Interaction Design
Enrico Coiera
Centre for Health Informatics
University of New South Wales,
UNSW NSW 2055, Australia
ewc@https://www.doczj.com/doc/4017814213.html,
Draft Date 30 April 2002
Abstract: This paper presents a framework for the design of interactions between agents, mediation by technological systems. The design of interactions within an organization is viewed from the point of view, not of the technology mediating the new interaction, but of the agents who are asked to use the new technology. Understanding the limits to individual agent resources allows an analysis of the impact that a new interaction will have in a given setting. When we look beyond simple interaction settings, we can use the notion of interaction equilibria to predict the impact of new information and communication technologies within an organization. Economic supply and demand curves, for example, may allow us to make both qualitative and quantitative predictions about technological adoption of communication systems.
Introduction
“In the next fifty years, the increasing importance of designing spaces for human communication and interaction will lead to expansion in those aspects of computing that are focused on people, rather than machinery. … The work will be rooted in disciplines that focus on people and communication, such as psychology, communications, graphic design, and linguistics, as well as in the disciplines that support computing and communications technology. … Successful interaction design requires a shift from seeing the machinery to seeing the lives of the people using it”.
Winograd (1997).
Traditionally information systems are designed around an idealised model of the task that needs to be accomplished, and failure in system performance is explained away by blaming human social and cultural “barriers” to technology adoption. In this world, the newly engineered system is always impeded by the barrier of human frailty. The design of the computational machinery is the scientific high ground, and understanding the mess of implementation in the real world is left to “soft” social science and happenstance.
But people are part of the system. The web of interactions needed to make anything work in a complex organisation always involves humans solving problems with limited resources and working around imperfect processes. Designing the technological tools that humans will use, independent of the way that the tools will impact the organisation, only optimises local task-specific solutions, and ignores global realities. The biggest
information repository in most organisations sits within the heads of those who work there, and the largest communication network is the web of conversations that binds them. Together, people, tools, and conversations – that is the “system”. Consequently, the design of information and communication systems must also include the people who will use them. We must therefore design interactions that reflect the machinery of human thought and communication, sometimes mediated by communication channels, sometimes in partnership with computational agents.
Interaction design is a newly coined discipline, and it focuses on constructing the ways people interact with objects and systems, and the product of interaction design is almost entirely the quality of the user’s experience (Crampton Smith and Tabor, 1996). In this world, the effectiveness of a piece of software is not an internal attribute of the software, but emerges from the way the software is interpreted by users, and that interpretation is dependent upon the user’s specific context, culture, knowledge and resources.
It has been stated that at the level of an individual user, interaction design will always be an ambiguous or subjective process for the designer (Crampton Smith and Tabor, 1996). In this paper I will argue that, from an organisational or large population perspective, the process of interaction design can be a much more principled one. Based upon models of the way populations of agents interact with each other, mediated by technology, but bounded by scarce cognitive and physical resources, we are able to model the effect of introducing new technologies, or indeed design technologies in the anticipation of their effect on populations of users.
This paper is the third in a series which sketch out a general theory of mediated agent interaction. In the first paper, the communication space was introduced as the major component of the overall interaction space for humans working in complex organizations, and the gap between the current focus on interactions outside of the communication space and the overall interaction needs of humans was highlighted. A framework for mediated interaction theory was also developed, based upon the notion of bounded mutual knowledge, or common ground, that is shared between communicating agents (Coiera, 2000). This was developed to help us decide whether an interaction was best served by communication or information technology. In the second paper, an examination of the costs of the grounding process lead to the view that grounding is a kind of cost minimisation process, based upon maximisation of agent utility, and resulted in cost-minimising equilibria developing between agents (Coiera, 2001). This is important because it suggests that, although there may be a wide variation in the nature of individual interactions, we can over time predict the characteristics of interactions at a population level, and design systems to effectively support the majority of interaction at the organisational level. In this third paper I will turn to the task of designing systems to support interactions, whether in the communication space or not, and specifically look at how we cannot design and implement technological systems to support individual interactions, without attempting to model the wider interaction space within which individuals operate.
Designing the interaction space
“In creating tools we are designing new conversations and connections”.
Winograd and Flores (1986).
Typically information system design occurs with a single task assumption, that the user is going to be wholly focussed on interacting with the system that is being designed. Such ‘in vitro’ laboratory assumptions do not translate ‘in vivo’ when individuals use systems in working environments. Since individual agents in an organization are working in a complex environment, they may at any one time be carrying out a variety of tasks and interacting with different agents to help execute those tasks. With the exception of environments where there is rigorous workflow control, this means that we cannot predict what interactions will actually be occurring at the same time as any interaction we specifically design.
While, we cannot predict every specific interaction that will occur in the real work environment, the design process can model the typical interaction space within which any new system will be introduced. The interaction space can be modelled to include the most important interactions that will be competing with the new designed interaction. Within the overall interaction space, the communication space forms a subset containing interactions that are informal person-to-person exchanges. We can also conceive of an information space, which by corollary contains those formally structured interactions that typically rely on information models and computational systems. I have argued elsewhere that the notions of the communication space and information space are probably driven by a dichotomy in communication and information technologies, and that at a more abstract level, they actually are not separate spaces but form different ends of a continuum of explicit task modelling (Coiera, 2000).
To construct the interaction space, we start with a general description of an interaction between two agents, which may either be human, or computational. An agent typically has a number of tasks that need to be carried out, and a pool of resources available to accomplish those tasks. An interaction occurs between two agents when one agent creates and then communicates a message to another, to accomplish a particular task. A mediated interaction occurs when a communication channel intermediates between agents by bearing the messages between them. For example, e-mail can be used to mediate an interaction between two individuals, just as can an electronic medical record, which is as much a service to communicate messages between clinical staff as it is an archival information repository.
The first step in modelling the interaction space surrounding a new interaction we wish to design is to note which other agents will be local to the new interaction, and then to examine the likely effects of any interactions they might have on the new interaction that is being contemplated. By doing so we enhance the chance that new interactions will succeed when they are eventually introduced into the intended interaction space.
In general terms the impact of one interaction on another may be to:
? Compete with another interaction as a direct substitute. For example, a user could use an on-line database to seek information, or instead, use the telephone to call a colleague to ask for the same information. The human-human interaction mediated by the telephone competes with the database to meet the user’s information needs.
? Compete with another interaction for the resources of an agent. An agent has limited resources and if they are expended on one interaction they may not be
available to another. For example, an information system may be well received in ‘in vitro’ laboratory tests, but when placed in a work environment we may find the users have insufficient time to use the system, because of competing tasks.
The new system is then ‘rejected’ not because it does not do what was intended, but because the impact of the real-world interaction space on its intended users was not modelled. Concurrent interactions can also subvert the execution of a designed interaction. For example, a user may be interrupted in the workplace and take up a new task, and not log off the information system they were using, causing a potential breach of security.
? Create new information transfer pathways, through a combination of interactions.
Each interaction connects agents, and each new interaction enables novel conversations between agents. If these combinations are not factored into system design, then the introduction of a system may produce unexpected results. For example, consider the interaction between a human agent and an electronic record system or EMR (which is represented as a computational agent). Computational agents that might co-exist with the EMR could include other applications like e-mail or a word-processor. If the design process fails to include these additional computational agents, then unintended interactions made possible via these other agents may subvert the original EMR design. For example, it may be possible for
a user to copy a section of text from the medical record to a word-processor,
where it can be edited, and then re-inserted into the EMR. However, since this interaction with the word-processor is not part of the original design scope, it may introduce problems. A user could inadvertently paste the text into the record of a different patient, and no formal mechanism would be in place to prevent this context-switch error. Similarly, text might be copied from an EMR that has been designed with powerful security features to prevent unauthorised access, and then copied into an e-mail message, which is insecure. In both cases, the co-existence of an unmodelled additional computational agents introduce interactions beyond the scope of the original system design, and permit behaviours which would be prohibited within the designed interaction, but which are permitted in the interaction space.
