation of this and other struc-tures (Bostrom, Clawson, and Anson 1991).

3. Appropriating GSS structures

The main thesis of this paper is that group development can be influenced by GSSstructures, if they are appropriated by the group. The process of appropriation may eitherbe faithful (i.e., in the spirit in which its developers intended it to be used) or ironic (i.e.,in contrast to the spirit in which it was intended to be used). Over time, as groups use andreuse these structures, they are transformed in new and innovative ways by group mem-bers; such a process of transformation has been described by Poole and DeSanctis (1990)using adaptive structuration theory, often referred to as AST (reviewed in our earlierpaper).

AST focuses on the process of appropriation without prescribing a specific sequence ofevents and, thereby, explains how groups use and reuse structures over time. Therefore, itcan be viewed as an umbrella theory that unifies the various models of group develop-ment. From an AST perspective, the richer collection of structures embedded in a GSSmay be expected to result in better developed groups vis-à-vis manual groups, if thesestructures are effectively appropriated. Over time, in general, one can expect better per-formance from GSS groups, although initially they may not perform as well as theirmanual counterparts because the process of technology appropriation takes time (Burkeand Chidambaram 1994).

Introducing a GSS may not immediately improve group performance; novice groups,new to group technologies, may need repeated exposure to a GSS before they can beginexhibiting performance improvements (Chidambaram and Bostrom 1993). Some GSSstructures may take longer to appropriate than others. For instance, structures that affectdecision quality, like extended information processing capacity, will take longer forgroups to appropriate faithfully than those structures that affect group creativity, likesimultaneity. Once groups have appropriated those structures of interest to them, they thenbecome capable of performing well.

So, groups need time to become comfortable using GSS structures and incorporatingthem into their group processes. As groups begin to appropriate the structures, the natureof these structures may also evolve. Further use and reuse may modify the structures evenmore. Hence, adaptation of structures to fit group needs can be described as an ongoingand recursive cycle of events. Initially, in GSS groups, the introduction of the technologyadds an alien element to normal group interactions. Participants who have not used a GSSbefore may be apprehensive about using it for group tasks. For instance, some participantsmay initially find it constraining to use the system for brainstorming while simultaneouslycurbing their tendency to talk. To reject well-established structures (like talking) andincorporate new ones (like anonymity) takes time. And, even with time, some groups maynot appropriate all new structures or may appropriate some structures in ways unintendedby the designers.

GROUP DEVELOPMENT (II) 239

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 9 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 10 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

Effective appropriation—in addition to time, faithful use, and positive attitudes—re-quires training and facilitation. Training groups on what structures to appropriate and howto incorporate these structures into their interaction processes are key elements in helpinggroups develop. Similarly, sociotechnical facilitation also can be a key to helping groupsdevelop into mature, high-performing teams (Bostrom, Clawson, and Anson 1991). Rec-ognition of impending crises, resolution of critical impasses, and helping groups incor-porate the technology are all important aspects of facilitation that can contribute to groupdevelopment. Effective facilitation support can also help groups balance their task andsocioemotional needs over time. As we discussed in the previous paper, such a balance iscritical to the development and growth of a group.

As users give new meaning to the technology and adapt it to their own particularcontexts, they also create an environment that can be conducive to group development.With training and facilitation, this environment, over time, can be used to help speed updevelopment and support various developmental aspects such as cohesion, conflict man-agement, and communication. Research is needed to identify and study these appropria-tion processes more directly and in greater depth. Future GSS studies, which employinteraction analysis and process tracing methods, need to adopt an Input R Process ROutput model that focuses on developmental processes, in addition to outcome or perfor-mance issues.

4. Propositions

The ideas already discussed of using technology to develop cohesive, productive teamsnow are distilled into a set of testable propositions. These propositions, which can assistgroup support researchers, users, and developers, explore specifically how the structuralfeatures of a GSS influence group development.

4.1. Propositions related to critical periods

Groups need time to adapt new technologies like GSS, adopt them to fit their uniqueneeds, and become adept at using them in their regular activities. This implies a “gestationperiod” for groups using the technology before they can begin performing well. Theprocess of adaptation can be explained as a series of overlapping (and often recursive)phases that include

Initial difficulty in incorporating the technological structures.Awareness about the potential of the structures.Incorporation of the structures into group processes.

During the initial phase, groups are often apprehensive about using the GSS as a vehiclefor group communication (Burke and Chidambaram 1994). Also, the structural restraintsimposed, such as limited verbal communication during idea generation and voting, are

240 CHIDAMBARAM AND BOSTROM

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 10 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 11 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

difficult for many members to accept in the beginning. Hence, groups may have someinitial difficulty incorporating the technology into group processes.

As mentioned earlier, introduction of GSS technology into group meetings is essen-tially an imposition of an artifact that is not easily understood by the group, at leastinitially. Studies show that there is often a delay in the faithful appropriation of techno-logical structures (Gopal et al. 1992–93). When a new technology such as GSS is intro-duced in a firm, training programs need to be designed to expedite the appropriationprocess. These programs should recognize the interaction of time and GSS structures. Forinstance, anonymity may be an important structure for groups to use early in their life, butas members become familiar with each other, it may become less important for the group.

Training and team building exercises, sensitive to the developmental needs of thegroup, are vital to helping groups effectively appropriate the various electronic structuresinto group processes.

4.1.1. The beginning: Initial meetings. As discussed earlier, the initial meetings of agroup are unique (Gersick 1989), as members attempt to familiarize themselves with eachother, establish group goals, and define the purpose of the group; they also have the leaststructure and are characterized by dependence on the leader, high expectations and anxi-ety, and occasionally low member participation (e.g., Bales and Strodbeck 1951, Mann,Gibbard, and Hartman 1967).

