Developing a Methodology for Designing Routine Information Systems Based on the Situational Theory of Action


of 9

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Developing a Methodology for Designing Routine Information Systems Based on the Situational Theory of Action
  DEVELOPING A METHODOLOGY FOR DESIGNINGROUTINE INFORMATION SYSTEMS BASED ON THESITUATIONAL THEORY OF ACTION Simon K Milton, Robert B Johnston, Reeva M Lederman, Vivienne Waller,Department of Information Systems, The University of Melbourne Abstract  Information systems are part of purposeful socio-technical systems and consequently theories of action may helpin understanding them. Current systems analysis and design methodologies seem to have been influenced onlyby one particular theory of action, which asserts that action results from deliberation upon an abstract representation of the world. Many disciplines have discussed an alternative ‘situational’ theory of action. Thereis no design methodology that fully supports designing systems reflecting the situational theory of action. Theaim of this paper is to describe the motivation for, and progress to date of research-in-progress developing adesign methodology based on concepts from the situational theory of action. This methodology for designingsituational information systems is being iteratively refined using a combination of case studies and actionresearch. This project is significant because many information systems fail in pressured routine environmentswhere we would argue that the situational theory of action provides a better description of purposeful activity.Key words: Situated action; Routine work systems; Systems development methodologies; Case study research; Action research. 1.   INTRODUCTION In the past 30 years a number of methodologies have been developed to assist information systems(IS) designers to produce and document systems designs. They can be referred to collectively asInformation Engineering Methodologies (IEM), with Structured Systems Analysis and DesignMethodology (SSADM, the British government standard) being a typical example. Thesemethodologies consist of prescriptive steps to be used in analysis of the task domain, and variousdesign aids used to represent and document the data models upon which databases are created and the processes that occur when transactions are recorded in the system. These methodologies have beenwidely disseminated through the IS profession by practitioners, consultants and educators. The abilityof these information systems to accomplish the goals set by those commissioning them is thus to alarge degree determined by the fitness for purpose of these methodologies.As a means of support for operations, particularly when work is complex, time-constrained but largelyroutine, systems designed using these methods are often ineffective, inefficient or not accepted by people using them. They work technically, but fail to support routines in the work context. For example, hospital systems often fail to support both the speed with which specialists and surgeons pass through wards visiting patients and the means to collect and communicate information vital for  patient care (Hardy and Drury 2001). Data collected is then not reflective of the activities undertaken by the specialists or of the decisions taken. There are consequent failures to manage or measure work  practices in the wake of the dirty data created (Lederman 2004). In a manufacturing context Johnston(1995) shows how this mode of system design marginalises the role of the environment (physical and social) of work which is key to understanding the efficiency of routines. An approach different from present systems analysis and design methodologies is needed in order to develop information systems better suited to pressured routine environments.In previous papers (Johnston and Milton 2001; Johnston and Milton 2002; Lederman, Johnston et al.2003; Lederman, Milton et al. 2004) we have argued that information systems are purposeful and thatmethodologies and tools used to build them should be analysed using theories of action, that is  theories concerning the nature and design of goal-directed systems. Theories of action have beenresearched in several disciplines (Brooks 1986; Agre and Chapman 1987; Suchman 1987; Hendriks-Jansen 1996; Johnston and Brennan 1996; Agre and Horswill 1997; Clancey 1997) where two main positions are found which we will call the “deliberative” and the “situational” theories of action. Thedeliberative theory of action posits that an agent acts in a goal-directed way by building and maintaining an abstract model or representation of the world and applying deductive processes to thisrepresentation to determine what to do. Contrastingly in the situational theory, agents respond more or less reactively to situations presented by structured environments without deliberation. The twotheories have quite different modes of representation and action selection. We have argued previouslythat Structured Systems Analysis and Design Methodologies are implicitly informed by thedeliberative theory of action, whereas the situational theory provides a better description of the modeof action that takes place in time-constrained, routine work environments.This paper describes a research project in progress which aims to develop a situational informationsystems analysis and design methodology explicitly informed by the situational theory of action for application in routine but pressured environments. In this paper we describe and justify the program of research that is in progress, the preliminary results of which will be available early in 2005. Theprogram is important to theory because it uses an approach to design that has not been