From CS 160 Fall 2011

Slides

Readings

Required:

• The role of ethnography in interactive systems design John Hughes et al., ACM Interactions, vol 2(2), 1995.

• On "Technomethodology: Foundational Relationships Between Ethnomethodology and System Design Paul Dourish, Graham Button, Human-Computer Interaction, vol 13, 1998

Videos

• Jane Fulton Suri on Ethnography and Design

• Genevieve Bell on Ethnography at Intel

Discussion Questions

• Contrast ethnographic inquiry with contextual inquiry.

• Compare Dourish and Button's description of "accounts" with user conceptual models we discussed earlier.

Your Response

Huan Ji 21:13, 10 October 2011 (PDT)

conceptual inquiry focuses on tasks, but ethnography is limited to realatively small scale and confined environment and focuses on internal relationship. Even though they are both connected with field work, they have different purpose. Ethonography needs more time and tends to research and ethical responsibilities, but conceptual inquiry will focus on short term and quick response and can be processed in laboratory.

Conceptual model is for users's behaviors, but account is defined as interation model and it's about the abstraction of the interactive systems. It "concerns the way in which that action is organized so that it can be made rational in particular circumstances" and is about the system's behaviors which accountabilities is not out of specific requests.

Frank 12:26, 10 October 2011 (PDT)

Contextual Inquiry is inherently a design based inquiry while ethnographic inquiry is rooted in sociological research. They both seek to understand their subjects on a sociological level, but the main difference lies in the scope and focus of the respective studies. Contextual inquiry generally is a relatively short and focused process where the goal is predefined whereas in an ethnographic study, the goal is simply to understand for the sake of understanding.

Accounts are defined by Dourish and Button as a self indicator of a system's status and scenario whereas conceptual models tend to be expressed before run-time. Conceptual models are used to assist in design and ideation while accounts can be considered a type of functionality in their own rights.

Eric Shih 12:40, 9 October 2011 (PDT)

Both ethnographic inquiry and contextual inquiry involve understanding situations/scenarios as they naturally occur, but ethnographic inquiry usually involves a much longer period of time, often months or years. Ethnography has not been tried as much as contextual inquiry, and as such, its potential beyond controlled, undisrupted environments is less clear. Both methods require collaboration between the incumbents and the studiers. Although ethnography typically involves a prolonged study, it can still be done in short periods (a few weeks) and produce results, dependings on the requirements of the study.

Dourish and Button describe an "account" to be produced each time a user interacts with a system in a setting and represents how they rationalize and interpret the system's behavior. They make the point of saying that the "creation of the account happens...in every circumstance in which the system is used," in order to indicate that the account is strongly correlated with the circumstance during which the system is used. They also go on to say that an account is "a run-time phenomenon, not a design-time one." In other words, accounts are models of a system's activity that the system itself provides to help a user understand what it is doing. Conceptual models, on the other hand, are pre-existing notions or beliefs a user has on how a system should work, which inform the user on how they think they should act. Accounts are more activity/circumstance driven, whereas conceptual models are more abstract and created over time.

Terri Yeh 16:33, 9 October 2011 (PDT)

Ethnography inquiry and contextual inquiry both seek to understand the domain for which the team is designing in its context. Researchers/ethnographers/engineers venture into an unfamiliar workplace to see how the current system is being used, and from there, come up with designs for the upcoming system. Ethnographic inquiry usually takes longer than contextual inquiry: months or years compared to days. It's a more involved study of the social components of the domain. And for the most part, ethnographic inquiry has only been employed in controlled environments, so we don't yet know how it will fare in everyday situations.

User conceptual models are user's understanding of how a system works. We may say it's the user's version of generalization/abstraction of the system over their various experiences with the same system. Forming a conceptual model of a system is an one-time thing; once the user has a working model, they are set on that model whenever they encounter this particular system. On the other hand, an account of a system is created in every circumstance in while the system is used. The effort is an on-going one instead of an one-time thing. Both an account and a conceptual model are generalizations of a system, but the former is a generalization over the system behavior and circumstance of usage in one single experience, whereas the latter is a generalization over system behavior in all encounters (most of the time disregards circumstances of usages)

Konstantin Rud 16:32, 9 October 2011 (PDT)

Ethnographic inquiry has it's roots in anthropology and sociology, whereas contextual inquiry is a HCI term for a similar thing. However, ethnographic inquiries tend to take a much longer time than contextual inquiry. They both have the purpose of observing a group of people for a period of time in order to determine how they think and are more detailed than a survey or a cursory observational period. Dourish and Button's "accounts" is an ethnomethodological term which they define as "a model of the system's activity offered by the system to account for and cast light on its own action." In other words this is different from a mental model in that an account exists in the system and is designed with the system. Another thing the author's point out is that an account occurs only during run time whereas a conceptual model always exists in the user and is the user's understanding of how the system works.

Jonah Turnquist 18:32, 9 October 2011 (PDT)

Ethnographic inquiry is similar to contextual inquiry but ethnographic inquiry is more culture based. It generally takes much long than contextual inquiry - usually months or years. They still accomplish the same thing in that they are made to learn about how the users interact. Ethnographic inquiry is less practiced than contextual inquiry.

Kevin Chung-Kai Wang 18:55, 9 October 2011 (PDT)

Both contextual and ethnographic inquiry tries to study consumer behavior when using designs in their natural environments, so the people studying them can tweak those designs to make them better to use. The difference between contextual and ethnographic inquiry is that ethnographic inquiry requires a longer time period that can span months or years. Ethnographic inquiry comes from sociology while contextual inquiry comes from HCI, and can be considered a less thorough version of ethnographic inquiry modified for the purposes of improving design rather than pure research.

Conceptual models are pre-existing understandings of how systems work or should work that are formed by the user of that system. Accounts are formed by the system that create new understanding of how the system works. Accounts are designed with the system and form with new experiences, while conceptual models are user generated and forms from long-established experiences.

Josef John 19:16, 9 October 2011 (PDT)

Ethnographic inquiry is newer and takes a lot more time and thus effort.

