MIS Problems and Failures: A Socio- Technical Perspective PART I

STS Perspective

MIS Problems andFailures: A Socio-Technical PerspectivePART I: THE CAUSESBy: Robert P. Bostrom

J. Stephen Heinen

AbstractMany of the problems arjd failures of ManagementInformation Systems (MIS) and ManagementScience/Operations Research (MS/OR) projectshave been attributed to organizational behavioralproblems. The millions of dollars organizationsspend on MIS and MS/OR development are of littlebenefit because systems continue to fail. Steps canbe taken to understand and solve these behavioralprobiems.

This article argues that in most oases thesebehavioral problems are the result of inadequatedesigns. These bad designs are attributed to theway MIS systems designers view organizations,their members, and the function of an MIS withinthem. I.e., systems designers' frames of reference.These frames of reference cause faulty designchoices and failures to perceive better design alter-natives. Seven conditions are discussed whichreflect current systems designers' points of view.

The discussion of these conditions demonstratesthe need to refrawe MIS design methodologywithin the Socio-Technicai Systems (STS) designapproach and change systems designers'perspectives. The STS approach is introduced as arealistic view of organizations and a way to changethem.

This article is the first of two to appear in conse-cutive issues of the MIS Quarterly. The purpose ofthis first article is to demonstrate the need for thesrs approaj:h. The second will present the basicconcepts and principles of the STS methodologyand how it can be utilized in the design of an MIS.

Keywords: MIS problems and fallurBS, behavioralprobiems, MIS design and Implementation,socio-technical design, systems designers

Categories: 1.3, 2.10, 2.40, 3.30, 3.50

IntroductionThe major reason Management InformationSystems (MIS) have had so many failures andproblems is the way systems designers vieworganizations, their members, and the functionof an MIS within them.' These views areimbedded in a design methodology or approachwhich guides development and implementationof an MIS. This article is the first of two which willappear in consecutive issues. The intent of thetwo articles is to provide the MIS practitioner andresearcher with an overview of a designapproach which is based on a more realistic viewof organizations. This design approach isreferred to as the Socio-Technical System (STS)design. STS is a fairly recent development in thequest for organizational systems which are bothmore satisfying to their members and moreeffective in meeting task requirements. Thisapproach is used for redesigning existing worksystems as well as for new site designs. Webelieve that the utilization of the STS approachwill solve many of the problems facing MIS andsubstantially reduce the number of MIS failures.'

The STS approach assumes that an organizationor organizational work system, e.g., a depart-ment, can be described as a socio-technicalsystem. In other words, a work system Is madeup of two jointly independent, but correlativeinteracting systems — the social and thetechnical. The technical system is concernedwith the processes, tasks, and technologyneeded to transform inputs to outputs. The socialsystem is concerned with the attributes of people(e.g., attitudes, skills, values), the relationshipsamong people, reward systems, and authoritystructures. It is assumed that the outputs of thework system are the result of joint interactionsbetween these two systems. Thus, any design orredesign of a work system must deal with both

1. Although we wilt use the term MIS, our argumentswould generally apply to any computer-based infor-mation systems effort The basic difference betweenMIS and other Information systems is Its decision-making support orientation.

2. Since the STS design approach Is an organizationaldesign technique, we would also stress itsapplicability for the design and management of anMIS department. Many of the organizational and othertypes of problems facing an MIS department could beeffectively attacked with the STS approach.

MIS Quarterly / September 1977 17

STS Perspective

systems in an integrated form. The purpose ofthis article is to demonstrate the need in MISdesign for the STS approach. The second articlewill present the basic concepts and principles ofSTS and how they can be utilized in the designof an MIS.In these articles, MIS is viewed as an interventionstrategy. To intervene is to enter into an ongoingwork system for the purpose of improving itsfunction. The terms interventionist, changeagent, and organizational designer will be usedinterchangeably. They all imply someone who ischanging, designing, or redesigning a worksystem. An MIS designer is a specific example ofan Interventionist. We will also use the termsintervention strategy, change program, and neworganizational design synonymously; they alldenote a planned change effort. An MIS is anexample of an intervention strategy. Whendiscussing concepts that have general appli-cation we will use the term change agent, andwhen discussing specific issues regarding MIS,the specific term systems designer will be used.The use of the these general terms is to stress tothe reader the change aspects of MIS implemen-tation and to make the point that MIS designcannot be isolated from organizational design. Itis important to note that the use of the terms"design" and "planned change" make explicitthe view that organizations are not "natural"phenomena but are artificial, man-madeinventions.

When one intervenes in a work system, twopotential Improvements are possible. The first isan improvement in task accomplishment, i.e.,improvement in productivity and/or quality of theproduct, reduced costs, etc. The second is animprovement in the quality of working life (QWL)of the work system's members. Historically, theidea of QWL has included only the issues ofwages, hours, and physical conditions. Theseissues are still included in any definition of QWL,but the concept is expanding to include otherconcerns such as meaningful and satisfyingwork, control and influence, and opportunitiesfor learning. Consistent with the expandingdefinition of QWL, wewill consider a high-qualityworklife as one that involves an interesting,challenging, and responsible job as perceived bythe job holder. We would be the first to admit thatthere are individual differences which makeunanimity about what constitutes a high or lowquality worklife almost impossible. It is important

then to be able to adapt as much as possibie thisgeneral definition of high quality worklife toindividual differences, The typical goal of anintervention into the technical system is animprovement in task accomplishment, whileinterventions which focus on the social systemtend to look for improvement in QWL. The STSapproach argues that any intervention must dealwith these goals simultaneously.

BackgroundLucas [40] summarized the results of hisempirical research involving over 2000 infor-mation system users in 16 organizations asfollows:

it is our contention that the major reasonmost information systems have failed is thatwe have Ignored orgaiiizational behaviorproblems in the design and operation ofcomputer-based information systems [40,p. 61.

The behavioral problem argument is supportedby other authors and researchers in the field ofMIS [2. 27. 3.1. 36. 49, 51, 55, 56. 64. 65]. Inaddition, support for the above thesis is growingin the related field of Management Science andOperation Research (MS/OR) [57. 59]. Thebehavioral problems range from outrightsabotage to non-usage of the MIS [23]. Theimplication of these findings is that if steps arenot taken to understand and solve these organi-zational behavioral/social system problems, themillions of dollars organizations spend on thedevelopment of information ^stems will be oflittle benefit because systems will continue tofail.

Many would argue that these social systemproblems are the result of the inflexibility ofcomputer-related technology. Further, theywould argue that this technology must beaccepted as a necessary requirement if we wantto maximize our wealth and comfort. People wiiladapt!

