Adaptive Structuration Theory Poole and Desanctis

Capturing the Complexity in AdvancedTechnology Use: Adaptive

Structuration Theory

Gerardine DeSanctis • Marshall Scott PooleCarlson School of Management, Information and Decision Sciences Department,University of Minnesota, 271 \9th Avenue South, Minneapolis, Minnesota 55455

Department of Speech Communication, University of Minnesota,271 I9th Avenue South, Minneapolis, Minnesota 55455

DeSanctis and Poole contribute to the organization sciences in two distinct ways. First, theyinsightfully probe and characterize the deep structures that exist within both the technological

artifacts and the work environments within which these artifacts are applied (within the context of agiven technology—group decision support systems). Second, they describe and illustrate innovativestrategies for collecting data on these structures. In doing so, the authors have laid an extremely strongfoundation for future scholarship exploring the "evolution-in-use" as well as the organizationalimpacts of advanced information technologies.

Robert W. Zmud

AbstractThe past decade has brought advanced information technolo-gies, which include electronic messaging systems, executiveinformation systems, collaborative systems, group decisionsupport systems, and other technologies that use sophisti-cated information management to enable multiparty partici-pation in organization activities. Developers and users ofthese systems hold high hopes for their potential to changeorganizations for the better, but actual changes often do notoccur, or occur inconsistently. We propose adaptive struc-turation theory (AST) as a viable approach for studying therole of advanced information technologies in organizationchange. AST examines the change process from two vantagepoints: (1) the types of structures that are provided by ad-vanced technologies, and (2) the structures that actuallyemerge in human action as people interact with these tech-nologies. To illustrate the principles of AST, we consider thesmall group meeting and the use of a group decision supportsystem (GDSS). A GDSS is an interesting technology forstudy because it can be structured in a myriad of ways, andsocial interaction unfolds as the GDSS is used. Both thestructure of the technology and the emergent structure ofsocial action can be studied.

We begin by positioning AST among competing theoreti-cal perspectives of technology and change. Next, we describethe theoretical roots and scope of the theory as it is appliedto GDSS use and state the essential assumptions, concepts,and propositions of AST. We outline an analytic strategy for

applying AST principles and provide an illustration of howour analytic approach can shed light on the impacts ofadvanced technologies on organizations. A major strength ofAST is that it expounds the nature of social structures withinadvanced information technologies and the key interactionprocesses that figure in their use. By capturing these pro-cesses and tracing their impacts, we can reveal the complexityof technology-organization relationships. We can attain abetter understanding of how to implement technologies, andwe may also be able to develop improved designs or educa-tional programs that promote productive adaptations.(Information Technology; Structural Theory; Technol-ogy Impacts)

1.0. IntroductionInformation plays a distinctly social, interpersonal rolein organizations (Feldman and March 1981). Perhapsfor this reason, development and evaluation of tech-nologies to support the exchange of information amongorganizational members has become a research tradi-tion within the organization and information sciences(Goodman 1986, Keen and Scott Morton 1978, Van deVen and Delbecq 1974). The past decade has broughtadvanced information technologies, which include elec-tronic messaging systems, executive infonnation sys-

1047-7039/94/0502/0121/S0I.25CopyrJBht © 1994. The Institute of Manasement Sciences ORGANIZATION SCIENCE/VO1. 5, No. 2, May 1994 121

GERARDINE DESANCnS AND MARSHALL SCOTT POOLE Adaptive Structuralion Theory

tems, collaborative systems, group decision supportsystems, and other technologies that enable multipartyparticipation in organizational activities through so-phisticated information management (Huber 1990,Huseman and Miles 1988, Rice 1984). Developers andusers of these systems hold high hopes for their poten-tial to change traditional organizational design, intelli-gence, and decision-making for the better, but whatchanges do these systems actually bring to the work-place? What technology impacts should we anticipate,and how can we interpret the changes that we observe?

Many researchers believe that the effects of ad-vanced technologies are less a function of the technolo-gies themselves than of how they are used by people.For this reason, actual behavior in the context ofadvanced technologies frequently differs from the "in-tended" impacts (Kiesler 1986, Markus and Robey1988, Siegel, Dubrovsky, Kiesler and McGuire 1986).People adapt systems to their particular work needs, orthey resist them or fail to use them at all; and there arewide variances in the patterns of computer use and,consequently, their effects on decision making andother outcomes. We propose adaptive structuration the-ory (AST) as a framework for studying variations inorganization change that occur as advanced technolo-gies arc used. The central concepts of AST, structura-tion (Bourdieu 1978, Giddens 1979) and appropriation(Oilman 1971), provide a dynamic picture of the pro-cess by which people incorporate advanced technolo-gies into their work practices. According to AST, adap-tation of technology structures by organizational actorsis a key factor in organizational change. There is a"duality" of structure (Orlikowski 1992) whereby thereis an interplay between the types of structures that areinherent to advanced technologies (and, hence, antici-pated by designers and sponsors) and the structuresthat emerge in human action as people interact withthese technologies.

As a setting for our theoretical exposition, we con-sider the small group using a group decision supportsystem (GDSS). A GDSS is one type of advancedinfonnation technology; it combines computing, com-munication, and decision support capabilities to aid ingroup idea generation, planning, problem solving, andchoice making. In a typical configuration, a GDSSprovides a computer terminal and keyboard to eachparticipant in a meeting so that infonnation (e.g., facts,ideas, comments, votes) can be readily entered andretrieved; specialized software provides decision struc-tures for aggregating, sorting, and otherwise managingthe meeting information (Dennis et al. 1988, DeSanctisand Gallupe 1987, Huber 1984). A GDSS is an inter-

esting technology for study because its features can bearranged in a myriad of ways and social interaction isintimately involved in GDSS use. Consequently, thestructure of the technology and the emergent structureof social action are in prominent view for the re-searcher to study. There currently is burgeoning inter-est in GDSSs and their potential role in facilitatingorganizational change. GDSS is a rich context in whichto expound AST, but the principles of the theory applyto the broad array of advanced information technolo-gies.

In this paper we outline the assumptions of AST anddetail a methodological strategy for studying how ad-vanced technologies such as GDSSs are brought intosocial interaction to effect behavioral change. We beginby positioning AST among an array of theoreticalperspectives on technology and change. Next, we de-scribe the theoretical roots and scope of the theory andstate the essential assumptions and concepts of AST.We summarize the relationships among the theoreticalconstructs in the form of propositions; the propositionscan serve as the basis for specification of variables andhypotheses in future research. Finally, we outline amethod for identifying structuring moves and presentan illustration of the theory's application. Together,the theory and method provide an approach for pene-trating the surface of advanced technology use to con-sider the deep structure of technology-induced organi-zational change.

2.0. Theoretical Roots of AST2.1. Competing Views of Advanced Information

Technology EffectsTwo major schools of thought have pursued the studyof information technology and organizational change(see Table 1). The decision-making school has beenmore dominant. This school is rooted in the positivisttradition of research and presumes that decision mak-ing is "the primordial organizational act" (Perrow1986); it emphasizes the cognitive processes associatedwith rational decision making and adopts a psychologi-cal approach to the study of technology and change.Decision theorists espouse "systems rationalism" (Rice1984), the view that technology should consist of struc-tures (e.g., data and decision models) designed to over-come human weaknesses (e.g., "bounded rationality"and "process losses"). Once applied, the technologyshould bring productivity, efficiency, and satisfaction toindividuals and organizations. Variants within the deci-sion school include "task-technology fit" models(Jarvenpaa 1989), which stress that technology mustmatch work tasks in order to bring improvements in

122 ORGANIZATION SCTENCE/VOI. 5, No. 2, May 1994

GERARDINE DESANCTIS AND MARSHALL SCOTT POOLE Adaptive Structuration Theory

Table 1 Adaptive Structuration Theory Blends Perspectives from the Decision-making Schooland the Institutional Schooi

Major Perspectives onTechnology andOrganizational Change

Characteristics of

Each Perspective Examples of Theoretical Approaches

Decision-making School

Social Technology School{integrative perspectives)

Institutional School

focus on technology engineering

hard-line determinismreiatively static models of behavior

positivist approach to research

ideographic, cross-sectional research designs

focus on technology and social stnjcture

soft-line determinismmixed models of behaviorpositivist and Interpretive

approaches are integrated

focus on social structurenondeterministic modeispure process modelsinterpretive approach to researchnomothetic, longitudinal

decision theory (Keen and Scott Morton 1978)task-technology "fit" {Jarvenpaa 1989)"garbage can" models (Pinfieid 1986)

sociotechnicat systems theory (Bostrom

and Heinen 1977. Pasmore 1988)structural symbolic interaction theory

(Saunders and Jones 1990, Trevino et al. 1987)Barley's (1990) application o( structuration theoryOrlikowski's (1992) structurational modeladaptive structuration theory

segmented institutional (Kling 1980)social information processing (Fulk et al. 1987.

Salancik and Pfeffer 1978, Walther 1992)symbolic interactionism (Blumer 1969, Reichers 1987)structuration theory (Giddens 1979) research designs

work effectiveness, and so-called "garbage can" models(Pinfieid 1986), which emphasize the timing of eventsand the need for technology to support informationscanning and information search activities.

Decision theorists tend toward an engineering viewof organizational change, believing that failure toachieve desired change reflects a failure in the technol-ogy, its implementation, or its delivery to the organiza-tion. Research hypotheses are grounded in eitherhard-line determinism, the belief that certain effectsinevitably follow from the introduction of technology,or more moderate contingency views, which argue thatsituational factors interact with technology to causeoutcomes (see Gutek, Bikson and Mankin 1984). Deci-sion theorists favor positivist research approaches thatmeasure—typically in quantitative terms—the effectsof technology manipulation on outcomes (Orlikowskiand Baroudi 1991).

Within the GDSS literature, technology designguidelines put forth by Dennis et al. (1988), DeSanctisand Gallupe (1987), and Huber (1984), and experimen-tal studies conducted by Jarvenpaa, Rao, and Huber(1988), Watson, DeSanctis and Poole (1988), and oth-ers (Connolly, Jessup and Valacich 1990, Gallupe,DeSanctis and Dickson 1988, George et al. 1990) ex-emplify the decision school perspective. This line of

research evaluates the effectiveness of GDSS technol-ogy by comparing groups given GDSS support withthose given manual or no decision structuring, or bycomparing groups given certain types of GDSS struc-tures with those given alternative designs of structures.In general, researchers expect GDSS conditions toyield more desirable outcomes than groups in otherconditions.

The decision school has yielded an extensive litera-ture on GDSSs and other advanced technologies, butthe approach has not produced a consensus on howthese systems should be designed or on how they affectthe people and organizations who use them.' For ex-ample, some researchers report that GDSS use im-proves group consensus and decision quality, whereasothers report the reverse (see George et al. 1990).Similarly, a number of studies have found differencesin attitudes or patterns of use of the same technologydesign across groups (e.g., Hiltz and Johnson 1990,Kerr and Hiltz 1982). Recently, decision researchershave tried to sort out GDSS impacts by isolating spe-cific features or properties of the technology for study.For example, Connolly et al. (1990) manipulatedanonymity and the evaluative tone of electronicallycommunicated comments and measured effects on ideageneration, solution quality, and satisfaction. Others

ORGANIZATION SCIENCE/VOI. 5, No. 2, May 1994 123


have considered the degree of "social presence" of theGDSS media (Hiltz and Johnson 1990); but these ap-proaches have led to mixed results as well, with valueson outcome measures begin improved in some casesand worsened in others (Jessup, Connolly and Galegher1990).

There is no doubt that technology properties andcontextual contingencies can play critical roles in theoutcomes of advanced information technology use. Thedifficulty is that there are not clearcut patterns indicat-ing that some technology properties or contingenciesconsistently lead to either positive or negative out-comes. Observed effects do not hold up robustly acrossstudies, and, even more disturbing, there is often sub-stantial variance in outcome measures within even onetreatment of any given study (e.g., Jarvenpaa et al.1988). To achieve greater consistency in empiricalfindings, decision school researchers advocate progres-sively finer, feature-at-a-time evaluation of technologyand more complex contingency classifications schemes(e.g., see Pinsonneault and Kraemer 1989, Valacich,Dennis and Nunamaker 1992). The diflRculty is, ofcourse, the repeating decomposition problem: thereare features within features (e.g., options within soft-ware options) and contingencies within contingencies(e.g., tasks within tasks). So how far must the analysisgo to bring consistent, meaningful results?

Researchers within the institutional school advocatea different approach: the study of technology as anopportunity for change, rather than as a causal agentof change (Barley and Tolbert 1988, Kling 1980,Perrow 1986). The focus of study for institutionalists isless on the structures within technology, and more onthe social evolution of structures within human institu-tions. Institutionalists criticize,decision theorists for the"technocentric" assumption that technology containsinherent power to shape human cognition and behav-ior; this assumption, they contest, leads to"gadgetphilia," an overemphasis on hardware and soft-ware and an underemphasis on the social practices thattechnologies involve (Finlay 1987, Markus and Robey1988). A strategic choice model is advocated Instead:technology does not determine behavior; rather, peo-ple generate social constructions of technology usingresources, interpretive schemes, and norms embeddedin the larger institutional context (Orlikowski 1992).Many institutionalist emphasize the role of ongoingdiscourse in generating social constructions of technol-ogy (e.g.. Barley and Tolbert 1988, Scott 1987), with aconsequent emphasis on human interaction (rather thantechnology per se) in studies of advanced technologyeffects.