What stuck out most to me out of all the readings / clips was Genevieve Bell saying "Chinese Malays" instead of Chinese Malaysians. Malays are an ethnic group in Malaysia (country). She meant Chinese in Malaysia, not people that are ethnically mixed Chinese and Malay.

== Hussein Omo Kadiri 13:41, 10 October 2011 (PDT) == 

Ethnograpic inquiry is used mainly in anthropology. It is when a person goes into a culuture and lives amongst the people in order to learn about the culture. Compared to contextual inquiry, ethnographic inquiry takes much longer to do. Another difference is that in ehtnographic inquiry, the person performs the activity while in contextual inquiry, the person is just observing. Both perform the same task which is understanding the user of the application.

Difference between Dourish and Button's description of "accounts" is that conceptual Models are user's perception of the system while accounts are accounts are system tools that aid the user in understanding that the system is doing.

Gong Cheng 19:43, 9 October 2011 (PDT)

Ethnograhic inquiry is similar to a subcategorie of a contextual inquiry since they both involve getting the information of the current scenarios and social backgrounds. However, ethnographic inquiry digs into the environmental/social influence of the target groups. Therefore, conducting an ethnographic inquiry takes longer time and requires further study into the target group. And such a technique is used less in the industry. However, in the end of the first reading, a quicker modification is suggested.

Conceptual models capture the entities and relationships between two different concepts. Therefore, it is created according to each person's understanding of the existing society and world. Accounts according to Dourish and Button stand for the entities/relationship formed when a new relationship is created. Furthermore, an account only exists when a new relationship is formed and is dependent on the each individual's understanding of that relationship.

Januardo Kusuma 20:48, 9 October 2011 (PDT)

Ethnographic vs Contextual Inquiry

Both of them are basically the same thing. The main difference is that ethnographic inquiry takes a lot longer to get done than contextual inquiry. Ethnographic inquiry method is based on sociology and/or anthropology field. Thus, we do not only focus on environment influence to the target users, but also the social influence as well in which contextual inquiry does not focus on.

Compare Dourish and Button’s description of “accounts” with user conceptual models we discussed earlier

Conceptual models are the pre-existing users’ understandings about a system. They use the system once, they get the general picture of the system, and it is going to set in their mind. Some would say it will last for a long time. Accounts are created not by the user, but by the system itself. The system that the users try is completely different from what they have in mind. Thus, account will always get created by systems that are not as expected as what the users think.

Guoxiong Xie 20:51, 9 October 2011 (PDT)

In the course of developing UI, both of Ethnographic inquiry and contextual inquiry require communication between potential users and developers. Both of them discourage isolating UI development from social context. However, ethnographic quiry in general requires a more lengthy investigation and iteration. Furthermore, the environment in which ethnography inqury is conducted is more confined compared to contextual inquiry. In ethnography inquiry, no intervention or distuption is allowed. Contextual inquiry is more task-based; when the potential user perform the task, interviewers ask them questions.

Conceptual models is what an individual already know. How people perceive the world and what they already know affect the affordances of an object. People are trained to act and think of things in a certian way. On the other hand, Dourish and Button's description of "accounts" is how users interpret a system's behavior every time users interact with a system. Unlike conceptual models, accounts are not permanent and generated everytime the system is used.

Warren He 22:12, 9 October 2011 (PDT)

Ethnographic vs. Contextual Inquiry

Consider an interview where an interviewer asks questions about the product. Contextual inquiry is sold as a departure from this interrogative model, and incorporates observation of actual product use. Ethnographic inquiry is like an extreme version of this, where there is no interference from the observing party at all. They take longer, and the results are less accessible.

Accounts vs. User Conceptual Models

Both of these concepts are related to a user's understanding of a system. A conceptual model is a device that generates predictions and aids a user in coming up with future actions. An account is more like a concrete detail of what is happening in the present or has happened in the past.

Conceptual models are more abstract, and are concerned with modeling behavior. Accounts are more concrete, and are concerned with capturing implementation detail.

Zheng He 22:54, 9 October 2011 (PDT)

• Ethnographic inquiry and contextual inquiry are similar, both aim for people's needs and focus on pratical solutions, but Ethnogphic inqury has a spicific target group, which is a specific ethnic that would be the essential difference. Using differenty inquiry methods would result differently. For example, Chinese people burn paper things for their ansistors, and paper money, sold in US, looks like US dollar that is the result of contextual inquiry and having paper things sold in US is the result of Ethnographic inquiry. Moreover, the time consuming of both inquiry methods varies and is uncomparible since many enthnographic reasearch findings are available either online or in libraries. For an instance, a American company wants to make a Kimchi, some Korean food, flavored instant soup product and they can done the ethnographic inquiry in a day by looking over wikipedia, where tells the company what people expect from the new product, and make a name like "papa kimchi" in korean. This method probably has been impelemnt for years since many instant soups taste the same disregared they are different flavors that may be the result of confusing of inquiries.

• Desicription of accounts and user conceptual models are focusing on different details when design a product. As mention in the file copy exmaple, when a copy action failed, conceptual models would consider "what resources are available to the user to understand what has happened and to understand what options are now available..."are important; accounts, on the other hand, would consider "the way in which people find, within the circumstances of action, the means to find that action rational" are important. As a result, Accounts would "s model provides a mechanism for dynamically relating manipulation of the abstraction to the detail of what is actually happening." unlike conceptual model, which "provides different abstractions for different circumstances."

Jie Min Wong 23:19, 9 October 2011 (PDT)

Like contextual analysis, ethnographic is a methodology to study user behavior to accurately pinpoint specific needs to be considered when implementing a system. The biggest difference is that ethnographic enquiry is usually conducted within a much longer time span. It is also less systematic compared to contextual analysis and is more focused on social interaction in a real-world setting.

Dourish and Button’s account refers to how a user makes sense of a system with respect to different circumstances. Account occurs from time to time when users interact with the system and is not a “one-off business”. The authors sum it up as a “run-time phenomenon, not a design-time one”. User conceptual models are users’ predefined notion of how the system works. It is something more internal to the user and has a looser tie to the system.

