Activity Theory: Applying the Theory to a Case
Are Solberg
Master student
Department of Informatics
University of Oslo
0 Abstract
CSCW (Computer Supported Cooperative Work) includes several different theories about how computer systems are to be viewed. Activity Theory (AT) is one of these theories. In this essay use AT to analyse a case about medical consultations, and examine what happens when paper medical records are replaced with computer records. The focus of this essay is on medical records as a tool in the medical consultation activity, and which implications replacing this tool has for collaboration.
Key words: Activity Theory, medical consultations, tools, paper records, computer records, collaboration, breakdowns
Citation: Solberg, A. (2002),”Activity Theory: Applying the Theory to a Case”
1 Introduction
This essay is written as a mandatory exercise in the course IN-CSCW (Computer Supported Cooperative Work) at the Department of Informatics, University of Oslo.
In this essay I use Activity Theory to analyze a case study. Activity Theory is one of several theories currently in use in the CSCW community. I have chosen to use Engestrøm (1990),Kuutti (1991) and Bardram’s (1997 and 1998) concepts in this essay. The concepts are used to analyze a case study about medical consultations, presented in Luff and Heath’s (1998)article “Mobility in Collaboration”.
The rest of this essay is organized as follows. In section 2 I present the key concepts of Activity Theory. Section 3 contains a brief description of the case. The analysis is presented in section 4. In section 5 I offer some concluding remarks.
2 Activity Theory: the key concepts
Activity Theory as a scientific tool has its roots in Soviet Psychology. However, it is gained users all over the world, and it has been used in several other disciplines such as CSCW. Kuutti (1991) defines Activity Theory as “a philosophical framework for studying different forms of human praxis as developmental processes, with both individual and social levels interlinked” (Kuutti 1991, p. 253).
A thorough description of all the concepts of Activity Theory (AT) is beyond the scope of this essay. Below I present the key concepts, based on Kuutti (1991) and Bardram’s (1997 and 1998) articles. The concepts presented are the ones I find to be most applicable to the cases that are analyzed later in this essay (section 4).
ATfocuses on activities as basic units of analysis. An activity is defined as a minimal meaningful context for individual actions. With such a definition it is possible to analyze individual actions and still maintain some contextuality. An activity consists of three main components; subject, object and community:
Subject:The subject can be an individual or a collective. All activities have one or more active subject, individual or collective; in addition they might have passive / non-active subjects. An active subject understands the motive of the activity, a passive subject does not (passive subjects might in fact be unaware of the activity all together). In his article, Kuutti (1991, p. 261) also accounts for a third type of subject; an expansive subject. An expansive subject, like an active subject, understands the motive of the activity; in addition the expansive subject is active in the development of the activity.
Object:All activities have a material object (material as it is understood in Marxist philosophy) and activities can be distinguished according to their objects. The existence of an activity is motivated by transformation of the object towards some desired state / outcome.
Community:All activities take place within an environment, and the subject therefore has to relate to a community. The community consists of passive subjects who share the same object. “AT describes cooperation as a collaborative activity, with one object, but distributed onto several actors, each performing one or more actions according to the overall and shared objective of the work” (Bardram 1998, p. 91)
There exist mutual relationships between each of these components, all of which mediate artifacts. The artifacts are either physical, e.g. a piece of paper, or abstract, e.g. a social norm. The central relationship, that between subject and object is mediated by tools, that between subject and community is mediated by rules and that between object and community is mediated by the division of labor. Figure 1 illustrates the main components, the relationships between them and the mediated artifacts.
Figure 1: “External” structure of an activity (Figure source: Kuutti (1991), p. 257)
The three main components and the mediated artifacts form what Kuutti (1991) calls an activity’s “external” structure. Bardram (1997) says that “[h]uman activity can be described as a hierarchy with three levels: activities realised through chains of actions, which are carried out through operations” (Bardram 1997, p. 141). Kuutti (1991) calls this hierarchy the activity’s “internal” structure. Activities are realized through a series of actions, and the same activity can be realized through different actions. Further, actions are realized through a series of operations, and the same action can be realized through different operations.Actions are controlled by the subjects’ conscious objective, while operations have no objective in themselves and are just required “steps” to perform an action.
AT also describes collaborative activities with a three level hierarchical structure; coordinated, cooperative and co-constructive collaborative activity. The three levels are closely linked to passive, active and expansive subjects, respectfully. I.e. in coordinated collaborative activities the subjects are passive, in cooperative collaborative activities the subjects are active, and in co-constructive collaborative activities the subjects are expansive. Essential to the hierarchical structure of collaborative activities is the transition between the levels, figure 2 illustrates this transition and the transition triggers.