Institutionalists began with the study of communitiesand society as a whole (Gidens 1979, Selznick 1969),but institutionai theory has been developed for organi-zations as well (Kling 1980). Theoretical perspectivesaligned with the institutional school in the study oforganizations include social information, processingtheory, which emphasizes the social construction ofmeaning (Fulk et ai. 1987, Salancik and Pfcffer 1978,Walther 1992); and symbolic interactionism, which fo-cuses on the role of communication in the creation andpreservation of the social order, i.e., roles, norms,values, and other social practices (Reichers 1987). Forinstitutionalists, the creation, design, and use of ad-vanced technologies are inextricably bound up with theform and direction of the social order. It follows thatstudies of technology and organizational change mustfocus on interaction and capture historical processes associal practices evolve. Process-oriented methods arefavored over outcome studies, and ideographic, inter-pretive accounts are preferred over nomothetic re-search designs (Barley and Tolberl 1988). Within theinstitutional school, technology is considered to beinterpretively flexible (Orlikowski 1992), and so analy-sis is the process of looking beneath the obvious sur-face of technology's role in organizational change touncover the layers of meaning brought to technology bysocial systems.

There is growing interest in institutional analyses ofadvanced information technologies, including GDSSs,though actual accounts are sparse (Barley 1986, Finlay1987, Markus and Forman 1989, Robey, Vaverek andSaunders 1989, Walther 1992)..These analyses describethe interplay between technology and power distribu-tion, politics, stratification, and other social processes.Institutional accounts of organizational change are in-herently less interested in the properties of technologythan in use of technology and the evolution of socialpractices. Consequently, the purely institutional ap-proach underplays the role of technology in organiza-tional change. A more complete view would accountfor the power of social practices without ignoring thepotency of advanced technologies for shaping interac-tion and thus bringing about organizational change.Such a view woutd integrate assumptions from thedecision-making and institutional schools and applyboth positivist and interpretive research approaches.^

2.2. An Integrative PerspectiveHow might the decision and institutional perspectivesbe integrated? Several theoretical views synthesize as-sumptions from these competing schools to form whatwe will refer to as the social technology perspective.

124 ORGANIZATION SCIENCE/VOI. 5, No. 2, May 1994


This third school of thought advocates "soft-line" de-terminism, or the view that technology has structures inits own right but that social practices moderate theireffects on behavior (Guetk et al. 1984). For example,sociotechnical systems theory argues that the impactsof advanced infonnation technologies depend on howwell social and technology structures are jointly opti-mized; technology adoption is interpreted as a processof organizational change (Bostrom and Heinen 1977,Hiltz and Johnson 1990, Pasmore 1988). Similarly,structuration theory, largely associated with Giddens'institutional theory of social evolution (1979), has beenapplied to explain organizational adoption of comput-ing and other technologies (Barley 1986, 1990, Or-likowski 1992, Orlikowski and Robey 1991, Robey et al.1989).

A third social technology mode!, structural symbolicinteraction theory, takes a more "micro" view, examin-ing interpersonal interaction that occurs via electronicand other new media (Saunders and Jones 1990,Trevino, Lengel and Daft 1987). The theory exploresthe inherent structure of technology more fully thanstructurational models, but it has been applied more tothe study of peoples' perceptions of technology than totheir actual behavior. Also, the theory does not explainthe dynamic way in which technology and social struc-tures mutually shape one another over time.

Adaptive structuration theory extends current struc-turation models of technology-triggered change to con-sider the mutual influence of technology and socialprocesses. AST provides a detailed account of both thestructure of advanced technologies as well as the un-folding of social interaction as these technologies areused. Its goai is to confront "structuring's central para-dox: identical technologies can occasion similar dynam-ics and yet lead to different structural outcomes"(Barley 1986, p. 105). To present the theoretical propo-sitions of AST, we focus here on small group interac-tion in the context of GDSS technology, but the con-cepts and relationships posited here could be appliedto other advanced technologies and other organiza-tional contexts. We consider both the structures ofGDSS technology and the structures realized in inter-action, but we particularly attend to the latter in thisexposition. We leave more in-depth analyses of GDSSand related advanced information technology struc-tures to other discussions (DeSanctis, Snyder and Poolein press, Huber, 1990, Silver 1991). The theoreticalpropositions presented here can be refined to formu-late specific research hypotheses, thus providing anempirical research agenda (e.g., see DeSanctis et al.1989, 1992, in press, Poole and DeSanctis 1992, Poole,

Holmes and DeSanctis 1991, Sambamurthy and Poole1992).

3.0. Propositions of AdaptiveStructuration Theory

AST provides a model that describes the interplaybetween advanced information technologies, socialstructures, and human interaction. Consistent withstructuration theory, AST focuses on social structures,rules and resources provided by technologies and insti-tutions as the basis for human activity. Social structuresserve as templates for planning and accomplishingtasks. Prior to development of an advanced technology,structures are found in institutions such as reportinghierarchies, organizational knowledge, and standardoperating procedures. Designers incorporate some ofthese structures into the technology; the structures maybe reproduced so as to mimic their nontechnologycounterparts, or they may be modified, enhanced, orcombined with manual procedures, thus creating newstructures within the technology. Once complete, thetechnology presents an array of social structures forpossible use in interpersonal interaction, including rules(e.g., voting procedures) and resources (e.g., storeddata, public display screens). As these structures thenare brought into interaction, they are instantiated insocial life. So, there are structures in technology, onthe one hand, and structures in action, on the other.The two are continually intertwined; there is a recur-sive relationship between technology and action, eachiteratively shaping the other. But if we are to under-stand precisely how technology structures can triggerorganizational change, then we have to uncover thecomplexity of the technology-action relationship. Thisrequires an analytical distinction between social struc-tures within technology and social structures withinaction (Giddens 1979, Orlikowski 1992, Orlikowski andRobey 1991). Then the interplay between the two typesof structures must be considered.

3.1. Advanced Information Technologies as SocialStructures

Advanced information technologies bring social struc-tures which enable and constrain interaction to theworkplace. Whereas traditional computer systemssupport accomplishment of business transactionsand discrete work tasks, such as billing, inventorymanagement, financial analysis, and report prepara-tion, advanced information technologies support theseactivities and more: they support coordination amongpeople and provide procedures for accomplishing in-

ORGANIZATION SaENCE/VoI. 5, No. 2, May 1994 125


terpersonal exchange. GDSSs, for example, provideelectronic paths for exchanging ideas among meetingparticipants and formulas for integrating the work ofmultiple parties. In this sense, advanced infonnationtechnologies have greater potential than traditionalbusiness computer systems to influence the social as-pects of work.

The social structures provided by an advanced infor-mation technology can be described in two ways: thestructural features of the given technology andthe spirit of this feature set. Structural features are thespecific types of rules and resources, or capabilities,offered by the system. Features within a GDSS, forexample, might include anonymous recording of ideas,periodic pooling of comments, or alternative votingalgorithms for making group choices. They govern ex-actly how information can be gathered, manipulated,and otherwise managed by users. In this way, featuresbring meaning (what Giddens calls "signification") andcontrol ("domination") to group interaction (seeOrlikowski and Robey 1991). A given advanced infor-mation technology can be described and studied interms of the specific structural features that its designoffers, but most systems are really "sets of looselybundled capabilities and can be implemented in manydifferent ways" (Gutek et al. 1984, p. 234). This varietyof possible implementations differentiates advanced in-formation technologies from their more traditionalcounterparts and is a driving force behind the need fornew research approaches, such as AST. Because of themany possible combinations of features, a parsimo-nious approach is to scale technologies among a mean-ingful set of dimensions that reflect their social struc-tures. Numerous dimensions for describing advancedtechnology structures have been proposed. For exam-ple. Silver (1991) characterizes decision support sys-tems in terms of their relative restrictiveness. The morerestrictive the technology, the more limited is the set ofpossible actions the user can take; the less restrictivethe technology, the more open is the set of possibleactions for applying the structural features. Advancedinformation technologies might also be described interms of their level of sophistication. For example,DeSanctis and Gallupe (1987) have identified threegeneral levels of GDSS: Level 1 systems provide com-munication support; level 2 systems provide decisionmodeling; and level 3 systems provide rule-writing ca-pability so that groups can develop and apply highlyspecific procedures for interaction. Finally, Abualsamh,Cariin and McDaniel (1990) and Cats-Baril and Huber(1987) characterize systems based on their degree ofcomprehensiveness, or the richness of their structural

feature set. The more comprehensive the system, thegreater the number and variety of features offered tousers. Scaling structural feature sets in terms of restric-tiveness, level of sophistication, comprehensiveness, orother dimensions, can be accomplished by consultinguser manuals, reviewing the statements of designers ormarketers of the technology, or noting the commentsof people who use the technology.

The social structures of an advanced informationtechnology also can be described in terms of their spirit(Poole and DeSanctis 1990). Spirit is the general intentwith regard to values and goals underlying a given setof structural features. Webster defines spirit as the"general intent" of something, as in "spirit of the law,"and we construe the spirit of a technology in the samesense. The spirit is the "official line" which the tech-nology presents to people regarding how to act whenusing the system, how to interpret its features, and howto fill in gaps in procedure which are not explicitlyspecified. The spirit of a technology provides whatGiddens calls "legitimation" to the technology by sup-plying a normative frame with regard to behaviors thatare appropriate in the context of the technology. It alsocan function as a means of signification, because ithelps users understand and interpret the meaning ofthe technology. Spirit can also contribute to processesof domination, because it presents the types of influ-ence moves to be used with the technology; this mayprivilege some users or approaches over others.

Spirit is a property of the technology as it is pre-sented to users. It is not the designers' intentions—these are reflected in the spirit, but it is impossible towholly realize their intents. Nor is the spirit of thetechnology the user's perceptions or interpretations ofit—these give us indications of the spirit but are likelyto capture only limited aspects. Spirit can be identifiedby treating the technology as a "text" and developing areading of its philosophy based on analysis of: (a) thedesign metaphor underlying the system (e.g., "elec-tronic chalkboard"); (b) the features it incorporatesand how they are named and presented; (c) the natureof the user interface; (d) training materials and on-lineguidance facilities; and (e) other training or help pro-vided with the system. Usually the best person to makethis reading is the researcher, who is able to consultwith designers, investigate the structure of the soft-ware, analyze training materials, study manners of im-plementation, consider a range of typical user interpre-tations, and triangulate among these sources of evi-dence. The researcher should consider the interpreta-tions of the spirit by users and designers insofar asthese can be used to crosscheck conclusions drawn


( ; E K A R D I N E DESANCTIS AND MARSHALL SCOTT POOLE Adaptive Structuration Theory

from analysis of artifacts. It is important to considermultiple sources of evidence to yield an interpretationof the spirit. No one source should be consideredprivileged.

The use of multiple sources of evidence lays openthe possibilities of contradictions; when these occur itsuggests that the system in question does not present acoherent spirit. For example, some technologies maypresent a clear, consistent spirit, whereas others maynot. The spirit is thus a variable for differentiatingadvanced information technologies. A coherent spiritwould be expected to channel technology use in defi-nite directions. An incoherent spirit would be expectedto exert weaker influence on user behavior. An inco-herent spirit might also send contradictory signals,making use of the system more difficult.

The nature of the spirit of technology can be furtherilluminated by exploring the analogy to legal gover-nance. Government institutions provide systems of lawthat can be described both in terms of their letters(e.g., statutes), which detail specific rules and resourcesfor social action, and their spirit, which is the historicalconsensus about values and goals that are appropriate(or legitimate) in society. At any given point in time,people may apply the letter of the law in ways that areconsistent or inconsistent with the spirit of the law. Inother words, spirit has the potential to be violated evenas the letter of the law is further developed or invoked.Whereas the letter of the law—like the features of atechnology—can be described in relatively objectiveterms, spirit is more open to competing interpretations.Early on, when a technology is new, the spirit of atechnology is in flux; spirit is put forth by the designersand is evident in their pronouncements (e.g., throughmanuals or marketing literature) about the values andgoals of the system and how it "should" be used.Organizations that subsequently adopt the technologyfurther contribute to the definition of the spirit (e.g.,through management pronouncements about the pur-poses of the system or through training programs).Once the technology is stable in its development andused in routine ways, the definition of spirit becomesmore stable; the spirit is less open to conflicting inter-pretations. For purposes of structural analysis, spiritcan be treated as the status quo, the researcher'scurrent interpretive account (based on multiplessources of evidence) regarding the values and goals ofthe technology.

When considering spirit we are more concerned withquestions like, "What kind of goals are being pro-moted by technology?" or "What kind of values arebeing supported?" than we are with questions like

Table 2 Example Dimensions for Characterizing the "Spirit"of an Advanced Information Technology'sSocial Structures

Dimension Description (reference)

Decision Process the type of decision process thai is beingpromoted; for example, consensus, empiri-cal, rational, political, or individualistic(Rohrbaugh 1989)

Leadership the likelihood of leadership emerging when

the technology is used; whether a leader ismore likely or less likely to emerge, orwhether there will be equai participationversus domination by some members(Huber 1984)

Efficiency the emphasis on time compression,whether the interaction periods will beshorter or longer than interactions wherethe technology is not used {DeSanctis andGallupe 1987)

Conflict Management whether interactions will be orderly orchaotic, lead to shifts in viewpoints or not,or emphasize conflict awareness or conflictresolution (Dennis et al. 1988)

Atmosphere the relative formality or informal nature ofinteraction, whether the interaction is struc-tured or unstructured (Dennis et al. 1988,Mantei 1988)

"What does the system look like?" or "What modulesdoes it contain?" Table 2 gives possible dimensions forcharacterizing the spirit of advanced information tech-nologies, particularly GDSSs. For example, a GDSSmay have a definable spirit with regard to the type ofdecision process that is promoted in a group; a certainstyle of leadership might be promoted by the system; orthe value of efficiency might be emphasized. DeSanctiset al. (in press-b) provide a method for scaling thestructural features and spirit of a GDSS based on bothdesigner and user perspectives.