We would disagree. Our basic premise is thatcomputer-related technology is essentiallyneutral: whether its application succeeds or failsdepends entirely on the decisions that are madeon how it shall be used.' This position is some-

3. This basic premise appiied to all technoiogies is oneof the basic buiiding biocica of the STS approach.

18 MIS Quarterly I September 1977

S7"S Perspective

where between two extreme positions, oneindicating that to maximize the use of tech-noiogy, and the other be wary of technology forit will only lead to disaster. The neutral positionbecomes clearer and the extreme positions tendto dissolve as one realizes that when technologyis mentioned, the reference is not to basictechnoiogy. but a technical design put togetherby designers for a given set of technicalrequirements. The problem is that the basictechnology is only one input into the designprocess. Other major forces which influence thetechnical design are the designer's knowledge,skills, values, and assumptions about people andorganizations. For example, a technical designwhich feeds back information from a machine tosomeone other than the person operating themachine says something about the designer'sassumptions about people. This feedback loop iscertainly not a technical requirement of thedesign. But once designed this way, it is passedon to the user as a technical requirement. Thus,most technical system design includes somesocial system design. Failure of designers andusers to recognize this fact leads to manydysfunctional consequences in the socialsystem.

These forces mold frames of reference whichserve as perceptual filters through which oneperceives the world and provides guides foractions. Thus, a work system designer's frame ofreference serves as a foundation for examining,understanding, and changing organizations. Interms of the design and implementation of anMIS. the system designer's frame of referencehelps determine the perceived design alter-natives and the chosen design alternative. Theyalso influence the perceived change strategiesand chosen change strategy. Design alternativesinclude combinations of hardware, software,operating procedures, work flow. etc. Changestrategies include decisions on issues such asuser involvement, user education and training,project team deveiopment, and implementationplans. Change strategies are important becausepeople respond to the manner in which changesare determined and implemented as much asthey do to the actual change 14, 19. 20, 51, 59].The current social system/behavioral problemsassociated with an MIS originate In the lack ofawareness of available design alternatives andchange strategies and faulty decisionsconcerning perceived options [see 31 for case

examples]. Both problems stem from the currentframes of reference of system designers.

This article describes seven conditions which arethe major causes of inadequate designs andunsuccessful change strategies, and whichcollectively reflect the existing frames ofreference. We wiil demonstrate the need for thediffusion of the STS design methodology intothe field of MIS by making the current frames ofreference explicit through the discussion of thaseven conditions. Figure 1 summarizes the basisfor this approach.

The term "systems designers" is being used toinclude all people who actually influence MISdesign decisions, i.e.. systems analysts, users,union officials, top management, computersystem designers, etc. As pointed out inCondition 2, today the dominant power resideswith the computer speciaiists, i.e., analysts,designers, programmers, etc.; they consciouslyor unconsciously direct the development anduse of computer technology in organizations. Itis very important not to place the blame for theinappropriate designs and change strategiesentirely upon the person who occupies thesystems designer's role. The formation of framesof reference, reflected in the conditions,described In Figure 1 is a very complex processwhich is not completely under the consciouscontrol of the person. This fact also implies thatpeople are not always aware of the content oftheir frames of reference. In addition, theresulting actions or behaviors of designers in agiven situation are not always based on theirframes of reference. For example, the organiza-tional reward system may force or supportbehaviors incongruent with the designer's frameof reference.

ConditionsCondition 1: Systems Designer'sImpiicit TheoriesSystems designers clearly make assumptionsabout people — e.g., "people are poor infor-mation processors;" about organizations —e.g.." information flow downward must becontrolled;" and about the change process —e.g., "we must get users to participate or theywilt not accept the system." Whatever the level ofsophistication, these assumptions or set of


STS Perspective

beliefs, taken as a set, could be described as thesystems designer's implicit theories aboutpeople, organizations, and the change process.These theories are "implicit" because systemsdesigners have never attempted to develop acarefully worked through, "logical" descriptionof their assumptions and beliefs. The designer'sassumptions about people and organizationsaffect which design alternatives are consideredand chosen. Additionally, the designer'sassumptions about how one should changeorganizations influence the change strategieschosen. The case study research of Pettigrew[53] and Hedberg, ef al. [31] clearly supportsthese conclusions.

No general theory will reflect precisely theimplicit theories of any single designer Never-theless, scholars have attempted to capture themain thrust of key concepts and assumptions.Two theories which scholars have beenpresenting are Theory X, traditional or machinetheory, and Theory Y, human resource theory.Theory X assumes a person is one who likesorder, wishes to work within tightly specifiedboundaries, and does not want to have a greatdeal of personal control over one's activities.Theory Y assumes a person is a responsible, self-achieving individual who can take full control ofone's work environment. Schein [58] points outthat there is a great deal of confusion aboutTheories X and Y. They are viewed as managerialphilosophies, management styles, types oforganizations, etc. Thus, we should stressagain that Theories X and Y are sets ofassumptions about human nature that a givenperson holds, consciously or subconsciously.

These basic assumptions about people shapeone's approaches to organizational designs,work and job designs, and change strategies. Forexample, a design based on Theory Xassumptions would tend to create a tightlystructured organization, with precise jobdefinitions and clear tines of hierarchicalauthority, emphasizing order and stability asnecessary to obtain technical efficiency. On theother hand, a design based on Theory Yassumptions would tend to create a flexibleorganization which has a great deal of self-direction and self-control at all levels becausethe integration of individual growth withtechnological Improvement is seen as the key

to organizational effectiveness.*

Which theories do systems designers hold?It is quite apparent that system designers ingeneral hold a Theory X view. The systemsdesigner's primary role today seems to be one ofservant to thetechnical system needs, consistentwith Theory X assumptions. Hedberg andMumford [30, 31] have empirically verified thatEuropean systems designers'operational modelof a non-management person and theappropriate organizational structure for non-management people is closer to Theory X thanTheory Y. Unpublished studies by Bostrom andby Taylor [61] have found in the U.S. a similarTheory X orientation toward management andnon-management persons.* Finally, one needsonly to examine the MiS literature to see theextent of the Theory X orientation.

An illustration of the systems designer'sTheory X view is the "operating unit" concept ofBoguslaw [6]. Boguslaw concluded from hisresearch that the "new Utopians" or systemsdesigners retain an aloofness from the humanand social problems by treating the humancomponent as just another operating unit withinthe system with people taking their placesalongside computers, display consoles, andother forms of system operating units. Takingthis view, the systems designer is the expertwho analyzes the problem, defines it, andprovides the solution. Any human problems aretreated by adjusting the operating units throughtraining, incentives, etc., to suit the technicalsystem.

4. It is beyond the scope of this article to give any morethan a brief description of the theories. For furtherdiscussion of the theories, see Miles [471, McGregor[451, Davis [15. 19] and Schein [581. Miles, in histreatment, adds a third major theory. HumanRelations. We did not include the Human Relationstheory because from a design framework it makes thesame basic assumptions as Theory X.