previouslyapplied to complex socio-technical systems. By undertaking this program, we expect to learn muchmore about the applicability of the theory of situated action in complex socio-technical cases. We alsoexpect to gain a deeper understanding of concepts such as situation, action, and routine that are centralto the theory of situated action. This paper presents the program for realising a methodology fordesigning situational systems. It is important for information systems practice because a case can bemade that it is the implicit commitment to the deliberative approach that underlies ineffectiveness of traditionally designed systems in routine environments and their frequent rejection by users. Amethodology for designing systems based on the situated theory of action should make a contributionto remedying this problem. Although open to refining through our research, we view action in routineenvironments as that where there is no choice for action selection except  where a routine breaks down.  In this paper we firstly discuss the deliberative theory of action that implicitly informs traditionalsystems design methodologies. We note that there is another theory of action (the situational theory)that appears to be more useful for explaining routine activity than the deliberative theory. We arguethat a design methodology (using the situational theory) is needed because the deliberative theory of action that underpins traditional methodologies is ontologically quite different from the situationaltheory of action that underpins routine activity. We describe a pilot case study we have undertaken andthen give a first cut of design methodology that has come from a study of situational action and partlyrefined through the pilot study. The research methods for the now in progress two-phase researchprogram for refining the initial design methodology are then described before we conclude. Theresearch program will run for just over three years and we intend conducting up to five iterative case studiesfollowed by a small program of action research.   2.   DELIBERATIVE ACTION AND TRADITIONAL SYSTEMSDESIGN   We have argued previously (Johnston and Milton 2001; Johnston and Milton 2002; Lederman,Johnston et al. 2003; Lederman, Milton et al. 2004) that existing information systems implicitlysupport the deliberative theory of action. According to this theory (Johnston and Brennan 1996), purposeful action proceeds by an agent building an abstract model of the external objective world from sense data and then reasoning about this model to determine actions that will achieve goals. Intraditional transaction-based information systems for example, ‘transactions’ are gathered thatrepresent changes in the world. Data models that correspond to the representation scheme are used todesign databases that are updated by the transactions (representing events in the world). In extremecases, such as Manufacturing Resource Planning, MRPII (Wight 1981), applications programs alsodeduce goal-attaining actions and human actors are only required to define the goal state, execute theactions in reality by following automatically generated schedules and provide ‘sense data’ byrecording transactions. More typically, applications programs help human actors to make decisions  (based on such a world model) by providing information about objects from reality and using datagained through transactions. Decision support systems are good examples of this type of system.Information engineering methodologies (IEM), share the ontological assumptions of the deliberativetheory, namely, that systems should represent the world in which the system acts in terms of external,independent and objective entities, properties and relations (Wand, Monarchi et al. 1995). Given thisfocus on symbol/object representation, use of these methodologies encourages designs for the socio-technical systems in which information systems form the representational scheme which mimic thedeliberative approach to action. 3.   SITUATIONAL ACTION AND ROUTINE SYSTEMS DESIGN Disciplines other than information systems have considered an alternative approach to theorisingabout purposeful action called the situational theory of action (Brooks 1986; Agre and Chapman 1987;Suchman 1987; Hendriks-Jansen 1996; Johnston and Brennan 1996; Agre and Horswill 1997; Clancey1997). In robotics specifically, this alternative theory has been motivated by the brittle performanceand computation intensity of artefacts based on the deliberative approach. The key to this alternativetheory is to provide an agent with largely reactive responses that utilise sense data obtainable directlyfrom the agent’s ground view of the world, and to introduce the agent’s goals and perspectiveexplicitly in the representation schemes implicit in the theory. In the situational theory, agents respond reactively to “situations” without deliberation. Situations are descriptions of the world centred on theagent and only include features of the world that are relevant to the agent’s purposes (Agre and Chapman 1987) and related actions. These features consist of the relations of things   to the agent givenits goals . Actions are selected from a repertoire used to respond to situations. This approach to actionselection leads to goal attainment only if the agent’s environment exhibits structure (“affordances”) toobviate the need to plan (Agre and Chapman 1987). An affordance is a structural aspect of theenvironment which makes it possible for an agent to reach a desired situation (a goal) by merelyreacting to its current situation. Analysis and exploitation of environmental structure is an important part of designing situated agents (Agre and Horswill 1992; Hammond, Converse et al. 1995; Horswill1995; Agre