Sylvia Lin 23:26, 9 October 2011 (PDT)

Contextual and ethnographic inquiry are both ways of gathering data to discover how people work and what they need. Communication is a necessary part of these processes, and developers must find users to study in order to gain insight as to how the user functions within a certain setting. Contextual inquiry is more established as a tool, while ethnographic inquiry is a more recent practice. The biggest difference between the two is that ethnographic inquiry requires a large amount of time. Ethnographic inquiry can also be used to focus on a more relaxed environment rather than very specific domains. One problem that ethnographers face that doesn't affect hte contextual inquiry process is the difficulties of conducting research outside of extremely controlled, small scale environments.

User conception models are what a user already understands about a system. On the other hand, the accounts described by Dourish and Button are not based on pre-existing ideas, but instead on creating new understandings. These are based on relationships formed between the user and the system. New accounts are created when the user interacts with the system, and are not permanent since they can be constantly generated every time the user uses the system.

Michael Chen 00:59, 10 October 2011 (PDT)

• Contrast ethnographic inquiry with contextual inquiry.
  • Ethnography and contextual inquiry both study the users in the natural contexts but differ in the amount of time they spend studying the subject. Also, the focus of ethnography has varied with purpose as it has been used also for sociological studies.
• Compare Dourish and Button's description of "accounts" with user conceptual models we discussed earlier.
  • Conceptual Models are what the users think the system will do for certain actions. These are preconceived notions that guide the user. Accounts, on the other hand, are what the system produce to help the user understand whats going on for that certain situation.

Heidi Galbraith

Ethnography focuses on the known culture of the subjects being studied. Based on this, designers try to make their product adhere to the expectations of the user. Traditional contextual inquiry involves studying the user in his environment, independent of any cultural knowledge. This does not have any prior knowledge or assumptions; via contextual inquiry, we hope to find out everything we need to know about the user's situation.

Dourish and Buton's accounts are described as individual interactions with an application. A new account is created each time a person uses an application, and these accounts contribute to an understanding of how the application works. The main difference between accounts and conceptual models as we have discussed is that a conceptual is one continuous entity which is changed and adapted over time, based on interactions with the world. Accounts on the other hand are each separate objects, and it is the compilation of many accounts which leads to generalized understanding.

Jian Chen 02:54, 10 October 2011 (PDT)

Ethnographic inquiry focuses on understanding natural settings/events, rather than the artificial or experimental conditions in those of contextual inquiry. Ethnographic inquiry requires more time and energy than contextual inquiry; we need to do conduct possibly many ethnographic studies depending on the debreifing meetings.

A user conceptual model is what the user think about the system (and his/her prediction on dealing with the system); however, an account is a model that describes how exactly the system functions (including every concrete details). It exists with the system (and one system usually consists of a lot of accounts, an account can be generated with every new set of user actions).

Dustin Shean 05:34, 10 October 2011 (PDT)

Ethnography seems a lot like contextual inquiry, the master apprentice approach, since it relies mainly on observation instead of data, and focuses on trying to get the observer to understand the user. However the ethnography approach seems more focused on cultural factors instead of the accomplishment of certain tasks.

This idea seems a lot more ambiguious. In summary an account is a continuous set of actions, where the actions and how the actions are preceived by others matter. I would imagine it would be harder to use this theory for HCI because there is not a specific task that you can observe, but rather is an event.

Phillip Carpenter 06:20, 10 October 2011 (PDT)

• Ethnography requires a lot more time for extended study than contextual inquiry does, while it may yield more concrete results, it is not as efficient as contextual inquiry, and often is over bearing in the time it would take compared a quick inquiry that needs to get done, which contextual can provide. Often times it seems that Ethnography would simply just be too much.

• It seems that the accounts which Dourish and Button describe are generated more from a result of interaction with a system, rather than mental conceptual models that the user develops on his/her own. Typically, a user generates the conceptual models that he/she holds over time form observation, but these "accounts" are generated more from direct interaction than mental observation.

BoYan Li 09:06, 10 October 2011 (PDT)

Ethnographic inquiry and contextual inquiry are with the same aim. They both are seeking the needs of people and getting the information of the current scenarios. The main difference is that Ethnographic would take a long period, like months to years, while contextual inquiry may take days to be done. Also, Ethnographic more concern about anthropology and sociology fields, while contextual inquiry comes from HCI.

Both account and conceptual model are the generalizations of a system. Since an account is designed with the system and formed with the user's single new experience. It's created while the system is used. On the other hand, conceptual mode is formed from user's long-time experience.

Ieong Chon Lo 09:14, 10 October 2011 (PDT)

Ethnography inquiry and contextual inquiry both have the same purpose, which is observing a specific user group for a period of time to study their needs and provide solution. However, ethnography inquiry will observe the users under no interruption. This is not doing pure interview and task-based observation to design what the users’ needs and wants. Also, ethnography inquiry will take longer than contextual inquiry, usually months or years. Therefore, ethnography inquiry is not practical in real industry, which cost a lot of time and money.

User conceptual model is users already know how the system works. Users have predefined knowledge of the system by past experience or observation. However, the account by Dourish and Button talks about the users will generate new understandings to the system every time they use it, so it is more based on direct interaction.

Tai Schuller 10:22, 10 October 2011 (PDT)

While contextual inquiry and ethnography may seem similar at first glance, one can find subtle differrences between the two methods of observation. Contextual inquiry is more task oriented, where the designers are only concerned with one worker at a time as they perform some task to be analized by the designers. Ethnography does look at how workers perform tasks, but in a broader context. The goal is to examine the social context of tasks to get a more accurate representation of how the task is performed in daily life. It takes into account interactions between office workers, and tries to design a system that will work within this social context.

A users conceptual model is a representation or abstraction of the system that is persistent within the user. The model is how a user makes sense of the system, and will bring this model to any other similar system. An account is more specific. According to Dourish and Button, "accounts are a run-time phenomenon", meaning that it is specific to each scenario a user may encounter.

Felix Wong 10:26, 10 October 2011 (PDT)

Both kinds of inquiry have the same purpose, to observe some user(s) and study their needs in some scenario. The main difference is that ethnography is mostly culture based. Ethnography usually has research that lasts many years, while contextual inquiries are generally a one-time occurrence. Ethnography is much more focused on research than contextual inquiries are. It is more detailed than a contextual inquiry. Ethnography is more focused on everything that a subject since the subject is being followed around for much longer than a few hours. Contextual inquiries are usually just a few hours long and is one time. Also, it is specifically for one context, so it is much more focused than ethnography.