Numerous group development studies (see our previous paper) suggest that groupsmeeting for the first time behave very differently from groups that have worked togetherbefore. However, past GSS research (e.g., Fellers 1989, Siegel, Dubrovsky, Kiesler, andMcGuire 1986, Turoff and Hiltz, 1982, Watson, 1987) has focused primarily on theone-time behavior of ad hoc groups. Ignoring the lessons from the group developmentliterature, could have produced the mixed results seen in GSS studies (Dennis, Briggs, andNunamaker 1990), as described in our earlier paper. To make research results morecomparable, the degree of development of experimental groups should be noted. A cleardistinction should also be made in the GSS literature between results obtained from usingad hoc groups in a single session and those from ongoing groups.

The group development literature suggests that the initial orientation phase (which mayextend beyond the first meeting) is critical to the long-term success of the group and setsthe tone for future actions. Lasting patterns of behavior are established during the veryfirst meeting of the group (Gersick 1989). Group members have various needs, some ofwhich may not be expressed, that should be met if productive patterns are to be established(Tuckman and Jensen 1977). Perhaps one of the most important needs during the firstphase is the social need of the members to get acquainted with each other. A GSS canfacilitate this process by “breaking the ice” and helping members introduce themselvesand highlight common interests. The anonymity offered by the GSS can reduce theinhibition of members and help create a congenial atmosphere where members can freelyexpress themselves without fear of censure.

Another characteristic of initial meetings is a lack of structure. As group membersslowly get acquainted, they are unsure of what is acceptable and consequently tend to betentative and cautious (Bales 1953). The imposition of structure by a GSS at this stage

GROUP DEVELOPMENT (II) 241

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 11 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 12 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

could enhance the group development process. The technology provides a structuredavenue for interaction where members need not be concerned about communicating con-troversial ideas or opinions early in the discussions. Hence, GSS designers, when devel-oping groupware applications, should provide users the ability to easily impose struc-ture—as through electronic agendas, straw poll mechanisms, and time-sensitive alarms—during the initial meetings of the group. Other examples of process structuring include theprovision of a set of integrated activities like issue identification, consolidation, andalternative evaluation.

Group development studies also suggest that groups may exhibit excessive reliance onthe leader for guidance during this stage. However, such reliance, if deemed unhealthy orunnecessary, can be reduced by improved GSS design. Features like agenda setting andmood meters, if incorporated into a GSS, can help reduce dependence on the leader byencouraging independence and individual participation. The role of the facilitator is criti-cal in reducing this dependency relationship and should be given considerable thought inusing GSS technologies.

We summarize this discussion of a group’s first meeting by Proposition 1.

Proposition 1. GSS structures can support the unique needs of a group’s initial meetingsby reducing inhibition, encouraging participation and providing structure.

4.1.2. Crisis points. As discussed previously, group performance and behavior changesover the life of a group. However, when such changes will occur and why they occur arenot quite clear. The punctuated equilibrium model suggests that changes in behavior occurat the midpoint of a group’s life (Gersick 1989). The growing awareness of time andrealization of the group’s impending demise are at the heart of such changes. On the otherhand, sequential models offer other instances of crisis points: downward swings in the lifecycle model, shifts from socioemotional to task-related activities in the equilibriummodel, and movement into the productive phase in the recurring cycle model.

No matter what model of group development one believes, they all identify, eitherexplicitly or implicitly, a few crisis points. To deal with such crises, groups need specialsupport. Two support features that could be used to help the transition of groups throughsuch periods are enhanced information processing support and sociotechnical facilitation.

The enhanced information processing support could help groups deal with knee-jerkreactions about the shortness of time left in a group’s life (or even in a single meeting).Such support could include continuous displays of time elapsed in a meeting, number ofmeetings remaining, record of past meetings, agendas of future meetings, and electronicreminders of priorities. Use of these enhanced information processing support featuresduring periods of high stress can help groups ease some of the tension associated withtime.

Many of these features, however, will have to be handled subtly by the facilitator. Thefacilitator has to be able to identify the onset of these crisis periods and take action to dealwith them effectively. Supporting the group with appropriate electronic tools and facili-tating the group’s transition from one period to another are vital aspects of the facilitator’srole. The facilitator should recognize that excessive use of these structures during “nor-

242 CHIDAMBARAM AND BOSTROM

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 12 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 13 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

mal” periods of development may have a detrimental effect on groups. GSS structures cansupport groups during their periods of crisis but, perhaps, be best appropriated whenaugmented with facilitation support.

In a very real sense, the facilitator’s responsibility includes identifying when a crisiswill occur and deciding what GSS structures to use during such periods. Missed crisissignals or inappropriate electronic tools can cause dysfunctional group development. Verylittle research has examined when crisis periods occur, why they occur, or what can bedone to support teams in transition through these periods. These questions are worthyareas of inquiry that future studies need to address.

We summarize this discussion by Proposition 2.

Proposition 2. GSS structures and facilitation support can help a group during crisispoints by helping to identify the onset of crisis and providing enhanced informationprocessing capacity.

4.1.3. The end: Final meetings. The final meetings represent another important temporalmilestone in the life of a group. They are the last phase of group development in whichthe behavior of the group changes. Members of some groups become preoccupied withthe issue of termination and may suffer from “separation anxiety”; also, the group be-comes capable of transmitting norms developed during its life to new members (Braaten1974). Recent studies (Gersick 1988) have shown that, during these last stages of agroup’s life, members become concerned with task completion. Moreover, for computer-supported groups, the process of appropriating the technology finally ends; if the processhas been effective, performance improvements are likely to occur (Burke 1994).