5. The research done by Bostrom was a pilot studyfocusing on a number of issues in this article. Thesample was 34 graduate students at the University otMinnesota who had an average of three yearscomputer systems design experience. This study willbe referenced throughout the paper.All of the data referenced represents computersystems designers' views. Hedberg and Mumfordindicata that the views on the usep side are moreTheory Y. The U.S. data does not support thisdifferenca found in Europe [see 471-

20 MIS Quarierly I September 1977

STS Perspective

SystemsDesigners'frames ofreference

cause

reflected in

Faulty designchoices; andthe failure toperceive betterdesign alternatives

to

Baddesigns

cause

Behavioralproblems

• leadIMIS problems

^ and failures

Seven Conditions:

1. "Implicit" theories held by systems designers about organizations,their members, and how to change them.

2. The concept of responsibility held by systems designers.

3. Limited conceptualizations of frameworks for organizational worksystems or user systems used by systems designers in the designprocess, i.e., non-systemic approach.

4. Limited view of the goal of an MIS implementation held by designers.

Failure of the systems designers to include relevant persons in thedesign referent group. Who is the user?

The rational/static view of the systems development process heldby systems designers.

5.

6.

7.

demonstratedthe needto;

The limited set of change technologies available to systemsdesigners who attempt to improve organizations.

a reframe MIS designmethodology within theSTS design approach; and

b. change systems designers'frames of reference.

Figure 1. The Rationale for the Soclo-Technlcal Desfgn Methodology

MIS Quarterly I September 1977 21

STS Perspective

Greater emphasis on user involvement in therecent MIS literature might at first seem toindicate a shift toward a Theory Y orientation;however, in most cases the approach is used togain acceptance by the users of the systemsdesigner's solution, rather than create aneffective reciprocal relationship between theMIS resource and the user system. Argyris [2]and Bostrom both found participation by userswas a means to gain acceptance rather thancollaborative problem solving.

Theories X and Y assumptions focus on themotivational patterns of individuals. Thetendency of systems designers to hold Theory Xassumptions is related to the implicit theoriesthey hold about people as informationprocessors. Two somewhat paradoxicalassumptions tend to be emphasized in theliterature:

1. Decision-makers are not getting enough ofthe right information.

2. The person is not every efficient informationprocessor.

Designs based on the "give them more" theoryhave been convincingly attacked by a number ofauthors [1. 29]. Thus, one no longer sees thisidea referenced in the literature, but manyapplications continue to appear in practice.

The research on human information processinggenerally supports the "inefficient" theory. Thisleads many MIS writers to advocate a designapproach which would automate as manydecisions as possible. Decision aids shouidbe supplied, if the decision is not fullyprogrammable. The difficulty with this designapproach is that it is congruent with a Theory Xview that a person is replaceable by a machine, itneglects the fact that persons have informationprocessing strengths. We would agree withHedberg [28] that an MIS design should be basedon a detailed knowledge of both the strengthsand weaknesses of human beings as informationprocessors. The division of labor between theperson and the computer in their respective rolesIn the MIS design should be guided by theprinciple of comparative advantages, keeping Inmind the motivational consequences.

implicit theories are a key input into theformation of the systems designer's frame of

reference. It will be virtually impossible toalleviate the remaining conditions discussed inthis section without first establishing anunderstanding of the basic assumptions.Theory X assumptions currently are reflected inMIS designs, design procedures, and designmethods. Based on available evidence fromother fields [15. 16. 35. 45, 47. 58] and problemsin MIS. there is a very strong indication thatthese Theory X assumptions are inaccurate.Thus, more effective MIS designs will result frommaking designers' implicit theories explicit anddetermining if their assumptions about theirclients are accurate and congruent. Knowledgeabout motivational patterns and informationprocessing capabilities of people need to bedeveloped and diffused to support this latterstage. •, i

Condition 2: Systems Designer'sConcept of Responsibility

The concept of responsibility is a very importantaspect in change theory, but it has received verylittle discussion in the literature.* The focus herewill be on the concept of responsibility as itapplies to one type of change, the introduction ofa new MIS. The critical question is. "Who isresponsible for the change effort?" Based onavailable research, the systems designers arethe ones taking responsibility for the systemdevelopment process [6, 31]. This concept ofresponsibility is congruent with the Theory Xassumptions that people will not take and do notwant responsibility. Therefore, some externalsource must take responsibilty for the client.

The difficuity with this whole approach toresponsibility from a change theory perspectiveis that it overlooks the fact that the changeagent can never truly assume the responsibilityfor another's change because only thaiindividual is capable of changing his/her ownbehavior. The change agent can only facilitate orinhibit individual change through the handling of

6 We cannot begin to develop the Importance of thisaspect In this article. It ts sufficient to say that thisIssue — "Who is responsible: change agent orclient" — will sooner or later have to be dealt with bysomeone in a change agent or helping role.


STS Perspective

the change programs.' Therefore, we wouldargue that an MIS change effort can besuccessful only if the client assumes theresponsibility for its success. The systemsdesigner then should act in such a way that theuser retains the responsibility.

Typically systems designers would denycompletely that they assume responsibility: "Ionly do what the customer tells me to do. Iimplement the values [implicit theories] ofsomeone else, rather than my own. And in theabsence of specific instructions, I use a guidelineof technical efficiency or cost or speed orsomething similar" [6, p. 198]. This responseraises a number of issues. First, it highlightsthe issue of "who is the user?" Discussion of thispoint is presented under Condition 5. Secondiy,it stresses the point that designers makedecisions based on the goal of technicaloptimization of cost speed, etc., which isdiscussed in Condition 4. Finally, the responseclearly recognizes that the user has theresponsibility for all components of the systemprovided that the user can specify them incomplete and rigorous detail; however, thedesigner assumes responsibility for ail items thatcannot be specified in this fashion. Since usersare not experts and are unable to specify detaileddesign guidelines, the designers end up takingresponsibility through their de facto decisionmaking.

The belief that the client should assumeresponsibility for the MIS change effort iscongruent with the Theory Y assumptions. Thisbelief further assumes that both the systemsdesigners and clients work collaboratlvely, butare able to identify and to understand situational

7. A similar argument is developed by Bowers, et al. [9]in their development of the Principle of Succession.The indirect nature of change is difficult to see incertain situations. For example, suppose a newmachine is brought in to improve task accomplish-ment, i.e.. more productive behavior on the part ofmachine operations. The new machine will force newbehavior? and constrain other behaviors of theoperators. But the operators may develop variousdefensive behaviors: a norm of low productivity,physical withdrawal, or absenteeism. The defensivebehaviors may cause the change effort to fail, l.e.. noimprovement or even a decrease in productivity.Thus, although the change, the new machine, appears:o have direct effects, the success of the change efforts still basically the responsibility of the operators,.«., luccess based on Indirect effects.

constraints which necessitate a noncollaborativerelationship.

Some attempts have been initiated to implementgreater user responsibility:

1. a) The user departments are given finan-cial responsibility for the MIS.

b) The systems analysts are placed withinthe user departments with primaryresponsibility for the final detaileddesign [53].

2. a) The user is madethehead of the projectteam.

b) A user steering commit tee isestablished.

c) User representatives are placed on theproject team.