A conceptual model is a user's own mental model of how some interface works. Accounts are similar in the sense that it is also a mental model. However, it is a model that occurs only during the usage of the interface. The account could be considered part of the interface, but a conceptual model is just what a user model's inside his or her head.

Phoebe 10:38, 10 October 2011 (PDT)

Both contextual inquiry and ethnography deal with observing people in their natural work environments in order to fully understand all the intricacies that go along with that work. They differ in the time they take and how hands-on the observer is. In contextual inquiry, the researcher might take a session or two to watch their subject work and ask them questions about what they are doing. The goal is to get the desired information in a relatively timely manner. With ethnography, on the other hand, the researcher spends an extended amount of time in the environment in order to understand everything that affects the subject's work. Even a "quick and dirty" session will take days, if not weeks to complete. In the end, the choice between contextual inquiry and ethnography boils down to one of quantity vs. quality.

Dourish and Button make the important point that "accounts" are comparable to run-time errors, as opposed to design errors, meaning that they depend on the specific circumstances at the time of testing rather than relying solely on what happened before then. According to them, a user forms an "account" when they actually interact with a system, and this "account" is subject not only to the design of the system, but also to the context in which they use it. The models we have been learning in class have suggested that people have a more abstract view of a system that does not depend on the specific circumstances of its use.

Konigswagger 10:41, 10 October 2011 (PDT)

Ethnographic inquiry per the article, "seeks to understand settings as they naturally occur, rather than in artifical or experimental conditions." This directly contrasts contextual inquiries, where the researcher works with the user to try to understand the user's work. Additionally results from ethnographc inquiries is readily usable by designers versus results from contextual inquiries which require modification and possibly further studeis. One of the key differences between ethnographic inquiry and contextual inquiry that is worth noting is the time that each takes. Ethnography is considered a "prolonged activity" so under time pressure, it is difficult to do an ethnographic inquiry (though evaluative ethnography attempts to solve this issue).

User conceptual models define a user's understands of a system. Dourish and Button describe the process that generates a understanding of a system through interaction with the system. This contrasts a user conceptual model where a user's understanding comes primarily from observation.

User conceptual model is users already know how the system works. Users have predefined knowledge of the system by past experience or observation. However, the account by Dourish and Button talks about the users will generate new understandings to the system every time they use it, so it is more based on direct interaction.

Stacy Hsueh 10:53, 10 October 2011 (PDT)

Ethnographic inquiry is a more in-depth study and analysis of a target user population than contextual inquiry. Its end goal is to understand the environmental, sociological, psychological factors that lead to certain human behaviors, which can serve as a valuable insight into how future users will interact with the product. Contextual inquiry is not nearly as involved. It finds just enough information about human behaviors to help create a successful product.

User conceptual models are based on preconceived notions of how a system works to different individuals. The idea of "account" described by Dourish and Button, on the other hand, refers to individual instances of personal experience with the system. In other words, once a conceptual model is formed, it does not change. However, an account is tied to the changing context and circumstances of each use, and is therefore uniquely defined by independent experiences.

Deryu0502 11:23, 10 October 2011 (PDT)

• ethnography is a long term research on the details of a person's bricolage background. Comparing ethnography to contextual inquiry is like asking how/why the person thinks that way to what a person thinks

• account is how to achieve that goal and conceptual model is the goal, thus in every instance the account will change while the conceptual model might not

Vincent Chiu 11:13, 10 October 2011 (PDT)

Ethnographic inquiry is generally done over a much longer period of time than contextual inquiry. It is also more of a cultural investigation to look at how people live over that time period, rather than contextual inquiry which specifically looks at how people might use a device. Accounts seem to be similar to conceptual models, but accounts are part of a system and are produced when a user interacts, while a conceptual model can exist outside the system inside a designer's mind as well as a user's mind. The accounts are created when the user actually uses the model rather than being a pre-existing understandings.

Jessica Pan 11:26, 10 October 2011 (PDT)

Ethnographic inquiry and contextual inquiry both look at the context of a problem; however ethnographic inquiry analyze more in-depth based on a target user population. Its goal is to discover and understand the cultural, environmental, sociological and psychological factors that lead to certain cultural human behavior. This is really important because certain solutions that would work in certain areas would not be possible in others.

User conceptual models are based on a how a system is perceived to word and thus do not change after its foramtion. D&B account refers to the personal experience with the system, therefore, this constantly changes.

Aaron Chiu 11:53, 10 October 2011 (PDT)

Ethnographic inquiry is analyzing how certain groups of people will react to the app/device, not unlike an anthropological study. The results of an ehtnographic inquiry may be different for each sub-group of people while a contextual inquiry should conclude the same.

Conceptual models won't change while accounts will.

Chetan Surpur 12:02, 10 October 2011 (PDT)

Ethnographic inquiry is similar to contextual inquiry in that they are both methods used to gain insight on how users think and work. However, they are different in that ethnographic inquiry is much more involved, and focused on gaining on understanding of the culture of a group of users as a whole, and is a much more long-term project than contextual inquiry (on the order of months or years rather than just days).

Dourish and Button's description of "accounts" involves a user's attempt to justify their actions and decisions soon after they make them. On the other hand, conceptual models are the user's mental model of how a system works. Conceptual models are more persistent and constant over time, while accounts are more on-the-fly and generated for individual actions and decisions.

Aaron Eidelson 12:03, 10 October 2011 (PDT)

Ethnographic and contextual inquiry are similar in that they both try to observe the user in their natural habitat. However, ethnographic unquiry looks at the user from a broader, more cultural, view. This method hopes to give some insite into why a user does what he or she does, not just that it occurred.

Conceptual models represent the user's inference of how their actions relate to the system's actions. On the other hand, accounts are a mapping of user actions to system actions. Accounts, in a sense, define the system.