An important research objective—yet to be studied—is to determine, if computer-supported groups can deal more effectively with the final meetings of their life thantraditional groups. Few attempts have been made to measure the impact of “separationanxiety” on the behavior and performance of computer-supported groups. Both lab andfield settings could be used to examine this and other issues related to the demise of thegroup. Mills (1967) suggests that anticipating and handling the reality of group death isan important issue to those who commit themselves to the group. Hence, this research willhelp firms understand the complex issue of dismantling a group and thereby assist themin using GSS to minimize the negative impact on managers, employees, and the organi-zation.

A question of critical consequence for many corporations is the perpetuation of cor-porate norms and the maintenance of organizational continuity. This is specially importantduring times of executive reshuffling, the CEO’s demise or retirement, succession battles,and takeover attempts. During these turbulent times that reflect the “unnatural death” ofgroups, a well-designed GSS can help reduce organizational instability. A GSS withaccess to historical data about how earlier groups solved problems may help in trainingcorporate teams and building organizational memory. Moreover, to help transitionalteams, the GSS could provide access to information about the outgoing group’s activities.An important point to note here (and one made earlier) is that different phases of groupdevelopment have different sociotechnical needs. Some GSS being designed today are

GROUP DEVELOPMENT (II) 243

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 13 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 14 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

based on the assumption that a single system can serve the diverse needs of groups atdifferent stages of development. GSS designers need to note the differential needs ofgroups at different stages of development.

We conclude the discussion of a group’s final meeting by Proposition 3.

Proposition 3. GSS structures can support the unique needs of a group’s final meetingsby helping to deal with separation anxiety and enhancing organizational memory.

4.2. Propositions related to developmental processes

An important issue addressed in our previous paper was what constitutes group develop-ment. Some observers of group behavior may tend to equate the longevity of a group withits development. However, spending time together is a necessary but not a sufficientcondition for group development. In addition to working together for a period of time,certain developmental processes also need to be present.3 These include the following:

• Despite differences among group members, the group is a cohesive unit capable ofworking together.

• While task-related conflict and disagreement may exist, group members have learned tomanage and redirect it to meeting group goals.

• Group members also have learned to balance their socioemotional needs with the taskneeds of the group.

• Communication is clear and open among group members and all issues related to thegroup are discussed candidly and freely.

• Members remain involved in group activities, despite disagreements, and participateactively in discussions; they do not disassociate themselves from the group.

Based on this synthesis, we will discuss ways in which GSS structures, if effectivelyappropriated, can support the developmental processes of groups. Each of the propositionsadvocated can be tested empirically in future studies of computer-supported teams.

4.2.1. Cohesiveness. Group development implies that group members, despite the exist-ence of conflict, are cohesive, capable of productive work, and well adjusted (e.g., Mabryand Barnes 1980, Seashore 1954). Members in such groups learn to handle individualdifferences and manage conflict effectively.

A review of the literature on group cohesion and conflict management indicates areciprocal causal link between these two variables; that is, cohesive groups are more likelyto handle conflict constructively and groups that manage conflict productively are morelikely to be cohesive in the long run (For a detailed discussion, see our previous paper).Both productive conflict management and group cohesiveness have been linked to anumber of positive outcomes, including a heightened awareness of problems, a proclivityto change, enhanced motivation, increased morale, better decisions, and greater creativity.Some consequences of low group cohesiveness are stress, misallocation of resources, and

244 CHIDAMBARAM AND BOSTROM

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 14 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 15 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

diminished performance. Cohesive groups tend to work harder to achieve group goals,communicate more evenly, exert stronger pressures on members to conform to groupnorms, and have higher job satisfaction than noncohesive groups (Seashore 1954).

In a longitudinal study of decision making groups with a limited life span, researchersexamined the impact of using GSS on group cohesion (Chidambaram et al. 1990–91).They found that GSS groups became more cohesive over time when compared to non-GSS groups. Results from this study suggest that GSS structures, like anonymity andenhanced information processing, are likely to improve communication and facilitateopenness, thereby increasing group cohesiveness. Future research could verify these find-ings in a field setting with organizational groups. An alternative approach would be toconduct a controlled lab experiment, using a repeated measures design. Decision quality,process satisfaction, and member participation also could be measured repeatedly to testfor changes over time. The findings from these studies will reveal how the degree ofcohesiveness and the ability to manage conflict are related to group performance and whatimpact computer support has on these variables. Very few studies in the GSS literaturehave examined these aspects using longitudinal designs.

The important issue of how a GSS can affect group cohesiveness over time can besummarized by Proposition 4.

Proposition 4. GSS structures such as anonymity can promote openness within a groupand thereby improve cohesiveness among its members.

4.2.2. Conflict management. As described in our previous paper, most groups passthrough a period of intense conflict during certain stages of the developmental process.Some sequential models identify this as the second stage of development, while nonse-quential models are less precise about when this occurs but nevertheless attest to itsexistence. The impact of a GSS on the management of conflict is an important issue someGSS studies (e.g., Miranda 1991, Poole, Holmes, and DeSanctis 1991) have examined.Results of these studies suggest that GSS structures, depending on how they are used, caneither help or hinder groups in handling conflict. More research needs to be done onisolating the effects of a GSS on the existence and management of conflict in groups.