Together, the spirit and structural feature sets of anadvanced information technology form its structuralpotential, which groups^ can draw on to generate par-ticular social structures in interaction. For example, arestrictive, level 2 GDSS with a spirit of high formalismand efficiency might be expected to promote ;i parsi-monious, step-by-step, data-oriented approach to groupdecision making. Group members might be expected tostick closely to the agenda and procedures provided bythe GDSS, with little room to diverge from the pre-scribed approach or to invoke decision structures otherthan those embedded in the GDSS. On the other hand,


GERARDINE DESANCTIS AND MARSHALL SCOTT POOLE Adaptive Slnicturation Theory

a less restrictive, level 1 system with an informal spiritmight lead to a looser application of the GDSS struc-tures to the decision process, with a relaxed atmo-sphere and a mixture of GDSS and other structuresappearing in the group's interaction. In sum, we pro-pose the following with regard to advanced informationtechnologies (AITs):

PI. AITs provide social structures that can be de-scribed in terms of their features and spirit. To the extentthat AITs vary in their spirit and structural features sets,different forms of social interaction are encouraged bythe technology.

3.2. Other Sources of StructureAdvanced information technologies are but one sourceof structure for groups. The content and constraints ofa given work task are another major source of struc-ture (McGrath 1984, Poole, Seibold and McPhee 1985).For example, if alternative projects are being priori-tized for budgeting purposes, then information aboutthese projects and standard organizational proceduresfor computing budgets are important resources andrules for participants as they undertake the prioritiza-tion task. Similarly, the organizational environmentprovides structures. For example, current pressures toreduce spending or circumstances that favor certainprojects over others may be brought into interaction asparticipants confront a budgeting task. Corporate in-formation, histories of task accomplishment, culturalbeliefs, modes of conduct, and so on, all provide struc-tures that groups can invoke, in addition to the ad-vanced information technology.

The structures provided by a technology may be useddirectly, but more likely they are invoked in combina-tion with other structures. The array of alternativestructures available to groups can affect which technol-ogy structures are selected for use, how the results areinterpreted, and how they are applied. AST is consis-tent with contingency theories in proposing that use ofadvanced information technologies may vary acrosscontexts:

P2. Use of AIT structures may vary depending on thetask, the environment, and other contingencies that offeralternative sources of social structures.

So the major sources of structure for groups as theyinteract with an advanced information technology are:the technology itself, the tasks, and the organizationalenvironment (see Table 3). As these structures areapplied, their outputs become additional sources ofstructure. For example, after the group enters data intothe GDSS, the information generated by the system

becomes another source of social structures. Similarly,information generated by applying task knowledge orenvironmental knowledge constitutes a source of socialstructures. In this sense, there are emergent sources ofrules and resources upon which people can draw associal action unfolds.

P3. New sources of structure emerge as the technol-ogy, task, and environmental structures are applied dur-ing the course of social interaction.

33. GDSSs in ActionThe act of bringing the rules and resources from anadvanced information technology or other structuralsource into action is termed structuration. Structura-tion is the process by which social structures (whatevertheir source) are produced and reproduced in sociallife. For example, suppose that a GDSS provides brain-storming and notetaking techniques (level 1 features,with tow comprehensiveness) which are highly 6exiblein their application (low restrictiveness) and that thesefeatures are preesented as promoting a spirit of effi-ciency and democratic participation. Structuration oc-curs when a group applies the brainstorming and note-taking techniques to their meeting, or strives for aspirit of efficiency or democracy.

When the social structures of the advanced informa-tion technology are brought into action, they may takeon new forms. That is, interpersonal interaction mayreflect rules and resources that are modified from theadvanced information technology. For example, when agroup uses voting rules built into a GDSS, it is employ-ing the rules to act, but—more than this—it is remind-ing itself that these rules exist, working out a way ofusing the rules, perhaps creating a special version ofthem. In short, the group is producing and reproducingthe GDSS rules for present and future use. Use andreuse of technology structures or emergent forms oftechnology structures lead, over time, to their institu-tionalization. When the technology structures becomeshared, enduring sets of cognitive scripts then thestructural potential of the GDSS has brought aboutorganizational change. Technology-triggered organiza-tional change thus takes time to occur, as technologystructures are produced and reproduced in interaction.

For analytic purposes, we can capture the structura-tion process by isolating a group's application of aspecific technology-based rule or resource within aspecific context and at a specific point in time. We willcall the immediate, visible actions that evidence deeperstructuration processes appropriations of the technol-ogy (Oilman 1971). By examining appropriations, we



Table 3 Major Sources of Structure and Examples of Each

Structure Source Definition Examples in GDSS Context

AIT (A) advanced information technoiogy including liardware,software, and procedures

AIT outputs data, text, or other results produced by the AIT(AO) software (ollowing input by group membersTask (T) task knowledge or rules; includes facts and figures,

opinion, folklore, or practice related to the task at hand

Task outputs the results of operating on task data or procedures:(TO) the results of completing all or parts of a taskEnvironment social knowledge or rules of action drawn from the(E) organization or society at large

Environmental the results of applying knowledge or rules drawnoutputs (EO) from the environment

keyboard input devices, viewing screens,group notetaking, voting modules, decision modelsdisplays of group votes, lists of ideas, opiniongraphs, modeling resultsa budget task, customary ways of preparingbudgets, specific budget data, budgeting goalsand deadlinesbudget calculations; the implications of certainbudget figures for other budget categoriesapplying a "spread the wealth" principleto budget allocation; applying a "majority rule"decision procedure to votes; reference tocorporate spending and reporting policiesimplications of corporate spending policies forthe budget process; the results and implicationsof applying a "majority rule" decision procedureto votes that have been taken

can uncover exactly how a given rule or resource withina GDSS, for example, is brought into action. Appropri-ation of GDSS structures is evidenced as a groupmakes judgments about whether to use or not usecertain structures, directly uses (reproduces) a GDSSstructure, relates or blends a GDSS structure withanother structure, or interprets the operation or mean-ing of a GDSS structure. GDSS structures becomestabilized in group interaction if the group appropri-ates them in a consistent way, reproducing them insimilar form over time. In the same vein, the groupmay intentionally or unintentionally change GDSSstructural features as it uses them; reproduction doesnot necessarily imply replication. For example, a groupwith a strict hierarchy of authority might blend thevoting module of an otherwise egalitarian-orientedGDSS with a structure of leader-directed choice. Theleader might state his or her position and then directothers to vote in its favor. Consequently, the votingfeature of the GDSS, when brought into action, ischanged from a mechanism for equal input to a mecha-nism for reinforcing leader directives.

In sum, the social structures available within ad-vanced information technologies provide occasions forthe structuring of action. As technology structures areapplied in group interaction, they are produced andreproduced. Over time, new forms of social structuremay emerge in interaction; these represent reproduc-tions of technology structures, or blendings of technol-

ogy-based with other structures (e.g., task and environ-ment). Once emergent structures are used and ac-cepted, they may become institutions in their own rightand the change is fixed in the organization.

P4. New social structures emerge in group interac-tion as the rules and resources of an AIT are appropri-ated in a given context and then reproduced in groupinteraction over time.

Appropriation and decision making proce.sses. Ap-propriations are not automatically determined by tech-nology designs. Rather, people actively select howtechnology structures are used, and adoption practicesvary. Groups actively choose structural features fromamong a large set of potentials. At least four aspects ofappropriation can be identified that illustrate variationin interaction processes. (In §4.1 we outline an ap-proach for analyzing these appropriation processes.)First, groups may choose to appropriate a given struc-tural feature in different ways, invoking one or more ofmany possible appropriation moves. Given the availabil-ity of technology structures, groups may choose to: (a)directly use the structures; (b) relate the structures toother structures (such as structures in the task orenvironment); (c) constraint or interpret the structuresas they are used; or (d) make judgments about thestructures (such as to affirm or negate their usefulness).Second, groups may choose to appropriate technology

ORGANIZATTON SCIENCE/VOI. 5, No. 2, May 1994 129


features faithfully or unfaithfully. The features aredesigned to promote the technology's spirit, but theyare functionally independent and may be appropriatedin ways that are not faithful to the spirit. Faithfulappropriations are consistent with the spirit and struc-tural feature design, whereas unfaithful appropriationsare not. Unfaithful appropriations are not "bad" or"improper" but simply out of line with the spirit of thetechnology. Third, group members may choose to ap-propriate the features for different instrumental uses,or purposes. For example, the group might use aGDSS to accomplish task activities, manage communi-cation and other group processes, or to exercise poweror influence (DeSanctis et al. 1992). The appropriationconcept includes the intended purposes, or meaning,that groups assign to technology as they use it. Byidentifying instrumental uses we can begin to under-stand not only what structures are being used and howthey are being used, but also why they are being used—the reasons or purposes for which groups elect tobring technology or other structures into action. Afourth aspect of appropriation is the attitudes thegroup displays as technology structures are appropri-ated, such as:(a) the extent to which groups areconfident and relaxed in their use of the technology(comfort); (b) the extent to which groups perceive thetechnology to be of value to them in their work (re-spect); and (c) their willingness to work hard and excelat using the system (challenge) (Billingsley 1989;Sambamurthy 1990; Zigurs, DeSanctis and Billingsley1990). These attitudes set the tone for applications ofthe technology and, in some measure, whether thegroup pursues its applications with sufficient vigor andconfidence to carry them ofT. Sambamurthy (1990)found that these three attitudes significantly influencedthe number of premises considered by planning groupsconducting a stakeholder analysis using a GDSS.

Appropriation processes may be subtle and difficultto observe, but they are evidenced in the interactionthat makes up group decision processes; appropriationsare, in essence, the "deep structure" of group decisionmaking. How group members appropriate structuresfrom technology or other sources will influence thedecision processes that unfold.

Decision theorists argue that advanced informationtechnologies, particularly GDSSs, are designed to over-come common diificulties, or "process losses," associ-ated with group interaction. The assumption is that useof GDSS features, such as input and exchange of ideas,computation and display of group member opinions,and quantitative decision models, will improve the pro-cesses and outcomes of group decision making

(DeSanctis and Gallupe 1987, Huber 1984). Decisionprocess improvements include, for example, expandedidea generation (Nunamaker, Applegate and Konsynski1988), more even participation by members in express-ing their opinions (Dennis et al. 1988), more effectiveconflict management behavior (Poole et al. 1991), moreeven influence by participants on the ultimate choicesmade by the group (Zigurs, Poole and DeSanctis 1988),and greater focus on the task relative to social con-cerns (McLeod and Liker 1989). Improvements in thesedecision processes are expected to lead to desirableoutcomes, such as efficient identification of choices(Nunamaker, Vogel and Konsynski 1989), accuratechoices or high quality solutions (Bui and Sivasankaran1990), high group consensus (Watson et al. 1988), andstrong commitment to implementing the group deci-sion (Dennis et al. 1988). To the extent that appropria-tions of technology structures vary over time or acrossgroups, decision processes and outcomes will vary aswell. Desired decision processes and outcomes are notguaranteed.

P5. Group decision processes will vary depending onthe nature of AIT appropriations.

Factors influencing the appropriation of structures.Although appropriation processes may not always beconscious or deliberate (Barley 1990), groups makeactive choices in how technology or other structuresare used in their deliberations. A given structure maybe appropriated quite differently depending on thegroup's internal system, which is the nature of mem-bers and their relationships inside the group (seeHomans 1950). Factors that might influence how agroup appropriates available structures include:

• Members' style of interacting. For example, an auto-cratic leader may introduce and use technology struc-tures very differently than a democratic leader(DeSanctis et al. in press-c; Hittz, Turoff and Johnson1981). Other stylistic differences, such as differences ingroup conflict management styles, may also influenceappropriation processes (Poole et al. 1991).

• Members' degree of knowledge and experiencewith the structures embedded in the technology. Forexample, understanding of possible pitfalls and prat-falls in the structures may contribute to more skillfuluse by certain members (DeSanctis et al. 1992, Pooleet al. 1991).

• The degree to which members believe that othermembers know and accept the use of the structures.The better known the structure is, the less membersmay deviate from the typical form of use (Vician et al.


GERARDINE DESANCTIS AND MARSHALL SCOTT POOLE Adaptive Slmcturation Theory

1992). This is consistent with the notion of "criticalmass" whereby the perceived value of a technologyshifts as it spreads rapidly through a community; lateradopters are influenced hy the values and behaviors ofearlier adopters and vice versa (Markus 1990).

• The degree to which members agree on whichstructures should be appropriated. There may be un-certainty about which structures are most appropriatefor thc given situation or power struggles over whichstructural features should be used. Greater agreementon appropriation of structures should lead to moreconsistency in the group's usage patterns (Poole, De-Sanctis, Kirsch and Jackson 1991).

These assumptions imply the following proposition:

P6. The nature of ALT appropriations will vary de-pending on the group's internal system.