Kim Nguyen 12:25, 10 October 2011 (PDT)

Ethnographic inquiry and contextual inquiry are similar in that both methods observe a group in their natural environments for a period of time to gain a cultural understanding and impact of an idea of interest. They differ because ethnographic inquiry is a much longer and less direct process. It is rooted in anthropology and mostly used in research. Contextual inquiry is a rapid, altered HCI version of ethnographic inquiry and used primarily to improve design.

Dourish and Button's "accounts" are formed by the system, for that particular use, in that particular time. Each use of the system is intended to create a new account. Conceptual models are created by users who form long-term experiences with the system through a person's understanding and experiences. Accounts exists only during runtime, whereas conceptual models always exist in the user's understanding of the system.

Nancy Wang 12:29, 10 October 2011 (PDT)

Like contextual inquiry, ethnographic inquiry involves research and testing on the users to help improve the design of their products. However, ethnographic inquiry takes much longer and involves more detailed studies, which can last for months or even years. It also focuses more acquiring the social organization of the work. Since we don't have that much time to do that much, we could use the "quick and dirty" method mentioned in the reading, which involves shorter amount of fieldwork to get the information.

Accounts are created when a new relationship is formed between the user and the system. In other words, the user conceives a new model of understanding of the system, which if different from conceptual models come from the user’s pre-existing understanding of the system. So the accounts can differ from what the user expected how the system works.

Michael Pack 12:31, 10 October 2011 (PDT)

Both contextual inquiry and ethnographic inquiry can be very helpful in designing new products. While they both provide a similar benefit to product design, the goal of contextual inquiry is only to aid in new product development, whereas the goal of ethnographic inquiry is to understand the relationships, dynamics, and importance between products and people. Ethnographic inquiry is an extremely in-depth study and analysis of the interactions people have with other people, their environment, and objects. This form of study can provide more benefits to product development than contextual inquiry, but at a much greater cost.

Accounts and conceptual models both describe the users experience and interactions with a system. Conceptual models describe pre-exisiting ideas of the system formed by the user. Accounts describe the model presented to the user by the system and learned through interaction. These are more specific to a given scenario. Through each experience the users conceptual model of the system changes.

Peter Tsoi 12:34, 10 October 2011 (PDT)

Ethnographic inquiry is more thorough of an examination of the target user group than contextual inquiry. The contextual inquiry, as the name might suggest, is more concerned with the task and the specific context that the user might be interacting with the product. One might describe contextual inquiry as task-based and over the span of a few minutes, hours or days. Ethnographic inquiry may take much longer, from months to years and examines a much broader social context that is not included in the task-centric contextual inquiry.

Minkai Ong 12:35, 10 October 2011 (PDT)

Ethnographic is a more general way to study individuals in a sociological setting. Compared to contextual inquiry, it is less technical and consumes more time.

Dourish and Button’s description of “accounts” refer to how people understand a behavior. It occurs every time people interact in a system to get a new understanding. User conceptual model is how a user thinks how a system works before using it.

Christopher Chen 12:41, 10 October 2011 (PDT)

Ethnographic focuses more on the social aspects of the studied environment and in general takes a longer period of time than contextual. In addition, it has only been used in controlled environments, and as such its effectiveness outside of those environments are largely unknown.

User conceptual model is referring to the preconception of how a system works that any user has before actually using the system. Dourish and Button's accounts are referring to users interacting with a system and gain a new understanding, one which perhaps later will have an effect on the user conceptual model.

Sorin Kim 12:36, 10 October 2011 (PDT)

Ethnographic and Contextual Inquiry try to do the same basic thing. Researchers study users in their natural environments in order to try to develop an understanding of current habits, problems, and potential solutions. The difference is that ethnographic inquiry has stronger roots in anthropology and pays much more attention to the social qualities and interactions of the subject. Ethnographic inquiry also takes much longer than contextual inquiry. Contextual inquiry can be completed in a short period of time, from a couple days to a couple weeks depending on the needs of the project, while ethnographic inquiry aims for a much more thorough understanding of the subject and can span months or even years.

Dourish and Button describe "accounts" as something that is created when a user interacts with a system. So accounts are kind of like logs of each interaction, used as samples to understand how a system behaves with its users. On the other hand, a user conceptual model is an idea of a system that exists before the user interacts with the system. This determines what the user thinks he knows about the system and how how the user will initial try to interact with the system.

Darin Fontes 12:50, 10 October 2011 (PDT)

While both ethnographic inquiry and contextual inquiry are used to learn about how people understand and interact with the environment, they do so in different ways. Contextual inquiry is more short-term and aims to learn how a certain user base will recieve an idea. Ethnographic inquiry aims to learn how an entire culture or society will receive an idea. The latter takes a much longer time to carry out, but it may yield much more information about cultures and societies as wholes, as opposed to a specific group of users.

A conceptual model is the understanding of how a system functions. It is what a user forms in order to relate the system to something familiar. An account is what the system itself promotes in order to show the user how the system should be used.

Harvey Chang 13:02, 10 October 2011 (PDT)

Ethnographic inquiry is similar to contextual inquiry in that they both attempt to study a users to learn what and how they work in their environment. The difference is that ethnographic inquiry seeks to analyze and understand the reasons why certain people behave the way they do. Ethnographic inquiry can be a long term field study of a target group, trying to see how the group lives and performs in their environment. Contextual inquiry generally is shorter study.

Dourish and Button's "accounts" describes how and why users perform certain actions to reach an end goal. Conceptual models describe how the user thinks something works. Accounts are immediate justifications for actions while conceptual models form from preexisting notions.

Scott Goodfriend 13:03, 10 October 2011 (PDT)

Ethnographic inquiry focuses on the social aspects of an organization, while contextual inquiry usually focuses on a process that might include only an individual or a group. Ethnography originates from studying small, isolated cultures — for example tribal communities — that had defined boundaries and groupings. Therefore, an ethnographer could view a closed-system of social interactions. In these studies, ethnographers could participate in the culture to see its affect on them. Ethnography for interface design deals with a newer field, communication ethnography that studies social interactions within a workplace or group environment. A large focus still remains on the social aspects, including language and nonverbal communication and norms. On the other hand, contextual inquiry focuses on a process the researcher is interested in evaluating. Therefore, a researcher will study how people accomplish the process and probe into their thoughts and reasoning for their actions. Ethnography focuses on "members' points of view" and on their "experience" rather than simply on their actions. Contextual inquiry focuses on the natural actions, and then probes users' points of view to understand why those actions were done.