? Support the new interaction by providing resources that are critical to its execution. For example, the designer of a medical record system usually focuses on sculpting the interaction between a single clinical user and the record.
However, other human agents also populate the EMR interaction space. The EMR user is often not the sole author of the content that is captured in the record, but is recording the result of a set of discussions with these other clinical colleagues. If the goal of designing an EMR is to ensure the highest quality data is entered into the information system, then it may be even more important to support the collaborative discussion between clinicians, than it is to engineer the act of record transcription into the system. Failing to model the wider EMR interaction space means we may over-engineer some interactions with diminishing returns, when we could be supporting other interactions that may deliver substantial additional benefit to our original design goals.
placed within a pre-existing interaction space. Other interactions that exist in the interaction space impact the new interaction in a variety of ways, and if they are ignored in the design process may have unexpected consequences when the new interaction is implemented.
Interaction dynamics and the competition for agent resources
Since it is not possible to predict every specific interaction within the space in which a designed system will operate, it may appear that there is little that can be done to account for all possible concurrent interactions. However we can be quite specific about the typical impact that concurrent interactions will be have on agents. As discussed above, one of the ways interactions impact each other is to compete for the resources of individual agents. Based upon cognitive psychological models, we should be able to say something about the cognitive resources available to human agents, the cognitive loads they will typically be under in a given interaction space, and the types of errors that may arise because of these loads. Consequently it should be possible to craft information or communication systems that are tolerant of the typical interaction load our users will be under.
f12 two interactions I1 and I2 compete for A f’s attention, and A f switches between them over time. Each switch is not instantaneous but requires a refocusing of attention and carries its own cognitive switching costs. Agent A1 makes two interrupting requests to A f during this period. The first request r1 succeeds in getting A f’s attention and causes a switch to interaction I1. The second request is not successful, because A f’s attention is fully loaded by the current interaction I2 and the request is missed.
We begin by noting that accomplishing a task normally consumes resources, and this also holds during an interaction. An agent has internal resources, which in the case of a human are cognitive resources like memory capacity and knowledge. These resources are limited, and in some cases will deplete with use. One consequence of operating with finite resources is that agents will need to minimise costs on resources and maximise the benefits of any action they contemplate.
The interaction between agents is thus dependent upon:
? The task at hand, which has a resource cost and an implied benefit upon completion, and which must be evaluated against other tasks the agent might have to determine whether it is worth doing,
? The internal resources available to the agent to attempt tasks, which include finite computational or cognitive resources,
? The external resources available to the agent within their interaction space, which includes communication channels and other agents. Such resources each have a cost,
a benefit, and limitations on their capacity.
? The impact of concurrent interactions.
The cognitive resources of a human agent includes prior knowledge, which is stored in long term memory (LTM), and attentional resources, which include working memory (WM). The cognitive capacity of individuals to successfully carry out a task are limited by the resources of attention, which are can store and process only a limited amount of data at any one time (Miyake and Shah, 1999). As attentional resources are loaded with new tasks, the capacity to enact further tasks is diminished, and current task execution may be compromised (Parker and Coiera, 2000).
As a consequence, the design of any single interaction needs to take account of an agent’s inherent capacity to carry out the task. However, in complex organisations, where we wish to introduce information and communication systems to support work, the design task is more complex. Specifically, each new tool introduces a new dialogue, a new interaction (Winograd and Flores, 1986).
We thus create a setting in which there are multiple possible interactions that could potentially occur at any one time, driven by the task at hand for individual agents. As a consequence, the agents involved in carrying out tasks both need to initiate interactions and are on the receipt of interruptions requesting interactions. The actual sequence of which interactions occur will be dictated by the circumstances local to each agent at that time, and multiple interactions may be maintained over time (Figure 2).
This multitasking setting has a number of consequences. Since a human agent’s cognitive resources are limited, the amount of attentional resources available to any single interaction will be reduced on average with each new interaction we introduce. It should thus come as no surprise that introducing a computer onto a Doctor’s desk will result in less attention being devoted to the patient. In one study, the presence of a computer during doctor-patient consultations had detectable effects on the focus of the doctor’s attention (Greatbatch et al., 1993). While they were at the computer, doctors confined themselves to short responses to patient questions, delayed responding, glanced at the screen in preference to the patient. The response delays probably were an effect of the costs involved in switching between interactions. We know that switching between tasks requires a suspension of the processing of the current task, and alterations to memory. Consequently, interruptions have the potential to disrupt memory processes, and lead to tasks being mis-executed, forgotten or repeated (Parker and Coiera, 2000). Interruptions, which are requests for a human agent to stop their current task and commence a new interaction, are common in clinical work places. In studies of hospital-based clinicians, interruption rates can be as high as 30% of the total time devoted to communication (Coiera et al, 2002). Consequently, the introduction of any new interaction into such organisations needs to recognise the existing interaction patterns of agents, and design the new interaction in such a way that it neither has a substantial deleterious effect on the existing interactions, as well as ensuring the new interaction will have sufficient attention from agents that it will be well executed.
When one task dominates attention, then other concurrent interactions may fail to get the attention they need. For example, studies have shown that when a doctor is interacting with a desk-based computer, they may not hear what is being said to them by the patient sitting opposite the desk (Booth et al, 2001). Sometimes, a single interaction so consumes a human’s attention, that all other interactions are unable to get their attention. In a study of the task-execution behaviour of anaesthetists, the study subjects were found to completely miss significant events displayed on the screen of their monitoring devices . It
was hypothesised that this was caused by the anaesthetist’s attention being swamped by other tasks (Coiera et al., 1996).
Requisite attentional resources
This leads directly on to the issue of how one crafts an individual interaction to ensure it minimise the load it imposes on the attentional resources of agents, as well as maximising the likelihood that it will be well executed. This reduces down to the following questions:
? What agent resources will be required to accomplish the new interaction?
? Given the existing work environment, what is the base-line cognitive load going to be on the agents who will be expected to participate in a new interaction? As a result, what agent resources are actually going to be available for the new interaction, given other likely co-current interactions?
? When resources are scare then either the new interaction or existing interactions will suffer. Consequently, given multiple possible interactions available to an agent, and the agent’s local priorities, where will be the agent’s focus of attention be?
? What will be the overall impact of introducing the new interaction into the existing interaction setting?
I have argued previously that we can model many of the resource requirements of a given interaction by considering the size of message that needs to be sent between agents, the channel that will be used to convey the messages, and the relationship between agents. I have also argued that we can best model relationship between agents as the degree to which they share models of the world, or have common ground (Clark, 1996; Coiera, 2000). This is because the greater the common ground is between agents, the more succinct and accurate the communication is between them.
How does common ground help us model attentional resources? We know from cognitive psychology that although working memory has a small and finite storage capacity to manage immediate tasks, long-term memory can substantially enhance its performance. An item in working memory may actually point to a complex schema in long-term memory, and with growing expertise individuals can substantially enhance task performance by building such schemas (Sweller, 1994). We can intuitively think of this as working memory using LTM schemas as ‘virtual memory’ to supplement its own very restricted resources.
We now can see how common ground enhances task execution. For a given task, the agent will need to send a message to another agent. The more the agents already share, stored as schemas in LTM, the less a specific message needs to contain, and the more that can be assumed. By simplifying the task of constructing and sending a message, common ground minimises the use of attentional resources.
We can formalise this discussion by nothing that for interacting agents a and b, a message of length m transmitted over a time t with a grounding efficiency of G E, the requisite channel capacity C is (Coiera 2001):
E G t
m C =[1]
Grounding efficiency is a measure of the resource requirements for an interaction between two agents, all else being equal, and can be thought of as the average length of messages sent between them divided by the ‘true’ message length. When a pair of agents shares much common ground they can exchange terse messages since much can be assumed, and as a consequence their interactions are shorter than the ‘true’ message, which needs to include everything needed for it to be understood.
