Task Analysis and Contextual Inquiry
From CS 160 Fall 2010
Slides
Readings
- Principles of Contextual Inquiry. Contextual Design. Chap 3. Beyer & Holtzblatt.
Optional Material (not required):
- Personas:
Discussion Questions
The chapter argues for the "master-apprentice" model during interviews. It also describes other relationships (like interviewer/interviewee) that interviews can fall into. What are potential disadvantages of the master-apprentice model? And conversely, what are potential advantages of other interview models?
Your Response
For this reading, we'd like you to insert your responses direction below by editing this page. Its a good idea to compose your reply with a text editor, and include it quickly. That helps avoids conflicts with other users editing at the same time.
Pp do that, first login by using your user name and password, edit the page and at the end of it put a whole line == ~~~~ == (literally) at the beginning of your submitted critique, so the wiki system will index, sign and date your submission automatically.
Richard Laroue 12:04, 14 September 2010 (PDT)
An apprentice spends too much time learning about how the worker actually does her work instead of learning just enough to support and improve her work. The apprentice needs to be able to serve individuals doing very different tasks in very different places. However, the master/apprentice relationship is still a very good start for a couple of reasons. 1. the worker (master) can talk while she works so that she feels like any other day. 2. sometimes a worker may not remember the details of her work until she is actually doing it. 3. conversation flows naturally; master describes her work as she goes and the apprentice asks questions as they arise. One relationship model that is a little more useful is the interviewer/interviewee model. This way, the interviewer can more easily gather information that is geared specifically toward supporting worker.
Geobio Boo 14:21, 14 September 2010 (PDT)
I really like the master apprentice model. For example, when helping people with homework, it forces me to re-think about what I'm doing, why, and point out things that I find important. Furthermore, the apprentice will notice things that I overlook (eg. because it's too common/natural for me), and I have the chance to think about it and explain what I'm doing. On the other hand, if I'm the apprentice, I can fully understand what they're doing, and I feel like with enough learning, I can actually repeat what they do--and have the subtle clues that go with it (examples from the book include the green stickers, voicemail while the computer boots, notes from the boss, etc). For most learning, the master apprentice model works, but I can see situations where it does not fit the need. For example, suppose that I need to start asking more questions or make vastly differe suggestions. A master/apprentice model just passes the information from one person to another, but does not focus on adding new ideas or focusing on different aspects that would otherwise be overlooked. And as a final statement for the reading: I really enjoyed it, and feel that it'll help me with understanding people and having productive conversations.
Alex Aberle 14:36, 14 September 2010 (PDT)
I think the hardest part about doing a good interview in this style wouldbe throwing away my assumptions. The section on "intrapersonal triggers" has to do with that, and I think as a software engineer it would be hard to avoid those problems. For example, being the "apprentice" to a lower-skilled "master" user, when you wrote the program in the first place. I would have in my own mind an idea of how I thought a task should go, and it would take some real effort to see a user "mis-using" the software and not tell them the "right" way to do it.
It seems to come down to this: the closer you work with the system in question, the less likely you are going to be able to conduct a useful and non-biased interview. It really would be better to interview somebody who uses a system that you've never touched before. That comes closer to a true master/apprentice relationship, and keeps prior assumptions out of the picture.
Frank Chew 15:19, 14 September 2010 (PDT)
Compared to the interview, which takes a day, the apprentice relationship may take a while. The apprecticeship model tilts power too much to the master-customer, whereas the interviewer wants data to feed the invention of a technology system which supports the work. The apprentice only learns one job, whereas the interviewer understands all the jobs, and comes up with a technology solution. The apprentice, by definition, does not design a technology system like the interviewer. The interviewer comes up with not only a technology solution that works but that is accepted by the customer and improves their efficiency.
Don Arboleda 17:14, 14 September 2010 (PDT)
The Master-Apprentice model narrows its view down to seeing what the customer does and how he does it. This can be very useful if the customer has a problem but does not know what it is specifically, as with the index anecdote given in the reading. However, because of the traditional heirarchy associated with the Master and the Apprentice, the inexperienced Master (pardoxical though it might sound) might flaunt his knowledge to the Apprentice and spew big words--they control the path of the conversation because they have the knowledge. At the same time, the Apprentice must make a commitment to the Master and learn the craft--perhaps more of the craft than is ultimately necessary.
Outside of the Master-Apprentice model, the two individuals (or teams) are plaed at equal standing. While the customer is still describing his actions, the interviewer can better lead the customer to fill in the gaps. The customer might need to explain rather than summarize, and the interviewer can help. The interviewer can interrupt the customer about a topic of discussion so the customer and interviewer both can think on the subject more clearly.