The existence of conflict has occasionally been mistaken for dysfunctional behavior.However, the work of Janis (1982), Putnam (1986), and Miranda (1991) have indicatedthat lack of conflict is not necessarily a desirable outcome in groups. Well-developedgroups, in fact, often exhibit conflict; but, this conflict is a positive expression of honestdifferences in opinion and positions expressed by group members (Miranda and Bostrom1993). Occasionally, however, negative conflict aimed at personalities and hostility di-rected at individuals may surface even in highly developed groups.

Negative conflict has been associated with low member satisfaction (Putnam 1986). Theincreased hostility among group members and the high levels of tension account for theirdissatisfaction with group processes. However, with GSS groups, it has been shown that,over time, greater acceptance of conflicting opinions (due to the anonymity of input) andincreased participation (due to the structured format) can help members deal effectivelywith conflict (Chidambaram et al. 1990–91). Therefore, computer-supported groups can

GROUP DEVELOPMENT (II) 245

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 15 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 16 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

(and often do) experience an increase in certain types of conflict (Miranda and Bostrom1993). However, over time, their ability to manage conflict effectively increases (Burke1994). AST may provide some insights to this apparent paradox: As groups wrestle withincorporating the structures offered by a GSS, conflict increases; however, as they suc-cessfully adapt these structures, like anonymity and electronic evaluation, to fit theirneeds, they learn to manage conflict better.

Existence of negative conflict also has been associated with the establishment of ahierarchy and the formation of subgroups or cliques (Bion 1961, Mann et al. 1967).Frequently, negative conflict is directed by some subgroups against other subgroups orcliques. Reasons for hostility toward other subgroups include the drive to revolt againstauthority, conflicting needs of members, need to maintain individuality, and desire to beappeased. Such hostility against subgroups tends to increase over time as hierarchies areestablished and coalitions are strengthened (Chidambaram 1989).

In GSS settings, however, the establishment of a hierarchy based on the social standingof group members becomes more difficult. It can be hypothesized that the anonymityafforded by the GSS will minimize the formation of cliques and subgroups as memberswith similar views cannot be easily identified. This is a testable hypothesis and can bestudied in lab and field settings. Using confederates to provoke “revolt” against authorityand using content analysis to understand the patterns of conflict over time could yieldvaluable results about conflict management in computer supported groups.

We synthesize our views about the impact of a GSS on conflict management by Propo-sition 5.

Proposition 5. GSS structures can help manage conflict and reduce interpersonal hostilityby redirecting the attention of group members to issues instead of personalities.

4.2.3. Balance between socioemotional and task needs. Group development is based onachieving and maintaining a balance between the socioemotional and task-related needs ofgroup members (Bales 1953, Bales and Strodbeck 1951). This concept of “equilibrium”dominated the work of Bales and Strodbeck and was the predominant theory of groupdevelopment for almost two decades. While some GSS studies have analyzed the inter-actions among group members on the basis of task-related and non-task-related comments(see McLeod 1992), the focus of these studies was not the issue of equilibrium. However,results from these and other studies suggest that GSS structures tend to reduce non-task-related (presumably, socioemotional) comments. Further research is needed to preciselyidentify the existence and nature of this need for equilibrium among computer-supportedand traditional groups.

Many GSS—like GroupSystems, Visionquest, and Meeting Ware—provide members avariety of tools that support both task-related and socioemotional interactions. Also, theglobal structures available with a GSS, like anonymity and simultaneity, depending on thecontext, can support either one or both interactions. For instance, a task-related maladythat frequently afflicts the creativity of traditional groups is production blocking. Suchgroups, by their very structure, restrict the number of members who can express theirviews simultaneously; in most cases, this means that only one person can talk at any given

246 CHIDAMBARAM AND BOSTROM

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 16 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 17 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

time. Such a restriction, however, limits the number of ideas that can be exchanged duringa meeting. As discussed earlier, a GSS can help participants exchange ideas simulta-neously, thereby reducing production blocking and increasing group creativity. Over time,as groups learn to effectively appropriate this structure offered by the GSS, their task-related performance is likely to improve.

In contrast, anonymity, another global GSS structure, can be used to improve socioe-motional interaction. Evaluation apprehension, personal inhibition, fear of social disap-proval, and the presence of authority figures often prevent members of groups fromexpressing their true feelings. However, a GSS provides members the ability to anony-mously express their views without recrimination. Over time, the ability to freely expressopinions will help strengthen the socioemotional bonds that well-developed groups pos-sess. Other structures, like mood meters, separation of group and individual work space,e-mail facilities, and announcement features, can also enhance socioemotional interac-tions.

Therefore, the GSS offers structures, if appropriated effectively, can support both thetask-oriented and socioemotional needs of a group. This dual support in turn can help thegroup maintain a balance between both needs and accelerate group development. Thisidea can be summarized by Proposition 6.

Proposition 6. GSS structures, depending on the context, can jointly support a group’ssocioemotional and task needs and thereby help maintain equilibrium.

4.2.4. Communication. Effective communication is a prerequisite for group development;the level of communication among members of mature groups is high and its nature open(Luthans and Larsen 1986, McGrath 1984). As members learn to work with each other,they begin to understand communication signals that normally would be missed by out-siders to the group. Also, prior research (as described in our previous paper) has shownthat well-developed groups are not apprehensive about discussing any issue relevant toachieving group goals. Hence, there is a sense of openness in exchanging ideas amonghighly developed groups.