Dourish and Button's accounts are the model that the device presents to the device's user, while a conceptual model is the user's personal mental model of the device. Therefore, accounts are part of the design of the device, especially the user interface. Designers and developers need to present a correct account, so that users can create a correct conceptual model to understand using the device comfortably and correctly.

David Powell

• Contrast ethnographic inquiry with contextual inquiry.
  • Contextual and Ethnographic inquiry are two ways in which one may acquire data which would offer an insight into ones natural behaviors. The primary contrast between Contextual and Ethnographic inquiry is the duration of the study. Ethnographic Inquiry is rooted in the social sciences (anthropology/sociology) which demonstrate the need for a thorough understanding of the demographic. Ethnographic inquiry seems too in depth to be beneficial to a User Interface designer, whereas, Contextual inquiry seems to fit designers needs better.

• Compare Dourish and Button's description of "accounts" with user conceptual models we discussed earlier.
  • Dourish and Button describe “Accounts” as the means in which a user makes sense of a system based on the users a priori knowledge of the system. With each new encounter the user makes slight modifications to their conceptual model of the system. The authors espouse that user establish relationships with systems. Furthermore, these relationships are temporary; and with new accounts with the system, new relationships are formed.

Mano Pagalavan 13:13, 10 October 2011 (PDT)

Ethnographic inquiry and contextual inquiry are very similar in that they both employ techniques to become familiar with a particular work setting and environment. While they're similar, ethnographic inquiry is a bit different in that it aims to learn how human's are affected by environmental, sociological, and psychological factors in these settings. However, it would usually take months and months to gather this information. By knowing this information, I would be able to design an application, keeping in mind the needs of the people that were observed.

Both accounts and conceptual models describe user experiences of a system, but each does so in a different way. Accounts, as defined by Dourish & Button, are system-generated experiences, ones that are never permanent. This way, the user never knows what to expect, but learns how the system works through interaction. Conceptual Models on the other hand are designed by the user, and presents a circumstance in which the user knows everything that the system can do from the start, and is always anticipating the next interaction.

Sakura Reyes 13:13, 10 October 2011 (PDT)

In comparison to contextual inquiry, which is centered on design, ethnographic inquiry is centered on the ‘social’ aspects of the subject, traditional for purposes of understanding rather than finding design goals. Furthermore, where contextual inquiries tend to return results that are very goal- and task-oriented, giving direct feedback to the designer, ethnographic results have tended to be produced in a lengthier format more suited to academic study or understanding of state rather than to giving the designer targets for their design. Owing to their differences in goals and results style, ethnography is also a much more time-consuming process, requiring several weeks or more of continuous presence from the researcher rather than the shorter sessions used in contextual inquiry. As a corollary of the ‘social’ focus of ethnography, studies may also be less focused, requiring the researcher(s) to observe not just one individual as tasks are performed, but many members whose activities relate to a given area or process of interest.

A primary difference between a conceptual model and an account is the association with the system in time – an account is something generated in advance, before the system is run, and which will likely persist across the user’s time using the system unless evidence appears that contradicts the model. By contrast, an ‘account’ is something that appears and is presented to the user at run time, in context to a particular instance of using one of the system’s functions. Several additional differences are now clear. While a conceptual model is created by the user, an account is something given to the user by the system to use in conceptualizing its function. Furthermore, while a conceptual model is a generalization of a given function that serves for all instances of that function and is implicit, existing only in the user’s mind, an ‘account’ would be a model that is instantiated separately for each use of a function, with information specific to that instance of use fitting into the account’s generalized model. It is also thus explicit, being generated and shown to the user by the system. The primary similarity is that both are models that allow users to achieve some understanding of how a system works and what functions it is performing ‘under the hood’, allowing evaluation and problem-solving during operation and failure conditions. The difference lies in where, when, and by whom the information is generated.

Liu Chao 13:29, 10 October 2011 (PDT)

for ethnography inquiry is a more systematic way of get information from the users and which also take much longer time. Moreover, ethnography inquiry may find some aspects that even the users didn't notice, becase users are used to such behavior and form experts' aspects, they may investigate more about how to improve the exist model. For contextual inquiry, it takes less time, give instance response and less cost.

as Conceptural model, which is the way of people think base on the previous experience. And they will use the model to evaluate the exist system, and try to learn the new system from the old model. While the "account", which is a concept which generate during the learnning process. It will change and enrich during the interaction with the system. The different between these two is how they are formed. One is base on the previous knowledge, another one is now.

Leslie Chang 13:36, 10 October 2011 (PDT)

Ethnographic inquiry is similar to contexual inquiry, but is based on sociology and cultural factors to understand the target users. Ethnographic inquiry also takes a much longer time to complete because understanding culture takes more time.

Conceptual models are the user's preexisting thoughts of how the system works before the user uses it. Accounts refers to when the users interact with the system and their understanding of it is changed each time.

Victor Tjhia 13:41, 10 October 2011 (PDT)

Ethnographic inquiry involves more social, cultural, and psychological aspects of the society and why they behave as they do. It usually takes a longer period of time (weeks to years) and can only be done effectively in a controlled environment; therefore any changes in the environment may affect the results of the inquiry.

Conceptual models are users’ pre-understanding about the system (how it should work, relationship of each elements, etc) while account is formed when a new system (entity relationship) is created. In another words, conceptual models are created by the users and account is formed by the system.

Jonathan Tien 13:42, 10 October 2011 (PDT)

• Both are important approaches, but ethnographic research is definitely broader in scope than a contextual inquiry. I like the contrast drawn that although contextual inquiry is quite a step up from simply interviewing customers, ethnographic research is even deeper in scope, often taking periods of years, and occurs in a natural setting, without any artificial construct set up around it. To me it seems like a purer form of research and ultimately may yield more significant and meaningful results, though it may not be the most practical approach in some cases

• The main difference to me seems primarily that user accounts are very concrete states existing at runtime, whereas user conceptual models are broader, more abstract.... conceptual models, that exist only as design ideas. You might be able to argue that user accounts are actually just realizations in the actual system of user conceptual models.