Steven Kisely 18:19, 14 September 2010 (PDT)
I find the master-apprentice to be a superb model for learning about the work structure of the customer. It allows us to see how the customer goes about their job. A possible problem with the master-apprentice model can be the customer. What if the customer is not a very good participant or chooses to gloss over questions. A great example is my father. He works on cars as a hobby and has built several. Asking him questions however can be problematic. If he is in the right mood he will explain how stuff works in detail, however when in the wrong mood asking him a question is pointless. He will either ignore you entirely or gloss over it. In general I think the master-apprentice model is best, however when we have a customer acting as above I think perhaps an interview may be best. If they are not willing to go into detail before while working. Perhaps sitting them down and using a questionnaire where you are in charge may be what is needed. We likely will not get the level of detail of a master-apprentice interview but at least we may have a better idea where the problems are.
Terrance Ng 18:23, 14 September 2010 (PDT)
Some problems with the master-apprentice model are as follows: the master-apprentice, as the article noted, could potentially take years, while a design team doesn't have that leisure. Also, as the apprentice concentrates on observing and learning from the master, there is little room for innovation in the workflow, unless the master realizes that something is extraneous or could be made more efficient, much like research scientist on page three of the article.
Since an interviewer is expected to ask questions, he is able to prompt the interviewee to fill in gaps in the explanation. However, as noted in the article, the ultimate goal is to create a sense of familiarity and comfort that allows for a natural relationship.
Benjamin Carpenter 18:53, 14 September 2010 (PDT)
Depending on how many users need to be interviewed, the master-apprentice model may be either a very fast way to learn about the user, or a very slow way if the user base is too broad. The article discusses how to interview the user, and which techniques to avoid, but time can be easily wasted if the wrong interviewees are chosen for this M-A model. On the other hand, it may be beneficial to spend extra time on this 'interviewing' portion of the design process, as it will save time later when you have to formulate a design and already have information from a broad array of your users. In addition, you can use these users to make your personas more realistic and keep track of similarities between different users that you interview.
James Yu 19:29, 14 September 2010 (PDT)
I think the negative to the master-apprentice model is it might allow the customer/master to do things that may not be relevant to the design. So the first formal interview part is necessary to make sure the customer focuses on what the design is looking for. Another disadvantage is usually apprentices study one thing for very long, while a designer needs to study many things in a short amount of time, so they are not exactly the same. The advantages to the interviewer/interviewee model is that it allows you to get direct answers to questions you as an interviewer is interested in. However, the answers may be missing context and be summary like.
Across The Pacific 19:50, 14 September 2010 (PDT)
The contextual inquiry described in today's reading strikes me as very similar to what a user experience researcher does: user studies. They both involve the designer (or the researcher in the latter case) observing and interacting with the user whose work we are interested in improving or making more convenient. In a user study, though, the user does not work in his/her usual work environment, but rather one that is set up by the researcher, and is observed while using the designed system/product. Both forms of interviews seem to be getting more and more attention from big tech companies, since the importance of user experience is becoming more evident with Apple's success.
Jonathan Look 20:55, 14 September 2010 (PDT)
Master Apprentice model disadvantages include, 1) interviewers cannot spend the time to learn the work like an apprentice, and 2) apprentices learn a specific job, but designers may need to be more diversified to address different projects on which they work.
I suppose an advantage for the guest/host model allows the interviewer to observe the customer with minimal interruption on the interviewer's part. This in effect places more weight on the interviewer to observe the customer and make note of many details, but could allow for a discussion at the office about how the customer works.
The expert/novice model can perhaps offer the advantage that allows the designer to interact with a problem that the customer finds difficult. In this way, the designer and customer can work through some problems in a work environment, allowing the designer to make note of specific steps or tasks taken that can later be expanded on or incorporated into a new system.
Sean Tai 21:22, 14 September 2010 (PDT)
One potential disadvantage to the master-apprentice model is that because the interview is interviewee-led, it can be unfocused and rambling, making it impossible for the designer to systematically conduct their research. Also, because the interview is focused on the interviewee’s performing of an activity, the interview may take a long period of time.
On the other hand, the interviewer/interviewee relationship can make for a more concise, focused interview. It may also make both participants feel more comfortable, as the setting is more formal and less invasive to the interviewee’s internal thought process.
Sui Kun Guan 21:26, 14 September 2010 (PDT)
The master/apprentice model has some advantages. For example, first, the apprentice can learn strategies and techniques by watch some instances of the task from the master. Second, when the master is doing the work and explaining the work to the apprentice, the master can also revisit all the work process and will not leave out some messages easily. Third, it can remind the past events for the apprentice. However, there are still few disadvantages for this model. For example, the apprentice needs to learn a lot of techniques from the master, and that sometimes will not get the most efficiency for the apprentice. On the other hand, in interviewer/interviewee model, interviewer can get more efficiency on learning the work because interviewer is going into the context of the interviewee and the interviewer can gather many useful information efficiently.
Vincent Rodriguez 21:48, 14 September 2010 (PDT)
A potential disadvantage of the master-apprentice model is that it might require a significant time investment on the part of the design team to completely understand the "master"'s work. It could also be the case that the chosen "master" turns out to be completely different from the majority of the user base and skews the whole interview. A potential advantage of the expert/novice model is that the design team could bring in new and perhaps better ways for the user to approach the problem. The expert could even potentially uncover problems that only someone with his expertise could have seen.