A few GSS studies have examined the impact of using computer-support on intragroupcommunication (e.g., Siegel et al. 1986). The results of these studies indicate that use ofa GSS is related to an increase in task-related comments. What is less clear, however, iswhether this pattern increases, decreases, or remains stable over time. Content analysissuch as Bales’s interaction process analysis can be used to examine the amount and typeof communication that occurs within groups interacting for long periods of time. Resultsfrom these studies would be of particular relevance in determining the extent of commu-nication support that a GSS should offer at the various stages of group development.

As described in our previous paper, newly formed groups generally communicatedifferently than older, more mature groups (also see, for instance, Bales 1953, Bales andStrodbeck 1951). Much of the communication in the early phases of development is aimedat building acquaintances, becoming familiar with other members’ views, and gettingoriented with the goals and purposes of the group. The communication during later phasesincludes giving and receiving opinions, expressing agreement or disagreement, and show-

GROUP DEVELOPMENT (II) 247

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 17 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 18 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

ing affection and solidarity. Groups at earlier phases may require different types of GSSsupport, like anonymity and system prompting to solicit members’ views, while moremature groups may require less intrusive, yet more robust types of GSS support, likeenhanced information processing. Different types of communication facilities, like accessto external communication links and online databases, also may be required at differentphases of group development.

Media choice has been a critical concern for observers of organizational communica-tion (Fulk and Steinfield 1990). In this context, recognizing when a face-to-face GSSversus a dispersed GSS should be used is not a trivial task. Groups at early stages ofdevelopment may require a face-to-face system that allows them to get oriented with eachother, develop trust, and understand task roles. However, members of well-developed,mature groups need not be subjected to the constraints of a same-place/same-time system.With increasing levels of communication and understanding, members of such groups canparticipate in dispersed meetings via distributed systems. While these are plausible as-sertions, they need to be tested by comparing longitudinally the communication patternsof groups using GSS in face-to-face and dispersed settings. We conclude our expectationsof computer-supported group communication with Proposition 7.

Proposition 7. The differential communication needs of a group at different stages ofdevelopment can be supported by different types of GSS media4 and GSS structures.

4.2.5. Involvement. Finally, the last aspect of group development is the ability of itsmembers to stay involved with the group’s tasks and goals despite differences of opinion.As discussed in our first paper, several authors have suggested that commitment to groupgoals and involvement in group tasks are fundamental to group development (e.g., Kor-manski and Mozenter 1987, Shaw 1981). Recent studies of GSS groups have confirmedthat involvement and cooperation, in the face of conflict and competition, are essential togroup development (Burke 1994).

GSS structures like process structuring, enhanced organizational memory, and anonym-ity are critical to members’ maintaining a continued interest in group activities. Processstructures permit the development of a common agenda and the self-determination ofprocedural details, which are critical for increasing involvement and commitment. Mem-bers who feel they can influence what goes on in a group are less likely to becomeapathetic. Anonymity is another GSS structure that can improve participation, at leastinitially. It minimizes the role of personalities and fosters an environment of trust; mem-bers can express differences of opinion without fear of reprisal or personal hostility.Hence, this structure, if appropriated by the group, can help members share their viewsopenly and remain involved with group tasks and goals.

Another GSS structure, enhanced organizational memory, also can improve involve-ment and commitment. Group memberships often change over the life of the group.Getting new members involved in group activities often can be a difficult and time-consuming process. The electronic memory provided by a GSS can serve as an ongoingrecord of group tasks. Training new members using this structure—on what the group hasachieved, how decisions were made and what members contributed—can spur enthusiasm

248 CHIDAMBARAM AND BOSTROM

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 18 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 19 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

and generate interest. Also, access to meeting records at any time and any place can helpkeep members involved in group activities, even when they are occasionally unable toattend meetings. We conclude our expectations with Proposition 8.

Proposition 8. GSS structures, by offering an open avenue for dialogue and a record ofpast accomplishments, can help improve members’ involvement in group activities.

5. Discussion of implications

The need for networked teams among today’s organizations is paramount. The “virtualcorporation” of tomorrow (“The Virtual Corporation” 1993) will rely on such teams,which may be formed and disbanded at will. To help these teams develop quickly intoproductive groups, GSS technology can be used as a catalyst. The discussion that followsexplores the implications of using GSS technology to help with group development. Italso examines the implications of using this technology for researchers of group behavior.

5.1. Implications for practice

Firms cannot simply install a GSS and hope to reap the benefits of the new technologyovernight. The “gestation period” for groups working with a GSS can last at least a fewmeetings before they become proficient in appropriating the necessary structures neededto develop into a cohesive team (Walther 1992). Depending on the type of technologyused and the task performed, even longer periods may be necessary before organizationalteams begin making real performance improvements. Such delays, if not anticipated andplanned for, indeed could result in significant organizational costs. Training, facilitation,practice, and advocacy are few of the tools available to firms that want to use thistechnology for helping groups develop into mature, productive teams. Whatever the spe-cific mechanism adopted, organizations hoping to reap the purported advantages of aGSS, need to pay careful consideration to helping groups quickly appropriate such tech-nologies.

One cautionary note is in order: AST suggests that groups, depending on their uniquecontext, may appropriate GSS structures ironically (i.e., contrary to the spirit in whichthey were intended to be used by the designers). In such an event, groups may eitherdevelop slower than normal or even regress over time. However, problems of ironicappropriation can be handled by using skilled facilitator services and developing well-tailored user support programs. If not explicitly recognized and dealt with, problemsarising from misappropriation of technological structures may have long-term negativeconsequences for team building and group development.