Figure 2: Transition between levels of a collaborative activity (Figure source: Bardram (1998), p. 92)
Another central concept in AT is that of breakdowns. Bardram (1998) defines cooperative breakdowns as breakdowns in the flow of work. He says that “[t]he importance of understanding cooperative breakdowns, however, lies in the importance of supporting such breakdowns in the design of computer technology” (Bardram 1998, p. 91). He says that if we need to understand the breakdowns, in order to be able to design for recovery from them. The reflections that cause a transition between to levels in figure 3 can be caused by a cooperative breakdown.
AT say that activities are under continuous development and that they “cannot be really understood without seriously analyzing the historical development which has led to their present state” (Kuutti 1991, p. 254). During this development the mediated artifacts are constantly being reconstructed. The development of an activity is uneven and discontinuous, and is driven by contradictions.
Engestrøm (1990) focuses on the tool. I will use some of his thoughts of tool in this essay. Engestrøm is the originator of the triadic structure shown in figure 1. He uses this structure to analyze a case; I use the same method in my analysis (chapter 4).
3 The cases: a brief description
“Mobility in Collaboration” by Luff and Heath (1998) addresses the question of how mobility can be supported within collaboration activities. In this section I will give a brief description of one of the cases that Luff and Heath use to illustrate the importance of considering the mobility when designing a CSCW system / tool.
Mobility
Heath and Luff introduces three types of (physical) mobility that are used to describe the way in which people interact with physical artifacts and other people: micro, local, and remote mobility. Micro mobility refers to the way in which small artifacts such as pieces of paper and books can be mobilized and manipulated in a relatively small, “at hand” domain. Local mobility refers to “real-time” interaction between people and technology within the same (delimited) location. Remote mobility allows both synchronous and asynchronous collaboration and information exchange between individuals in motion at different locations.
Medical consultations
A medical consultation can be said to be physical meeting between a patient and his / her doctor (General Practitioners GP) in the GP’s office. The consultation has, in essence, the form of a conversation between the GP and the patient. Each patient has his / her own medical record. During the consultation theGP reads from this record to check what has done at prior consultations, which medicines the patient has been given, etc. The GP also makes notes to the record concerning the present state of the patient, the medicinesgiven, etc.
The medical record has historically been a paper document. Luff and Heath (1998) “note that the paper record supports both synchronous and asynchronous collaboration, and collaboration between both doctors and other professionals, and between the patient and the doctor” (Luff and Heath 1998, p. 306). Since the mid-1980s, much effort has been put into replacing paper documents with computerized records, however, paper documents are still used by many GPs. Luff and Heath claim that the reason why paper documents are still in use is the micro-mobility functionality of the paper document.
4 Applying the theory
In this section Activity Theory is used to analyze the two cases. I use the same method as Engestrøm (1990) uses, meaning that I will try to structure the activities in the cases into triadic structures.
In section 3 I gave a brief description of the case. In this section I will use Luff and Heaths (1998) account of the case; in addition, I will use my own understanding of at medical consultation and its context. The analysis consists of three parts: analysis of the case with paper records, analysis of the system with computer based records and analysis of the changes in the activity.
Part 1: Paper documents
I begin my analysis by identifying the three main components and the mediated artifacts of the activity, figure 3 illustrates the situation:
Figure 3: Triadic structure of consultation activity with paper records
Subjects:The activity has two active subjects; the GP and the patient. Both the GP and the patient are subjects, because they both participate in transforming the object. In addition, they both are active subjects, because they both know the motive of the activity.
Object:The object is initially the undiagnosed patient. Through the activity the object is transformed into a diagnosis (of the patient, hence, the patient is diagnosed) and a treatment plan. I therefore choose to say that the object at any place in time is the diagnosing of the patient. The object in this case is abstract, not a physical entity that is changed, but rather a lack of certainty (undiagnosed patient) at first and certainty (diagnosis and treatment plan) in its final state.
Community:The testers; lab personnel, specialist doctors, nurses, etc., and other doctors in the clinic are passive subjects in the activity. However, they share the same the object as the GP and the patient, and are therefore part of the community.
Tools:Many tools are important in the process of diagnosing the patient (the subject-object relationship). The doctor checks the patient’s paper record to get an overview of the patient’s medical history. The record is also important for the GP to get an impression of who the patient is, if he / she doesn’t remember the patient from the last consultation. In a case presented by Engestrøm (1990, p. 178) a doctor says that “it’s awfully important to be able to check whether I’ve seen the patient or not”. He / she checks the record using his / her algorithms and routines for reading. By algorithms and routines, I mean that each doctor has different ways of reading the record and they focus on different things in the record. The communication between the GP and the patient uncovers the patient’s reasons to visit the GP and the patient’s hypothesis of what is wrong with his / her health. During this communication the GP make notes in the paper record and (possibly) share the content of the record with the patient. The paper record is a vital part of the communication; e.g. the GP refers to it by pointing at it and it is moved around the desk. If the GP finds it necessary examinations and tests are taken, either by the GP or by other testers. The GP might communicate with his colleagues, if he / she has problems diagnosing the patient. The paper document is also important in this communication, as it can be moved around the office and clinic when the doctor wants help from other doctors at the clinic. Based on information from the other tools, the GP uses his / her explanatory model of illness to diagnose the patient. This model consists of the GP’s knowledge of medicine, and books that he / she can check for references about symptoms, applicable medication, etc. Using this tool, the GP might need to (re-) use the other tools, e.g. perform more tests.