3. Power is shared during the system develop-ment process, a collaborative designapproach [28, 36, 41].

4. Ongoing management of the system is doneby the user.

Approaches 1 and 2 and other similar structuraltechniques may be useful in certain situations,but they do not attack the responsibility issuedirectly. Approach 4 is too late for the implemen-tation of responsibility. Once the system designis completed, there is typically very little that canbe done about it for a period of time.

Note that options 2a and 2c are structuralmechanisms for implementing some type ofsharing of power between users and systemsdesigners during the development process."1 hey were not included under 3 because in mostsituations which stress user participation, usersdo not significantly influence the chosendesign alternative. The major reasons for usingthese mechanisms appear to be the acceptanceof the new MIS by the users and/or the gatheringof information in order to design the system. Anexample of this latter approach is shown in

8- It is important to note that there are really three partiesthat have influence or power in the design of an MISBesides the user and the designer, thera is thecomputer equipment manufacturer whose technicaldesigns have a large influence in the ultimate designof an MIS. Users and designers need bettermechanisms to infiuenca the design of componentsof systems supplied by computer manufacturers.

MIS Quarterly I September 1977 23

STS Perspective

Argyris' [2] research on an MS/OR project team.The team continually attempted to elicit infor-mation about decision-making from the linemanagement. The line management felt atremendous need to be open with the MS/ORteam, but saw the team as unrevealing andsecretive in their use of the information given.The MS/OR team viewed itself as a rationalreform group that would supplement the"inadequate" line managers with efficientsystems.

Even if the situation existed for a truly colia-borative design, there still wouid be a majorproblem in the complete focus on the systemdesign process. Problem solving takes place inthe language and abstractions of the computerspecialists and leads to the need for educatingand training of user representatives in computerand systems design skills. Given this focus, theonly way one can have a truly collaborativedesign is to train the user to be an expert so thathe/she can help the computer specialist createdetailed design plans. However, as Hedbergargues, "including employee representatives inactual design activities and training them tobecome good systems analysts may be a wasteof time and skills" [28. p. 219J. As an alternative.Hedberg [28] proposes that users focus on thestrategic design which precedes the systemdesign stage, formulating and reformulating thegoals and policies which guide the systemsdesign activities.

Hedberg's approach affordsthe best mechanismfor implementing user responsibility. The majortask of the designers is to translate the output ofthe strategic design process into technicallyoperational solutions. The problem today Is thatthe strategic design process is not clearlyrecognized. Reference to anything resembling astrategic design process is clearly absent In theliterature and/orthesystem life cycle models andproject management schemes that are used Inpractice. Thus, the strategic decisions usuallyare dealt with implicitly during the systemdesign process, giving systems designers thedual role of policy maker and interpreter [31].This observation helps explain the apparentparadox that designers do not perceive them-selves as decision-makers or organizationaidesigners [30. 31]. They feei they have very littleinfluence in the design process and few alter-

native design choices. The explanation Is thatdesigners' existing frames of reference implicitlydetermine the domains of tower level decisions.Improved design methods must aim at makingstrategic criteria and rules explicit and subjecttodesign.

Currently, systems designers are takingresponsibiiity for MIS designs, but they do noteffectively use the power it provides. The reaipower resides in the prevailing frames ofreference which implicitly guide many designdecisions. The reciprocal part of the problem isthat users let designers take the responsibility,although on the user side there seems to be anapparent paradox. Guthrie [27] in his empiricalresearch on 2000 middle managers In Canada,found that 80% of the responding managerswanted to have a iot of influence in the MISdesign and implementation, but were unwillingto devote the study, time, and effort required tomake their participation meaningful. The largeuser commitment is a resutt of the reliance ontechnical models for user/designer interaction.More attention needs to be given to the strategicdesign phase and the appropriate sharing ofmodels, assumptions, and goals between usersand designers in order to develop meaningfulcollaboration. MIS designs, in our opinion, havea high probability of failure if the users do notassume responsibility for them.

Condition 3: LimitedFrameworks — Non-systemicViewThe third condition responsible for inadequatedesigns was the limited conceptualization ofwork systems/user systems used by systemsdesigners in the design process. The primarytargets of an operational information system ordecision-support MIS are the decision-making,data collection, data manipulation, and datatransmission tasks of the work system. Uponcompletion of an analysis of informationrequirements and flow, the MIS Is designed toreallocate data processing and decision-makingtasks between people and computer-relatedtechnology, and to create new tasks and modifyold ones to support this reallocation. This limitedfocus on decision-making and data processingtasks is refiected in the traditional framework

24 MIS Ouarterly I September 1977

Perspective

SocialSystem

TechnicalSystem

Structure

People

Technology

Tasks

MIS(Direct)

Figure 2. The Interacting Variable Classes Within a Work System

which views work systems in an informationprocessing system. This framework is presentedin almost every book dealing with MIS. Anexample is the book by Blumenthal [5] in whichhe concludes that the systems designer shouldconceptualize a work system as a set of com-plementary or interdependent informationsystems.

The limited focus on particular changes in thetask and technology variables leads the systemsdesigner to ignore the fact that these changescause more changes within other variables inthe work system. These other changes arelabeled secondary changes or effects, becausethey were not given primary consideration in theMIS design. Substantial, changes in the workrelationships among people accompanychanges in task structure. The strongassociation between these work relationshipschanges and the attitudes, motivations, and theinterpersonal behavior of the individuals withinthe system [3, 4] is of particular importance(3, 4]. An equally important relationship existsbetween these people variables and the changesin the task structure [ 12,17,19,25.34,38]. Thesesecondary changes in the work relationships andpeople variables are as important as the changes

in the task and technology variable. All of thesetypes of changes should be designed tocomplement and reinforce each other as shownin Figure 2. The Mumford [49] and Whisler [65]studies of operational information systemsIllustrate the importance of secondary effects.Some of the direct and secondary effects foundwere:

DIRECT

D1. Modification of old tasks or the creationof new tasks which lack interest andchallenge.

D2. Many decisions made by clerks andtheir supervisors have been automatedor passed up to middle management.The focus of decision-making movesupward.

SECONDARY

51. The direct changes indicate a de-enrichment of the supervisor and

. . Clark jobs which can lead to towermotivation, decreases in job satis-faction, etc.

52. Clerks seem to work less at their ownpace and are more tied to deadlines forcomputer runs.


STS Perspective

53. Clerks and supervisors tend to commu-nicate less with each other, whilemiddle managers increase theirinteractions with each other.

54. Strong trend toward increased centrali-zation of control.

These secondary effects ciearly indicate a lowerquality of working life for the cierks and super-visors. Argyris [2] discusses the possiblesecondary effects of a decision-supported MISIntervention. He concludes that a managerwhose decisions and alternatives are prescribedby sophisticated decision-making aigorithmsmay feel his/her action alternatives reduced andopportunities for effective, satisfying perform-ance diminished or iost.