If appropriated effectively, various GSS structures serve three important functions re-lated to group development: they can be used to move groups faster and more efficientlyto the productive phases of development; minimize some of the detrimental aspects ofgroup development like hostility toward authority, increased negative conflict, and the

GROUP DEVELOPMENT (II) 249

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 19 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 20 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

formation of subgroups and cliques; and foster positive aspects of group development likethe perpetuation of group norms, facilitation of organizational continuity, and the inclu-sion of new members to a group.

With the growing diversity of GSS products and services, organizational teams have thevery real prospect of using computer support not only to facilitate traditional groupactivities like decision making but also to manage the process of group development. GSSdesigners, in addition to other design factors (see, for instance, Huber 1984, Kraemer andKing 1986), should recognize that groups at different stages of development have differentrequirements. For instance, groups at later phases of development are likely to requiremore sophisticated features, while groups at earlier phases will require greater structure tofacilitate group communication and decision making. Moreover, dispersed GSS systemsmay be used more effectively during later stages of group development than in earlierstages. Understanding the group development process is critical to developing a GSS thatmeets the real needs of users. A GSS should be designed to satisfy not only the initialneeds of users but also to support the ongoing processes related to group development.

5.2. Implications for research

An important implication for GSS researchers is that group behavior and performancemay differ considerably over the life of the group. Thus, while little difference might existbetween traditional and GSS groups at the start of their lives, their profiles of developmentover time may vary dramatically. This suggests that GSS researchers must be cautious ininterpreting the behavior of ad hoc groups meeting for the first (and in some cases theonly) time. To have a more complete understanding of the impact of a GSS on groupbehavior and performance, longitudinal studies should be undertaken. The cost in terms oftime and resources is much greater for repeated measures designs, but the payoff isobvious. Recent GSS research has started to recognize the importance of longitudinaldesigns in studying group behavior (e.g., Burke and Chidambaram 1994, Chidambaramand Bostrom 1993, Gopal et al. 1992–93).

Researchers also need to be aware that the same group is capable of making high- orlow-quality decisions, taking more or less time to reach consensus, and being satisfied ordissatisfied with its performance based on its stage of development. If single session labexperiments are used to study groups, then the level of the groups’ development must bemeasured and reported to make results comparable across studies.

As discussed in this and our earlier paper, group history and individual expectationsinfluence the behavior of established groups. On the other hand, the behavior of ad hocgroups, which have never worked together before and will never work together again (afteran experiment is over), are influenced only by the immediate incentives offered. Thesebehavioral differences suggest that researchers should be cautious when extrapolating theresults from ad hoc groups to established groups. The eight propositions developed in thispaper are a valuable starting point for future research aimed at examining the existence,magnitude, and causes of differences between ad hoc and established groups.

250 CHIDAMBARAM AND BOSTROM

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 20 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 21 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

These propositions should be tested with real groups in organizations over severalmonths, or perhaps years, of GSS use to help determine whether behavioral changes andperformance gains (or losses) are temporary or long-lasting phenomena. How groupsactually appropriate specific technologies can best be studied by analyzing interactionprocesses among group members. These processes can be captured on video- or audio-tapes and analyzed using one of several protocols developed by group behavioral re-searchers (e.g., see Jessup 1987). Such protocol analysis will yield valuable insights intovarious aspects of how groups develop; for example, how group members interact overtime with each other and with the technology, what impact training has on this interactionprocess, how mature groups appropriate technologies compared to naive ones, and howrepeated exposure to different technological structures affect conflict.

The role of time within and between GSS sessions, referred to in the propositions, is anunexplored area worthy of attention. The social entrainment model discussed in our earlierpaper has found support for interactions between time limits and task type, not only fortask performance but also for interaction processes. Hence, a GSS can affect the timing ofcommunication activities and thereby influence the synchronization (or entrainment) pro-cesses in groups. In other words, the GSS can be viewed as an external pacer (like astopwatch) that produces temporally influenced behaviors (like following an agenda).Future GSS research needs to draw upon this emerging research on the social psychologyof time to explore the interaction effects of task, technology and time.

Integrating time-based theories with adaptive structuration theory (AST) could alsohelp us better understand computer-supported group developmental processes. While theresearch literature has demonstrated independently the usefulness of the concepts ofappropriation (from AST) and entrainment (from time-based theories) for explainingvariations in group behavior, it has not tested both concepts jointly. This combined viewwill help answer questions such as what causes changes in synchronization patterns overtime and how entrainment mechanisms are altered repeatedly.

6. Conclusion

Group development requires time. However, a group’s chronological life span alonecannot determine the extent of its development. In other words, not all groups that haveworked together for a long period of time can be considered well developed. As mentionedpreviously, spending time together is a necessary but not a sufficient condition for groupdevelopment. While time plays a critical role in the development of groups, this papersuggested that technology can also play an equally critical role.

Specifically, in this article—the second of two—we described five GSS structures thatwere expected to influence group development. We proposed that, over time, groups thateffectively appropriated these structures would exhibit patterns of performance and be-havior different from those groups that either did not have computer support or failed toappropriate these structures effectively. We developed a set of eight propositions thathypothesized how these structures could help groups develop into cohesive, high-performing teams. In conclusion, the two articles in this series explored the interaction of

GROUP DEVELOPMENT (II) 251

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 21 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 22 SESS: 12 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

time, technology, and teams. The issues raised by these papers can help researchers,designers, and users of GSS understand and support the continually changing needs ofwork groups.

Notes

1. The term manual is used in this paper to denote traditional, non-computer supported meeting environments.2. For a description of what constitutes effective appropriation, please see our previous paper, “Group Devel-

opment (1): A Review and Synthesis of Development,” which appeared in issue 2 of this volume.3. For a detailed account, see our previous paper.4. Such as face-to-face GSS in the early stages and dispersed GSS in later stages.