and Horswill 1997). An activity in this theory is a grouping of situations and associated actions that together lead to a reliable reaching of a desirable situation.We can see the differing role of representation in the situational theory: Situations are agent-centred and intention-laden views of the world for which a symbol/object isomorphism is neither possible nor necessary. The situated agent is not aware of individuated objects in its environment but rather of aspects of situations which betray opportunities for action. To the extent to which these aspects need to be represented in action selection the representations are, according to Agre and Chapman (1987),indexical (agent-centred) and functional (specific to the agents purpose).The situational theory has a different position on intentionality also. In the deliberative theory goalsare desired future states of the world and consequently the apparatus of world modelling is necessaryfor the agent to act intentionally in the present. In the situational theory, goals are achieved by anorientation of action in the present to likely goal attainment provided by awareness of affordancesspecific to that goal in the immediate environment. Thus, activities are sets of situational responsesthat are about  goals rather than driven by goals.Finally, the ontology that underlies the situational theory differs from that of the deliberative theory.Because the deliberative theory relies on deduction of actions from future objective states of the world it is forced to commit to an ontology of things, properties, relations and world states. The deliberativetheory begins with things and properties and derives action as a secondary concept: a particular kind (deliberate) state change. By contrast the situational theory has action, activity (extended episodes of  purposeful action) and affordances (possibilities of action provided by structure of environments) asits grounding concepts. Things are apprehended by a situated agent only through their role in affordingcertain actions, that is, as ‘equipment’. This suggests that a system design methodology built upon thesituated theory of action should focus on entirely different entities than one built upon the deliberativetheory.Clearly the situational theory of action has much more in common with routine behaviour than thedeliberative theory of action. This is because when engaged in routine behaviour one is interested in  how an activity is progressing and acting accordingly rather than returning to first-principles and  planning. Following a period of gaining experience in recognising situational cues, an agent uses thosecues to notice what action is appropriate according to the status of the activity. Consequently,methodologies for designing systems to support routine work would be better built upon the situated theory of action. Our research brings to information systems ideas from diverse disciplines whereauthors have argued that the "deliberative" theory of action and the "situational" theory of action, areradically different. The situational theory of action has had no significant influence on design of information systems despite there being proof-of-concept through effective and novel systems designsin robotics and artificial intelligence. 4.   METHODS FOR CREATING SITUATIONAL SYSTEMS There are three ways in which a situational system is brought into being. Firstly, a situational systemcould  evolve so that agent actions and the effects of actions knit perfectly with environment and situations to make activities reliable. Evolved biological organisms are excellent examples of evolved situational systems (Juarrero 1999) In many cases, such as in social activities, the activity and itsenvironment may have co-evolved. Secondly, an agent may learn an activity by seeing the effect of actions in specific situations. In this case trial and error is used to find the action rules that best exploitthe structures in the environment, but also environments might be chosen because of their particular affordances for action. Thirdly, and this is the approach we propose for information systems, a systemcan be designed  so that actions taken in response to situations will have desired effects. Depending onconstraints, either or both the action rules and the environment structures will be deliberately designed to ensure the reliability of an activity. It is for this purpose that we propose our methodology, and it isa distinctive feature that “environmental engineering” is part of it. Some level of iteration is alsoneeded. It is on this third basis that we have built an initial cut of the methodology which we detail below.We have used the situational theory of action, as it is discussed in robotics and other disciplines, todetermine the concepts central to an agent-centred situational system design methodology and thesteps that are likely. Whereas, the deliberative theory suggests information systems design should emphasise modelling the world using objects, properties, relations and states, and deduction uponthese models to determine action (such as decision support and planning), the situational theory should make central the notions of an activity, situations that comprise activities, actions that are largely areaction to situations and affordances that show the possibility for action, and hence also characterisethe situation. The situational theory would emphasise the importance of proper structuring of environments of action, which is largely ignored in the deliberative approach.Having identified the key concepts important for routine activity we undertook a pilot case study inorder to clarify the concepts and to construct a starting design methodology. We expected to learnabout the roles of the different concepts and how they are used to construct a design methodology. 