As MiS practitioners and researchers, we prideourselves on being very system-oriented people.It should be clear by now that this is not the case;we are not very systemic. We do not understandthe relationship between the variables wemanipulate and other variabies within the worksystem. This statement is not intended to negatethe importance of the information processingview. Many researchers and practitioners acrossmultiple disciplines have concluded thatinformation is the critical resource on whichorganizations operate. Our point is that ifsystems designers want to improve theinformation processing system of an organi-zation, they must incorporate their informationprocessing view into a more complete systemicframework.

Condition 4: Limited GoalOrientation — Optimizing theTechnical SystemThe non-systemic view held by systemsdesigners reflects their servant role to thetechnical system. Thus, we would expect to findthat the designer's view of the goal of an MISwould be the typical goal of an interventionwhich focuses on the technical system, i.e.,improvement in task accomplishment. Eventhose changes In the social system explicitiydesigned into the MiS are attempts to improvetask accomplishment through better control ofthe variations in the technical system. Aninstance of this situation would be a

performance-monitoring MIS where a designchoice must be made to feed back dataconcerning individual performance andperformance-related problems either to amanager or to the employee performing thetasks. Current frames of reference identify themanager as the most efficient choice, i.e.. abetter control mechanism. Our point is that thischoice is made without concern for quality ofworking life issues. It is also worth noting thatthe research literature would certainly notsupport the choice of the manager in thissituation either from an improvement in taskaccomplishment, better control.orOWL [10.13].The MIS literature clearly indicatesthedominantgoal of an MIS intervention is technicalsystem optimization. Empirically this fact hasbeen demonstrated by Hedberg and Mumford[30. 31 ] in Europe and by Bostrom and by Taylor[61] in the U.S. These researchers found thatsystem designers mentioned most frequentlythat "greater efficiency" and "better informationfor management" were the prime contributionsof an implementation of an MIS. In addition, onecan examine almost any text and find authoritiesin the field stating such things as:

The computer resource exists solely to helpataff offices and operating units to executetheir responsibilities tjetter through cheaperprocessing of data, more efficient organi-zation of information systems, and procure-ment and development of information that istoo expensive to obtain otherwise. Theresource has no reason (or existing exceptto provide such services, and these servicesshould result in greater profits. In short, theresource has • purely economic purpose.[20. p. 68]. (Emphasis In quote by authors.)

MIS Is still a relatively young technology.Besides the notable exception of operationalinformation systems [7. 40, 43. 44, 46, 48. 49. 56,60, 65]. little field research exists on theeffectiveness of MIS interventions in terms oftechnical system optimization. Some of theindirect effects of these operational informationsystem interventions were cited in Condition 3.The effects ciearly indicated a lack of concern forthe improvement In the quality of working life forcertain users, clerks, and supervisors. Thissituation is usually rationalized by systemsdesigners as being inevitable due to thecherished assumption that the two goats are at

26 MIS Ouarterly I September 1977

S7S Perspective

opposite ends of the same continuum, i.e.,productivity/efficiency vs. quality of working life!Some of the learnings and tentative conclusionsfrom international experiments indicate thatviewing productivity/efficiency and quality ofworking life as opposing points of view is aninappropriate concept [18], They are notopposite ends of a continuum, but two differentscales. Enhancing one does not necessarilyrequire diminishing the other. Given theappropriate organizational design, experienceshows that both can increase together ri3 la35]. ' '

Even though designers are operating under thedominant goal of technical system optimization,all empirical effects in terms of quality of workingiife are not negative; however, negativeoutcomes dominate the literature. The indirecteffects cited in Condition 3, although negativefor clerks and supervisors, appear to be positivefor middle managers — jobs were enriched.Other examples of positive effects are discussedby Lucas [39] and demonstrated in the researchby Mueller [48], Mann and Williams [43], andMarencQ [44]. However, the conclusions of thesestudies are obscured by what appears to beinappropriate ongoing management of the neworganizational design. Although social systemimprovement was not planned, its potential isclearly shown in the above empirical research.Also, European researchers [31] recently foundthat there were a number of unexplored, possiblealternative designs for existing MIS designsthat would have improved not only taskaccomplishment, but also quality of working life.These results support the opinion that bothgoals can increase together for an MIS inter-vention.

Further evidence of the importance of MISIntervention on social system change can befound in the research by Taylor [61]. Heconcluded that the direct effect of a technicaisystem intervention on the social system seemsto create constraints on employee behavior. Hefound, on the other hand, the technical systemintervention provides an "unfreezing force" onthe social system. There seems to be a freedomprovided by the new technology to seek newways of behaving. This new freedom may make aplanned social system change possible.Technical system Intervention can actuallyfacilitate social system change.

MIS development efforts must take seriously thetask of using computer technology to broadenthe concept of organizational efficiency toinclude OWL issues. The social systemimprovement needs to be considered jointly withthe technical system improvement in the designof an MIS.

Condition 5: Limited DesignReferent Group — Who is theUser?Tichy [63] in his empirical study of 91 changeagents classified systems designers as "analysisfor the top" (AFT's). This classification resultedfrom the commitment by AFT's to aid, adviseand design for the heads of systems. The"secondary effects of an MIS interventiondiscussed in Condition 3 further support thelimited referent group position. Their resultsindicate that middle management was theprimary referent group and received the mainbenefits of the system while the secondary usersof the system, the supervisors and clerks, wereignored. This lack of concern for secondaryusers resulted in a de-enrichment of thesupervisor and clerks' jobs. We are left with theconclusion that systems designers designsystems for a very small set of primary users,usually managers who receive the outputs of theinformation system.*

The significance of the limited referent groupcondition is that even if the systems designer nolonger operates under conditions 1, 2. 3. and 4.he or she may do so for only a select group ofpeople, the primary users. The secondary usera,the clerks, data collection people, customers,etc.. on the periphery of the MIS are ignored inthe design. This situation is very unfortunate forthree reasons. First, the secondary users areusually those people who have continual day today contact with the system. Second, the sucessof the secondary users doing their jobs in manycases determines how successful the MIS will be.Third, the design of the MIS critically affects thejobs of the secondary users. Little attention has

9. it shouid be noted that the Introduction of databasetechnoiogy in the U.S. is forcing a larger referentgroup orientation in certain areas. But this largerreferent group stiii remains a set ol primary usera.


STS Perspective

been paid to the alternatives available for thesejob designs.

A related problem is that an MIS often feedsinformation to people who cannot take action onthat information [10. 21. 33]. The magnitude ofthis probiem grows larger as one looks at thefindings of Cummings. ef al. [13] in theircomprehensive review of 2000 references on theempirical literature on job satisfaction, industriaiorganization, and productivity. They found thatinformation/feedback appears to have thegreatest impact on productivity. They argue thatthe probable reason for the increase in produc-tivity is that information/feedback providesworkers with an indication of performance-related problems and, more importantiy. pro-vides them with meaningful feedback on theirown performance. Therefore, if the information/feedback is not given to the people who canrespond to it with action, little improvement inorganizational efficiency will result.

Systems designers must take a "total system"referent group orientation in order to improvetheir designs. Total system orientation impliesthat consideration be given in some degree to aiipeople affected by the design.