Chris Song 21:53, 14 September 2010 (PDT)
One of the potential disadvantages of the master-apprentice model is that you cannot tell whether there needs to be a radical change in the way customer goes about her work. Even if there is a set routine the customer follows daily, maybe there is a potential for you to design a product that completely revolutionize the way she does things, yet can be learned rather quickly. By following the master-apprentice model, you try to first observe the customer, then improve or fix the problem. But if there is a problem that comes from the methodology already in use by the customer, then it may not help very much to improve a failing system. In this case, more time should be spent to determine how fast the customer can adapt to new ways.
In the case described above, it may be more productive to try scientist-subject model. You observe the customer. But instead of finding what she is doing and trying to understand her, you can focus on what she’s doing wrong to create the problem in the first place. However, I believe this kind of model should be used secondarily to the master-apprentice model because it’s obviously very difficult to come up with a revolutionary and easy-to-adopt design.
Daniel Yoo 22:04, 14 September 2010 (PDT)
The potential disadvantage of the master-apprentice model is sometimes people do not even know how to do their jobs themselves, and they depend on someone instead. The time that consume for studying differs from master and apprentice. On the other hand, the advantage of the master-apprentice model is the model can effectively collect data from getting the direct answers from the interviewee or may happen from the interviewer. People get more aware of everything when someone is watching over other or taking an action together, so they tend to do better at work. Master-apprentice model is good for learning about the work structure between client and the customer.
Chao liu 22:12, 14 September 2010 (PDT)
One thing about the master-apprentice model is that, the designers are the “masters”, which means, designers may unconsciously import their ideas into the apprentices’ minds and “force” the apprentices think in the same way. But the whole idea of let various people test the app is to find the potentially fault of the application. Of course, when the designers show the application to the apprentices, it is a process of analyzing the app again. And teach the customers how to use the app. Sometimes the customers are not familiar with the app and think it’s a “bad” design, in fact, each app is target for certain group of people, but when we want to make it for more people, other groups of people’s opinions need to be considered. One more thing about the master-apprentice model is. When the apprentice doesn’t understand some function of the app, the master can help him/she to figure it out. And take my personal experience, when I think one app lost me, I will never use it again. The good part of interviewer/interviewee is… it’s more customer-centered. If the customers don’t know how to use the app, it’s a wrong design or at least it’s not a convenience design. The interviewee will be naturally responses to the app and give some questions which the designers hard to think about, such as “why the button is placed here?” or “I can’t see the number in the text box”. Something maybe very small but it really affect the users’ convenience.
Andy Lin 22:18, 14 September 2010 (PDT)
In my opinion, I think that master-apprentice model might be a great model. As the lecture and previous readings mentioned, it is important that after letting the interviewees know what interviewers want to know, interviewers should just stop talking. However, interviewees may not be well prepared for the interview, or they may not able to give a well-organized structure about the things they do every day. As a result, master-apprentice is good for these purposes. However, as the article points out, it takes years for an apprentice to learn from master, and a design team typically has much less time to do the interview. Consequently, I believe this is the major disadvantage of this model.
Derrick Tao 22:31, 14 September 2010 (PDT)
The master-apprentice model is a really good way to conduct interviews because it allows the developer to have to watch how the customer interacts with the product. The interviewer must act like an apprentice and ask why the customer, who is now the master, how and why they choose to make those actions. The possible downfalls include having a customer that does only gives a short response without really trying to work with the product. When this happens, it becomes more of an interviewer/interviewee because the customer might be waiting to be told how to use it or can't figure out how to use it. Some of the advantages of this model is that you can explain the process to the user who might not get it right away. This, however, seems to have more disadvantages than the apprentice master model.
Theron Ji 22:50, 14 September 2010 (PDT)
One foreseeable disadvantage of the master-apprentice model is that the master may feel more self conscious during the process. While the article discusses how this could be a benefit (shedding light on aspects of the process that have always been taken for granted), I can see how this could be a downside in the sense that this is more unnatural to the master. They may not behave normally, and may try to change their process by the very fact that you're watching them. For the interviewer/interviewee model, this can have good results if run correctly. Good interviews usually are two-way conversations, instead of the awkward, one-sided one given in the example. Good questions will often lead to topics that the interviewee would find interesting, so in this case the conversation would continue, and could perhaps lead to insights that wouldn't have been discovered before.