Kate Feeney 13:49, 10 October 2011 (PDT)

• Ethnographic inquiry and contextual inquiry are very simular. They both involve gathering data by observation and questions. Ethnograpic inquiry is newer and less well tested. It aims to describe and understand the user fully. Contextual inquiry aims more toward understanding the task the user is doing.

• Dourish and Button's description of "accounts" is different than contextual model because accounts can evolve. Contextual model explains what the user already knows. While 'accounts" look more at how peoples knowledge changes every time they do a task.

Evan Kawahara 13:51, 10 October 2011 (PDT)

• Ethnography sets its goals up in a sociological context - that is, ethnography aims to observe activities as social actions within a strictly social domain. Ethnography is observation in a purely social form, there is no artificial constructs in ethnography and there is no experimental conditions to uphold. Through ethnography, an observer can describe the social setting as it is perceived by the those who are inherently active in the social domain. In much the same way, contextual inquiry seeks to observe "users" in a natural setting to gain a better understanding. Therefore, both ethnography and contextual inquiry set out with the purpose of observing "users" in their respective domains to gain understanding of natural flow and activities. Ethnography, however, are much less rigid in its methodology: often ethnography requires a great deal of time to gain comprehensive understanding of the domain, and because its design options are much less clear, it is often hard to convert ethnographic work to design implementation, whereas the entire purpose of contextual inquiry is to investigate the context of an activity in order to influence specific design choices. In addition, it seems that contextual inquiry requires more interaction between the "user" and the researches (the researcher often interjects with questions for clarification or deeper understanding), while ethnography is more conducive to keeping this interaction minimal and there is as little interference as possible.
• Earlier in our discussion of user conception models, we discussed that a user conception model is a model composed of what users already know/understand about a system from observation or inherent knowledge. Dourish and Button, however, describe accounts as models formed after direct interaction and use of a specified system. Dourish and Button's version of accounts are accounts that are generated in each instance of interaction. So when a user interacts with the system he forms an account, but this is just for this instance of usage. It becomes a different matter, as Button and Dourish present their accounts in the light of direct usage and interaction, while user conception models are mentally conceived models based off of the system's traits, observations, and common knowledge, but not necessarily direct interaction.

Catherine Callaghan 13:53, 10 October 2011 (PDT)

Ethnographic inquiry and contextual inquiry are processes at different points along a continuum between product-focused design and people-focused design. Contextual inquiry is somewhere in the middle of this continuum: fieldworkers go out into the place of work and observe the users and the social interaction, but they also ask questions specifically about the product design. Ethnographic inquiry is closer to the people-focused end: the fieldworker mostly just observes the users and focuses on understanding the nature of work and the social environment in which it takes place. Ethnographic inquiry also takes longer than contextual inquiry, often being a continous process spread out over months or even years. A contextual inquiry, in contrast, can be completed in several sessions that will each probably fall within a day. Both processes have merits, but ethnographic inquiry is probably more beneficial for very large projects with a big budget, long timescale, and a complex product that will be used in a social setting.

Dourish and Button make a clear distinction between their "accounts" and the user conceptual/mental models that are commonly referenced in user interface discussions and that we talked about in class. They define accounts as models of system that are presented by the system itself, to the user, as an explanation of what is happening inside the system. Contextual/mental models, on the other hand, are models of a system that the user creates for themselves to explain how a system works. Also, accounts, unlike conceptual/mental models, are created alongside the system.

Omar Rehmane 13:54, 10 October 2011 (PDT)

Both ethnography and contextual inquiry are tools to determine how best to design something. Ethnography uses a sociology base to get a wider and longer-term study. Contextual inquiry is far less resource and time intensive, and therefore a lot more doable, especially in a fast-moving world.

An account is an actual, single experience, while a conceptual model is what a user thinks an account will be like. Reducing the gap between the two will make the design more intuitive, and may draw more users to stay.

Filbert Hansel 13:59, 10 October 2011 (PDT)

While contextual inquiry focuses on learning from certain individuals about the act or mastery of certain skills, ethnographic inquiry focuses more on the social context and “real world” placement of software design within the individuals/groups that may not be taken into account as much in contextual inquiry. Aside from this definition, the respective times that each inquiry takes is different in that ethnographic inquiry may take much longer compared to contextual inquiry, although the “quick and dirty” version may suffice.

Dourish and Button describe accounts as the reflective self-representations that systems offer of their own activity. Accounts happen as a system is being used by users, and provided by the systems. User conceptual model exists beforehand, provided by the users, that is about to affect the system being used.

LilithSchneider 14:00, 10 October 2011 (PDT)

• Contextual and ethnographic inquiries both focus on tasks that potential users already complete. Contextual inquiry is more of an interview method of gathering information. The subject of inquiry is questioned at the same time as the relevant tasks are being executed. Ethnographic inquiry on the other hand involves much more observation over a significantly longer period of time. The subject of the inquiry is observed over months or years doing the tasks in question many many times.
• While conceptual models are approximations of how users behave, the "accounts" discussed by Dourish & Button correspond to real data gathered from actual users. A designer might consider a contextual model while building a prototype, but "accounts" can only be created when the design is tested on real users.

Yongjin Jin 14:01, 10 October 2011 (PDT)

While both contextual inquiry requires the developer to go to the actual work environment, the data gained from it could possibly end up being vastly different. While contextual inquiry is about studying how the user themselves work, ethnography is more of how the social environment itself affects the user. This is the main reason why it takes much longer than contextual inquiry. While contextual inquiry mostly focuses on the user to a great extent, ethnography expands the domain to the social environment as well, creating much more complexity and data.

Accounts as described by Dourish and Button are "explanation of the system's behavior." This is in a way an opposite to conceptual model. Accounts are more like experiences by trial and error that explains why the system's behavior is like this at the moment. On the other hand, conceptual model is, the image of how the system should work based on previous experiences. In a way, conceptual model is like cumulative experience that leads the user to the current model they have, while accounts are more like understanding formed only by current experiences through trial and error.