Condition 6: Rational/Static Viewof the Systems DevelopmentProcessThe traditional approach to the developmentprocess of an MIS treats design and implemen-tation as a rational, systematic process thatproceeds in a static environment. The organi-zation is assumed to have a set of weil-definedinformation processing problems which the MISdesigner expertiy analyzes. A rationai decision-making process which examines alternativedesigns in terms of the goals of the system isfollowed to select the specific design. Implemen-tation of this expert solution is then assumed totransform the organization immediately into amore effective state. This viewpoint overlooksmany of the dynamic properties of the environ-ment in which the design decisions are made. Itcan be labeled rational/static.

Attempts at what is seen as rational, decision-making In the development process frequently

become distorted and overlaid by political orpower Issues which are not always recognizednor made explicit [26. 54. 55]. Poiitical issuesarise due to conflicts of interest among variousgroups [2.39.55] and/or anticipated shifts in thebalance of power caused by the implementationof an MIS [37. 39.55]. There is no correct or idealsolution in the development of an MiS. Each setof involved parties will have its own preferences.This means that the constraints and alternativedesigns emerge only in the social interactionamong the peopie who represent different roiesand positions of power. The choice betweenalternatives quite often is based on the mobili-zation of power resources by individuals [26, 37.55].

Also unrecognized in the design and impiemen-tation of the MiS is the fact that the organizationcontinues to change and does not remain insome suspended state. The implementation isnot immediate and several transitional statesmay be passed through by the organizationduring the impiementation process. Moreattention needs to be paid to the adjustment andon-going management of the new MIS design.The actions carried out during the designprocess may also createorganizational changes.For example, gathering data from users mayraise expectations or create fears. Dickson andPowers [24] found that users undergo significantlearning during the design process. Similarlearning may also occur for systems designers.This learning may be reflected In new problems,alternatives, criteria, constraints, etc. Theimplication is that the design process is a fluid,iterative process and not a linear sequence ofsteps as in the rationai/static view.

The rational/static view could be useful in short,simple MIS projects, but it is rather obvious thatthis view does not accurately reflect the MISdevelopment process in real life. Yet mostorganizations use project pianning and controltechniques based on this view. In addition, mosttextbooks and training materials emphasize thisview. Based upon this, it is no surprise to find thatcurrent project control techniques can bedysfunctional to project success [24]. We are notnegating the importance of project planning andcontrol techniques or training. We are merelystressing that the current emphasis on therational/static view must be replaced by a more


STS Perspective

realistic view of the political/dynamic dimen-sions of MIS development.

Condition 7: Limited ChangeTechnologiesTichy [63] found in his study that systemsdesigners use a limited set of change tech-nologies. Conditions 2 - 6 cannot be effectivelyresolved with the limited change technologies atthe disposal of systems designers. Systemsdesigners essentially draw from MIS and MS/ORtechnology.

An integration between the MIS technology andother technologies is needed to produce a moreeffective intervention. It should be clear fromour previous discussion that the techniquesshould be those that tend to facilitate theimprovement of both social and technicalsystems. These techniques coupled with an MISintervention would allow us to pursue the goal ofjoint optimization. For example, the problems ofde-enriching jobs might be solved through jobdesign techniques, such as job enrichment, jobenlargement, etc. [12. 17. 19. 25. 34]. Theineffective work organizations supportinginformation systems might be improved throughthe use of autonomous or semi-autonomousgroup structures [19, 32]. In addition, many ofthe techniques that focus on the social systemcould be utilized during the developmentprocess. For example, the interpersonalproblems of the MS/OR project team identifiedby Argyris [2] might be eliminated through teambuilding interventions. The problem of gettingmore participation by secondary users in theMIS design process could be handled by usingthe survey-feedback method [4]. Thesetechniques along with many other interventionsare part of the field known as OrganizationalDevelopment (OD). The OD practitioner, usuallya behavioral or social scientist, is simplytranslating what is known about people andorganizations from the behavioral sciences intoapplicable programs whose primary interventionpoint is usually the social system.'" A furtherdiscussion of integration of MIS and OD can befound in Bostrom and Heinen [8].

Successful integration of the change tech-nologies requires new roles for both the systemsdesigner and the behavioral scientist. First, theymust develop some type of mutual design theoryso that they can effectively communicate witheach other. Second, the design effort must be atruly collaborative effort." This means that thebehavioral scientist whose past role has beenprimarily to help implement predesignedinterventions must assume a new design role.Third, for a collaborative design effort to besuccessful, each party will have to learn moreabout the other's change technologies. Thus, thebehavioral scientist needs to have a good under-standing of MIS related technology while thesystems designer must understand theusefulness and consequences of 0 0 typeinterventions.

These new roles offer a new challenge to boththe systems designer and the behavioralscientist. The new design role for the behavioralscientist offers the enormous challenge ofhelping to find systematic ways of introducingquality of work considerations into a sophisti-cated technical design. The systems designerson the other hand, will have their skills andcreativity severly tested by the necessity ofdeveloping a substantial number of alternativetechnical designs which allow more choices insocial system design within which quality ofwork considerations and constraints can beevaluated. For an excellent description of thiscollaborative process and the challenge of thenew roles, see the short case history by Lupton[42] describing the design of a new manu-facturing process by engineers, behavioralscientists, and users.

ConclusionWhen examining the social system/behavioraiproblems associated with the introduction of anew MIS, three questions become crucial:

1. What are the human or behavioralproblems?

2. What are the causes of the behavioralproblems?

10. The most notable exception to this is the job designand redesign techniques which focus on the tasks inthe technical system.

1t. For a more detailed discussion of this point, seeMumford. 8f«A [52] orCiark (11].


S7S Perspective

3, How can the causes be eliminated to solvethe behavioral problems?

MIS practitioners and researchers spent a greatmany years focusing on question 1, with somevery futile attempts at question 2; this couldbe called the story telling phase. This focus onthe story telling phase was due to inadequateframes of reference to help better explain anddeal with these situations. Today, we seem tohave switched our focus to question 3 without acomplete understanding of the answer toquestion 2. This instant solution phase, focusingon question 3, has resuited from the develop-ment of techniques that appear to solve some ofthese behavioral problems. User participationand the utilization of Job-enrichment schemesare examples of these techniques. This type ofapproach is piecemeal and ignores thecontingencies in each unique situation. Themissing link is an effective design approach,such as the STS design methodology.

Our focus in this articie has been on question 2.It was argued that the social system problemsare the result of inadequate designs andunsuccessful change strategies which arisefromthe failure of systems designers to perceiveavailable opportunities. Designers' existingframes of reference must be made explicit inorder to understand and change this situation.Seven conditions were identified and discussedas the major causes of the inappropriate designsand change strategies to aid this process. Thegoai was to demonstrate the need for thediffusion of the STS approach into the field ofMIS.