Appropriation and decision making outcomes. Themodel presented in Figure 1, which summarizes therelationships discussed in this section, has importantimplications for the study of AIT effects on organiza-tional change. A major implication of PI through P6 isthat clearcut predictions about how AIT structures willbe appropriated, or what the ultimate outcomes of thatappropriation will be, are difficult to formulate. Thestructural features of the technology, along with thetask, the organizational environment, and the group'sinternal system, act as opportunities and constraints inwhich appropriation occurs. In general, we would ex-pect desired decision processes to be more likely toresult when appropriation patterns take on the follow-ing properties: (a) appropriations are faithful to thesystem's spirit, rather than unfaithful; (b) the numberof technology appropriation moves is high, rather thanlow; (c) the instrumental uses of the technology aremore task or process-oriented, rather than power orexploratory-oriented; and (d) attitudes toward appro-priation are positive, rather than negative. These con-stitute an idealized profile of appropriation by thegroup. To the extent that appropriation diverges fromthis ideal, desired group decision processes may notoccur. Improvement in decision outcomes, in turn, willemerge only if the group's decision processes are suit-able for the task at hand (e.g., greater participationand productive information sharing for idea generationtasks; systematic reasoning and resolution of stake-holder conflicts for planning tasks). Thus there is a"double contingency":

P7. Given AIT and other sources of social structure,n, • • • ni., and ideal appropriation processes, and deci-

sion processes that fit the task at hand, then desiredoutcomes of AIT use will result.

If group interaction processes are inconsistent withthe structural potential of the technology and sur-rounding conditions, then the outcomes of group use ofthe structures will be less predictable and, on thewhole, less favorable. There is a dialectic of control(Giddens 1979) between the group and the technology;technology structures shape the group (PI), but thegroup likewise shapes its own interaction (P6), exertingcontrol over use of technology structures and the newstructures that emerge from their use (P3). Organiza-tional change occurs gradually, as technology struc-tures are appropriated and bring change to decisionprocesses. Over time, new social structures may be-come a part of the larger organizational life (P4). Thechange is evidenced in group decision processes (e,g.,methods of idea generation, participation, or conflictmanagement). In this way, advanced information tech-nologies can serve to trigger organizational change,although they cannot fully determine it.

4.0. The Analysis of Structurationin GDSS Use

The AST perspective of technology and organizationalchange implies a research agenda that investigate allaspects of the model presented in Figure 1. To illus-trate such an agenda we will consider GDSSs in a smallgroup context, but our analytic strategy could be ap-plied to other advanced information technologies andsettings as well. Figure 2 summarizes our proposedstrategy. Steps 1 through 10 in the figure represent adiachronic analysis of structuration, examining the de-velopmental path of technology use for a given groupover time. The diachronic analysis can be repeated fordifferent types or levels of technology support, yieldinga synchronic analysis. For example, we might comparegroup interaction processes with GDSS versus noGDSS support, or GDSS versus some manual form ofsupport; level 1 versus level 2 types of GDSS supportcould be compared as well. In the same way, thediachronic analysis can be applied to compare groupsor clusters of groups within or between organizations,yielding parallel analyses. Diachronic, synchronic, andparallel analyses are important, complementary ap-proaches to understanding technology-triggered orga-nizational change (Barley 1990). A complete researchagenda should include all of these approaches,Diachronic analysis is particularly crucial to under-standing the adaptive process by which technology


GERARDINE DESANCTIS AND MARSHALL SCOTT POOLE Adaptive Structuraiton Theory

Figure 1 Summary of Major Constructs and Propositions ot AST

Structure of AdvancedIntormation Technology0 structural features

restrict! venesslevel of sophisticationcomprehensiveness

0 spiritdecision processleadership

efficiency

conflict managementatmosphere

Other Sources of Structure0 task

0 organization environment

P1

P2

Social interactionAppropriation of Structures0 appropriation moves0 faithfulness of appropriationo instrumental useso persistent anitudes

toward appropriation

P5

Decision Processeso idea generation0 participationo conflict managemento influence behavioro task management

Group's Internal System0 styles of interacting

0 knowledge and experiencewith structures

0 perceptions of others' knowledge0 agreement on appropriation

Decision Outcomeso efficiencyo qualityo consensus0 commitment

P7

Emergent Sources of Structureo AIT outputs0 task outputs0 organization environment outputs

New Social Structures0 mies0 resources

Figure 2 General Analytic Strategies tor Assessing the Constructs and Propositions ot AST

DiachroniC Analysis , , Synchrnnlr

CO

"55_>CO

I For a given group and AIT;

1. Doscrlbs the atructura

ofthaAtT,

2. Desciibs other availablestructures.

3. Describe ihe groupcomposition.

4. Develop hypotheses aboutAIT appropriation.

5. Assess extent of AITappropriation, degree ottalthful use, types ofinstrumental uses. ar>dattitudes towardapproprtatton.

6 Develop hypotheses aboutdecision processes.

7. Assess decision processes.8. Deveiop prsdictions about

decision outcomes andnew sodal structures.

9. Assess decision outconws,

10. Describe new sociai structures.

For a second group:

10,

AIT1 vs. AiT2 vs. AiTn | AIT vs. manual support vs. no AIT


GEEURDINE DESANCTIS AND MARSHALL SCOTT POOLE Adaptive Structuration Theory

Structures are incorporated into interaction, so we willfocus on diachronic techniques in detail and provide anillustration of how such an analysis might be under-taken.

4.1. Diachronic AnalysisFor a given group and technology, a clear understand-ing of the structural features and spirit of the technol-ogy must first be articulated (Figure 2, step 1). Thisunderstanding can be gleaned from manuals, discus-sions with designers, observation of the system itself,reports from users, and so on. Such a descriptionshould be more systematic than a simple description offunctions or interface characteristics; it should scalethe technology along meaningful, comparable dimen-sions (such as those in Figure 1 and Table 2) thatreflect the spirit and the structural feature set. Acareful analysis of the structure of the technology yieldsinformation about the kinds of social interaction andoutcomes that the technology is likely to promote.Silver (1991) and DeSanctis et al. (in press-b) illustratehow decision support technologies can be described instructural terms.

Other sources of structures can be similarly de-scribed (Figure 2, step 2). For example, what socialstructures are provided by the task(s) the group con-fronts? And what structural potentials exist within theorganizational environment? Tasks can be described interms of complexity, richness, or conflict potential(McGrath 1984). The organizational envrionment mightbe scaled in terms of complexity, formalization, ordemocratic atmosphere (Collins, Hage and Hull 1988),By scaling sources of social structure along a meaning-ful set of dimensions, hypotheses about the degree of"fit" between technology and other sources of struc-ture can be identified. Most likely, high task-technologyfit will be associated with greater AIT appropriationmoves, more faithful appropriation, and more positiveattitudes toward appropriation. Assessment of thegroup's internal system, such as their degree of experi-ence in working together or with the AIT, their domi-nant style of leadership, or their agreement with re-spect to the purpose of the AIT or how it should beused, can also lead to hypotheses about AIT appropra-tion (step 3), For example, in the case of a GDSS,greater experience with using the technology, greateragreement about how the system should be used, and amore participative style on the part of the leader,might be expected to lead to greater and more faithfulappropriation moves (step 4).

Assessment of appropriation processes is at the heartof the analysis (step 5 in Figure 2), Appropriation

analysis tries to document exactly how technologystructures are being invoked for use in a specific con-text, thus shedding light on the more long-term processof adaptive structuration (i.e,, the formation of newsocial structures). Discourse is the object of study. ASTfollows the tradition of structuralism in assuming thatlanguage is reflective of cultural evolution and can beinvestigated scientifically (Thompson 1981). Conversa-tions, announcements, documents, and all forms ofwritten and spoken speech are of potential interest tothe investigator. Appropriation analysis examines howtechnology and other sources of social structure arebrought into human interaction through discourse. Suchan analysis can be undertaken at one of three generallevels: micro, global, or institutional. At each level, thefour aspects of appropriation identified earlier can beexamined: (a) appropriation moves, (b) faithfulness ofappropriaton, (c) instrumental uses, and (d) attitudestoward appropriation. Appropriation analysis can logi-cally begin at the microlevel, since it is in specificinstances of discourse that the formation of new socialstructures begins. Written or spoken discussion aboutthe technology is particularly important since this isevidence of people bringing the technology into thesocial context. From there, appropriation analysis canproceed to higher levels, global and institutional. Theresearcher can proceed from a microlevel, then to aglobal level, and finally to an institutional level ofanalysis, progressively investigating more and morestrata of the technology's role in organizational change.Lower levels of analysis help to explain changes thateventually are evident at the institutional level. Fur-ther, lower ievels of analysis can help to explain whytechnology brings change in some contexts (e.g,, insome groups) but not in others. Over time, institu-tional-level appropriation affects micro-level appropri-ation, and vice versa. Engaging in multiple levels ofanalysis can yield ideas for improving technology de-signs or the conditions under which they are used.Table 4 shows how appropriation analysis for AITstructures might be undertaken at the three levels,

4.1.1 Microlevel analysis, examines the appropria-tion of technology structures as it occurs in sentences,turns of speech, or other specific speech acts, in thecase of GDSS use, microanalysis might study the speechacts of group members, or sequences of speech acts,that occur during a computer-supported meeting. Tomake the analysis systematic, the range of possibleappropriations can be identified and speech acts thenclassified according to that scheme. An a priori set ofpossible appropriations of technology structures cues

ORGANIZATION SCIENCE/VO1, 5, No. 2, May 1994 133

GERARDINE DESANCTIS AND MARSHALL SCOTT POOLE Adaptive Stmcturalion Theory

the observer on "what to look for"; the interpretivedemands of the research, though not eliminated, aresubstantially reduced. Table 5 illustrates a straightfor-ward approach to identifying group response to AITand other structures, starting with the four generaltypes of appropriation moves idenitifed earlier andthen describing subtypes within each of these. Anygiven speech act in the group may include one or moreof these appropriation moves. For example, consideran excerpt of discourse among five people who areusing a GDSS in a face-to-face meeting, as shown inTable 6(a). Each move to appropriate structures can bedescribed in terms of the source of structure (Table 3)

Table 4 Three Levels of Appropriation Analysisfor AIT Structures

Level of Analysis Unit of Analysis

Micro speech or

other acts

meeting phases

Global enlire meeting

multiplemeetings

Institutional multiple groups

across

organizations

Aspects of Appropriation

appropriation moves

(types and subtypes);faithful vs. unfaithful

appropriation;

instrumental uses

of struciures;

attitudes toward structures

dominant appropriation moves;degree of faithful appropriation;

dominant instrumental uses;persistent attitudes toward

structures;relatively stable patterns ofappropriation, in terms

of moves,

degree of faithful use.instrumental uses,

and attitudes

predominant typesof moves in the

business unitor type of user group;

degree to which faithful useis widespread;

typical instrumental usesamong the studied groups;

dominant attitudes;commonalities

and differences inappropriation moves,faithful use.instrumental uses.

and attitudes

across organizations

and the appropriation type and subtype (Table 5). Inthis way, actual appropriation of structures can bedocumented as they occur in discourse. New structuresthat emerge in the group, such as outputs generated byuse of the technology or the results of applying taskknowledge, can also be noted and their approprationdocumented. For an example, see Table 6(b). The goalis to identify (a) what staictures are being appropriatedand (b) how they are being appropriated. Interpretiveschemes, such as those in Tables 5 and 6 make theanalysis systematic and allow comparisons of appropri-ation over time or across groups.

Note that our interpretive scheme includes a distinc-tion between faithful and unfaithful appropriation ofstructures. Within the interpretive scheme in Table 5,an unrelated substitution (2c) and a paradoxical combi-nation {3b) are unfaithful appropriations. Unfaithfulappropriations are judged by reference to the spirit ofthe technology; combinations which meid structuresthat are incompatible with each other or with the spiritare unfaithful. Unfaithful appropriations are importantto track because they help to explain how technologystructures do not always bring the outcomes that de-signers intended. Instrumental uses that technologystructures serve for the group can also be examined atthe microlevel. For example. Table 7 outlines possibleinstrumental uses that we have observed in our studiesof GDSS use (e.g., DeSanctis et al. 1992, in press-a).Instrumental uses are not always obvious in just a fewspeech acts. Typically these are revealed through anal-ysis of meeting phases, or extended periods of dis-course. For exampie, in the illustration given in Table6(a), the instrumental use appears to be task-oriented;the group is using the GDSS voting function as ameans of assessing member priorities on projects. Theremay be multiple instrumental uses implied in any onephase of technology use, and several types of uses mayoccur over the course of an entire meeting.

The fourth aspect of microlevel analysis is the atti-tudes the group displays as technology structures areappropriated. Three important attitudes that we havestudied in our research are the extent to which groupsare comfortable, value, and feel challenged as theyappropriate the technology. (See §3.3 for definitions ofthese attitudes.) These or other attitudes of interestcan be measured via observer ratings or retrospectivelyvia self reports of group members. (See Billingsley 1989and Sambamurthy 1990 for examples.)