Rules:There are several rules which governs the relationship between the GP, patient, and community. There are laws and medical standards that govern the relationships, e.g. test sent by the GP to a lab are to be handled in a special way. There are also norms, e.g. agreements between the GP and the lab about how long time it should take to perform a given test.
Division
of labor:GP vs. testers; as mentioned, the doctor might choose to perform tests himself, or delegate this activity to testers. The division of labor is here important, if the testers e.g. contaminate a test this might influence the test results, and result in the GP making an erroneous diagnosis. In addition, “[d]iscontinuities or gaps in the communication between the patient’s various providers often manifest themselves in the form of missing documents which would be vitally needed in order to complement the patient’s accounts” (Engestrøm 1990, p. 184).
Above I have illustrated a medical consultation with the use of paper medical records. I have tried to show that the paper record is an important tool in itself, and that it also is important in combination with other tools. Below, in part 3, I describe the changes in the activity if this tool is replaced by another one.
As I mentioned in the description of the various components and mediated artifacts above, the activity has both active and passive subjects. These subjects collaborate at different levels according to Bardram’s (1998) hierarchical structure, figure 2. The GP and the patient who are both active subjects cooperate; they share the same objective which they try to reach by working together. The relationship between the GP and the testers is a coordination. The GP and the testers haven’t got the same objective for their actions. Since the testers don’t share the objective of the medical consultation activity, they are passive subjects in this activity. What the testers objective is we can’t know from the case alone, it might be to just perform the test and to return the result (making this an activity for them, and them active subjects in this activity), or something completely different. The relationship between the GP and the other doctors at the clinic (his colleagues) is also a coordination, since they don’t share the same objective. Since they have different objectives, there is also a coordination between the patient and the person(s) who perform test on him / her. In some cases you might say that more of / all the subjects share the same objective, e.g. if the patient is related to the tester and other doctors who know that they are looking at his test or record. However, these are extreme cases an irrelevant in the discussion, because of the choice I’ve made of active and passive subjects.
There are several possibilities of cooperation breakdowns in the activity. I have already mentioned that it is essential that the GP and the testers communicate well. If the communication is disrupted or fails, this might be a cooperation breakdown. The same will be the case with the communication between the GP and other doctors. The breaking of rules or norms might also result in cooperation breakdowns. The communication between GP and patient might also result in a breakdown, e.g. if the doctor talks to a grown patient as if he / she were a child.
Part 2: Computer records:
As with the activity containing the patient’s paper record as a tool, I start this part with identifying the main components and mediated artifacts of the activity. In order not to repeat myself, I will focus on the artifacts that are different from those in part 2. Figure 4 illustrates the situation:
Figure 4: Triadic structure of consultation activity with computer records
Subjects:GP and patient.
Object:Diagnosing the patient; initially undiagnosed patient which is transformed into a diagnosis and a treatment plan.
Community:Testers and other doctors.
Tools:The GP uses the computer to access the patient’s computer record. He / she then uses the record to get an impression about who the patient is, and to get an overview over the patient’s medical history. The GP checks the record using his / her algorithms and routines of computer use. When the doctor reads the computer record he / she searches for specific things in the record, and does this in his / her own way. The communication between the GP and the patient is the GP’s possibility to uncover the patient’s own thoughts about what “is wrong” with him / her. During the conversation, the GP reads from and makes notes to the computer record. The GP might in some cases turn the computer screen or laptop towards the patient, in order to share the record with the patient, but it is often difficult for because of the layout of the GP’s office. Luff and Heath (1998, p. 307) say that “It is difficult to position a workstation, both keyboard and monitor in such a way that the doctor can maintain a principal orientation towards the patient whilst reading o[r] entering information, and the physical separation of the area where text is entered from where it is seen, undermines the doctor’s ability to momentarily enter data”.Examinations and tests might be taken either by the doctor or by other testers. The GP might communicate with his colleagues, if he / she needs some help in diagnosing the patient. The medical record is often important in this communication. In order for the colleague the see the record, the GP can either print out a paper copy, bring the laptop to his colleague, ask the colleague to come and look at his computer screen, or make the record available for the colleague to read electronically (e.g. making the file / record public, or sending it with an e-mail). All these options are rather cumbersome, and might complicate or hinder the communication. The GP uses his / her explanatory model of illness to diagnose the patient. This may result in the (re-)use of other tools.