The problem of inadequate designers' framesof reference is not unique to MIS. and tendsto bea universal problem with change programs in theU.S. This problem is reflected in the piecemealapproach in which people try to changeorganizations. Job enrichment experts tend toview a work system as a set of jobs. TheirInterventions focus on enriching a particular setof jobs. They fail to realize that one person'senrichment is often another's impoverishment.Production engineers view a work system as atechnical production system. They faii to realizethat their designs contain many social systemdesign decisions which are made by default or asa byproduct of optimizing some technicalvariable. Many organizational development (OD)

practitioners see a work system only as a socialsystem. They accept the technical system asgiven and then try to adapt the social system to it.This type of OD approach deals only with a smallportion of organizational iife. All of theseapproaches are appropriate and contain usefultechniques and interventions. But theseapproaches are marginally effective becausethey faii to take into account all ofthe importantelements of the work system.

What is needed is a more reaiistic view oforganizations embedded in a solid designmethodology through which various inten/en-tions can be integrated into effective changeprograms. The STS approach argues that anydesign/redesign of a work system must dealjointly with the social and technicai systems.The application of the STS approach to MISdesign provides an adequate frame of referenceto eliminate the seven conditions outlined. TheJoint focus on social and technical systemsimplies that the STS design approach facilitatesefforts in dealing with the technicai system. Forexample, one of the major problems in MISdesign is determining information needs of awork system. The STS approach via its technicalsystem analysis techniques provides a means fordirectly determining certain information needs.

The success of the application of the STSapproach to MIS can only be determined throughits monitored use in a variety of organizationalsettings and situations. The STS designapproach must be diffused to MIS practitionersand researchers before this can be done. As afirst step in this diffusion process, oursubsequent articie to appear in the next issue ofthe MIS Quarterly will present the basic conceptsand principles of the socio-technical systemdesign approach, demonstrating how they applyto MIS design.

References(11 Ackoff. R. L -Management Misinformation Systems,"

Management Science. Vol. 14. No. 4. December 1967.121 Argyris. C. "Management Information Systems: The

Challenge to Rationality and Emotionality." Manage-ment Science. Voi. 17, No. 6. 1970. pp. B275-B292.

(31 Beer, M. The Technology of Organization Develop-ment." in Handbook of Industriai and OrganizationaiPsychology. M. Dunnette (ed.). Chicago: RandMcNally. 1976.


STS Perspective

[4] Beli. C. and French. W. Organization Development:Behavioral Science Interventions for OrganizationImprovement, Prentice-Hall, inc., Englewood Cliffs,New Jersey, 1973.

[5] Blumenthal. Sherman C. Management InformationSystems: A Framework for Planning and Develop-ment, Prentice-Hail, inc.. Engiewood Ciiffs, NewJersey, 1969.

[6] Boguslaw. Robert. The New Utopians: A Study of thaSystem Design and Social Change, Prentice-Hall,Inc.. Englewcod Cliffs. New Jersey, 1965, 213 pages.

[7] Borodin, A. and Gotlieb, C. C. "Computers andEmpioyment," CACM, Vol. 15, No. 7, 1972. pp. 695-702.

18] Bostrom. Robert P. and Heinen, J. Stephen. "Organi-zation Development: its Relationship to MiS,"Midwest AIDS Proceedings, 1974. pp. P1-P12.

[9] Bowers, David G.. Franklin, Jerome L, and Percorella,Patricia A. "Matching Probiems, Precursors, andInterventions in OD: A Systemic Approach," TheJournal of Applied Behavioral Science, Vol. 11. No. 4.1975, pp. 391-409.

110] Cherns. Albert B. "The Principles of Socio-TechnicalDesign," Human Relations, Vol. 29, No. 8, 1976.pp. 783-792.

[11] Clark, Peter A, Organizational Design: Theory andPractice, London: Tavistcck, 1972.

(12] Cooper. Robert. Job Motivation and Job Design,London: Institute of Personnel Management 5Winsley Street, Oxford Circus, WIN 7A0, England.1974.

[13] Cummings, T. G., Malloy, E. S., and Glen. R. H. "Inter-vention Strategies for Improving Productivity and theQuality of Work Life," Organizational Dynamics,Summer 1975.

t14] Davis, G. B. Management Information Systems:Conceptual Foundations, Structure and Develop-ment, McGraw Hill, New York. 1974.

[15] Davis. L. E. 'The Coming Crisis for ProductionManagement: Technology and Organization," inter-national Journal Production f^esaarch. Vol. 9, 1971,pp. 65-82.

[16] Davis, L E. "Job Satisfaction Research: The Post-Industrial View." Industriaf Relations. Vol. 10. No. 2,May 1971, pp. 176-193.

(171 Oavis, L E. and Taylor. James C. Design of Jobs.Penguin Books. London, 1972.

£18] Davis. L E. and Trist. E. L "Improving the Quality ofWork Life: Experience of the Socio-TechnicalApproach." A background paper commissioned byH.E.W. for a report on work In America, June 1972.

[19] Davis. L. E.. Cherns, Albert B.. and Associates, Ouaiityof Working Life, Volume One: Problems, Prospects,and The State of the Art, Free Press, New York. 1975.

[20] Davis, L. E., Cherns, Albert B., and Associates. Ouaiityof Working Ufe, Voiume Two: Cases and Commen-tary, Free Press, New York. 1975.

[21] Davis, L E. Talk given at National AIDS Conference,

November 1975.[22] Dearden. J, and Noian. R. L "How to Control the

Computer Resource," Harvard Business Review,November/December 1973, pp. 66-78.

[23) Dickson, G. W.. Simmons. J. K.. and Anderson, J. C."Behavioral Reactions to the Introduction of a

Management Information System: Some EmplricaiObservations," Management information SystemsResearch Center Working Paper 69-3. University ofMinnesota. 1969.

[24J Dickson, Gary and Powers, Richard. "MiS ProjectManagement: Myths, Opinions, and Reality." Cali-fornia Management Review. Voi. 15. No 3 Sprtno1973.

[25] Ford. R. Motivation Through Work ttseff. AmericanManagement Assoc.. New York, 1969.

[26] Gibson, Cyrus F. "A Methodology for ImplementationResearch." in Implementing Operations Research/Management Scienca, Randall L. Schultz and DennisP. Slevin (eds.), American Elsevier Publishing Com-pany. New York, 1975.

(27] Guthrie, Art. "Attitudes of the User-ManagersTowards Management Informat ion Systems,"Management Informatics. Voi 3, No. 5 1974 pp 221-232.

(28) Hedberg, Bo. "Computer Systems to Support Indus-trial Democracy." in Human Choice and Computers.E. Mumford and H. Sackman (eds.), AmericanElsevier Publishing Co., New York, 1975.

[29] Hedberg, Bo, "Towards A Credo of MIS Deslgn,"International Institute of Management, Berlin,Preprint l /73-5i May 1973.