In sum, microlevel appropriation analysis consists ofidentifying types of appropriation moves, distinguishingbetween faithful and unfaithful appropriation, and ex-amining the instrumental uses and attitudes group


GERARDINE DESANCTIS AND MARSHALL SCOTT POOLE Adaptive Structuralion Theory

Table 5 Summary of Types and Subtypes of Appropriation Moves

Appropriation Moves Types Subtypes Definition

Direct Use(structure ispreserved)

Relate toOtherStructures

(structure maybe blendedwith anotherstructure)

1. Direct appropriation

2. Substitution

3. Combination

4. Enlargement

5. Contrast

a. explicit

b. implicitc. bida. partb. related

"c, unrelateda, composition

*b. paradox

c. corrective

a. positive

b. negative

a. contrary

;;onstrain theStructure

(structure isinterpreted orreinterpreted)

Express ;JudgmentsAbout theStructure

6, Constraint

•

7. Affirmation{structure is accepted

•

8. Negation(structure is rejectedor ignored)

9. Neutrality

b. favoredc. none

favoredd. criticisma. definition

b. command

c. diagnosis

d. orderinge, queries

f. closureg. status

reporth. status

requesta. agreement

b. bid agreec agree

rejectd. compliment

a. reject

b. indirect

c. bid reject

openly use and refer to the structureuse without referring to the structure {e.g., typing)suggest use of the structureuse part of the structure instead ol the wholeuse a simitar structure in place o( the structure at handuse an opposing structure in place of the structure at handcombine two structures in a way consistent with the spirit of bothcombine contrary structures withno acknowledgement that they are contraryuse one structure as a correctivefor a perceived deficiency in the other

note the similarity between thestructure and another structure via apositive ailusion or metaphornote the similarity between thestructure and another structure via anegative allusion or metaphorexpress the structure by noting whatit isn't, that is, in terms of a contrasting structurestructures are compared, with one favored over the othersstructures are compared, withnone favored over the otherscriticizing the structure, but without an explicit contrastexplaining the meaning of thestructure or how it should be usedgiving directions or ordering others ,to use the structurecommenting on how the structure is working,either positive (H-) or negative (-)specifying the order in which structures should be usedasking questions about thestructure's meaning or how to use itshow how use ot a structure has been completedstate what has been or is beingdone with the structurequestion what has been or isbeing done with the structureagree wth appropriation of the structure

ask others to agree with appropriation of the structureothers agree to rejectappropriation of the structurenote an advantage of the structure

disagree or otherwise directlyreject appropriation of the structurereject appropriation of the structureby ignoring it, such as ignoring another's bid to use itsuggest or ask others to reject use of the structureexpressing uncertainty or neutralitytoward use of the structure

*These represent unfaithful appropriations. All others are faithful appropriations.


. V


Table 6(a) An Illustratiort of MIcrolevel Analysis of Appropriation

Sources ofStructure^

Appropriation GroupMember Speech or Other Action

Explanation cfAppropriation Move

A-T

AA

A

A

A

A

A

A-T

A

A

A-T

3a

1C

6b6d

7a

8c

6b

6c(-)

6h

5d

6a

6a

5b

6a

1b

3 Well, look— let's vote, let's vote onour priorities for these projects.

1 Why don't we use the votingtacility...

3 (interrupting) Let's rank the alterna-tives and then vote.

5 OK, let's rank them.

2 I don't see why we are ranking thealternatives

3 Just - everyone go ahead and do it.

2 I still haven't got an ansvi/er to myquestion

2 Are we ranking the alternatives?

4 We already know —I already knowwhat everyone's priorities are onthese projects,

5 Because the software is built forthis.

3 We don t know everybody -somebody might be thinking differ-ently than... you know... (fades)

2 What is this going to show us thatwe don't already have in the budgetproposals?

3 Nol everybody is voicing their opin-ions, and I want to clarify exactlywhere everyone stands.

all (everyone inputs / keys into theGDSS)

The voting feature of the GDSScombined with the priofitization goalof the budgeting task.A suggestion is made to use astructural feature of the GDSS.Member 1's suggestion is ignoredand an order for using the GDSSstructures is proposed.Member 5 agrees with the appropri-ation move made by member 3.Member 2 disagrees with the ap-propriation move and asks others toreject it.

Member 3 commands member 2 tofollow the appropriation move.The proposed appropriation moveis criticized,

A query on what is being done withthe GDSS structure.The idea of using the GDSS to dothe task is criticized.

M explanation for the proposedappropriation of Ihe GDSS is given.Further justification of the appropria-tion move is given.

The GDSS and task structures arecompared, with the task informationfavoredAn explîation for the proposedappropriation of the GDSS is given.

Group members use the GDSS.

A refers to the advanced information technology, in this case a GDSS. T refers to the task. See TaWe 3.See Table 5 for definitions of appropriation moves.

that calegorization of appropriation moves is made not only on the basis of the text transcript of the group's interaction, but also onlistening to the discourse and observing the group. Hence, inferences about the intent of Ihe speaker are being made.

members apply to technology structures. Appropriationmoves associated with individual speech acts, whencompiled across meeting phases or entire meetings,may reveal dominant patterns of appropriation in thegroup. (For an illustration, see Poole and DeSanctis1992.)

4.1.2. Global level appropriation. By identifying themost persistent types of appropriation moves made by

a group over a period of time, microlevel appropriationanalysis can be extended into the global level of analy-sis. Global analysis examines conversations, meetings,or documents as a whole, rather than isolating thespecific acts within them. In the GDSS setting, globallevel analysis might consider appropriation across thecourse of an entire meeting, or a series of meetings.This can be done by collapsing data obtained fromspeech acts or multiple meeting phases over long peri-

136 ORGANIZATION SOENCE/VOI. 5, No. 2, May 1994


Table 6(b) An

Sources ofStructure^

AO

AO

Illustration of Microlevel Analysis of the Outputs of Appropriation

Appropriation

Move^

ia

7a

GroupMember

3

5

Speech or Other Action

OK (looking at votes on large

screen), so two of us are adamantlyagainst funding the Pierrson plan,

but on all the other projects we

basically agree.

That's right.

Explanation of

Appropriation Move

The outputs of the GDSS are explic-

itly used.

Member 5 agrees with the appfopri-

1a

ib

8a

1a

EO

2 The Pierson plan is the most re-

searched and carefully planned pro-posal I've seen in a long time. Lookat the customer support figures onpage 5.

all (all look at documents; silence andshuffling of paper)

4 I don t know,

it's just not done around here. Theidea of using customer-based in-centives is against our corporatepolicy, in my opinion.Not i( you apply the policy to includepotential customers, not just exist-ing customers.

ation of the AO structure.

Task information and materials are

explicitly used and referred to.

There is implicit use of the taskstructures.Disagreement with Member 2's ap-propriation of the task reference to astructure of the organization (what isgenerally done and not done)

Member 2 applies the outputs ofexternal (organization) structure to

disagree with the appropriation ofthe external structure.

'AO refers to outputs of the advanced information technology, in this case a GDSS. T refers lo the task. E refers to the external environment. EO

refers to outputs trom use of an external structure. See Table 3.

^See Table 5 tor definitions of appropriation moves.

ods of time. Alternatively, segments of interaction canbe studied at systematic intervals, such as the' start,middle, or end of each meeting, or throughout a sam-pling of meetings. The goal here is to identify system-atic patterns in the way a given group appropriatestechnology structures, including dominant appropria-tion moves (types and subtypes), degree of faithful orunfaithful appropriation, and the instrumental usesand attitudes associated with the appropriation pro-cess.

Some previous research has attempted to identifyglobal appropriation. For example, DeSanctis et al.(1992) identified three types of appropriation patternsbased on instrumental uses across multiple meetings ofseven groups using a GDSS: (a) pure task and processgroups, (b) social and power-oriented groups, and (c)mixed groups. The group's dominant type of instru-mental use was found to relate to: their overai! amount

of GDSS use; who initiated system use in the group;observers' ratings of group comfort toward the technol-ogy; and members' expressed sentiments toward thesystem as they used it. Billingsley (1989) has developeda method for coding global appropriations from groupinteraction with a GDSS. Her coding process involvestwo "sweeps" through videorecordings of meetings. Inthe first sweep, coders classify one-minute segments ofinteraction for: (a) the specific task for which the groupis using the GDSS; and (b) whether the use in questionis faithful or unfaithful. In the second sweep, 15-minutesegments are coded for: (c) degree of challenge and (d)comfort with the system.

4.U. Institutional level appropriation. Appropria-tion analysis at the level of the institution as a wholerequires longitudinal observation of discourse aboutthe technology, with the goal of identifying persistent



Table 7 Instrumental Uses, or Functions, of AIT Appropriation

Instrunnent Use Definition Includes Does Not Include

Task

Process

Power

Use of the AIT to facilitate substantivework on agenda-setting, problem def-inition, solution generation, or othertask-related operationsUse of the AIT to manage communi-cation and other group processes

Use of the AIT by a group member loinfiuence others' Ihinking or to movethem forward in their work

Social Use of the AIT lo establish or maintain

social relationships among members,such as to joke, laugh, or tease oneanother

Individualistic Use of the AIT by an individual purelyfor private reasons, such as to takepersonal notes or to explore systemfeatures

Fun/Exploratory Use of the AIT for its own sake, withno specific goal in mind other than to"play" or "understand how the sys-tem works"

Confusion Use of the AIT during a period of

disorientation, or where there is noclear focus of attention in the group

uses where the group first decidesthe activity they will undertake, thenmoves to the AIT to facilitate accom-plishment of the activitywhere the group is on a tangent, orfloundering about how to proceedand then looks to the AIT to help themdecide how to proceeduse where the user(s) deliberatelyintended to affect the general discus-sion or other's opinionslaughing and joking together whileentering information on the AIT ordiscussing outputs; shared jokes inthe context of AIT useindividual task-related or fun/explo-ratory uses of the AIT

laughing at incorrect or inept uses;using the AIT to make others laugh;most or all members are involved

multiple conversations or simultane-ous AIT uses in the group with nocommon goal or focus

uses where the group looks to the AITto determine how they should pro-ceed.

where the group first decides an activ-ity, or how to proceed, and then looksto the AIT to accomplish the activity

use which is not intended to influencethe group

socializing that has not been broughtabout by, or directly involves, use ofthe AIT

individual uses that are used to influ-ence others (as in Power uses)

exploratory uses that are conducted

by one person (as in Individualistic)

disorientation periods where the AITis not being used or referred to, orperiods where use is clearly for fun/exploratory purposes

patterns across business units (e.g., production versusmarketing), users types (e.g., management versus union;men versus women), or organizations (e.g., manufactur-ing versus service firms). As at other levels, the analysisaims to identify how technology structures are directlyused, interpreted, combined with other structures, andso forth; but at the institutional level the goal is toidentify persistent changes in behavior following intro-duction of the technology, such as shifts in how prob-lems are described, decisions are made, or choiceslegitimated. In the case of GDSS, example questions ofinterest include: What kinds of tasks tend to be com-bined with GDSS uses in this business unit or organiza-tion? Have GDSS structures, such as a democraticspirit or specific decision techniques, been widely in-corporated into organizational meetings? Are thesestructures being applied even when the technology isnot available? Has extensive GDSS use led to in-creased task and process-orientation in meetings, andless socialization, fun, or confusion in meetings? Arethere fewer power moves in meetings since the GDSS

has been adopted, or more? What kinds of attitudestoward the technology are being promoted in organiza-tional training sessions? What are the dominant atti-tudes among users of the system? In our research wehave just begun to study appropriation at the institu-tional level, electing instead to start with microlevelanalysis. Barley (1990), Barley and Tolbert (1988), andRobey et al. (1989) provide institutional-level analysesof technology effects that would be useful to re-searchers interested in structuration accounts of ad-vanced information technologies in organizations.

4.2. Analytic StrategyIn sum, assessment of appropriation processes (Figure2, step 5), whether at the micro, global, or institutionallevel, can be accomplished via a procedure such as thefollowing:

(1) Begin by documenting an interaction sequence,such as a group conversation, meeting or other timeperiod in which the advanced information technologywas present and available for use. For nnicrolevel anal-



ysis, a verbatim transcript is needed. For global levelanalysis, a detailed description of the sequence ofevents may be sufficient. For institutional analysis,samples of conversations, memos, announcements, orother documents may be necessary.

(2) For each speech or other action, identify thegroup member(s) initiating the appropriation and thesource(s) of the structure being appropriated, such asthe AIT (A), task (T), environment (E), or an output ofone of these (AO, TO, or EO) (Table 3).

(3) Classify each act into one or more interpretivecategories of appropriation, such as those given inTable 5.

(4) Identify the instrumental uses of technology ap-propriation (Table 7); this can be done for each speechact, grouping of speech acts, or other meaningful unitof analysis.

(5) Parse the interaction sequence into meaningfulphases of appropriation; these may be delineated interms of AIT use/nonuse, faithful use/unfaithful use,task uses/nontask uses, or any other meaningfulmethod of parsing the interaction. Descriptive observa-tions (made by the researcher or informants) can begiven for each sequence, applying the various dimen-sions given in Figure 1.

(6) Systematically reduce the data to a manageableform (Miles and Huberman 1984). Data reducing cantake the form of deriving frequencies of interpretivecategories (steps 2, 3 and 4). Even more informative isto construct a concise, qualitative map of each meetingor other segment of discourse, along the lines de-scribed by Krippendorff (1980). The map consists of asynopsis of the group's discussion on the right half ofeach page, with descriptions and code letters on theleft half denoting phases of appropriation; code letterscould be used, for example, to locate every speech actor phase involving a combination of A and E structuresor extended periods of unfaithful appropriation. Pooleand DeSanctis (1992) provide an illustration of thisprocedure at the microlevel.

(7) Identify dominant types of moves and persistentpatterns of instrumental uses and attitudes for theinteraction sequence of interest. This may be compiledfor a single meeting, or in the case of global or higherlevels of analysis, for multiple meetings or other formsof discourse. This can be done by computing summarydescriptive statistics for interpretive scheme data (seeDeSanctis et al. (1992) for an illustration), and/or byapplying techniques proposed by Miles and Huberman(1984) for collapsing qualitative data.

These procedural steps are similar to those followedby Courtright, Fairhurst and Rogers (1989) in their

interpretive analysis of interaction patterns. Based onthe patterns of appropriation that emerge in the analy-sis, specific hypotheses about decision processes can bedeveloped (Figure 2, step 6). Existing approaches areavailable to study the group's internal system, decisionprocesses, and decision outcomes (Figure 2, steps 7-9).For example, there are rating scales for assessing styleof interaction, decision quality, and commitment(Gouran, Brown and Henry 1978); models for calculat-ing evenness of member participation (Watson et al.1988) and consensus (Spillman, Spillman and Bezdek1980); and coding schemes for assessing confiict man-agement (Poole et al. 1991). influence behavior (Putnam1981), and task management (Poole et al. 1990). Docu-mentation of new structure formation (Figure 2, step10) will require longitudinal observation of the groupand identification of persistent use of the technology-based structures in the group or organization at large.