References

Anson, R. G. (1990). “Effects of Computer Support and Facilitation Support on Group Processes.” Unpublisheddoctoral dissertation, Indiana University.

Applegate, L. M., B. R. Konsynski, and J. F. Nunamaker, Jr. (1987). “A Group Decision Support System for IdeaGeneration and Issue Analysis in Organization Planning,” Journal of Management Information Systems 3,5–19.

Bales, R. F. (1953). “The Equilibrium Problem in Small Groups,” in T. Parsons, R. F. Bales, and E. A. Shils(eds.), Working Papers in the Theory of Action. Glencoe, IL: The Free Press.

Bales, R. F., and F. L. Strodbeck. (1951). “Phases in Group Problem Solving,” Journal of Abnormal and SocialPsychology 46, 485–495.

Bion, W. R. (1961). Experiences in Groups, 2nd ed. New York: Basic Books.Bostrom, R. P., V. K. Clawson, and R. Anson. (1991). “Training People to Facilitate Electronic Environments.”

University of Georgia Working Paper.Braaten, L. J. (1974). “Developmental Phases of Encounter Groups: A Critical Review of Models and a New

Proposal,” Interpersonal Development 75, 112–129.Burke, K. (1994). “Developmental Processes and Performance in Groups Using an Electronic Meeting System:

A Comparison of Face-to-Face, Synchronous and Asynchronous Meeting Environments.” Unpublished doc-toral dissertation, University of Hawaii.

Burke, K., and L. Chidambaram. (1994). “Developmental Differences between Distributed and Face-to-FaceGroups in Electronically Supported Meeting Environments: An Exploratory Investigation,” Group Decisionand Negotiation.

Chidambaram, L. (1989). “An Empirical Investigation of the Impact of Computer Support on Group Develop-ment and Decision Making Performance.” Unpublished doctoral disseration, Indiana University.

Chidambaram, L., and R. P. Bostrom. (1995). “Group Development (I): A Review and Synthesis of DevelopmentModels.”

Chidambaram, L., and R. P. Bostrom. (1993). “Evolution of Group Performance over Time: A Repeated Mea-sures Study of GDSS Effects,” Journal of Organizational Computing 3(4), 443–470.

Chidambaram, L., R. P. Bostrom, and B. E. Wynne. (1990–91). “A Study of the Impact of Group DecisionSupport Systems on Group Development,” Journal of Management Information Systems 7(3), 7–25.

Connolly, T., L. M. Jessup, and J. S. Valacich. (1990). “Effects of Anonymity and Evaluative Tone on IdeaGeneration in Computer-Mediated Groups,” Management Science 36(6), 689–703.

Delbecq, A. L., and A. H. Van de Ven. (1971). “A Group Process Model for Problem Identification and ProgramPlanning,” Journal of Applied Behavioral Science 7(4), 466–492.

Dennis, A. R., R. O. Briggs, and J. F. Nunamaker, (1990). “A Comparison of Ad Hoc and Established Groupsin an Electronic Meeting System Environment,” Proceedings of the Twenty-third Annual Hawaii InternationalConference on System Sciences 3, 23–29.

252 CHIDAMBARAM AND BOSTROM

Kluwer Journal@ats-ss3/data11/kluwer/journals/group/v6n3art2 COMPOSED: 04/24/97 4:03 pm. PG.POS. 22 SESSION: 10

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 23 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

Dennis, A. R., and R. B. Gallupe. (1993). “A History of Group Support Systems Empirical Research: LessonsLearned and Future Directions,” in L. M. Jessup and J. S. Valacich (eds.), Group Support Systems: NewPerspectives, pp. 59–77. New York: Macmillan.

DeSanctis, G., and R. B. Gallupe. (1987). “A Foundation for the Study of Group Decision Support System,”Management Science 33(5), 589–609.

Fellers, J. (1989). “The Effect of Group Size and Computer Support on Group Idea Generation for CreativityTasks: An Experimental Evaluation Using a Repeated Measures Design.” Unpublished doctoral dissertation,Indiana University.

Fisher, B. A. (1970). “Decision Emergence: Phases in Group Decision Making,” Speech Monographs 37, 53–56.Fulk, J., and C. Steinfield. (1990). Organizations and Communication Technology. Newbury Park, CA: Sage

Publications.Gersick, C. J. G. (1988). “Time and Transition in Work Teams: Toward a New Model of Group Development,”

Academy of Management Journal 31(1), 9–41.Gersick, C. J. G. (1989). “Marking Time: Predictable Transitions in Task Groups,” Academy of Management

Journal 32(2), 274–309.Gopal, A., R. P. Bostrom, and W. W. Chin. (1992–93). “Applying Adaptive Structuration Theory to Investigate

the Process of Group Support System Use,” Journal of Management Information Systems 9(3), 45–69.Hoffer, J. A., and J. S. Valacich. (1993). “Group Memory in Group Support Systems: A Foundation for Design,”

in L. M. Jessup and J. S. Valacich (eds.), Group Support Systems: New Perspectives, pp. 214–229. New York:Macmillan.

Huber, G. P. (1984). “Issues in the Design of Group Decision Support Systems,” MIS Quarterly 8, 195–204.Jablin, F. M., and D. R. Seibold. (1978). “Implications for Problem-Solving Groups of Empirical Research on

‘Brainstorming’: A Critical Review of the Literature,” The Southern Speech Communication Journal 43,327–356.