Melissa Lim 23:18, 14 September 2010 (PDT)
Overall, the decision regarding which method to use is contingent on a.) the researcher's goal of the study and b.) what the target group's needs are/how customers are using the product. There is not one universal method that works for all situations. Although the master-apprentice model gives insight to the context of what the user does before, during, and after product use, it also has its disadvantages. First of all, if the researcher fails to establish a comfortable environment for the master to work in, the master may hesitate to give information and feel too pressured or scrutinized to share. Second, having to describe each action in detail slows the user down or may lead the participant to go off in a tangent and momentarily abandon the task at hand. The interview model has advantages when the researcher wishes to determine the schedule of events; for example, a researcher performing task analysis will want to have a controlled study consistent across all users. The interview technique allows for a more controlled approach and is particularly useful when performing a direct comparison of reactions to specific tasks across users.
Christine Lu 23:58, 14 September 2010 (PDT)
One flaw to point out in the master/apprentice model is the closed-minded way of learning. The apprentice learns the tasks that the master performs, and perhaps, with questions, can go a little beyond. However, oftentimes an equally important view is the theory of what the master is doing. I think that going beyond the step-by-step instructions and moving into the realm of "why?," "how come?," and "so what?" will give us even more insight, and an out-of-the-box view of the subject in question. This might then allow for more creativity in design and spark new innovative ideas that wouldn't have been unearthed by watching the master perform a task. In effect, the key idea is to keep probing and asking questions, which is what the rest of the chapter talks about. We want to explore all spheres of the user's brain and determine what causes his actions, why he thinks the way he does, and how this could relate to our design. In this sense, we can flesh out the master/apprentice model to give us more information than if the master were to simply dictate his actions.
Courtney Wang 00:49, 15 September 2010 (PDT)
While the master/apprentice model is a great way to learn about the tasks that the customer performs and the needs that you as the designer must address, there are a few shortcomings to this interview method. This method does not address tasks that may have a high roll-over rate, i.e. a task that requires ease for anyone to pick up immediately, something that anyone at a company must be able to do (making a copy from a copier for example). The master/apprentice model only allows interaction with one person at a time, and it is presumed that this person is an expert (master) at the task. Getting only this one view through the master/apprentice model is not enough to gauge the design required to simplify a task with quick roll-over rates for customers. Some potential advantages of the scientist/subject model is that it leaves you very removed from the customer in some ways. It's important as a designer not to forget the fact that you are designing to simplify a task for all people doing it, not just one. In the master/apprentice model, you might get too focused on how one master does it; with the scientist/subject model you can focus a bit more on the task and the actions.
Karl He 00:51, 15 September 2010 (PDT)
Whether the master-apprentice model is disadvantageous would depend on what you as an interviewer are planning to do. If you are designing something to suit the user's needs as they currently are, great. You will learn a lot about how the user does his work. However, the more you are trying to renovate or improve how the user does things, the more the master-apprentice model prevents you from innovating and rather locks your mindset into what the user usually does.
An advantage of stricter interview models is the ability to drive things where you want them to go. With a master-apprentice model, you may spend a long time watching the user do stuff that ends up irrelevent. With the strict interview model, you can get straight to the point.
Raymond Williams 01:39, 15 September 2010 (PDT)
When your actions are based mostly on observations (as the actions of an apprentice would be), you are susceptible to losing structure and digressing from more than that of an efficient plan would allow.
You won't always have the opportunity to learn (or demonstrate) by doing, so one must be able to give an adequate representation of oneself by other means (as the text describes "...discussing his skill in the abstract"). This is actually quite a useful skill in the real world. I would rather hire the person that can tell me exactly what they can do for my company than the person that shows me a few random skills.
Furthermore, a master's technique is often unique to him/her. In my experience, it's exactly that uniqueness that has set the master ahead of everyone else.
Applying this to a customer/developer relationship - there is no unique master amongst them. Instead, a persona representing a portion is the better model to follow.
The text mentions an expert/novice model in which there is a clear distinction between the roles of either side (as does the interviewer/interviewee). I find that this restricts people from feeling the freedom to react how they would naturally. We must always strive for natural reactions. The guest/host is perhaps the closest thing to a partnership, but it's still too restricted.
A true partnership which likely be the best model to follow. All parties feel more comfortable, and as such, more ideas and solutions will emerge.
Adam Vogt 03:06, 15 September 2010 (PDT)
I can see two problems with looking at a customer as a master craftsman. First, master craftsmen are experts at their jobs they are an elite subset of a larger working group. While they may have extensive knowledge of the job they also may have developed shortcuts that would be inappropriate for those less experienced in their field. The second problem is the while they are experts in their field there is a good chance they may have developed one or two bad habits on the job that would be bad for an apprentice to learn.
One benefit of the Interviewer/interviewee is that it gives you an opportunity to answer specific questions that may not easily present itself during a normal work day. A good design will not only perform during normal day to day operations, but also during those moments when work becomes a little irregular. It may not be possible to observe this, but with a little forethought, and good questions, a designer can try to take these circumstances into account.
One benefit of the Expert/novice relationship is that designs that are necessarily more complex may be difficult for a customer to walk through without help their first time. That doesn’t mean the interface is bad, it could be that it is actually considerably more efficient after a customer has received some amount of training.