4.3. An IllustrationTo illustrate the use of our analytic strategy for study-ing appropriation, we compared two groups that usedthe same GDSS for prioritizing projects for organiza-tional investment. We applied the interpretive schemesgiven in Tables 3, 5, and 7 to verbatim transcripts ofone decision-making meeting for each group. Since theschemes account for group members' intentions withrespect to interactions with others, as much as theparticular words or expressions used, categorizationwas done using both a written transcript and an audiotape of the meeting.'* Consistent with Krippendorff s(1980) approach, after initial categorization and againafter development of phasic maps, we met to compareresults (see Gersnick (1988) for a similar approach).We discussed discrepancies until agreement could bereached, referring to the audio tapes as necessary. Thisprocess produced a final set of categorizations and adescriptive map for each meeting. Next we computedquantitative summaries of appropriation moves anddeveloped descriptive accounts of each meeting. Sam-ples of micro and global analyses for our two illustra-tive groups are shown in Figures 3(a) and 3(b). Thisrepresents a diachronic analysis for each group and aparallel analysis as groups are compared.

Following the model given in Figure 1, both groupshad similar inputs to group interaction. The sources ofstructure and the group's internal system were essen-tially the same in each group, except that group 1 had amember who was forceful in attempting to direct oth-ers and was often met with resistance. Figure 4 pre-sents descriptive summaries of our appropriation anal-ysis for each group. Notice that group 2 spent much

ORGANIZATION SaENOs/Vol. 5, No. 2, May 1994 139


Figure 3a An Illustration of Micro and Global Appropriation Analysis: Group 1

tauracUon tmatyiU

Atuiuto umrtQDSSi

OK. 'mam bmMiatt u ika ammunaj'

o*!' (• ODas ( w m ) Im

A-la

A7i

A.hAO.|«

AlaAU

«!.•>••

iftltBuii ID I ODSI haan)'•Ming dumnvii'Wui. can-i n lua cu ilw (b« </itBM oi Ihc tame) temtOK. "difininf kwl vmmg ^baniflb'1 tfai^l Ihnk we can l^ttMrwij taiadiB pojacM J « i riHî luH Ioi Ibt cnHru iw'n vug lo wtJuaiB tfHI BMI I I doHs'i pnoaad lopsdlir if ôiJHH ttlan ai4 don'i I H H aa^ aa E S H H .

©

A a AOma™ - M. E. « ririilMHiwi (iiirliiil»n A wiib r ar E) i

al I W M U U J ffnpBMUtm - 1

wwicri ODSI. m o d n u ofmton.

qr AO • 7<a4B ^HoM IS CHitnuii (te cote) II

. B. ot oanbuutn (iBd>adii« A vub T a E)

wiih « « • panaî of aaAmimM H eanAn. ! •«<• • •

more time than group 1 defining the meaning of thesystem features and how they should be used relativeto the task at hand; also, group 2 had relatively fewdisagreements about appropriation or unfaithful ap-propriation. In group 2 conflict was confined to criticalwork on differences rather than the escalated argu-ment present in group 1. Although two members ofgroup 2 were dominant in initiating appropriationmoves, making participation in discussion somewhatunevenly distributed, there was an atmosphere of re-spect for difi'erences among members. The result wasthat the decision process in group 2 was more consis-tent (than group 1) with the spirit of the GDSS. Moreproductive conflict and task management in group 2,relative to group 1, resulted in a relatively efficientmeeting and high post-meeting consensus.

Overall, the illustration highlights how AST conceptscan shed light on the process of advanced technologyuse in group interactions. Although the same technol-ogy was introduced to both groups, the effects were notconsistent due to differences in each group's appropri-ation moves. Group 2's appropriation patterns were

more "ideal," so decision processes and outcomes weremore desirable than in group 1.

4.4. Measurement IssuesWe offer our analytic strategy as a starting point fromwhich other research can proceed. Appropriation pro-cesses are complex and subtle, so measurement ap-proaches are tricky, to say the least. Because the im-plied meaning of action is critical to appropriation,strict coding schemes are less informative than morequalitative interpretive schemes. Whereas codingschemes interpret utterances according to a standardset of rules and classify them into a relatively small setof a priori categories, interpretive schemes, such asthose in Tables 5 and 7, infer actors' intentions byapplying a framework that relies as much on speakers'intentions as on literal words or expressions used(Poole, Folger and Hewes 1987). Interpretive schemesare difficult to program, or automate, and so are ex-traordinarily labor-intensive. As in ethnography andconversational analysis, classification rests heavily onthe researcher's logic, and, because a single utterance



Figure 3b An Illustration of Micro and Global Appropriation Analysis: Group 2.

iBtcracthn tmtijtt* MMlIng ptlBM

GnupmantMT

IrumunoiuJ iDTGDSS CDSS luucuiiai

I

AO-lcAO-7I.A-IIA l *A-lb

AO-AO-1*

AO-ls,A-li:AO. I*. A-l«AO-AO-Sb

E-1*AO-lb

AO-lt

WdL 1 tei'i k m .OK. !«'• na- <BHD1 (• OD3S r~l»).

•a Ikal 4i*)f a x kx* i i iba

WdL I d i M gnc *B<1 Osaa ((mi numbai <rf O U B Uu Aa GDSS •oam) on pany mwdi ibc taai*Y«a. iboH bav« UH •*« tii|>iTr (wtt^Jiu)Sa lai'f dalala lumbK caia wa4 ww^l ihAn agunWhr wnild you • • !» u ^ thxT[ Ihink ii't atpvuHii Ihal Iha mijcvujr erf d fed Ihuinciviauii buiuHai u frHrt itopa^tM itun dinolyhalpm^ Itea Mtmimnily H • vtiola^ b«A L IhaUL llul clkun§ rîu'n mimizit H Ibai Ibva H jumf ID ba 1

Sa iwvtiAB back la ibsa faapijad pvDjiEU flbvboudyAiabar ow. towarwf u u * . • (OMf lo b« baa^nil•HI only H buaoHHt but 4koTbM'i if jnu uaiaH ibaa'l • ocUa down ^ a o .It if l i , dort'i fiai faal Ibu VBfTN a tguUy Ta • pgiM * a a I*, tailY w can M p Hna g( (ba paapla H U M d iba DBabiN am (U o( ibe p a c ^ ill irf Iba loMOK. m ihau ua OK. Wa * D U U man « u> a»luatii>|oui tb*naiiv« MMjbt * * flhouU Ma likaia prajacM.

®

t A ST AO BIDMI - MpopDiucii erf T. E. <v nmwbiBiboii (niduliBf A aiidi T nr B) i

Nianbs a/ unfkidlAJ «{^nîvufi(aia ^ 0

Dofaman niiuJn i . muiaiia raapaa. muda—ia diallai^

For tht CBtlra ant lD i «>Dcamiuia ippvûiKai imwaa

I D( A or AO - , nr A oaâ damud u gavouM (}* <>o*i): M

pvpsnua irf T. e. « eonbiniiHai (iacludiaa A wtA T s E) iNunts ll wif iilblU ifiinipnatiaii - 1

K4 O O U :

or action may carry multiple meanings, it may beclassified into more than one category. Although vali-dation might be achieved by asking informants of thescheme's adequacy, more often validity is achievedthrough researchers' ongoing dialectic over specificclaims. Analytic criticisms of Searle's analysis of theconstitutive rules for the performance of speech acts(Frank 1981; Levinson 1981) illustrate this form oftheory testing. The debate over the adequacy of aninterpretive scheme is advanced largely through thepresentation of examples and counterexamples thatillustrate potential advantages or problems. Indeed,in interpretive analyses there Is an implicit belief thatthe knowledge the investigator is unearthing throughthe identification of formal properties may be beyondthe informants' expressive capacity. In sum, althoughwe can argue the validity of our interpretive schemesbased on case illustrations and the scheme's ability topredict group consensus (as in Poole and DeSanctis1992), a continued dialectic among scholars interestedin appropriation analysis is perhaps more important.

Finally, it is important to keep in mind that just astechnology impacts are not pure and are mediated by acomplex web of forces (Kling 1980), interpretiveschemes—however rich and sensitive to subtle mean-ings—cannot be all-encompassing. As representationschemes, they have the problems of reductionism thatplague nearly all behavioral measurement. On the otherhand, comprehensive, clean prediction of structuraleffects on interaction or behavior outcomes is not thegoal. Our interest is in describing appropriation pro-cesses with sufficient refinement so that we can gainmeaningful (though not perfect) insight into the con-nection between technology and action.

5.0a ConclusionBusiness professionals, researchers, and social eom-mentators often express disappointment with the factthat advances in computing technology have notbrought about remarkable improvements in organiza-tional effectiveness. Why is it that technology impacts



Figure 4 Sample Descriptive Analyses

Group 1

GDSS appropriation11K gimp relied heavily on Ihe CDSS lo direct iu ducuiiioni. fineconiuliint ihc GDSS fetluitt tnd then dccidinj hov lo proceed. Metnbertttaiud 1^ defininf iheii deaiion problem in Ihe GDSS uut [hen enteredcriicru for cvaluiung Itieif projecti. They did nol me weighting, nting. orvoung fcaluna to cittbliih the icluive unpoiunce of crileiu. Sevenlmanbcn conluied ihe mcuiing irKl ciptbilitiei of the 'cntena" teituie inthe lyuem (3b in Tible S). There wu • good deal ol mitliled lubilitubon(!c) ol CiDSS iinicturei; one member in pamcular kcp> luggeiling ihii thegroup uie tx not uie GI]SS itnicturei biied on fuilty undentaiding of thelyiiem. incofreclty exlrapoUting from other lynemi or enpehencei (2c): ilirvtnl poiMi he directed the groop to uie ceiuin fealurei th*l could IKM•cconimoiliie iheir wort ictiviiiei. Memben had pn>blBni coordinaiingidu enuy IDIO ihe GDSS; a long leriei d command! (6b). itaiuirtquettj(6h), tnd lUtui icporti (6g) reflected the difficulty the group h*d incoordinatint their ettont. There were penodi ot high diiigreement amongihe memben (lirge numben of 8i, 8b. and 8c codei). but they did net haveinwbie openting ihe lyllan (6c-), nor did Ihey critidic ii (5d).

Group 2

GDSS appropriationThe gioup begin by enleruig • luk problem italement into the lyitem andthtn uiuig Ihe "critena" fcalure to bninilomi *ay» o( evaluating of thepro)ecu under coniidenlion. Next, memben evaluated the criteri* uiing *we)ghting acheme an4 diicuiied their agnemenu uid diiagreemcnu abouithe cnleria and ihe weight valuei. Ai in Gioup 1. the pmpomon oC A uidAO movei wai tguiie high, inificaing lubitamuJ approjinauon a* the GDSSdumg Ihe ineetuig; however. Group 2 ipem much marc Ume deflning ihemeaning of the lyiletn fiaiuret and how Ihey ihnuld be uted rclaove U> theuik u hand <6«). Group 2 huj linle trouble coordinaiing lyium UK andhad icLalivety fe* diiagiccmenti about ippnjprialion or unJaithful•pfircâiiani. Member) ilepfied in icidily to help each other in lyitemopenlion ihitiugh commmd.reiponie icquencei (6b followed by 7a).R»ther than htvmg ihe lyitein drive the gnx^ pnxsit, memben tended lofim decide on t oouru of action ind then look U> the lyuem to helpoiccute ihe actron. Tliough not alwiyi high in comfon wuh the lochnoktr,they eihitited high reipect and • lense of challenge toward uiing the•yuan. Abo, there wu lubtuntial blending ol lytum ouputi (AO) wiihtalk and uiemal ttnicturct. niher than lole dixuiiion al one or ihe olha.

Decision processes and outcomes

Thii gnwp uted Ihe GDSS a gical deal and, tlttxxigh iherc were periodi olconfuiion in uiitrumenlal uie, monben eiiiibiled coiiiitcnlly poailiveauiludei towanJ Ibc technology. Given ihii pattem ot •ppropnaiicn. wewould eipeci ihe group to have fairly poiiuve amtudet luward the GDSS mthe end of the meeting, which Ihey did. In lermi of deciiion pniceiiei, ihegroup wu able u> genenli ideal readily, but became one memberdofnmaied in ippnifinauon movei. paruciptuoi wa< not even. Membeiteipreiseii high diiagreement with one another about the ideal theyjertenled via the technology. Cmflicl wai quilc high and ihe group had(..ifficully managing iti taik. mmg the technology a> m uiiuument olpi --ei) mon than fa laik aimi. Theie inlenctiun paiienu led lu anextremely long meeung. niher ihan an efficient one, and reiulied in mixedteelingi about the quality of the gnx^'i Tuial deciiiun Tlie group did nolconverge in iheir viewpomti at a retult of iliac meeting, allhough iheygained giealer urtdentanding of each ochei'i poiitioni un iituei.