[30] Hedberg, Bo and Mumford. Enid. "The Design ofComputer Systems: Man's Vision of Man as anIntegral Part of the System Design Process." mHuman Choice and Computers, E. Mumford and H,Sackman (eds.), American Elsevier Publishing Co..New York, 1975,

[31] Hedberg, Bo, Mumford. Enid. Anderson, Niels. 9t al.Computer Systems, Work Design and Job Satisfac-tion, Pergamon Press, New York, (to be available Fall1977).

[32] Herbst, P. G. Autonomous Group Functioning: AnExploration in Behavior Theory and Measurement,Tavistock Publications, London, 1962.

[33] Herbst, P. G. Socio-Technical Design: Strategies inMuitidisciplinary Research, Tavistock Publications,London. 1974, 242 pages.

[34] Herzberg F., Paul, W. J., and Robertson, K. B. "JobEnrichment Pays Off," Harvard Business Review,January/February 1968.

[35] Katzell. R. A. and Yankelovich. D. "ImproviogProtfuctivity and Job Satisfaction," OrganizationalDynamics, Summer 1975.

(36] KIrng. R. Towards a Person-Centered ComputerTechnology," Proceedings of tha ACM NationalConference, 1973. pp. 367-391.

(37] Kting. R. "Computers and Social Power." Compufersand Society, Vol. 5, No. 3, Fall 1974.

[38] Lawler, Edward E.. III. Motivation in Work Organiza-tions, Brooks/Cole Publishing Company. Monterrey,California, 1973.

[39] Lucas, Henry C. Computer Based InformationSystems in Organizations. Science ResearchAssociates, Palo Alto, California. 1973.

[40] Lucas. Henry C. Why Information Systams Fail,Columbia University Press. New York, 1975,


STS Perspective

[41] Lucas. Henry C. Toward Creative Systems Design,Columbia University Press. New York. 1975.

(42] Lupton, Tom. "Efficiency and the Quality of Worklife:The Technology of Reconciliation." OrganlzationafDynamics, Autumn 1975.

[43] Mann, F. C. and Williams. L. K. "OrganizationalImpact of Whrte-Collar Automation." Industriai Rela-tions Research Association Proceedings. 1959.

[44] Marenco. C. "Gradualism, Apathy, and Suspicion in aFrench Bank." in Office Automation. W. H. Scott (ed.),OECD, Paris. 1965.

(45] McGregor. D. The Human Side of Enterprise.McGraw-Hill Book Company, New York, 1960.

[461 Meyers, Charles A. (ed.). The Impact of Computerson Management. The MIT Press. Cambridge, Mass.,1967.

(47] Miles, Raymond E. Theories of Management: Impli-cations for Organizational Behavior and Develop-ment, McGraw-Hill Book Company. New York, 1975.

[4B] Mueller, E. Automation In an Expanding Economy.Institute for Social Research, Ann Arbor. Michigan,1969.

[491 Mumford. E. and Banks, O. The Computer and theClerk, The Villafield Press, Slasgow, England, 1967.

[50] Mumford. E. and Ward, T. B- Computers: Planningfor Peopie. B. T- Badsford Ltd., London, 1968.

151] Mumford, E., Mercer. D.. Mills. S.. and Weir, M. "TheHuman Problems of Computer Introduction."Management Decision. Vol. 10. No. 1, 1972, pp. 6-17.

(52] Mumford. E., Mercer, D.. Mills. S.. and Weir, M."Discuss the Suggestion that No Project Team Con-cerned with the Implementation of the MIS Should Bewithout a Trained Social Scientist." The ComputerBuiietin. July 197i pp. 356-358.

[531 Nolan. Richard L. and Gibson. Cyrus F. -Managingthe Four Stages of EDP Growth." Harvard BusinessReview, January/February 1974. pp. 76-88.

[54] Pettigrew, Andrew. 'Towards a Political Theory ofOrganizational Intervention." Human Reiations, Vol.28, No. 3. pp. 191-208.

(55] Pettigrew, Andrew and Mumford, Enid, implementingStrategic Decisions, Longman, New York, 1975.

(56] Pollock. Ken. e( al. improvements Needed inManaging Automated Decision-Making by Com-puters Throughout the Federal Government,Report to the Congress by the Comptroller Generalof the United States (FGMSD-76-6). 1976. availablefrom U.S. General Accounting Office, DistributionSection. P.O. Box 1020. Washington. D.C., 20013.D.C.. 20013.

(571 Rubenatein. A. H.. Radnor, M.. Buker, N.. Heiman. D.,and McCally, J. B. "Some Organizational FactorsRelated to the Effectiveness of Management ScienceGroups In Industry." Management Science. Vol. 13,No. 8. April 1967. pp. B-508 - B-518.

(58] Schein, Edgar H. "In Defense of Theory Y," Organi-zationai Dynamics. Summer 1975. pp. 17-30.

(59) Schuttz, Randall L and Slevln. Dennis P. (eds.).Implementing Operations Research/ManagementScience. American Elsevier. New York. 1975.

[60] Sollenberger, Harold M. Major Changes Caused byImplementation of a MIS, Research Monograph 4.National Association of Accountants. 1968.

[61] Taylor, James C. Technoiogy and Planned Organi-zationai Change. CRUSK: Institute for SocialResearch. The University of Michigan..Ann Arbor.Michigan, 1971. 151 pages.

(62] Taylor. James C. "A Report of Preliminary Findingsfrom February 1976 Work Organization Pilot Study."Center for Quality of Working Life, UCLA. LosAngeles, California. 90024.

(631 Tichy. Noel M. "Agents of Planned Social Change:Congruence, of Values, Cognitions and Actions."Administrative Science Quarterly. Vol. 19. No 2.1974, pp. 164-182.

(64] Tomeski. Edward A. "Building Human Factors intoComputer Applications: The Computer ProfessionMust Overcome a 'Jackass Fallacy!" ManagementDatamatics. Vol. 4. No. 4. August 1975. pp. 115-120.

[65] Whisler, T- L. The Impact of Computers in Organiza-tions, Praeger, 1970,

About the AuthorsBob Bostrom Is an Assistant Professor of MIS Inthe Department of Operations and Systems(Management at the Graduate Schooi of Busi-ness, indiana University - Bloomlngton. He hasbeen Involved with STS design approach tor overtwo years through collaborative efforts withthe Center for Quailty of Working Life (CQWL) atUCLA. The CQWL is a major source of STSknowledge In this country. He hoids a B.A.and MBA from Michigan State University, an MSin Computer Science from SUNY at Albany anda Ph.D. In MiS from the University of Minnesota.He also has extensive consulting experienceespeclaliy in the areas of database managementand MiS design. Professor Bostrom Is currentiyengaged in research reiated to the STSapproach.

Steve Heinen Is an Assistant Professor ofManagerlai Psychology at the University ofMinnesota. Steve has been Involved withproblems of organizationai change and develop-ment and has consulted with several companiesIn Minnesota and Michigan on more effectivemanagement of peopie. Steve has an AB degreefrom Xavier University of Ohio and an MA andPh.D, in psychoiogy from Michigan StateUniversity.


Documents

MIS Problems and Failures: A Socio- Technical Perspective PART I