Brian Maissy 03:31, 15 September 2010 (PDT)
The article makes a very good case for the master/apprentice interaction model, and addresses most of the potential problems I saw with it. One risk is seeing too much detail, getting lost and not be able to tell what's relevant. However, too much might be better than too little, it's the job of the interviewer to focus appropriately. It is possible that the customer might not be comfortable with the apprentice interaction, and resultantly change their behavior with you around, but even so this model might still be more accurate than others. It is also possible that the interviewer might not understand if what the customer is doing is do technical, but its the interviewer's job to direct the customer's explanation by asking the right questions. Sharing interpretations and broadening focus can mitigate this issue.
The interviewer/interviewee model can be useful if the designer needs aggregate, summarized information, or answers to specific questions. The weakness of this model can be lessened by asking questions such as "What is important to you in the system?", or the benefits of both models can be reaped by using the technique of withdrawal and return. The expert/novice model also has something to contribute. The designer often has more knowledge and insight about the product and the technology than the user, and the customer might be able to give more useful feedback if they are aware of the abilities and limits of the designer and his domain. This is especially true when the customer understands his own workflow and problems very well. Maybe spending a little bit of time in this interaction at the beginning of the contextual interview could help enhance the customer's contribution to design ideas.
I think overall, the master/apprentice model is a good primary strategy, but flexibility is also important, and the designer should be capable of using multiple interaction models and switching between them at will. A good engineer uses his tools effectively.
Jeremy Sasson 03:41, 15 September 2010 (PDT)
The master-apprentice model has a few potential disadvantages. First, it puts too much of an assumption that the 'master' is exactly that- a master at his craft. They could not be as gifted as you assume, and thus the apprentice could learn some incorrect things. It is also possible that an experienced person has gotten so good at things that he takes shortcuts and does not give a complete effort in his task, teaching faulty protocol to the apprentice. I think one of the main potential advantages of other interview models, such as the interviewer/interviewee model is that the interviewer can really extract the information that is needed form the interviewee. Ultimately, though, there must be some hybrid of the many different relationships to attain the most effective interview.
Sara LaVigna 04:36, 15 September 2010 (PDT)
Disatvantages are that the interviewers are not learning about the work in order to to do it but rather in order to get data to feed invention of a system that supports the work. Because interviewers have a different goal than the best use for the model, they have to adjust the master-apprentice model for their purpose, which is to meet the needs of a whole market or department - a much more varying task that the apprentice's single job. During the partnership phase, sometimes the power is shifted a little too much into the hands of the customer (master) because it is they who decide what to do during the interview and what to say, which is too limiting for an interviewer understanding work practice. Another disatvantage is that people might react differently in this situation since they have never had this much attention given to them for their job. Apprentices create close partnerships between master and apprentise and the master becomes invested in making sure the apprentice gets it right - which means some information could be lost or skewed in the process.
Interviewer/Interviewee has the potential atvantages that in order to form good questions, the interviewer has to do a lot of research beforehand and if they can think on the spot, their questioning can allow the questions to evolve into a more wholistic understanding of the problem. Expert/novice has the potential atvantage that if the customer looks to the interviewer as the expert, when they have trouble, they will go to the designer and explain it. Although it's nice to see how they would go about figuring it out, this relationship lends itself well to the good explanations of frustrations or confusions. Guest/host has the potential atvantages that it might make some customers feel more comfortable in the designer's presence, especially when they are cultural considerations and the customer's country dictates that newcomers arereceived as guests, at least at first.
Albert Tseng 04:56, 15 September 2010 (PDT)
This article formalizes some elements of successful designer-customer interviews and inquiries. Some potential disadvantages of the master-apprentice model include 1) the sheer length of time that must be spent to be a dutiful apprentice; 2) the heavy influence of a master's idiosyncracies on an apprentice, which may negatively affect a design; and 3) the preclusion of workflow improvements as made by the designer -- i.e. the designer designs to the current, incumbent mode of workflow.
Other models offer these potential advantages: 1) the interviewer-interviewee model provides a structure where information can be systematically gathered, with point-blank questionaires directly procuring answers and 2) the expert-novice model allows a more organized inquiry that allows the designer to focus on important areas quickly and efficiently.
Bichen Wang 06:39, 15 September 2010 (PDT)
The master-apprentice model seems to overemphasize the work itself. The whole point of such a setup originally was to teach apprentices their future job. The customer should not need to actually know how exactly the product works. The products themselves should be intuitive enough (or learned through practice) to be used by the customers without a techie having to iterate through the steps. Although it can be quite useful for the “master” or designer in certain setups shown in examples in the reading, a lot of the times it just seems like a giant lecture; the customer might not appreciate a relationship like this. In the same sense, a lot of the other more traditional methods seem to have a lot better public appeal. For example, the guest/host relationship sounds extremely polite. Perhaps it is not the most efficient method, but it would keep company-customer relations a lot looser and perhaps more friendly. Also, sometimes it could be good to just stop and ask questions (interviewer/interviewee). He definitely has a point that going to this model will fail, but every once in a while, a quick little questionnaire might make some aspects of what the customer wants more explicit.