Decision processes and outcomesThe group wai agreeable and approached it> taik in a lerioui. maner of factmanner. They took a itep-by.iiep approach to the deciiion proceai. fintentering ideal into ihe GDSS uid ihen u>ed vanoui voting melhodi loevaluate iheir ideu. Memben brainitoimed in thii faihion for crucria toevaluate projecti for [unding. Allhough iti dectiion ilept were •imiUi' loGroup 1. there wu much gnaier agncmeni an appropriation in thii gnup.There wu leu repeution. of backtiscking. at iiqn in Giwip 2 ai Iheypfoceeded ihrough the deaiian prooeii imocihly Ccnflicl wai confined locriDcal wori cn difference! niher than eKalaird aigumoiL I'wo membenwere more doninanl than oiiien in iniliaung appmphatian movei. makingpailicipalion in the diicusiion ume what tineven (wtlh lame membenHying leu ihui oihen). Ncvcnheleit. thei« wai an Mmopherc of rtipectfor differtncet between memben, yielding a deciiion proceii ihat wuccniiilent with the ipiiit cf ihe GDSS ln addition to prodiKlive conflictmanagement, ihe gro^ engaged in good taik management aa membert TintdiKuiied iheir objectivei ind deciiicn pnxcii and ihen invoked ihc GDSSto ficiluate their worii, Tiiete deciiion proceitei reuilted in inmeeting ind iltong poit-aieeling conteiuui.

are often more subtle than dramatic? Positive in someorganizations, yet neutral or even negative in others?Fresh theoretical approaches are needed to shed newhght on these old questions. Structuration models areappealing because they emphasize the interplay be-tween technology and the social process of technologyuse, illuminating how multiple outcomes can resultfrom implementation of the same technology. Becausethe new structures offered by technology must beblended with existing organizational practices, radicalbehavior change takes time to emerge, and in somecases may not occur at all. Structuration models gobeyond the surface of behavior to consider the subtleways in which technology impacts may unfold. Limita-tions of structuration models to date have been theirweak consideration of the structural potential of tech-

nologies in general and advanced infonnation tech-nologies in particular, their exclusive focus on insti-tutional levels of analysis, and reliance on purelyinterpretive methods. To yield useful knowledge fororganizations, structuration-based theories of teehnol-ogy-induced change must devise detailed models ofgroup dynamics and a set of methods for directlyinvestigating the relationship between structure andaction (Barley and Tolbert 1988). In this paper we haverefined structurational concepts to the realm of ad-vanced information technologies, integrated conceptsfrom the decision-making school with structurationconcepts, and demonstrated how structuration can bestudied within an empirical program of research.

To summarize, AST argues that advanced informa-tion technologies trigger adaptive structurational pro-

142 ORGANIZATION SCIENCE/VO1. 5, No. 2, May 1994


cesses which, over time, can lead to changes in therules and resources that organizations use in socialinteraction. Change occurs as members of organiza-tional groups bring the structural potential of thesenew technologies into interaction, appropriating avail-able structures during the course of idea generation,conflict management, and other group decision activi-ties. Group members can opt to directly use technolog-ical features, relate the structures to other structures,constrain or interpret the structures, or make judg-ments about the structures. The impacts of the tech-nology on group outcomes depend upon: the structuralpotential of the technology (i.e., its spirit and structuralfeatures), how technology and other structures (sueh aswork tasks, the group's internal system, and the largerorganizationai environment) are appropriated by groupmembers; and what new social structures are formedover time. Appropriations which initially occur in mi-crolevel interaction eventually may be reproduced tobring about adoption of teehnology-based structuresacross multiple settings, groups, and organizations.

One strength of AST and the method outlined hereis that they facilitate analysis of between-group differ-ences. To determine whether advanced informationtechnologies have the deterministic effects that deci-sion theorists hypothesize or the emergent effects envi-sioned by institutionalists, it is necessary to assesswhether between-group differences are significant. Tous it seems most likely that there will be some variationin the strength of the two types of effects across organi-zational contexts. In some organizations, norms andthe power structure may be crystallized so that ad-vanced information technology effects will appear to bedeterministic; most groups will use the technology in asimilar fashion and the interaction system will be regu-larized such that similar outcomes will ensue for aligroups. At the other extreme there may be organiza-tions which are so fluid that a wide variety of technol-ogy uses and impacts occur. In the middle range, theremay be organizations that experience some variety inoutcomes but enough commonality to detect patterns.

A second strength of AST is that it accounts for thestructural potential of technology and at the same timefocuses squarely on technology use as a key determi-nant of technology impacts. Technologies differ in thesocial structures they provide, and groups can adapttechnologies in different ways, develop different atti-tudes toward them, and use them for different socialpurposes. AST expounds the nature of social structureswithin advanced information technologies and the keyinteraction processes that figure in their use. By captur-

ing these processes and tracing their impacts, we canreveal the complexity of technology-organization rela-tionships. We can attain a better understanding of howto implement technologies, and we may also be able todevelop improved designs or training programs thatpromote productive adaptations.

AST can also enhanee our understanding of groupsin general, not just those using technology. The majorconcepts of AST, as illustrated in Figure I, a)ver theentire input ~* process -• output sequence that Me-Grath and Altman (1966) and Hackman and Morris(1975) advocate as an organizing paradigm for groupresearch. AST provides a general approach to thestudy of how groups organize themselves, a processthat plays a crucial roie in group outcomes and organi-zational change.

Several avenues of study are important at this point.First, the theory and measurement approaches laid outin this paper ean be further developed. We presentedmajor concepts for the study of technology-inducedchange and stated seven propositions regarding rela-tionships among these concepts. Refinement of theseconcepts and articulation of specific research hypothe-ses is the next step. We outlined a general analyticstrategy for applying AST and illustrated its applica-tion to the study of GDSSs in small group settings. Ourreseareh strategy could be speeifled in more detail andtested for its usefulness across a range of advancedinformation technologies and organizational contexts.Because GDSSs make structures particularly salientand manipulable, they are excellent test cases for re-seareh on group structuring behavior; but settings otherthan GDSS use by small groups must be examined ifthe power of AST is to be fully explored. AST assumesthat although structural change lies below the surfaceof decision making, it can be eaptured in interpersonalinteraction, at micro, global, and institutional levels.For each level we offered illustrative variables andmeasurement approaches. But speciflc variables andmeasurement will depend, of course, on the particularteehnology, context, and interaction processes oi" inter-est to the researcher. A critical challenge is to system-atize the research so that technologies and interactionprocesses can be meaningfully assessed and compara-tive measurement is possible. To organize the interpre-tive process of studying strueturation, we devised elab-orate schemes (e.g.. Table 5) and simpler schemes (e.g..Table 7) for categorizing appropriation and its subpro-cesses; we acknowledge that there is a tradeoff be-tween comprehensiveness and parsimony, and simpleschemes may do as well as elaborate schemes. Devel-


GERARDINE DESANCTIS AND MARSHALL SCOTT POOLE Adaptive Siructuration Theory

opment and debate about ways to codify the socialstructures of technology and action would appear to bea healthy agenda for researchers.

In addition to these theoretical and method issues, asecond direction for research is to directly test theexplanatory and predictive power of AST. AST positsthat four major sources of structure (technology, task,environment, and the group's internal system) affectsocial interaction which, in turn, is the key determinantof social outcomes (such as decision efficiency, quality,consensus, etc.). Empirical tests of these relationshipsand of the evolution of new social structures areneeded. Further, AST rests on assumptions that aresimilar (e.g.. technology is socially constructed) anddifferent from (e.g., appropriation is the critical processin social constructionism) other emergent models.Studies which clarify and empirically test the validity ofassumptions that underlie emergent models in general,not just AST, would be especially helpful to our under-standing of advanced information technologies andtheir use in organizations.

Finally, the link between technology-triggeredchanges at micro, global, and institutional levels can bestudied. Individual studies tend to target one level ofanalysis, rather than multiple levels; and theoreticalexpositions tend to be unilevel as well. AST focuses oninterpersonal interaction and so is amenable for studyat multiple ieveis. Pursuit of methods to link study ofinteraction at, for example, the small group level, withinteraction that occurs in organizational units, the or-ganization at large, or even outside of the organization,will strengthen research on organizational change andthe role of technology in change processes. Such analy-ses wiil serve to further link inquiry in informationsystems and organizational communication to the largeand growing study of advanced information teehnolo-gies.

AcknowledgementsThe authors wish to thank the three anonymous reviewers and theSenior Editor for detailed guidance during several revisions of thismanuscript,

This research was supported by National Science Foundationgrant SES-8715565. The views expressed here are solely those of theauthors and not of the research sponsor.

Endnotes'See Grief (1988), Jessup & Valacich (1993), and Pinsonneault &Kraemer (1989) for reviews of GDSS literature and analyses ofconflicting findings.Several writers recently have called for the development of integra-

tive theories and methods (Lee 1991; Orlikowski 1992).The term group is used in our discussion to refer to two or more

people who interact with one another in the context of the advanced

information technology; dyads, small or large groups, departments,and organizations are included.' In fact, we applied the same schemes to an additional 16 groups,with each of us (as researchers) categorizing the speech or other actsof all 18 meetings. The estimate of intercoder reliability for thecategorizations, based on a sample of 225 codes and assessed withCohen's Kappa, was 0.92 for structure source (Table 3) and 0.84 forthe nine major categories of appropriation moves (Table 5). Rawpercentage of agreement between two coders on appropriation movesranged from 60% to 90%. The results of this more extensive analysisare given in Poole and DeSanctis (1992).

ReferencesAbualsamh, R. A., B. Carlin and R. R. McDaniel, Jr. (1990), "Prob-

lem Structuring Heuristics in Strategic Decision Making," Orga-nizational Behavior and Human Decision Processes, 45, 2,159-174.

Barley, S. R. (1986), "Technology as an Occasion for Structuring:Evidence from Observations of CT Scanners and the SociaiOrder of Radiology Departments," Administrative Science Quar-terly, 31, 78-108.

(1990), "Images of Imaging: Notes on Doing Longitudinal FieldWork," Organization Science. I, 3, 220-247.and P. S. Tolbert (1988), "Institutionalization as Structuration:

Methods and Analytic Strategies for Studying Links betweenAction and Structure," paper presented at the Conference onLongitudinal Field Research Methods for Studying Organiza-tional Processes. Austin, TX, Sep 14-16, 1988.

Biilingsley, J. (1989). "An Analysis of Small Group Appropriation ofDecision Structures in a Computer-assisted Group DecisionSupport Environment." paper presented at the InternationalCommunication Association Annual Meeting, San Francisco.CA, May 1989.

Blumer, H. (1969). Symbolic Interactionism: Perspectives and Method,Englewood Cliffs. NJ: Prentice-Hall,

Bostrom, R. P. and Heinen, S. J. (1977), "MIS Problems andFailures: A Socio-technical Perspective Part II: The Applicationof Sociotechnical Theory," MIS Quarterly, 1, 4, 11-28.

Bourdieu, P. (1978), Outline for a Theory of Practice. Cambridge:Cambridge University Press.

Bui, T. and T. R. Sivasankaran. (1990), "Relation between GDSSUse and Group Task Complexity: An Experimental Study."Proceedings of the 23rd Hawaii International Conference on Sys-tem Sciences. January 1990, pp. 69-78.

Cats-Baril, W. L. and G. P. Huber (1987). "Decision Support Sys-tems for Ill-structured Problems: An Empirical Study," DecisionSciences, 18. 3, 350-372.

Collins, P. D., J. Hage and F. M, Hull (1988), "Organizational andTechnological Predictors of Change in Automaticity," Academyof Management Journal, 31. 3. 512-543.

Connolly. T.. L. M. Jessup, and J. S. Valacich (1990), "Effects ofAnonymity and Evaluative Tone on Idea Generation in Com-puter-mediated Groups," Management Science, 36. 6. 689-703.

CourtHght, J. A., G. T. Fairhurst and L. E. Rogers (1989). "Interac-tion Patterns in Organic and Mechanistic Systems." Academy ofManagement Journal, 32, 4, 773-802.

144 ORGANIZATION SOENCE/VOI . 5, No. 2, May 1994


Dennis, A. R., J. R. George, L. M. Jessup, J. F. Nunamaker and D.

R. Vogel (1988), "Information Technology to Support Elec-

tronic Meetings," MIS Quarterly, 12, 4, 591-624.

DeSanctis, G., M. D'Onofrio, V. Sambamurthy and M. S. Poole

(1989), "Comprehensiveness and Restriction in Group Decision

Heuristics: Effects of Computer Support on Consensus Decision

Making," Proceedings of the Tenth International Conference on

Information Systems, Boston, MA, December 1989, 131-140.

R. B. Gallupe (1987), "A Foundation for the Study of Group

Decision Support Systems," Management Science, 33,5, 589-609.

M. S. Poole, H. Lewis and G. Desharnais (1992), "Using

Computing in Quality Team Meetings: Some Initial Observa-

tions from the IRS-Minnesota Project," Journal of Management

Information Systems., 8, 3, 7-26,

, , G. W. Dickson and B. M. Jackson (in press-a), "An

Interpretive Analysis of Group Support System Use," Journal of

Organizational Computing.

, J. R. Snyder, and M. S. Poole (in press-b), "Tbe Meaning of

the Interface: A Functional and Holistic Evaluation of a Meet-

ing Software System," Decision Support Systems: The Interna-

tional Journal.

Feldman, M. S. and J. G. March (1981), "Infonnation in Organiza-

tions as Signal and Symbol, Administrative Science Quarterly, 26,

2, 171-186.

Finlay, M. (1987), Powermatics, London: Routledge and Kegan Paul.

Frank, D. (1981), "Seven Sins of Pragmatics: Theses about Speech

Act Theory, Conversational Analysis, Linguistics and Rhetoric,"

in H. Parret, M. Sbisa and J. Vershueren (Eds.), Studies in

Language Companion Series, Vol. 7: Possibilities and Limitations

of Pragmatics, 225-236, Amsterdam: John Benjamins B. V.

Fulk, J., C. W. Stcinfield, J. Schmitz and J. G. Power (1987), "A

Social Information Processing Model of Media Use in Organiza-

tions, Communications Research, 1987, 14, (5), 529-552.