Janis, I. Victims of Groupthink, (1982). 2nd ed. Boston: Houghton Mifflin.Jessup, L. M. (1987). “Group Decision Support System: A Need for Behavioral Research,” International Journal

of Small Group Research 3(2), 139–158.Jessup, L. M., and D. A. Tansik. (1991). “Decision Making in an Automated Environment: The Effects of

Anonymity and Proximity on Group Process and Outcome with a Group Decision Support System,” DecisionSciences 22(1), 266–279.

Kayser, T. A. (1990). Mining Group Gold: How to Cash in on the Collaborative Brain Power of a Group. ElSegundo, CA: Serif Publishing.

Kormanski, C., and A. Mozenter. (1987). “A New Model of Team Building: A Technology for Today andTomorrow,” Developing Human Resources.

Kraemer, K. L. and J. King. (1986). “Computer-Based Systems for Group Decision Support: Status of Use andProblems in Development,” Proceedings of the Conference on Computer Supported Cooperative Work, Austin,TX, 353–375.

Luthans, F., and J. K. Larsen. (1986). “How Managers Really Communicate,” Human Relations 39(2), 161–178.Mabry, E. A., and R. E. Barnes. (1980). The Dynamics of Small Group Communication. Englewood Cliffs, NJ:

Prentice-Hall.Mann, R. D., G. S. Gibbard, and J. J. Hartman. (1967). Interpersonal Styles and Group Development. New York:

Wiley.McGrath, J. E. (1984). Groups: Interaction and Performance. Englewood Cliffs, NJ: Prentice-Hall.McLeod, P. L. (1992). “An Assessment of the Experimental Literature on Electronic Support of Group Work:

Results of a Meta-Analysis,” Human-Computer Interaction 7, 257–280.Mills, T. M. (1967). The Sociology of Small Groups. Englewood Cliffs, NJ: Prentice-Hall.Miranda, S. M. (1991). “Cohesiveness and Conflict Management in Group Decision Support Systems.” Unpub-

lished doctoral dissertation, University of Georgia.Miranda, S. M., and R. P. Bostrom. (1993). “The Impact of Group Support Systems on Group Conflict and

Conflict Management,” Proceedings of the Twenty-sixth Annual Hawaii International Conference on SystemSciences 4, 83–94.

GROUP DEVELOPMENT (II) 253

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 23 SESSION: 5

5% 50% 90% 100%

JOBNAME: Kluwer Journals − GR PAGE: 24 SESS: 5 OUTPUT: Mon May 12 08:55:25 1997/data11/kluwer/journals/group/v6n3art2

Nunamaker, J. F., A. R. Dennis, J. S. Valacich, D. R. Vogel, and J. F. George. (1993). “Group Support SystemsResearch: Experience from the Lab and Field,” in L. M. Jessup and J. S. Valacich (eds.), Group SupportSystems: New Perspectives, pp. 125–145. New York: Macmillan.

Poole, M. S., and G. DeSanctis. (1990). “Understanding the Use of Group Decision Support Systems: TheTheory of Adaptive Structuration,” in J. Fulk and C. Steinfield (eds.), Organizations and CommunicationTechnology, pp. 173–193. Newbury Park, CA: Sage Publications.

Poole, M. S., M. Holmes, and G. DeSanctis. (1991). “Conflict Management in a Computer-Supported MeetingEnvironment,” Management Science 37(8).

Putnam, L. L. (1986). “Conflict in Group Decision Making,” in R. Y. Hirokawa and M. S. Poole (eds.),Communication and Decision Making. Beverly Hills, CA: Sage Publications.

Seashore, S. E. (1954). Group Cohesiveness in the Industrial Work Group. Ann Arbor: University of MichiganPress.

Shaw, M. E. (1981). Group Dynamics: The Psychology of Small Group Behavior, 3rd ed. New York: McGraw-Hill.

Siegel, J., V. Dubrovsky, S. Kiesler, and T. W. McGuire. (1986). “Group Processes in Computer MediatedCommunication,” Organizational Behavior and Human Decision Processes 37, 157–187.

Tuckman, B. W., and M. A. C. Jensen. (1977). “Stages in Small Group Development Revisited,” Group andOrganizational Studies 2, 419–427.

Turoff, M., and S. R. Hiltz. (1982). “Computer Support Group versus Individual Decisions,” IEEE Transactionson Communications 30(1), 82–90.

Van de Ven, A. H., and A. L. Delbecq. (1974). “Nominal Versus Interacting Group Processes for CommitteeDecision-Making Effectiveness,” Academy of Management Journal 14(2), 203–212.

Van de Ven, A. H., and A. L. Delbecq. (1979). “The Effectiveness of Nominal, Delphi and Interacting GroupDecision Making Process,” Academy of Management Journal 17(4), 605–621.

Business Week “The Virtual Corporation,” (1993). no. 3304 (February 8), 98–102.Walther, J. B. (1992). “A Longitudinal Experiment on Relational Tone in Computer-Mediated and Face-to-Face

Interaction,” Proceedings of the Twenty-fifth Annual Hawaii International Conference on System Sciences 4,220–231.

Watson, R. T. (1987). “A Study of Group Decision Support System Use in Three and Four Person Groups for aPreference Allocation Decision.” Unpublished doctoral dissertation, University of Minnesota.

254 CHIDAMBARAM AND BOSTROM

Kluwer Journal@ats-ss5/data11/kluwer/journals/group/v6n3art2 COMPOSED: 02/27/97 9:18 am. PG.POS. 24 SESSION: 5

5% 50% 90% 100%