Mark Wei 07:21, 15 September 2010 (PDT)
One possible problem with the master/apprentice relationship is that much of what the master (customer) says will be of no use to the apprentice (designer). Like many other people have noted, this relationship counts on the customer to be a "master" at his task, which is not true in many cases. Plus, different customers will have different ways of doing the same task, not because of any specific reason, but because of their own idiosyncracies. Finally, this method depends on the observer/designer to quickly note patterns and interrupt the customer with questions that reveal more detail. But often, the customer will be ploughing through his task quickly, and it would be very difficult to know when to interrupt. Interrupting the customer often will disrupt his work flow and you may not be able to observe what the customer would do if he was alone.
Alan.choi 09:28, 15 September 2010 (PDT)
Potential disadvantages of the master-apprentice model fall with the "masters" themselves. It's hard to generalize an entire customer persona with a few people who may do things differently than most. And since it takes up so much time, you're limited in terms of who you can reach out to. Additionally, I think that it's just really hard for a person to be able to become natural in a situation like this without a lot of practice.
The interviewer-interviewer model has the advantage of the question process. The interviewer can basically ask the questions that they are looking for, but that's also a disadvantage because they control the interview basically, so they kind of have tunnel vision and don't get the full picture that they might with the master-apprentice model.
Danica Shei 09:34, 15 September 2010 (PDT)
One of the downfalls of the master/apprentice relationship is that the focus may be hard to maintain once the customer starts going on a tangent. It may be difficult for the apprentice(interviewer) to take back the focus because he/she may not be aware that the customer has taken the wrong direction or is giving details to the wrong part of their work routine. Thus, when the customer is given too much control and becomes too enthusiastic, the interaction becomes pointless because you need the apprentice bringing attention to what originally needed to be fixed. That is why the interview/interviewee model does not have to be sacrificed - as long as the interviewee does not take away too much control, the interviewee would not be intimidated and gloss over or summarized critical areas of their work routine. Also, if the questions are specific enough, it should prompt the interviewee to describe it at great lengths. However, the apprentice master does have it benefits and i believe the best woudl be a combination of the master/apprentice and the interviewer/interviewee.
Evan Rosky 10:18, 15 September 2010 (PDT)
The master-apprentice is limiting for a designer. While good for a worker to learn how to do the job, the designer wants details of what is done in the process. The master describes what they are doing which allows the interviewer to get a look inside their particular job, but without much inquiry. On the other end of the spectrum, there is the interviewer/interviewee model that puts the interviewer in charge. While the interviewer can get the information they want, that isn't necessarilly what they need (unless they ask the right questions). A mix of both (like partnership) is probably the most effective. Allow the master to work and describe what they are doing, but with the right questions at the right times, the interviewer can likely extract the most useful information. In other words, the right questions might be inspired by some particular behavior or practice of the master that is only brought out if the master is allowed to work in his/her normal routine.
Alexander Wong 10:36, 15 September 2010 (PDT)
I found the master/apprentice model very interesting. In fact, the idea that the relationship can drive discussion is an excellent insight. Over the past year, I have been designing a system to help my sister manage her company's operations. Whenever we sat down to talk about her operations, we would always get bogged down by tangents and abstractions. The reason why it was so difficult to communicate was most likely because we did not have a master/apprentice model. More than that, because her operations where overseas, I never observed her and her staff's actions as they happened. The best I could do was ask for a summary. Next time, I will fly there and observe.
Kyle Gorlick 10:46, 15 September 2010 (PDT)
Some disadvantages of the master-apprentice model are that the customer may not be a master. The customer's actions may have been developed off of a poor current design, and thus rely on poor design choices. It also gives total control to the customer, whereas a partnership model would allow the interviewer to more freely question what provokes the customer's actions.
Karthik Jagadeesh 11:15, 15 September 2010 (PDT)
The Master/Apprentice relationship is an interesting way for interviews to be conducted, because it allows the interviewer to delve into many details of the customer’s life, which may otherwise have been overlooked. The one problem with this sort of discussion is that the interviewer is not able to drive the conversation and so they can’t really find out what they want. Another major problem is that it might be hard for the customer to treat you as an apprentice when they know that you are an expert in the design field.
On the other hand if you use the interviewer/interviewee model then the interviewer can ask the exact questions that they are looking for and get responses to it. Also the expert/novice method kind of fuses both of these two previous methods, which is nice since you can direct the conversation as the customer shows you their daily routine.
Samantha Paras 11:18, 15 September 2010 (PDT)
I think the most interesting part of the entire Contextual Inquiry practice is the fact that the customer/user can't really put into words their needs, and it is your job as the "apprentice" to delve deeper and figure out what is needed. Even when the user is more direct and says "I want this", they really may mean something else. I agree with the authors when they say that you must follow the user around in their environment and observe their tasks instead of just interviewing them. Showing a process instead of explaining it is a lot more helpful. I can see why this will be important when we start conducting research/tests for our applications.
I think one of the problems with the master-apprentice method is that it could over constrain the problem. By seeing/learning the way the master does things, you might subconsciously be adding constraints to the problem instead of looking at the problem with fresh eyes. In an interview, the user might not have the perspective to tell you what they really need. This means it could be helpful to have someone that does not know the problem so well and is free of constraints that might be preventing a better solution.
Avery Gee 11:27, 15 September 2010 (PDT)
I think one of the problems with the master-apprentice method is that it could over constrain the problem. By seeing/learning the way the master does things, you might subconsciously be adding constraints to the problem instead of looking at the problem with fresh eyes. In an interview, the user might not have the perspective to tell you what they really need. This means it could be helpful to have someone that does not know the problem so well and is free of constraints that might be preventing a better solution.
That said, I think that in general the master-apprentice relationship is pretty good. Other than the problem mentioned above, I think that it's only other downside is time. A potential benefit to other methods is that if the problem has a known solution, it is very quick to fix if someone just tells you what they want to happen. You could implement it and reiterate if necessary. Fail quickly.
Arthur Huang 11:33, 15 September 2010 (PDT)
The article highlights many advantages of the master-apprentice model for interviews, but there are still some flaws. To be a true "master" of their field, the designer should be a top expert already, and thus would have most likely developed many shortcuts or unintentionally speed through important details that the apprentice should pay attention to. This model also requires a lot of time, something the design team may not be able to afford.
Despite all that, this is a very strong model because it is fluid and allows the apprentice to quickly learn by seeing how a professional would do the work. The scientist-subject model is also rather productive, because the designer is observing the customer and seeing where the customer is having trouble. This is helpful because those areas are most likely also where the majority of the customer base will have problems with the product.
Seng Heng 11:34, 15 September 2010 (PDT)
I found it surprising that the reading discouraged the interviewer-interviewee model. A question such as, "What do you find most frustrating about this tool," may be a more efficient way to discover design flaws than simply observing a user's convoluted and strange methodology to use a tool. On the other hand, I also found it surprising that at one point people contemplated simply sitting someone down in a conference room and interviewing them, as opposed to watching them work; the danger of hearing simply what the user believes you want to hear would undermine the basis for user-oriented design.
Also, after facts have been gathered, what is the correct interpretation? The chapter suggests that the user/customer always has the proper interpretation of the data, but this seems to preclude anything but small evolutionary steps, since users simply want what they're used to, in most instances.
Robert Connick 11:45, 15 September 2010 (PDT)
The master-apprentice model relies on the fact that the customer is an expert at doing his own job. But although the customer's expertise is the whole point of using the model, it could be too focused for some design problems. The customer is not just an expert in what he does; he is an expert in how he does it. Stay close to the boundaries of the customer's job, and it can provide valuable insight into the specific needs and problems that he faces. But stray too far away and at best you will get an idea of how well he would transition to your completely new system. Consider introducing a secretary to a Dvorak keyboard for the first time. No amount of partnership or work structure analysis will change the fact that the secretary has to learn the new system before using it effectively or reaping any of its benefits. In this situation, a scientist-subject relationship might be more appropriate, because that's really what would be going on. I don't use Dvorak, so maybe it's a bad example. But the idea is that the design that we're exploring is so radically different from how the customer works that he has no expertise at all.
Aaron Loessberg-Zahl 11:52, 15 September 2010 (PDT)
I really like the master/apprentice model of contextual inquiry, but I an see where there could be problems. As the text mentions, you have to be careful with this method that you do not let the "master" customer take complete control of the interview and prevent you from asking questions to clarify the customer's actions. Also, as the interviewer, you have to be careful not to give the customer too much help with using a system, as that will defeat the purpose of determining the effectiveness of a design. On the other hand, the standard interview relationship model can be beneficial for getting preliminary information from the customer that may not otherwise be revealed.
Overall, I feel that the master/apprentice model is an effective one for getting user feedback on a design, and I hope to use it in the process of designing our project.
Richard Nguyen 12:07, 15 September 2010 (PDT)
I've never actually thought about it but I find that I do put myself in the master/apprentice model sometimes when I'm discussion new ideas or projects with my friends. They usually want to get feedback and insight from a fresh pair of eyes to see how effective their idea is and what can be improved on. The problem is when they try to talk about the idea/project because they're already very well acquainted with it so they gloss over important details, such as who's going to be using the product, what the product is geared for, components of the product, etc and instead just throw the over arching idea at me which becomes very difficult to understand and analyze. So I usually try to have them walk me through what they want to do and how it's going to be used, step by step and in this way it becomes more clear what can be improved upon. Definitely the biggest problem here is the abstraction of their ideas rather than the actual user experience/interaction.
Yue Chang Hu 12:37, 15 September 2010 (PDT)
Although I really like the "master-apprentice" model of interviews, however there are still some minor flaws and they are too time consuming, and will be alot of work for the designers because they will have to become an apprentice interviewer for all their interviewee. In addition, it also lacks a little bit of additional creativity. I think that designers should uses the "master-apprentice" model for the benefit that it will guide them to learn the routine of the master's work but they should also once in a while uses a different form of interview techniques so that they can add a little more creativity to their knowledge. For an example, the interview to list all the needs and requirement is not totally bad(although might be a little boring if done too much) because one will get to know all the special task that the master will do and this will definitely strength their design as the process goes.
Asa Zernik 12:38, 15 September 2010 (PDT)
The Master/Apprentice model, as described, seems mostly aimed at overcoming the gap between how people describe their work after the fact, and how they actually do it. When describing a task, people generally skim over details they consider to be "obvious" and focus instead on the broad structure of the task. (In other cases, they will not even be able to describe broad structure after the fact). In doing so, they may skip over sub-tasks that actually take up most of the time involved, but that in the haze of memory do not appear as significant. The Master/Apprentice model both brings up details for the apprentice to watch as they happen and refreshes the master's memory - by bringing up glossed-over details, introducing a more accurate sense of time spent, and allowing the apprentice to redirect the master's attention to important details that are usually passed over on autopilot.
Sung Ma 12:39, 15 September 2010 (PDT)
There are many ways of interviewing and these ways have different advantages and disadvantages depending on situations. The "master-apprentice" model seems really good when a person needs to transfer his or her idea to another who is new to this information. The "master-apprentice" technique will give oneself a thorough knowledge, but it will take up some time and result of it depsends on the "master." In "master-apprentice" model, the master is one in charge; however, in "interviewer/interviewee" model, it is vice versa. Unlike "master-apprentice" model, the goal of "interviewer/interviewee" model is to quickly access the problem or issue. The process should never goes through all the information since it is assumed that there are some basic knowlege. Both models have characteristics that could be advantageous or disadvantageous depending on situation.
Anthony Puccinelli 12:43, 15 September 2010 (PDT)
The article was interesting in how it argued that people, especially customers, can always be pushed into playing a certain role by acting a certain role yourself. I feel like this is probably a valid argument since the customer is a willing participant in your inquiry, and if so, the master/apprentice model seems like a good way to go. One advantage is that the model is based on a time-tested relationship (master/apprentice relationships have been successful for much of man's history!). This model also lets the customer remain comfortable as they don't feel the pressure of being talked-down to and are able to get into the thick of their normal work process instead of abstractly trying to explain it through an interview. However, the master/apprentice model is probably harder to pull off as a designer, since your natural inclination is to show off your current design and be proud of features you've implemented/the technical know-how required to do so. So the biggest problem with the master/apprentice model is the natural pride of designers!
James Butkovic 12:45, 15 September 2010 (PDT)
In beginning contextual inquiry, a master-apprentice situation is a good starting point. Constantly probing for feedback from someone who's analyzing something will yield more rich data than if you asked them for a summary after their analysis. Inquiry should be done where the analysis is taking place. The way that the data is interpreted is critical in influencing the design.
Tiago Bandeira 12:47, 15 September 2010 (PDT)
The master apprentice model is great for learning a profession in it's entirety. However, it suffers from being incredibly time consuming. The interviewer/interviewee model can be better at gathering specific information pertaining to a task. Therefore, it can be much faster in certain instances when the interviewer does not want to know everything.
Robin Liu 12:51, 15 September 2010 (PDT)
One of the major problems with the master-apprentice model is the length of time it may take for the apprentice to learn a significant amount from the customer in order to adequately improve design. The interview may not always be informative, as there are no set criteria beforehand. Another disadvantage is that data from the interview might not easily aggregate for meaningful statistics. Similarly, it is also difficult to get a diverse range of opinions, due to the lengths of the interviews. There are numerous advantages to the master-apprentice model. If the interviewer is astute and thorough, he would be able to better understand the user’s need by observing them and asking questions. This is a marked improvement over traditional interviews, as customers can often take for granted their everyday environments and contexts, instead focusing on the parts of their tasks that are the most salient. The interviewer-interviewee model’s main advantage is that it is fast and easily scaled. An interviewer can determine what information he wants beforehand, and conduct all the interviews based on sets of similar questions. This allows the interviewer to collect a broad range of data.
Simran Chaudhry 13:09, 15 September 2010 (PDT)
Master-apprentice. Major advantages of Master-Apprentice: see the user in their environment, document natural interactions. Major disadvantages of Master-Apprentice: takes time to document, could be that watching how things are currently done introduces a bias for the status quo, rather than introducing new ideas. Other relationship models may not have these disadvantages: e.g. Expert/novice model, which is similar to what Apple used. Rather than asking the customer what they want or how they currently interact with tech, they introduced their own design -- creating a completely novel product, without the time consuming effort of learning from/watching with everyday customers.