Gallupe, R. B., G. DeSanctis and G. W. Dickson (1988),

"Computer-Based Support for Group Problem Finding: An

Experimental Investigation," MIS Quarterly, 12, 2, 277-298.

George, J. F., G. K. Easton, J. F. Nunamaker, Jr. and G. B.

Northcraft (1990), "A Study of Collaborative Group Work with

and without Computer-Based Support," Infonnation Systems

Research 1, 4, 394-415.

Gersnick, C. J. G. (1988), "Time and Transition in Work Teams:

Toward a New Model of Group Development," Academy of

Management Journal, 31, I, 9-41.

Giddens, A. (1979), Central Problems in Social Theory, Berkeley, CA:

University of California Press.

Goodman, P. S. (Ed.) (1986), Designing Effective Work Groups, San

Francisco: Jossey-Bass.

Gouran, D. S., C. Brown and D. R. Henry (1978), "Behavioral

Correlates of Perceptions of Quality in Decision-Making Discus-

sions," Communication Monographs, 45, 51—63.

Grief, I. (Ed.) (1988), Computer-supported Cooperative Work: A Book

of Readings, San Mateo, CA: Morgan Kauffman.

Gutek, B. A., T. K. Bikson and D. Mankin (1984), "Individual and

Organizational Consequences of Computer-Based Office Infor-

mation Technoiogy," in S. Oskamp (Ed.), Applied Social Psy-

chology Annual: Applications in Organizational Settings, 231-254,

Beverly Hills, CA: Sage.

Hackman, J. R. and C G. Morris (1975), "Group Tasks, Group

Interactions Process, and Group Performance Effectiveness: A

Review and Proposed integration," in L. Berkowitz (Ed.), Ad-

vances in Experimental Social Psychology, Vol. 8, New York:Academic Press.

Hiltz, S. R. and K. Johnson (1990), "User Satisfaction with Com-

puter-Mediated Communication Systems," Management Science,

36, 6. 739-764.

., M. Turoff and K. Johnson (1981), "The Effects of Human

Leadership and Decision Support Feedback on Group Problem

Solving in Computerized Conferences: A Controlled Experi-

ment," Computerized Conferencing & Commun. Cen., New

Jersey Institute of Technology, Newark, NJ.

Homans, G. (1950), The Human Group, New York: Harcourt, Brace.

Huber, G. P. (1984), "Issues in the Design of Group Decision

Support Systems," MIS Quarterly, 8, 3, 195-204.

(1990), "A Theory of the Effects of Advanced Information

Technologies on Organization Design, Intelligence, and Deci-

sion Making," Academy af Management Reiiew, 15, 1, 47-71.

Huseman, R. C. and E. W. Miles, (1988), "Organizational Communi-

cation in the Information Age: Implications of Computer-Based

Systems," Journal of Management, 14, 2, 181-204.

Jarvenpaa, S. L. (1989), "The Effect of Task Demands and Graphical

Format on Information Processing Strategies," Management Sci-

ence, 35, 3, 285-303.

, V. S. Rao and G. P. Huber (1988), "Computer Support for

Meetings of Groups Working on Unstructured Problems: A

Field Experiment," MIS Quarterly, 12, 4, 645-666.

Jessup. L. M.. T. Connolly and J. Galegher (1990), "The Effects of

Anonymity on GDSS Group Process with an Idea-generating

Task," MIS Quarterly, 14, 3, 313-321. |

and J. Valacich (Eds.) (1993), Group Support Systems: New

Perspectives, New York: Macmillan.

Keen, P. G. W. and M. S. Scott Morton (1978), Decision Support

Systems, Reading, MA: Addison-Wesley.

Kerr, E. B. and S. R. Hiltz (1982), Computer-mediated Systems:

Status and Evaluation, New York: Academic Press.

Kiesier, S. (1986), "The Hidden Messages in Computer Networks,"

Harvard Business Review, February, 46-59.

Kling, R. (1980), "Social Analyses of Computing: Theoretical Per-

spectives," Computing Surveys, 12, 61-110.

Krippendorff, K. (1980), Content Analysis: An Introduction to Its

Methodology, Beverly Hills, CA: Sage.

Lee, A. S. (1991), "Integrating Positivisl and Interpretive Ap-

proaches to Organizational Research," Organization Science, 2,

4, 342-365.

Levinson, S. (1981), "The Essential Inadequacies of Speech Act

Models of Dialogue," in H. Parret, M. Sbisa and J. Vershueren

(Eds.), Studies in Language Companion Series, Vol. 7: Possibili-

ties and Limitations of Pragmatics, 473-492, Amsterdam: JohnBenjamins B. V.

Mantei, M. (1988), "Capturing the Capture Concepts: A Case Study

in the Design of Computer-Supported Meeting Environments,"

Proceedings of the Conference on Computer-Supported Coopera-

tive Work, 257-270, Portland, Oregon, September 26-28.

ORGANIZATION SaENCE/VoI. 5, No. 2, May 1994 145

GERARDINE DESîNCTIS AND MARSHALL SCOTT POOLE Adaptive Stmcluration Theory

Markus. M, L. (1990). "Toward a "Critical Mass" Theory of Interac-

tive Media." in J. Fulk and C. Steinfeld (Eds.), Organizations

and Communication Technology, Beverly Hills, CA: Sage,

194-218.

and J. Forman (1989), "A Social Anaiysis of Group Technology

Use," UCLA Information Systems Working Paper #2-90.

and D. Robey (198fi), "Information Technology and Organiza-

tional Change: Causal Structure in Theory and Research," Man-

agement Science, 15, 5, 583-598.

McGrath, J. E. (1984), Groups: Interaction and Performance, Engle-

wood Cliffs. NJ: Prentice-Hail.

and I. Altman (1966), Small Group Research: A Synthesis and

Critique of the Field., New York: Holt, Rinehart & Winston.

McLeod. P. L. and J. K. Liker (1989), "Computer Meeting Technol-

ogy: Effects on Group Process and Outcomes," unpublished

working paper. University of Michigan, Ann Arbor, MI.

Miles, M. B. and A. M. Huberman (1984), Qualitatiiv Data Analysis:

A Sourcebook of New Methods, Beverly Hills, CA: Sage.

Nunamaker, J. F.. L. M. Applegate and B. R. Konsynski (1988),

"Computer-aided Deliberation: Model Management and Group

Decision Support," Operations Research, Special Issue on Deci-

sion Support Systems, 36, 6, 826-848.

D. Vogel and B. Konsynski (1989), "Interaction of Task and

Technology to Support Large Groups," Decision Support Sys-

tems, 5, 2, 139-152.

Oilman. B. (1971), Alienation: Marx^s Conception of Man in Capital-

ist Society, Cambridge: Cambridge University Press.

Oriikowski, W. J. (1992), "The Duality of Technology: Rethinking

the Concept of Technology in Organizations," Organization Sci-

ence, 3, 3. 398-427.

and J. J. Baroudi (1991), "Studying Information Technology In

Organizations: Research Approaches and Assumptions," Infor-

mation Sy.<!tems Research, 2. 1. 1-28.

and D. Robey (1991), "Information Technology and the Struc-

turing of Organizations," Information Systems Research, 2, 2,

143-169.Pasmore, W. A. (1988). Designing Effective Organizations: The So-

ciotechnical Systems Perspective, NY: John Wiley.

Perrow, C. (1986), Complex Organizations: A Critical Essay, 3rdedition. New York: Random House.

Pinfieid. L. (1986), "A Field Evaluation of Perspectives on Organiza-

tional Decision Making," Administrative Science Quarterly, 31,

365-388.

Pinsonneault, A. and K. L. Kraemer (1989), "The Impact of Techno-

logical Support on Groups: An Assessment of the Empirical

Research," Decision Support Systems, 5. 2, 197-216.

Poole, M. S. and G. DeSanctis (1990), "Understanding the Use of

Group Decision Support Systems," in J. Fulk and C. Steinfield

(Eds.), Organizations and Communication Technology, Beverly

Hills, CA: Sage. 173-193.

and (1992), "Microlevel Structuration in Computer-sup-

ported Group Decision-making," Human Communication Re-

search, 91, 1, 5-49.

, G. DeSanctis, L. Kirsch and M. Jackson (1991), "An Observa-

tional Study of Everyday Use of a Group Decision Support

System," presented at the Twenty-Fourth International Confer-

ence on System Sciences, Kauai, HI January 5-7.

, J. P. Folger and D. E, Hewes (1987), "Methods of Interaction

Analysis," in G, R. Miller and M. Roloff (Eds.), Explorations in

Interpersonal Communication, 2nd edition, 220-256, BeverlyHills, CA: Sage.

M. Holmes, R. Watson and G. DeSanctis (1991), "Conflict

Management in a Computer-supported Meeting Environment,"

Management Science, 37, 8, 926-953,

, D. R. Seibold and R. D. McPhee, (1986), "Group Decision-

making as a Structurational Process," Quarterly Journal of

Speech, 71, 74-102.

Putnam, L. L. (1981), "Procedural Messages and Small Group Work

Climates: A Lag Sequential Analysis," in M. Burgoon (Ed.),

Communication Yearbook, 5, pp. 331-350, New Brunswick, NJ:

Transaction Books.

Rice, R. and Associates (Ed.) (1984), The New Media: Communica-

tion, Research, and Technology, Beverly Hills, CA: Sage Publica-

tions.

Rohrbaugh, J, (1989), "A Competing Values Approach to the Study

of Group Decision Support Systems," Proceedings of the

Twenty-Second Annual Hawaii Intemational Conference on Sys-

tem Sciences, 4, 158-166.

Reichers, A. E. (1987), "An Interactionist Perspective on Newcomer

Socialization Rates," Academy of Management Review, 12, 2,

278-287.

Robey, D., K. A. Vaverek and C. S. Saunders (1989), "Social

Structure and Electronic Communication: A Study of Computer

Conferencing." paper presented at the Hawaiian International

Conference on Information Systems, Kona, HI.

Salancik, G. R. and J. Pfeffer (1978), "A Social Information Process-

ing Approach to Job Attitudes and Task Design," Administra-

tive Science Quarterly, 23, 224-253.

Sambamurthy, V. (1990), "Supporting Group Performance during

Stakeholder Analysis: The Effects of Alternative Computer-

based Designs," unpublished Ph.D. dissertation. University of

Minnesota.

and M. S. Poole (1992), "The Effects of Variations in Capabili-

ties of GDSS Designs on Management of Cognitive Conflict in

Groups." Information Systems Research, i, 3, 224-251.

Saunders, C. and J. W. Jones (1990), "Temporal Sequences in

Information Acquisition for Decision Making: A Focus on

Source and Medium," Academy of Management Review, 15, 1,

29-46.

Scott, R, (1987), "The Adolescence of Institutional Theory," Admin-

istrative Science Quarterly, 32, 493-511.

Selznick, P. (1969), Law, Society and Industrial Justice, New York:

Russell Sage Foundation.

Siegel, J., V. Dubrovsky, S. Kiesler and T. W. McGuire (1986).

"Group Processes in Computer-mediated Communication," Or-

ganizational Behavior and Human Decision Processes, 37,

157-187.

Silver, M. S. (1991), Systems That Support Decision Makers: Descrip-

tion and Analysis, New York: John Wiley.

Spillman, B., R. Spillman and J. Bezdek (1980), "A Fuzzy Analysis of

Consensus in Small Groups," in P. O. Wang and S. K. Chang



(Ed.s.). Fuzzy Sets: Theory and Application to Policy Analysis and

Information Systems, 291-308, NY: Plenum.

Thompson, J. B. (Ed.) (1981), Paul Ricoeur: Hermeneutics and the

Human Sciences, London: Cambridge University Press.

Trevino, L. K., R. H. Lengel and R. L. Daft (1987), "Media Symbol-

ism. Media Richness, and Media Choice in Organizations,"

Communication Research, 14, 5, 553-574.

Vaiacich, J. S., A. R. Dennis and J. F. Nunamaker, Jr. (1992),

"Group Size and Anonymity Effects on Computer-mediated

Idea Generation," Small Group Research.

Van de Ven, A. H. and A. L. Delbecq (1974), "The Effectiveness of

Nominal, Delphi, and Interacting Group Decision Making Pro-

cesses," Academy of Management Journal, 17, 4, 605-621.

Vician, C , G. DeSanctis, M. S. Poole and B. M. Jackson (1992).

"Using Group Technologies to Support the Design of "Lights

Out" Computing Systems: A Case Study," in J. DeGross (Ed.),

Proceedings of the International Federation of Information PrO'

cessing Working Group 8.2.

Walther, J. B. (1992), "Interpersonal Effects in Computer-mediated

Interaction: A Relational Perspective," Communication Re-

search. 19, 1, 52-90.

Watson, R. T., G. DeSanctis and M. S. Poole (1988), "Using a GDSSto Facilitate Group Consensus: Some Intended and UnintendedConsequences," MIS Quarterly, 12, 3, 463-477.

Zigurs, I., G. DeSanctis and J. Billingsley (1990), "Attitudinal Devel-

opment in Computer-supported Meetings: An Exploratory

Study," Journal of Management Information Systems, 7, 4, 51-70.

, M. S. Poole and G. DeSanctis (1988), "Computer Support of

Group Decision Making: A Communicat ion-based

Investigation," MIS Quarterly, 12, 4, 625-644.

Accepted by Robert W. Zmud, received August 8, 1989. This paper has been with the authors for four revisions.


Date post:	07-Nov-2014
Category:	Documents
Upload:	marcus-adomey
View:	53 times
Download:	0 times

Adaptive Structuration Theory Poole and Desanctis

Documents