Jay Chen 14:01, 10 October 2011 (PDT)

Ethnogrpahy and Contextual inquiry both have the purpose of trying to understand the user's thoughts and actions in a natural setting. However, they differ in the way the research is implemented. Ethnography is a fully immersive experience that extends a long period of time where a observer looks at social actions within the natural setting. There is no experiments or artificial constructs within ethnogrpahy. The observer tries not to disturb the natural flow of user. However, in contextual inquiry, we seek directed and specific answers about how a user would interact in regards to a specific activity. There is often questions and interactions between the user and the observer/researcher. Ethnography is more complete, however it can also be impractical.

The conceptual model we discussed earlier is a model that is composed of what the users already know and how to build upon that inherent knowledge. Dourish and Button describes accounts as something that occurs while the user is using the system, and that account is modeled after the user's interactions. This is different from our old model that is based on the traits of the app before the user actually interacts with it as opposed to the Dourish and Button model that involves the user's actual interaction.

Markus Geyer 14:06, 10 October 2011 (PDT)

Ethnographic and contextual inquiries are similar in that they both seek to learn something about their user and apply it to the product. Ethnographic inquiry focuses more heavily on studying the user and their interactions with other workers in an undisturbed field. It also takes a long time to complete and ethnographic study. Contextual inquiry on the other hand focuses more strongly on quick methods to look at the issues the user has and on how to fix those problems.

Accounts and conceptual models are both forms of understanding a system. Conceptual models are a user’s beliefs about how to interact with a certain system that form over a long period of time. Accounts on the other hand focus more heavily on the way actions are organized within the system to make it clear for the user.

Allan Yu 14:08, 10 October 2011 (PDT)

Ethnographic inquiry and contextual inquiry are quite similar. They are both founded on the principle that there is a social aspect to usability that needs to be considered for an interface to be successful in the field. Both divert from simple task analysis in that they are trying to get a better understanding of the social environment by recording and observing the user base. Ethnography is a much more involved process that is more long-term, employed for months and years, while contextual inquiry can be done in a matter of days and weeks.

User conceptual models are models that users have procured from being accustomed for a long time, while accounts are models that are created within a user by the system. These accounts are made by the system every time the user interacts with it.

Ivan Motyashov 14:16, 10 October 2011 (PDT)

Contextual inquiry and ethnographic inquiry both attempt to be non-interventionist ways of studying users and their needs through observation of their activities in context. Contextual inquiry, however, seems much more goal-oriented and aims to provide specific information relation to the design and implementation of a product. Ethnographic inquiry, on the other hand, is more concerned with the social aspect and, as such, provides more abstract, more general, and more elusive results, not always directly applicable to design or engineering. There is also a difference in time frames: contextual inquiry is usually completed in hours; ethnographic inquiry may take months or years.

The conceptual model is something users mainly construct themselves, based on previous experiences -- it is a mental model they develop over time that represents their idea of how an application works (for example). The user plans his use of the system using this conceptual model. An account, on the other hand, is a model that the system presents to account for its activities.

Manduo Dong 14:23, 10 October 2011 (PDT)

Ethnographic inquiry and contextual inquiry are similar in a way that they both study the behavior of the user when interact with products. However ethnographic inquiry no only require a longer time to conduct (although not all ethnography are. For example, "Sanity Check".), but also focus more on the socially organized setting.

Conceptual model is users' understanding of how interface works. The "accounts" of ethnography introduced in the article more focus on users' "rational social behavior." In addition, ethnomethodological accounts are used when there is "greater disconnection" between its work and the design.

Chien Ting Tang 14:26, 10 October 2011 (PDT)

Contextual and ethnographic inquiry are pretty similar. They both emphasize on the importance of social impects on user interface. However, ethnographic is takes more effort to observe and to use on an UI design. A conceptual model is what people predict things will be based on what they already know. Account model is more practically focus on what really happened. And how people react to what happened.

Wei Jiang 14:27, 10 October 2011 (PDT)

Whereas contextual inquiry is normally conducted for a short period of time to understand a user's basic work flow for a certain task, ethnographic inquiry is conducted over longer periods of time to understand cultural, sociological basis for certain decisions or actions. Ethnographic inquiry seeks to understand more extensively about the users' cultural inclinations.

Dourish and Button's "accounts" refers to concrete snippets of a state of the system, as they occur, real time. In contrast, a conceptual model is a metaphor we develop for the entire system.

Katelyn Sills 14:37, 10 October 2011 (PDT)

While both ethnographic inquiry and contextual inquiry both seek information about how the potential user operates, ethnographic inquiry places more importance on being a “fly on the wall,” and not disrupting the normal work environment of the user. Because of this, ethnographic inquiry is generally more time consuming (unless it is the “quick and dirty” type) since you don’t have the luxury of asking pointed questions to prod the user along.

Accounts are the glimpses into the workings of the system that the system provides the user. This is different than the user conceptual models, which may exist regardless of the output from the system. Dourish and Button give the example of a file transfer failing at 40%. The account in this example is the explanation that the system gives for why the file fails, and what state the file is in now. Is 40% of the file on the hard drive? Etc.

Benjamin Hsieh 14:41, 10 October 2011 (PDT)

Ethnographic inquiry is based on research done over a seemingly more thorough and long-standing process involving deep analysis of cultural and sociological issues. On the other hand, contextual inquiry seems to provide a more in-the-moment type of information based on a survey of a certain sample size of target users.

"Accounts" as described by Dourish and Button refer more to a user's persona at any given moment during the execution of the process, whereas the user conceptual model is developed before any actual implementations.

Victor Krimshteyn 14:50, 10 October 2011 (PDT)

Whereas contextual inquiry tries to understand what design goals should be on a surface-level, ethnographic inquiry goes deeper and tries to understand the culture and social aspects where the product will be used more deeply. Therefore, ethnographic inquiry requires a much longer period of time.

Whereas conceptual models stand alone from the product being used, accounts only exist when the user is actually using the product. An 'account,' as opposed to a conceptual model, is how a user understands what's going on in an application directly, rather than how the user thinks of the app abstractly.

Matthew Leung 15:53, 10 October 2011 (PDT)