Consequently, for interactions, which by definition require that agents are exchanging messages to accomplish a task, Equation 1 should provide us a basis for modelling the requisite resources to accomplish an interaction. It provides us with observable measures of internally unmeasured resources.
We next note that the likelihood that an agent will be able to handle an interaction if given it, or even recognise the request for an interaction, will be determined by the resources available to the agent. If the agent’s cognitive load is high on a current task then new tasks may be missed, poorly understood, or poorly executed. The threshold for a successful interaction is thus when the available agent capacity exceed the requisite channel capacity ie:
With Equation 1, we can connect together the triad of channel characteristics, context-specific message requirements and the nature of the relationship between communicating agents. It allows us to make three different types of interaction design inference:
? Channel-centric inference : For a given pair of agents and a message, we can specify channel capacity requirements.
? Relationship-centric inference : For a given agent, channel and time resource, we can say something about the common grounding that would help select an appropriate second agent to communicate with.
? Task-centric inference : Time can be considered a key task-limiting resource. For a pair of agents and a given channel, we can say something about the complexity of messages that can be exchanged between them over a given period of time.
The last of these suggests that we can generalise the previous equation further to incorporate other resources. Resource-specific versions of the equation can be crafted, substituting bandwidth for other bounded resources like time, monetary cost, utility of knowledge received etc.
Global interaction design
Fundamentally, every agent makes some assessment of the costs C(x) and benefits B(x) of a potential interaction x , and when the benefits exceed costs, elects to interact (Frank, 1998) i.e.:
if B(x) > C(x) then do x else don't
If a number of interactions are available to the agent, it will probably choose the one that it perceives to deliver the maximum cost-benefit trade-off. What emerges then is a picture
req
avail C C ≥
of a series of cost-benefit trade-offs being made by individual agents across different possible interactions. The goal for the agent is to accomplish its tasks through frugal utilisation of its resources, maximising the utility of any given interaction, through a maximisation of benefit, and minimisation of cost.
As an interaction designer looking at an organisation, it will not be possible to model all possible interactions nor determine what the outcome of individual agent interaction choices will be. What we would like, however, is to determine the overall state of interactions within an organisation, and make design choices that reflect the current interaction load. This will allow us to design new interactions that suit the current context, as well as predict how likely agents in the organisation will choose the new interactions.
Consequently we would like to be able to make statements about the likelihood of an interaction occurring, given the organisational context. Specifically, the probability that an interaction will occur is driven by the difference between cost and benefit. For most agents, this decision is based upon evidence of past interaction choices. We can say that, for an agent which has a history of n previous interactions of type x , that the probability of it carrying out x again next time should be related to the average overall benefit B av in the past i.e.
[2]
However, an agents available capacity C avail to execute a new interaction also determines whether a new interaction will be taken up or succeed. Consequently we can say that the likelihood that an interaction will occur is a function of the agent’s estimate of its benefit as well as the agents current capacity to handle it i.e.
P(interaction) = f(C avail,, B av )
We now set up a situation where we are measuring populations of interactions across groups of individual agents in an organisation, to determine likely future choices. This would seem to be a fraught task, given the complexity of the systems we are trying to model. However, it is likely that agents will behave in fairly stable ways over time, because each interaction is not an independent event , but depends upon past and future interactions. For example, an agent may have invested in developing a relationship with another in the past, or expended resource learning to use a piece of software, which benefits a present interaction.
Similarly, the anticipation of a future interaction may alter an agent’s behaviour in the present. For example, if the agent is going to have a series of similar interactions, then it may be cheaper in the long run to establish a high level of shared common ground, and minimise grounding during individual interactions. One of the consequences of such grounding behaviour is that, over time, agents should choose a degree of shared common ground with other agents that will minimises future interaction costs, but will still be sufficient to accomplish the goal associated with interactions.
n
x C x B B n
i i i
i av ))()((1∑==?=
The two agents in an interaction either give or receive ground, and the decision to participate in the interaction occurs when both agree that the ‘price’ or benefit of doing so is mutually acceptable, much like the way two agents would agree to exchange other goods of value. This leads to a hypothesis called the law of the mediated centre (Coiera, 2001) which states that communicating agents will be driven to an intermediate level of model sharing with other agents over a series of interactions, where the total costs of maintaining shared models and communicating models at interaction time are minimised. The mediated centre represents an interaction equilibrium point.
The economic interpretation of interaction
The grounding equilibrium is one example of agent’s adapting their interaction behavior over time, and they should also make similar adaptations to other aspects of interactions, such as choice of interaction task, other agents and channels. If agents do display such equilibrium behaviors over time, then we have a powerful set of analytical methods to assist in the modeling of organisational interaction to assist design. Specifically, we can look to other disciplines that have had the same task of inferring the likely outcome of the multiple, often conflicting individual choices of agents, yet which collectively coalesce over time into definable population behavioral equilibria.
Specifically, we can look to disciplines like economics that offers an array of analytical techniques to predict decision equilibria. A number of different economic methods seem applicable to the task of predicting interaction outcomes including supply-demand analyses, barter modeling, game theory and computational simulation methods. To illustrate the nature of this type of analysis, the remainder of this paper will present a supply-demand analysis of interaction, and illustrate its potential to support interaction design with in organizations.
Since we are dealing with the choices to initiate or receive an interaction by free-willed and independent agents, under conditions of scarcity of resources such as time, we can think of interacting agents as behaving according to economic principle of supply and demand. Sending agents are 'buyers' with the need to interact, and receiving agents 'supply' them with their time and information by making themselves available for the interaction. There is thus at least an implicit, and sometimes explicit, negotiation by agents over their willingness to interact with each other, based on their personal assessments of cost and benefit from past interactions and current goals.
We can thus structure the supply and demand for interactions by graphing the quantity of product 'sold' - in this case the number of interactions made - against the price of interacting for the sender and receiver (Figure 3). A key idea of economic analysis is that these curves represent the emergent behaviours of populations of individual agents. Without necessarily understanding all the local decision criteria adopted by any individual agent, we can still make robust predictions about how a group as a whole will behave.
The curves indicate that when interactions are cheap, then senders are willing to make many of them. Conversely, when the cost of an interaction is high, senders only make a few. For receivers, the cost increases with the number of interactions, as interrupted tasks and newly acquired tasks compound the demands on their resources like time. Sender and
receiver curves may not always have this simple shape, but so long as the curves are non-identical and intersect, the following analysis should generally hold.
The point at which supply and demand curves intersect represents the point of market equilibrium, where most buyers and sellers have maximised their outcomes. Any transactions occurring at prices away from the equilibrium result in unhappy buyers or sellers, who then drive the market back to equilibrium as they try and minimise their dissatisfaction with the outcome (Frank, 1998).
This is clearly a simplification of what occurs in reality, which we will retain for the present discussion. However, just as some simple economic analyses assume agents have perfect information about the market place to decide what to buy and sell, we are making a similar assumption that our interacting agents have perfect information about which interactions to choose to make or receive. Yet we have already seen that under conditions of high cognitive load, some agents may be ‘blinded’ to the available interactions around them. Economic methods for handling situations in which information is ‘asymmetric’ between agents should also allow us to incorporate such details in interaction equilibrium models.
Getting Specific about interaction costs
We can assume that agents are willing to collapse their cost and benefit assessments into a 'price' at which they are willing to interact. In traditional economics, price is usually expressed in monetary units, but in this case the price of an interaction is a complex amalgam of personal costs and benefits. Amongst these, time is an easily quantifiable and identifiable component of the overall costs and benefits. As we saw earlier, message length with respect to a specific task is a proxy measure for cognitive effort, mediated by the available common ground.
We can now be more precise about the notion of interaction costs, based upon the general model of agent interaction presented in the previous section. Specifically, the channel across which an interaction occurs, and the state of grounding between agents are the two major external costs that shape the interaction curves for a given task.
Thus, it is cheaper to converse with an agent that understands the issues being discussed, compared to one that needs to be filled in with the necessary background examples. However, developing that background knowledge is not a free process, and in many situations an agent may feel that the cost of doing the work in building prior common ground is not worthwhile. We can summarise by noting that grounding costs relate to: ? the grounding efficiency of the interaction, which is reflected in the cost of constructing a message given the relative ground, and the extra ground that needs to be transmitted within an interaction to facilitate message interpretation.
? the cost of building and maintaining common ground prior to the interaction.
These two costs are traded off over time to arrive at the equilibrium point of the mediated center.
Similarly, there are costs associated with the channel chosen to mediate an interaction. Some channels may allow rapid and high-quality transmission of a message, but be very expensive. Others may be slow but cheap, and in many circumstances sufficient for the task at hand. Channel costs are numerous and include (Coiera, 2001):
? the actual price charged for using the channel
? the effect of noise on the transmitted message
? the effort involved in locating channel access
? the number of connection attempts required before another agent replies
? the opportunity costs involved in message delay due to the channels transmission characteristics like bandwidth and latency
? and similarly, the time it takes to transmit a message
Figure 3 - Supply and demand curves for interactions. Demand is generated by a population of potential sending agents, and supply provided by a population of agents who are potential receivers of those interactions.
The interaction equilibrium point in Figure 3 should represent the point in an organisation at which most senders and receivers are making mutually acceptable use of the communication system. Thus, for a given organisation, it's communications infrastructure and practices, the interaction equilibrium point represents the optimum communications outcome. However, while an organisation over time should reach equilibrium, and that equilibrium represents the optimal use of the existing communication system, the wider impact of that equilibrium may not be ideal. For example, the level of calls in a hospital may not be ideal if the traffic at equilibrium is adversely affecting patient care.
Predicting the impact of a new interaction class
Changing the communication infrastructure or practices in an organisation will alter caller and receiver costs and benefits when making calls, and thus alter the shape of the supply and demand curves, resulting in a new equilibrium. Thus the communication goals for any organisation are twofold. Firstly, they should free up the 'communications market', so that individuals can make optimal personal use of the existing system. Secondly, given other organisational goals and constraints, the ideal level of interaction should be estimated. Then, the underlying structure of the communication infrastructure and the way it is used should be altered to shift the equilibrium towards the ideal.
The notion of freeing up the communication market deserves some further comment. In one clinical user study, people were characterised as acting selfishly, reasoning only locally about their personal costs and benefits when interacting with others (Coiera and Tombs, 1998). This was portrayed in the study as being in some sense a negative characteristic of the study subjects' behaviour. However, in the present analytic framework, such locally motivated selfishness does not preclude a system moving to the optimal allocation of resources. In fact, it ensures it.
Thus, if we believe in this “free market” model, an organisation shouldn't use the tactic of fixing the costs of calls by edict e.g. limiting the time or number of calls. Individuals will simply behave in ways that circumvent these rules (Frank, 1998). What we can say is that if we want resource allocation that is close to ideal, we should not focus on the behaviours of individuals, who will always seek to optimise their individual circumstances, but we should seek to improve the underlying communication market's structure.
An individual's assessment of the costs and benefits of initiating and receiving interactions over the long run determine how they use a communication system. The only avenues for changing the resulting equilibrium level are to either change the decision-making process of individuals or the costs and benefits that drive those decisions. New technologies introduced into an organisation could plausibly do either of these two things. By making it easier to make a call, for example, a new technology will reduce costs and increase the number of instances in which people judge it worthwhile to call. Alternatively, by making hidden global costs and benefits explicit, a visualisation of the call traffic levels within an organisation may influence individuals to alter their local decision criteria.
Even though the supply and demand model of call behavior presented so far is very simple, it is enough to permit us to model the effects of technologies, and predict the communication patterns that will emerge in an organisation once a particular new technology is introduced.
Example 1 - Introducing mobile telephones into the workplace
What would the effect be of giving hospital workers mobile telephones? Using the current framework, introducing mobile phone can only either alter the supply of calls being answered or the demand to make calls. In fact, simple analysis suggests that it is the supply that is mainly altered. When someone carrying a mobile phone is called, there is an increased likelihood that they will answer, compared to someone who's phone is in an office or responds to a pager by finding a shared phone.
Figure 4 - Introducing mobile phones increases supply of calls, decreasing cost of calling and increasing the number of calls made.
So, we can say that the supply of calls received increases, and this has the effect of shifting the supply curve to the right (Figure 4). Consequently, the quantity of calls made will also increase, as the equilibrium shifts. This is because callers now find it cheaper to make a call in terms of the time taken for a call to be answered, and the number of call retries, increasing the overall level of interruption in an organisation.
Example 2- Asynchronous message services
So far the analysis has looked at the effects of technological interventions directly affecting equilibrium by changing characteristics of call supply or demand. In contrast, asynchronous services like voicemail or e-mail do not alter the supply of calls nor the underlying demand for calls to be made. Rather we can consider asynchronous services as direct product substitutes, competing directly with synchronous calls to satisfy some underlying communication need.
In this case, individuals make local cost and benefit assessments about whether to use a synchronous or asynchronous service to satisfy their communication need. When it is perceived that one type of service is cheaper and as effective as another, individuals will over time adopt the most economic modality of communication.
It is likely that there may be some inherent biases in this decision making, for example a habitual preference for synchronous systems simply because they were introduced first into the communication service 'marketplace'. Certainly, there has been an observed bias to synchronous systems in the example data that suggests that individuals considered synchronous methods to be better than the existing asynchronous ones (Coiera and Tombs, 1998).
If an asynchronous system were introduced into an otherwise synchronous dependent environment, and the new system was perceived to often be a cheaper alternative, then it would steal some of the existing synchronous market share, and have the effect of reducing the overall call demand. The net effect would be a decrease in the level of interruption and as well as call cost (Figure 5).
In practice, while asynchronous systems might be cheaper to use, for example not having any line busy or phone unanswered costs, they may have other hidden costs that make them unattractive. For example, the lack of feedback about whether a message has been received, read or acted upon may mean that even if there is no immediate need to complete a task, a caller cannot readily consider that sending an asynchronous message completes the a communication task. The costs here include checking back to see if the
Figure 5 – Asynchronous message services reduce call demand, resulting in a decrease in calls made and call cost.
Example 3 - Call filtering services
A system that permits a call receiver to automatically filter an incoming message can be built by detecting caller attributes such as past call history, telephone number, or receiver attributes such as current task, location and so on. Messages are filtered if they fail to meet specified attribute values for allowed calls. Such filters are commonly available today to help manage the growing amount of email many individuals have to manage as part of their organisational duties.
To a first approximation we can treat a call filter as a barrier, letting in selected messages and excluding the remainder. So, without changing the inherent demand for making calls, it does restrict supply. Call receivers thus limit the number of calls they take, at the increased cost of having to specify the behaviour of the call filter on their communication system, and perhaps also at the cost of the filter failing in some circumstances and blocking important messages. Equally, callers have to work harder to get a call through a filter on average as they build up a model of its behaviour and try and ensure their call meets the criteria for call success.
increasing the cost per individual call.
Consequently filtering services decrease the number of calls made, but at an increased cost per call for callers and receivers over time (Figure 6).
The effect of introducing more than one technology
The current qualitative analysis could now be transformed into a quantitative one, for example explicitly measuring costs to those making calls, and those receiving them. However, as we have seen, the qualitative argument is already sufficiently powerful to draw some robust conclusions. The analysis becomes more difficult however, when more than one new technology is introduced.
In particular, when opposing qualitative directions of change in supply or demand occur, we cannot say what the overall trend will be. Thus, introducing mobile phones and asynchronous messaging will decrease overall interaction costs, but we are unclear what the effect will be on the total number of interactions at the new equilibrium (Table 1).
Number of calls Call cost
Mobile telephony Up Down
Call filtering Down Up
Asynchronous messaging Down Down
Table 1 - Qualitative changes in call costs and overall call traffic
To resolve the issue, the opposing effects need to be relatively ordered in magnitude (Coiera, 1992). So, if a call filtering service was coupled with asynchronous messaging, we could have opposing effects on call cost, but reinforcing effects reducing the number of calls made. One would need to measure the cost of using the filtering service, and the changed demand on calls with competing asynchronous systems, to resolve the issue. Even simple order of magnitude measurements may be sufficient.
Supporting the economic analogy for interaction
The central idea presented in the last section is that we can profitably draw an analogy between the communicative interactions between agents and their economic interactions. More specifically, we can adopt some of the analytic framework of neoclassical economics to help in the analysis of general agent interactions. Interactions have been presented here as a set of ongoing adaptations by individual agents to optimise their overall resource utilisation and benefit. A similar argument has been made by Cliff and Bruten in economics, where they argue trading behaviours can also be considered to be an example of long-run adaptive behaviour (Cliff and Bruten, 1998).
Specifically, the conception of communicative interaction between agents as trading may appear problematic for a number of reasons:
? There is no competitive marketplace within which agents can trade. For example, an agent may only want to interact with another specific agents. How can that be competitive, when no other agents get the opportunity to interact? This criticism is answered by noting that there is no necessity for inter-agent competition implied in the interaction equilibrium model. Rather, the competition occurs between the various channels, ground and messages the agent can bundle together during interaction. One could even envisage an interaction equilibrium emerging with two agents over time.
Certainly, experiments in auction trading have shown that very few agents are needed to replicate normal market equilibrium (Smith, 1962; Cliff and Bruten, 1997). Having said this, the opportunities for interaction are enlarged when there is inter-agent competition, so it is a sufficient rather than necessary condition for the emergence of interaction equilibria.
? Money is not exchanged during an interaction, so there is no external measure of utility, nor compensation available for engaging in the interaction. A simple response to this is to simply note that in many cases a true price can be attached to an interaction, and money exchanged. For example, providers of information services charge for answering specific questions. In most cases, however, agents certainly do not charge each other for interaction. They do however make judgments about the cost and benefits of engaging in the interaction. In such circumstances, when costs and benefits are not readily reflected in monetary terms, economists use what is called
a 'reservation price' for the activity (Frank, 1997). This is the amount of money at
which an agent would no longer be indifferent between doing and not doing something. So, people can be characterised as making decisions 'as if' everything has
a price attached. In fact, we often consider the time spent in telling someone
something as a loosely transferable thing, for example after explaining something to someone, you can consider that they now 'owe you one'.
? Agents may not be able to make free choices to interact, nor benefit from those interactions. For example, working within an organisation, an agent may be compelled to answer telephone calls, thus having neither direct choice in the response, nor direct benefit. In response, one could simply observe that many buyers in the economic marketplace are themselves constrained intermediaries. For example,
a buyer for a clothing store may gain no specific benefit form purchasing stock for the
store, nor have the choice to elect not to purchase stock. Economists thus need to invoke several models of self-interest (Frank, 1997). In one, the self-interest standard
of rationality, agents only consider the costs and benefits that directly accrue to themselves. In the present-aim standard of rationality by contrast, they simply act efficiently in satisfying their current goal eg working for an organisation, they will maximise results for the employer. Both are suitable foundations for developing economic models. One can also note again that even when the choice to interact and select a channel is removed, an agent still has choice about message constructing and grounding. One can choose to answer a question to the best of our abilities, or do the minimum possible.
? There are limited choices of channels, agents, and messages in many organisations that prevent optimal interaction equilibria developing. For example, an organisation may only supply telephones, or limit the time available to answer a telephone call. It is indeed hard to argue against the existence of such structural impediments in some organisations. In practice we approach communication in more or less optimal ways, depending upon organisational communication processes, and mechanisms for making communications decisions. However, just as a command and control economy doesn't preclude the existence of a free market in other countries, a restricted interaction environment in some organisations doesn't preclude their 'market' being freed up or other organisations adopting a different model. And the more freely determined agents are in their interaction choices, the more 'efficient' the 'market' should become.
? In a free market, no individual intentionally enters into a loss making deal. They enter any trade with the expectation of some benefit. However, when an agent answers a telephone, for example, the receiver has restricted choice in answering, and knows nothing about the 'deal' being offered which could be costly. This criticism is actually directed at the phone as a 'trading' mechanism, rather than the notion of interaction as trading. When agents answer a telephone, they indeed do not have information about the proposed interaction. But we could construct scenarios in which recipients do have the choice to answer, and information about the 'deal' being offered. For example, an answering machine can be set to screen calls. Based upon the 'bids' of people talking to the machine, an agent can choose to answer or not.
There is even greater freedom in choosing to interact with other communication services like e-mail. However, it is only recently that we have the means to do so.
This makes for interesting predictions about the optimising benefits of communication services that permit freer choice (e.g. call screening or e-mail) compared to those that make choice opaque such as the telephone
Conclusion
In this paper I have show how we can approach the design of interactions within an organization by viewing them from the point of view, not of the technology mediating the new interaction, but of the agents who are asked to use the new technology. Understanding the limits to individual agent resources should allow an analysis of the impact that a new interaction will have to a given setting. When we look beyond simple interaction settings, we can use the notion of interaction equilibria to predict the impact a new interaction class will have on the interactions within an organization. Economic
supply and demand curves, for example, may allow us to make both qualitative and quantitative predictions about technological adoption of communication systems.
Rather than focusing solely on characteristics of individual technologies, or psychological and social issues, these are combined to explain the overall decisions of individuals using technologies. In particular, we have seen that the supply and demand curves represent the emergent behaviours of populations of individual agents. Without necessarily understanding all the local decision criteria used by any individual, we can still make robust predictions about how a group as a whole will behave. Future work will extend this analysis to a broader technological framework, and show how a deeper analysis of local communication behaviours can help us generate economically in spired population-level explanations of information and communication technology use.
INTERACTION DESIGN: Industrial Design in the Information Age Elaine Ann Director,Kaizor Innovation Abstract:This paper introduces a newly developed discipline in the U.S.and Europe:Interaction Design.As we enter the Information Age,products are no longer only electrical and mechanical,but also include computing and networked capability.Designing products highly interactive in nature becomes much more complex than before going beyond the traditional realm of Industrial Design.Moreover,the fundamental definition of"a product"is being challenged and requires a fundamental shift in thinking as well as new work methods.How people interact with products,systems or environments and its social and cultural impact is what Interaction Design is concerned about. Keywords: Interaction design,user experience,networked products,interdisciplinary,industrial design. 1Introduction Traditional products are mechanical and electrical like toasters,shavers,walkmans etc.With today's increased computing power,miniaturized chips and the advent of the Internet,this drastically alters the meaning of traditional products.Industrial Design has always dealt with how people interact with things,designing for a product's form factor,ergonomics,psychonomics etc. Computing and networked products introduces a new dimension of interactivity beyond with its physical form,but extends to the digital arena.Now that most products are embedded computing with complex interaction,what should industrial designers design in the information and digital age?How should industrial designers innovate for these new breeds of products? 2Product nature redefined in the Information Age 2.1Products are software driven Traditional products are mainly physical in nature and design constrains are governed by principles of physics and mechanics.Today,many products are in fact microcomputers in disguise with computing capability storing more than40GBs of digital information and ever increasing processing speed.Products become hybrid in nature with both hardware and software components.For example:traditional walkmans have evolved into digital MP3music players,mechanical cameras into digital cameras,cordless phones into cellular phones etc. What defines a product is not so much dependent on the hardware when a products'function resides more in its software capability.(Figure1)
Part 2 Design practice Chapter 4 Usability paradigms and principles Overview Designing for maximum usability is the goal of interactive systems design. Examples of effective strategies for building interactive systems provide paradigms for designing usable interactive systems . The evolution of these usability paradigms also provides a good perspective of the history of computing. Abstract principles offer a way of understanding usability in a more general sense, especially if we can express them within some coherent catalogue. 4.1 Introduction As we noted in chapter 3 ,the primary objective of an interactive system is to allow the user to achieve particular goals in some application domain ,that is the interactive system must be usable .The designer of an interactive system ,then ,is posed with two open questions: 1.How can an interactive system be developed to its usability? 2.How can the usability of an interactive system be demonstrated or measured? There are two approaches to answering these questions. The first is by means of example , in which successful interactive system are commonly believed to enhance usability and ,therefore , serve as paradigms for the development of future products . The second approach is more theoretically driven , deriving abstract principles for effective interaction from knowledge of the psychological , computational and sociological aspects of the problem domains . These principles direct the design and evaluation of a product from its onset. This destinction between paradigms and principles is an important reflection on the history of HCI as a discipline . We believe that we now build interactive systems which are more usable than those built in the past . We also believe that there is considerable room for improvement in designing more usable systems in the future . As discussed it Chaper 2 , the great advances in computer technology have increased the power of machines and enhanced the bandwidth of communication between humans and computers . The impact of technology alone , however , is not sufficient to enhance its usability . As our machines have become more powerful , the key to increased usability has come from the creative and considered application of the technology to accommodate and augment the power of the human . Paradigms for interactive have for the most part been dependent upon technological advances and their creative application to enhance interaction . Principles for interaction are independent of the technology ; they depend to a much greater extent on a deeper understanding of the human element in the interaction . The creative development of paradigms for interaction , which we discuss in Section 4.2 , is the main reason we are able today to build more usable systems . The problem with those paradigms is that they are rarely well defined . It is not clear how they support a user in accomplishing some tasks . As a result , it is entirely possible that repeated use of some paradigm will not result in the design of a more usable system . Derivation of principles for interaction , discussed I Section 4.3 , has usually arisen out of a need to explain why a paradigm is successful and when it might not be . Principles can provide the repeatability which paradigms in themselves cannot provide . However , in defining these principles it is all too easy to provide general and abstract definitions which are not very helpful to the designer . Therefore , the future of interactive system design relies on a complementary approach . the creativity giving rise to new paradigms should b strengthened by the development of a theory which provides principles to support the paradigm in its repeated application . 4.2 Paradigms for interaction As we have said , the greatest advances in HCI have come by way of exploratory and creative design . In this section , we investigate some of the principal advances in interactive designs . What is important to notice here is that the techniques and designs mentioned are recognized as major improvements in interaction , thought it is sometimes hard to find a consensus for the reason behind the success . We will discuss 14 different paradigms in this section . They do not provide mutually exclusive categories , as particular systems will often incorporate ideas from more than one of the following
ISSN 1923-1555[Print] ISSN 1923-1563[Online] https://www.doczj.com/doc/4017814213.html, https://www.doczj.com/doc/4017814213.html, Studies in Literature and Language V ol. 7, No. 1, 2013, pp. 22-26 DOI:10.3968/j.sll.1923156320130701.3085 Classroom Interaction and Second Language Acquisition: The More Interactions the Better? ZHAO Congmin [a],* [a] School of Foreign Languages, North China Electric Power University, Beijing, China.* Corresponding author. Received 17 November 2012; accepted 6 January 2013 Abstract This paper attempts to describe the relationship between interaction and SLA in the classroom. Research findings tend to point to the conclusion that more involvement in interaction does not ensure better achievements. This conclusion points to the importance of looking at classroom interaction (CI) and second language acquisition (SLA) holistically. Learners learn by engaging in interactions per se but also by listening to interactions. The implication for classroom pedagogy is that the teacher should not encourage more interactions single-mindedly but base his decisions of varying the dimensions of CI on a host of factors. Key words : Interaction; Observable participation; Unobservable participation; Eavesdropping ZHAO Congmin (2013). Classroom Interaction and Second Language Acquisition: The More Interactions the Better?. Studies in Literature and Language , 7(1), 22-26. Available from: https://www.doczj.com/doc/4017814213.html,/index.php/sll/article/view/j.sll.1923156320130701.3085 DOI: https://www.doczj.com/doc/4017814213.html,/10.3968/j.sll.1923156320130701.3085 According to Ellis (1994), the classroom provides the L2 researchers with three perspectives of study: comparative method studies, the study of the effects of formal instruction, and classroom interaction (CI) studies (p.565). Among the three, it is the last which attracts researchers’ prolonged interest. The reasons are simple and evident: firstly, interaction is “the fundamental fact of classroom pedagogy …everything that happens in the classroom happens through a process of live person-to-person interaction” (Allwright, 1984, p.156). Secondly, interaction plays an important role for second or foreign language acquisition (SLA/FLA). It provides the opportunity for the obtaining of comprehensible input and the production of pushed output which are crucial for the internalization of language knowledge. Language serves for communication and the acquisition of a language is generally fulfilled in the interaction with others. In interaction with others one learns to use language and resultantly to modify and expand the IL system. Classroom interaction, compared with interaction in the naturalistic environment, presents different patterns of interaction with different characteristics, such as teacher-fronted interaction and small group work. The functioning of different types of CI has close connection with the organization of classroom activities which is generally divided into three broad stages: presentation, practice and production, and this fosters the types of interaction to happen and student participation influences the real occurrence of interaction. This paper aims to identify the relationship between interaction and SLA in the classroom setting, that is, how these interaction opportunities bear on the learner’s language acquisition; is it true that learners who actively initiate interaction and negotiate meaning are better achievers than those that are not? Or in other words, is it the more interactions the better? 1. THEORETICAL BACKGROUND The introduction of the interactive approach into classroom learning and the study of CI is largely attributed to social interactionism which emphasizes the role of other speakers around the language learner by means of interaction. 1.1 Social Interactionism Actually, the realization and recognition of the role of interaction for language learning are recent events. According to Richards and Rodgers, “Interaction has been central to theories of L2 learning and pedagogy since the 1980s” (p.22).
主题公园创意景观规划设计 主题公园型旅游综合体创意建筑景观规划设计是一种人造旅游吸引物，按照一个或多个特定的主题进行规划设计，采用现代化的科学技术和多层次空间活动的设置方式，成为集聚诸多娱乐内容、休闲要素和服务接待设施于一体的现代旅游场所。 主题公园是现代旅游业在旅游资源的开发过程中所孕育产生的新的旅游吸引物，1955年，美国著名动画片制作家沃尔特·迪斯尼在洛杉矶建造迪斯尼乐园，标志着主题公园的诞生距今已经50多年的历史。中国主题公园发展自1983年因拍摄四大名著而建设的大观园、西游记宫，至今已发展到第三代主题公园产品。
中国主题公园发展历程图 (一)开发条件 1、资源条件 主题公园是一种人造旅游吸引物，围绕一个或多个特定的主题，采用现代化的科学技术和多层次空间活动的设置方式，形成聚集诸多娱乐内容、休闲要素和服务接待设施于一体的现代旅游场所。 创意性、具有启示意义的主题是主题公园的灵魂，是主题公园区别于其他商业娱乐设施的根本特征。因此，主题公园的主题选择就是其核心的资源。成功主题公园的运作经验表明，主题公园的主题必须鲜明，针对特定的细分市场，满足特定客源的需求。主题结构可以是一个主题多个次主题，也可以是一园多个主题。 其次，由于主题公园所提供的产品是一种以旅游方式被消费的文化产品，从这个意义上来说：主题公园是一种特殊的以旅游为经营形式的文化产品制造商。同时，文化的特性决定了文化产品应该具有鲜明特色，而且这种特色应该具有可以被欣赏的群众基础，就需要对基底的文化进行挖掘和适应，避免文化的滥用。 迪斯尼乐园在世界上很多国家都获得了巨大成功，可是在法国一度遭到失败，这正好印证了主题公园是以旅游方式经营文化产业的判断。美国式的快餐文化与法兰西文明在文化价值取向上的差异导致游客对主题公园产品的选择差异。本质上是一种文化认同风险所导致的运作失败。
METHODS Estimating measures of interaction on an additive scale for preventive exposures Mirjam J.Knol ?Tyler J.VanderWeele ?Rolf H.H.Groenwold ?Olaf H.Klungel ? Maroeska M.Rovers ?Diederick E.Grobbee Received:30July 2010/Accepted:4February 2011/Published online:23February 2011óThe Author(s)2011.This article is published with open access at https://www.doczj.com/doc/4017814213.html, Abstract Measures of interaction on an additive scale (relative excess risk due to interaction [RERI],attributable proportion [AP],synergy index [S]),were developed for risk factors rather than preventive factors.It has been suggested that preventive factors should be recoded to risk factors before calculating these measures.We aimed to show that these measures are problematic with preventive factors prior to recoding,and to clarify the recoding method to be used to circumvent these problems.Recoding of preventive factors should be done such that the stratum with the lowest risk becomes the reference category when both factors are considered jointly (rather than one at a time).We used data from a case-control study on the interaction between ACE inhibitors and the ACE gene on incident https://www.doczj.com/doc/4017814213.html,e of ACE inhibitors was a preventive factor and DD ACE genotype was a risk factor.Before recoding,the RERI,AP and S showed inconsistent results (RERI =0.26[95%CI:-0.30;0.82],AP =0.30[95%CI:-0.28;0.88],S =0.35[95%CI:0.02;7.38]),with the ?rst two measures suggesting positive interaction and the third negative interaction.After recoding the use of ACE inhibitors,they showed consistent results (RERI =-0.37[95%CI:-1.23;0.49],AP =-0.29[95%CI:-0.98;0.40],S =0.43[95%CI:0.07; 2.60]),all indicating negative interaction.Preventive factors should not be used to cal-culate measures of interaction on an additive scale without recoding. Keywords Interaction áPreventive factors áRelative excess risk due to interaction áSynergy index Introduction Interaction refers to the situation where the effect of one exposure on a certain outcome is different across strata of another exposure.This means that if interaction between two exposures is present,these exposures are not inde-pendent in causing a certain outcome.A classical example is the interaction between smoking and asbestos on the risk of lung cancer [1].The presence and direction of interac-tion depends on the scale,e.g.additive or multiplicative,that is used.Interaction on an additive scale means that the combined effect of two exposures is larger (or smaller)than the sum of the individual effects of the two exposures,whereas interaction on a multiplicative scale means that the combined effect is larger (or smaller)than the product of the individual effects.A number of epidemiologists have argued that biologic interaction should be assessed on an additive scale rather than a multiplicative scale [1–6].Interaction on an additive scale can be calculated using relative risks and different measures quantifying this interaction have been described,such as the relative excess risk due to interaction (RERI),the proportion attributable to interaction (AP),and the synergy index (S)[7].Provided M.J.Knol (&)áR.H.H.Groenwold áM.M.Rovers áD.E.Grobbee Julius Center for Health Sciences and Primary Care,University Medical Center Utrecht,PO Box 85500,3508GA Utrecht,The Netherlands e-mail:m.j.knol@umcutrecht.nl T.J.VanderWeele Departments of Epidemiology and Biostatistics, Harvard School of Public Health,Boston,MA,USA e-mail:tvanderw@https://www.doczj.com/doc/4017814213.html, O.H.Klungel Division of Pharmacoepidemiology and Pharmacotherapy,Utrecht Institute for Pharmaceutical Sciences,Utrecht University,Utrecht,The Netherlands Eur J Epidemiol (2011)26:433–438DOI 10.1007/s10654-011-9554-9
交往焦虑量表 Interaction Anxiousness Scale (Leary 1983c） 一、简介 交往焦虑量表(Interaction Anxiousness Scale，IAS )用于评定独立于行为之外的主观社交焦虑体验的倾向。 IAS含有15条自陈条目，这些条目按5级分制予以回答。（1：一点儿也不符合我；5: 非常符合我）。条目是根据下述两个标准选出的：(1)：涉及主观焦虑（紧张和神经症）或其反面（放松、安静），但并不涉及具体的外在行为。(2)：条目大量涉及意外的社交场合。在这些场合中个体的反应取决于在场其它人的反应，或受其影响（与之相反的，例如公开演讲场合）。量表历经四阶段从最初的87条中选出了现在的15条。其总评分从15（社交焦虑程度最低）到75（社交焦虑程度最高）。 在美国，在不同地区对大学生进行各种规模的测试时，IAS的均值及标准差是相当稳定的。来自三所不同大学的1140名受检者的均值为38.9 (SD=9.7) (Denison大学Texas大学及Wake林业大学）。 二、信效度 内部一致性：量表所有条目与其它条目的总数相关系数至少为0.45，Cronbachα系数超过0.87。八周的重测相关系数为0.80 IAS与其它测量社交焦虑及羞怯量表高度相关(r IW>0.60) (Jones，Briggs及Smith，1986，Leary与Koualski 1987）。 此外，IAS与在真实交往中的自陈焦虑相关良好。与低得分者相比，高得分者陈述在人际交往之前及之中都更加焦虑及缺乏信心，并关注在交往中别人怎样看待他们。在交谈中也更多的感到抑制。别人也认为他们表现的显得较为紧张及缺乏信心(Leary；1983c，1986b)。高得分者还耽心别人如何评价其外表(Hart、Leary及Re jeski，1989)。得分与在面对面的交往时的心率增加有关。IAS评分与社交回避及抑制量表正相关(Leary，Atherton，Hill及
龙源期刊网 https://www.doczj.com/doc/4017814213.html, 初探创意产业园的景观规划 作者：叶柳萍 来源：《建筑建材装饰》2017年第01期 摘要：近年来，在中国许多城市不断出现了创意产业，形成了当下特有的“文化创意产业园”，本文以福大怡山文化创意园项目为例，首先阐述文化创意产业园的概念和特征，着重论述了基于高校旧址的创意产业园的景观规划，提出了在文化创意产业区景观设计中文脉精神和创意氛围的重要性，提出了一些关于创意产业园的景观设计思路，以供参考。 关键词：创意产业园区；创意氛围；景观规划 中图分类号：TU986.2文献标识码：A文章编号：1674-3024（2017）01-84-02 1.1创意创业园概述 1.1概念及特征 在工业高度发达的今天，人类社会日益呈现出个性化，人性化和多元化的趋势，同时，文化和艺术在经济发展的过程中也越来越起到不可忽视的作用。创意产业是一种新兴的产业类型，作为其载体的创意产业园区也应运而生，并迅速发展起来。创意园区的环境景观设计是反映创意园特色的组成部分之一。因此，创意园去的景观设计思路和注意事项对于创意园区景观营造创意氛围和发扬艺术魅力有重要意义。 创意园区的的建设依托各类创意产业，以高科技手段支撑，文化艺术与经济发展结合，具有时代特色。强调以创意型文化为主题，是创意园区最鲜明的特征之一。 1.2国内创意园区的发展及存在的问题 当前，国人许多经济中心城市整积极的推进创意产业的发展，大力推进创意产业的空间集聚，构筑创意产业园区。近年来，以一线城市为先涌现了一批具有开创意义的创意产业集聚区，如：北京的798、上海的田子坊、赤峰路建筑设计一条街、广州红专厂创意生活区等等，这些园区或依托城市中废弃的工业建筑，或是依托大学作为科技和人才的聚集地，因其空间宽敞和廉价租金，成为创意产业的滋生之地。 现在全国的文化创意园如雨后春笋般遍布各大城市，除了规划建设的以外，很多地方甚至将原来的开发区和科技园区等更名为创意园。这其中存在着大量的鱼目混杂，尽管社会各界纷纷提出质疑，但是不少地方的创意园仍纷纷立项，从而导致创意产业园的建设存在以下问题： （1）定位模糊，缺乏科学管理
Interaction Layer
GENy
? 2011. Vector Informatik GmbH. All rights reserved. Any distribution or copying is subject to prior written approval by Vector. V1.01 2011-05-18
Agenda
INTRODUCTION TRANSMISSION WITH INTERACTION LAYER RECEPTION WITH INTERACTION LAYER STATE MACHINE OF THE INTERACTION LAYER FUNCTIONS OF INTERACTION LAYER GENY - GENERATION TOOL
? 2011. Vector Informatik GmbH. All rights reserved. Any distribution or copying is subject to prior written approval by Vector. Slide: 2
Q
Communication Needs for ECUs
Typical ECUs Dashboard ABS Dashboard
CAN
Engine Speed
ABS
Information exchange between the ECUs Engine Speed (ES): rpm (revolution per minute)
? 2011. Vector Informatik GmbH. All rights reserved. Any distribution or copying is subject to prior written approval by Vector. Slide: 3
Q
艺术型文化创意产业园公共空间景观设计初探 发表时间：2017-10-11T12:09:35.157Z 来源：《基层建设》2017年第14期作者：蒋娜[导读] 摘要：在创意经济时代，推动经济增长的主要因素不再是技术与信息，而是创意和创新。我国的文化创意产业已进入快速增长期，产业领域 浙江安道设计股份有限公司浙江杭州 310000 摘要：在创意经济时代，推动经济增长的主要因素不再是技术与信息，而是创意和创新。我国的文化创意产业已进入快速增长期，产业领域、文化领域对其集聚区的研究很多，但针对园区公共空间景观方面的研究还很欠缺，艺术型文化创意产业园是其中活动最为艺术，场所最为个性，社会效益最为突出的一类。 关键词：文化创意；公共空间一艺术型文创园公共空间概述文化创意产业园概述 1.文化创意产业的概念及特点 继农业经济以自然资源、工业经济以资金设备为重要资源之后，知识经济使智力、创意和人才作为经济资源获得了空前重要的战略地位。文化创意产业指的是文化内容通过创意手段，并取得经济效益的所有生产和销售与精神消费品相关的产业，具有高创意、高附加值、高流通、高渗透性等特点。 2.文化创意产业园区的概念及功能 中国当前的文化创意企业，几乎都是小型甚至微型的，有的只不过是工作室或创意个体，在这种情况下，文化创意企业集聚起来形成集群，可以大大减少成本，降低风险，提高效益。同时，许多资源集中到一起，可能催生出更优秀的创意。创意产业园作为文化创意产业集群的外在形态，是一个界定了的、具备一定规模的特定区域，有较为完善的公共设施、社会网络和管理系统，以创造性智力劳动为主。 二艺术型文化创意产业园的社会功能及特点 1.艺术型文创园的社会功能 文化创意产业园区作为后工业时代的一种产业集聚形式，按照产业链的模式形成利益共同体以创造竞争优势是其最基本的经济功能。与此同时，艺术型文化创意产业园还具有更重的文化事业色彩，其社会功能包括：（1）培养、保护艺术人才。 艺术人才是艺术创作中最活跃、能动性最强的因素，在艺术型文创园区内保持艺术人才层次的多元性有助于艺术人才业务能力的培养和提升，艺术门类的不同有助于彼此间相互借鉴，年龄层次的完整性有助于实现继承与创新，从整体上促进艺术家群体的孵化。园区内沙龙、展览和研讨会的定期举行能够促进艺术家之间的交流与学习，其他业态的存在也帮助艺术家成长。园区服务平台可为维护艺术家的人身权利和经济权利提供法律咨询和保障，尤其是知识产权保护方面。 （2）平衡、繁荣艺术生态。 艺术型文创园的社会价值并不是体现在能够成就几位艺术巨匠，而在于这些或成功或平凡的艺术家们独特的生活方式、高价值的精神创作可为园区及周边区域营造浓厚的文化氛围。园区能够帮助艺术家摆脱单纯创作的栓桔投身于完整的艺术生态链条中，在比较和学习中直观地审视自己的作品，在“往复上升”的过程中获得教益并继续创作。 （3）提升城市品位，增强国家软实力。 艺术型文创园作为艺术家集聚的场所，是城市中艺术类人才和文化设施的高密度区域，有完整的艺术家培养及推广机制，文化艺术氛围浓厚，能够对城市居民起到艺术熏陶和审美教育的作用。园区在一定意义上代表了城市艺术发展的现有水平，塑造城市形象的同时也为城市旅游业提供了一个新的增长点。当下，现代艺术占据着中国艺术的高地，我国传统的艺术门类和艺术特质在逐渐被消解，艺术型文创园可以通过繁荣艺术事业增强国民的文化认同感和自信心。 三艺术型文化创意产业园公共空间的界定创意产业园的物质空间具有多样、模糊、功能复合性的特点，其基本使用功能大致可分为工作、交流、展演、休憩、商业和其他六类。而创意产业园的公共空间指的是面向公众开放的承载各种活动的空间，即在园区用地范围内，为现有企业驻户及周边市民提供交往、展示、休闲、消费、娱乐等功能的物质空间，包括各种室内、半室外、室外空间。本文中的公共空间景观设计重点指室外空间，既可能存在于园区的建筑群体之间，也有可能与周边建筑外部空间产生部分重叠，或与街道、广场、绿化等城市开放空间产生重叠。具体实例的室外公共空间的范围可以通过其位置、功能、与内部空间的关系来界定。根据功能特征可以分为广场、内街、庭院、屋顶平台、绿化等。 四营造创意氛围 1.规划多义性的公共空间 创意的产生离不开人与人的思维交流、人与自然的情感交流及人与网络的信息交流，这些交流可以随时产生和中止，不受时空的限制。多义性公共空间具有功能的多元和功能转化的灵活，能够满足不同活动的要求，可以提升人气和活力，实现创意的互动。在具体设计手法上，可以增加不同空间活动时的视线联系，提高到达公共空间的便利程度，适度扩大公共空间的规模尺度，为不同时段的使用提供物理条件，为容纳多种活动设计可变的分隔界面等。 2.保留场所记忆 当公共空间的设计形式、语言来自于园区的历史和地域文化，或者艺术家们的生活方式时，会有助于艺术家们生成场所归属感，令参观者感受到独特的场所气质。在具体的设计手法方面，可以选择代表的构筑物、植物、小品进行遗迹保留，还可以经过富有创意的加减设计、具象抽提、符号化等方法进行形式保留，在操作时可以充分调动园区艺术家的创造热情，为其提供装饰表达的余地。 3.引导自主观察 自主观察行为有助于创意的产生，通过物质空间的营造可以引导观察行为，激发好奇心理，丰富主体感受。在具体设计手法上，可以增加迂回以创造活泼的行进路线，丰富层次以加强空间的深邃感，设置一些“安全点”以及活动座椅以鼓励休息人群观看周边活动。 五展现艺术氛围 1.运用多元化的艺术语汇