5.   CASH ENGINEERING: A PILOT CASE STUDY We applied the methodology to a pilot case study in a compressor manufacturer called CashEngineering in Richmond, Australia. The methodology was applied to the design of a routine manualair compressor manufacturing system at Cash Engineering. In 2002 and 2003, the managing director of Cash consulted with one of the authors throughout the design of a manufacturing system for a newrange of air compressors and the resulting manual control system strongly reflects the principlesespoused by the situational theory (Johnston and Milton 2002; Lederman, Johnston et al. 2003;Lederman, Milton et al. 2004).The re-designed system, known as the Cash Compressor System, is for production control in a smallfactory of four staff manufacturing about 200 air compressors a year. The system has a white board that represents non-routine aspects of the compressors being made. There are no computers in thefactory. What is interesting is how little information is represented on the white board withoutcompromising control or efficiency.  The factory is designed so that the person taking orders on the telephone in the middle of the factoryhas full view of all available stock on shelves lining the walls. The main components of the systeminclude a white board of open customer orders and the physical parts of the air compressors that, bytheir construction, implicitly contain information about their own method of manufacture. Theinformation on the whiteboard is job-specific including name of client, and options such as colour, and compressor motor size. The system has been designed deliberately in this way to reduce the need torepresent things.Manufacturing commences when the order is received by phone and a line order is added to the white board. Major parts for making the customer’s compressor are checked for availability visually, and if need be, ordered on a one-off basis. Floor stock of other parts is also maintained by reordering based on visual clues. The machine assembler then takes a machine base and begins construction, referringto the white board only for order-specific information that is not part of the standard assembly routine.What is interesting in the Cash System is what is not represented. There is no detailed informationabout how to construct the machine: the machine acts as its own “jig” through devices for guiding atool or part to a specific place. Employees have learned the limited number of techniques used withthe “jig”. There is no parts-list or inventory of parts: the availability and quantity of parts holdings areclearly seen on the shelfs. The only recorded requirements-related information is in the reference tonon-standard choices on the whiteboard.If this system were to be explained by the deliberative theory of action, representation would includedetailed information about each compressor being manufactured. This is not the case at Cash wherevery minimal information is kept explicitly on the white-board. No rules can be found to enable aworker to take the individual parts and assemble a compressor. Instead we see the next action beingafforded to the worker by the structure of part-manufactured machine. Only a limited range of choiceis available. When there is a choice, the white-board tells them the option to be selected based on thecustomer’s desires. Similarly, deliberate structuring of the plant environment allows parts to bereplenished with little deliberation or planning.What the Cash system shows is that with careful design of the work environment (including the objectof manufacture itself) reliable goal attaining activity can be managed with very parsimoniousrepresentational systems of a type rather different to those envisioned by current information systemsdesign theories. Cues for action selection are largely provided by the affordances of the environmentand the machine design. The white board represents activities and the choices needed at key situationsin the activity. But the representation is more fragmentary and ephemeral than a world model. This is possible because to a great extent “the world (is) its own model” (Brooks 1991) in the Cash system.The system at Cash Engineering is manual. However, there is no reason why the principles uncovered at Cash Engineering and emerging from the situational theory of action cannot be applied to socio-technical systems involving information and communications technology. It is likely that ubiquitouscomputing and mobile devices will play a role in structuring an agent’s environment and allowing for  better communication of situations to and between human actors. 6.   AN INITIAL SITUATIONAL DESIGN METHODOLOGY Reflection on the re-design of the manufacturing system at Cash Engineering which was stronglyinformed by our understanding analysis of the implication of the situational theory of action for information system design has led us to formulate an initial design of methodology for designingsituational systems. The methodology has six steps:1.   Identify the multiple agents and their specific environments that constitute the totalsituational system. Situational systems of any complexity will consist of a multiple of interacting agents (human and technical) each situated in their own unique environment.2.   Identify the activities needed in the situational systems in pursuit of specific goals.  3.   Analyse activities of agents into the situations, their aspects, and actions constitutingeach activity . Activities can only work if an agent is able to notice when it is in a particular situation and are able to act routinely.
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks