Conceptual Models I

From Cs160-sp08

Jump to: navigation, search

Lecture on Jan 31, 2008

Slides

Lecture Video: Windows Media Stream Downloadable .zip

Contents

Readings

Brian Taylor - Jan 29, 2008 03:01:14 pm

Wow, apparently I did not know how a thermostat actually controls one's home heater or air conditioner. Just the other day, my roommate told me, "Turn the thermostat up all the way. It'll heat up faster. Afterwards, we we'll lower the setting to around 70 degrees". I dutifully followed along, fully believing that heating our apartment worked according to that design. It is really funny how large an effect the user interface can have on a person's understanding of how a device should really work. In the past, I never thought about how a person develops a conceptual model of how something he or she uses works. I would have simply placed buttons on the interface, and let a person choose options from that, and attempted to hide as much as possible about the inner workings from the user. But it seems that a vague high-level concept (conveyed through the UI) of how the device actually performs its task can actually ensure that a user will properly use the device. Ultimately, I also really enjoyed Norman's discussion of the many items that often come with terrible interfaces, such as his refridgerator, office phones, and even doors.

And from now on, whenever I push on a 'pull' door, I'm going to blame the design.

Chris Myers - Jan 29, 2008 06:15:08 pm

In response to Brian Taylor, that is a common misconception with HVAC thermostats. Somehow people think the heat output is variable. It would be a lot more energy efficient if it was designed that way, but the controls for variable heating/cooling are too expensive. I think another reason might be that when you are really cold, 90F actually seems like it would be comfortable. I've noticed that the misconception doesn't apply to ovens, where the same technical concept applies. Although in that case it is actually a good idea to over crank your oven when preheating, since a lot of heat will escape when you open the door.

I've met a few different kinds of people. There are those that only use a tool as it was marketed to be used. There are others that use tools in ways they shouldn't be used in, and fail miserably or break them. And then there are people who use a certain tool in a new way that works well, but not intended by the design. I had a friend who used Excel to build really cool game maps. It may not have been the best choice of program, but the large grid afforded him the ability color large regions neatly.

Gary Miguel - Jan 29, 2008 07:21:50 pm

The discussion in the reading about conceptual models was extremely interesting, especially when you think of devices in terms of their abstractions. In the refrigerator example, the conceptual model indicated by the device was misleading and incomplete. This was because the controls were set up in such a way that hid the details of the functioning of the refrigerator from the user. On the face of it, hiding how a machine actually works from the user is not a bad thing. Much of human technology works only because of layers of abstraction, each layer hiding something about its workings from the person interfacing with it. The problem with the refrigerator was that it not only hid the details of its operation, but also provided no simple way to set up the machine. If the refrigerator only had one thermostat, but it somehow provided an interface that gave the user the illusion of it having two, then there would be no problem.

So the challenge is in giving the user the cleanest, most logical interface possible, even if this interface does not match up completely with what is actually going on.

I found the examples with the telephones very instructive. Designers need to remember that just because something is simple (eg, press R then *66) does not mean it is easy for people to learn, figure out, or remember.

Khoa Phung - Jan 29, 2008 09:48:33 pm

This article is awesome. I was never really aware of the paradox of technology and design simplicity. However, reading this article made me conscious that I actually do not actually know how to use a lot of features from my phone at work. It has a lot of functions as mentioned in the text and a lot more buttons, but I do not dare to mess with them when I am on the phone because of the possibility of ending the call by accident. Being aware of this issue will help me to design interfaces with a little more caution and not to add too many features just to make a product better. I also find the natural design for doors fascinating. It is such a simple design for such a simple object, yet there are still a lot of doors that cause me to get confused. I have not put much attention and value to a good design, but this article makes it very obvious.

Benjamin Lau - Jan 29, 2008 05:59:13 pm

In POET, Donald Norman points out that although it's true that human error plays an important role in daily problems, much of the time poor design plays an equal if not greater role. By reworking and refining the design, we can significantly decrease problems with everything from doors to nuclear power plant controls. This is a much more productive approach than blaming "human error"-- an unavoidable element in real life. I liked Norman's focus on case studies to illuminate his principles regarding visibility, cues, feedback, constraints, mappings, and affordances. In one of his case studies, there is a point where he runs into some people who apparently didn't read the instructions for a telephone. At first, most people would be inclined to blame these users, but I think Norman hit it right on the nail-- the user shouldn't have to read a manual to understand the operation, at least not for simple objects we encounter in our daily lives.

I also liked Norman's discussion of "conceptual models" which I first learned about in depth in CogSci 100 and 101 and it's nice to find some practical applications. I can see how affordances are critical for the proper simulation of a device in the user's mind, and how this is necessary for the user to correctly operate it. Without obvious, visible affordances, the user has few constraints to act upon and cannot establish a causal relationship between his actions and the results. So to me, this is an argument that interface design, besides taking into account representative tasks, should also sketch out a conceptual model (like what Norman did on P.15) and verify that the test users get this model out of the affordances of the design. A good start would be to use physical analogies and cultural standards to create an initial mockup.

Michael So - Jan 30, 2008 11:23:50 am

I liked how the author talked about real-life everyday objects like the door, the telephone, the fridge, etc. I agree with the points he said that make a good design (i.e. visibility, feedback, etc.) Everyday objects should be simple and intuitive to use. Having one button only do one function is an example of simplicity versus having them have multiple functions such as the projector or telephone the author talked about. Visibility is a good thing because it helps make learning new devices easier. Like that door the author talked about; if you see where the hinges are, you would know whether to push or pull.

It's also interesting how the author talked about how we should put the blame on the bad design of some object rather than the human using that object. If objects were like that phone the author described, where to do something was pushing some arbitrary buttons and where there is no visible feedback, then I agree the object's bad design deserves blame when something goes wrong. A good design would prevent those errors because the person would actually be able to know how to operate it easily.

Michelle Au - Jan 30, 2008 12:15:02 pm

Norman's comparison between the telephone and the automobile was very interesting. Before, I had always gone along with the idea that smaller sets of functions and controls would be easier to learn. However, I see now that this thought was partly due to my assumption that the mappings between features and controls would remain around a 1:1 ratio, which is what Norman emphasized as good design. Norman also points out that in a good design the mappings between functions and controls are visible. Car controls are displayed with images, labels and colors to visually illustrate what function the control provides. Many telephone systems, on the other hand, do not have such illustrations and instead require the user to memorize different sequences of numbers associated with the function. At first glance, it is almost impossible for a user to figure out how to put a phone on hold without looking at the manual. However, a user can guess how to turn on the heater in a car by moving the control to the red area.

These two points: single control to single function mappings and visible, natural mappings are the most important elements to good design. Take for example the cockpit of an airplane. There is an unbelievable number of controls in the cockpit, each mapped to a function. However, the mappings between function and control are not very visible, and requires extensive training and certification in order to use them. In this case, both single and natural mappings are required to have a good design.

Bruno Mehech - Jan 30, 2008 12:27:03 pm

One of Norman's most interesting points is the importance of feed back to the user. I think that though the phone systems he describes are clearly badly designed they could eventually be learned if people put some effort into it. Just like cashiers at grocery stores eventually memorize the most common produce codes, a user of the telephone system could eventually memorize the codes for thee more commonly used functions. The problem though is that without any feed back even if the user knew all the codes for all the functions they wanted to use, it would still be very hard to use since the user would have no idea of whether or not he did the thing he actually wanted to do, which would also make it really frustrating to try to learn how to use it in the first place so that the codes could be memorized.

Henry Su - Jan 30, 2008 02:10:46 pm

I really enjoyed Norman's discussion on the faulty designs of many everyday things. The land-line office telephone described seemed particularly awful. I for one hate memorizing arbitrary numbers. Norman makes a good point that even if the telephone were cluttered with numerous other buttons, if each button served one function, the design would be much more intuitive. If the designer must minimize the number of buttons (for cost reasons?), one way would be to make the "codes" more intuitive. For example, to check the minutes on my cellphone, I dial #646#, which at first seems random, but actually is #MIN#, according to the letters beneath. Not that this is the best design, but it's better than something arbitrary.

Norman makes a point that even though products today offer many more features than in the past, their operation complexity also increased to the point that many people don't know how to use most of the features. This problem is often caused by designers over-zealous about competing on the technological side, but failing to grasp the fact that just because he or she is able to operate the product, doesn't mean general users can figure it out. This is why usability studies need to be done.

Eric Cheung - Jan 30, 2008 02:43:31 pm

One thing I was curious about as I read the chapter was at what point do you not blame the design and blame the user instead. There are probably people out there that are confounded by the simplest of designs. Most of the examples Norman mentions are kind of extreme as examples of bad design. I found his points about how people keep buying products with a lot of features even though they don't know how to use them rather interesting. It does seem to be quite true that a lot of people buy things more for their feature lists than whether or not they'll actually put in the time to learn how to use all those features. This is especially true for watches; the majority of people seem to only use them to tell the time mainly because it's so hard to figure out how to use the buttons on the sides.

Benjamin Sussman - Jan 30, 2008 03:48:43 pm

This summer my Grandfather got a new Cell Phone (despite my clear logic as to why it was a bad idea) and requested my help to teach him how to use it. In POET, Norman often mentions the word Natural however this idea does not remain consistent over the generations, and thus reveals its slippery nature. All of the ideas which are natural to me where unbelievable to my grandfather. Trivial things like "return to the home/start screen" were concepts that were outrageously difficult for him to swallow. He would ask questions like "Now how do I just make a call now?!" after I had just explained what the red button had done. I think these massive discrepancies in what is considered natural are what make design so difficult, especially on large scale products which affect users with a range of skills and experiences. POET mentions many ridiculously faulty designs but it doesn't get at some of the more complex aspects of the issue where different individuals disagree (sometimes greatly) on what is visible or what mapping is appropriate.

Bo Niu - Jan 30, 2008 06:01:06 pm

In The Psychopathology of Everyday Things, Norman has illustrated on both good and bad designs with the examples of poorly designed doors, telephones etc and well designed auto car chair controller. The idea of affordance was mentioned over and over again within this chapter and from the examples listed in the book, it seems that the only difference between a good design and a bad one is the difference in whether having affordance that relates the interface with the functions of the item.

Katy Tsai - Jan 30, 2008 06:23:01 pm

I found it interesting how Norman pointed out the importance of design by highlighting the faults in everyday items that we all take for granted. The were objects that we all know and use daily, objects that we feel are pretty intuitive, but the process to get there seems far from simple. His idea about feedback also made me realize how important it was to me how objects respond and interact with the user (i.e. how easy it is to text with my cell phone, how the keys feel when I type on my laptop, etc.)

I was also surprised by Norman's ideas regarding mapping. I always favored simplicity in a product's interface, but Norman's idea that as the number of functions exceed the number of controls, a product becomes more and more complex. This brings up the challenge of balancing function and control, and realizing when to cut down on functions that people probably won't use.

Gerard Sunga - Jan 30, 2008 06:36:18 pm

Norman's constant emphasis on the importance of a clean, simple design was fairly interesting. This idea seems to follow the current trend of technology towards simplicity, especially with the advent of the iPod and the Wii as well as the rising popularity of Ubuntu and the Mac platform as a Windows alternative. The problems of a lack of simplicity are evident in his many stories of poor interfaces,especially his friend being unable to open a door or the struggle to handle the projector. However, as others have mentioned, what is the proper balance for functionality and simplicity?

Another interesting part of the reading I found was the difficulty in drifting too far from a "successful" design. I feel this makes it difficult to innovate (especially via a complete overhaul) and results in repeated designs with little improvement (especially the iPod line prior to the release of the iPhone and its non-phone equivalent as well as the perennial sports game titles).

David Jacobs - Jan 30, 2008 07:09:29 pm

Although I agree for the most part with Norman's assertion that interfaces with 1:1 control to function mappings are better, I have to say that it doesn't seem quite so practical for all devices. In large devices like cars and buses, there is plenty of room for hundreds of controls, each clearly marked with its corresponding function. In mobile devices, like a digital watch, however, there is simply no way to include additional functionality with 1:1 controls without creating a device bristling with buttons and displays. Consider my digital watch, the Casio Telememo 30 (Here is a picture of it).

As far as digital watches go, its pretty well equipped. It's got two alarms, a super nifty phone number memory thing; it even has a four function calculator. Enabling these interactions requires tons of buttons. However, in 99% of my interactions with the thing, I use maybe 3 of them -- the light, a mode button, and the start/stop button for timing things. For the most part anytime I need a calculator or phone number, I turn to my cell phone, since the interaction is easier, despite the specialized buttons. This begs the question, *why include so many functions?* Outside of the geek factor, I could get by just fine with a simpler watch. What I'm trying to get at is even though overloading buttons can be confusing, it has to be balanced against how often a particular feature is used. Do I really save more time setting my watch time twice a year with a numeric keypad vs. being able to read it more quickly all year (with a potentially larger display)?

Adam Singer - Jan 30, 2008 08:51:10 pm

I've always thought creating a website or writing a book that could help non-technical people understand how to operate technical things like computers and televisions. I think I might be on to something. When reading Norman's encounters with poorly designed phones, projectors, and VCR's, all I could think of was my parents. I get at least 2 to 3 calls per day with "tech support" questions from them. So not only do I have to walk them through some technical problem, but I have to do it over the phone. This adds to their frustrations and creates some of my own.

Norman's work really opened my eyes to the possibility that the frustrations my parents and I have been experiencing all these years isn't (totally) because of technical ineptitude. Rather, flaws in the designs of technically oriented products themselves can often be to blame. With that in mind, I also agree that increased complexity usually results in more (or more complex) controls, but even the controls of relatively simple devices, such as toasters and microwaves, can be unnecessarily confusing. It is this tradeoff - between complexity of features and simplicity of design - that caused me to take a class in user interface design.

Hopefully, after the semester is over, the world will become a little more obvious with my help and the help of my peers.

Yunfei Zong - Jan 30, 2008 09:11:14 pm

The author's use of witty humor made this a very interesting read. The presentation of comical failures in modern electronics to reinforce the correct way of thinking was ingenious. However, I think he failed to see the obvious solution in the two temperature control knob dilemma. All he had to do was turn one knob to max and the other to the min, record the internal temps, and then switch the 2 controls [turn the max to min and the min to max]. If he turned the cooling knob to max, and the valve knob to min, then either the freezer or fresh food side would be rediculously cold. If he turned the cooling knob to min, and the valve knob to max, then both sides would be warm.

I also seemed to think that his way of explaining well-designed objects to be a bit biased. For example, he explained that the scissors were well-designed because they were simple to use and the 2 holes were obviously just large enough for your fingers. True, but he only knows this because he grew up using scissors. Someone without a lifetime's worth of experience using scissors might think differently. If everyone grew up with the phone that had the R button, I'm sure everyone would know how to use that too.

Alex Choy - Jan 30, 2008 10:03:13 pm

I thought that Norman's focus on everyday things was really good. After looking around at things in my apartment, I realized that I do not know how to use every feature of some of my belongings, such as my cell phone or microwave. Something that stuck out to me was the fact that the "R" button was left on the British Telecom Telephone. The "R" button was placed apart from most other keys, making it seem like a special/important button (like something to be used only in emergency). Despite having no practical use, engineers left it there because it existed in an older version. Even when asked about what it does, they will only remove it if "nobody can think of an example" of something the button can do. While leaving implementations in the current design from previous versions can be a good thing to do in certain situations, it is not a good feature here.

I am probably one person that always turn air conditioner or heater higher or lower than the temperature I really wanted first then tune it down to make things go "faster". What I thought is heater or conditioner will go slower when things getting closer to the "goal" point, so I wanted them go straight pass that point with full speed. Anyway, enough of heaters. The idea of "natural" design seemed really interesting to me. However, I think its implication is rather limited. Not only each people could have different standard of "natural", some of the design is really hard to have a "natural" definition. For example, what is a "natural" thing to do if you want to play a music? Push a button? Tune up volume? insert a CD? For something to be natural to every user is actually quite hard. And what is a music record? Someone may have CD in mind, some others may have tape in mind and people like us probably have a digital file in mind. Natural is quite different depends on everyone's life experience. Even back and forward, we almost always assume left is back and right is forward because how our language goes from left to right, but many other language actually goes into opposite way. Specify the user group seems much more important to me.

Yang Wang - Jan 30, 2008 10:10:46 pm

I am probably one person that always turn air conditioner or heater higher or lower than the temperature I really wanted first then tune it down to make things go "faster". What I thought is heater or conditioner will go slower when things getting closer to the "goal" point, so I wanted them go straight pass that point with full speed. Anyway, enough of heaters. The idea of "natural" design seemed really interesting to me. However, I think its implication is rather limited. Not only each people could have different standard of "natural", some of the design is really hard to have a "natural" definition. For example, what is a "natural" thing to do if you want to play a music? Push a button? Tune up volume? insert a CD? For something to be natural to every user is actually quite hard. And what is a music record? Someone may have CD in mind, some others may have tape in mind and people like us probably have a digital file in mind. Natural is quite different depends on everyone's life experience. Even back and forward, we almost always assume left is back and right is forward because how our language goes from left to right, but many other language actually goes into opposite way. Specify the user group seems much more important to me.

Hsiu-Fan Wang - Jan 30, 2008 10:07:41 pm

One of the most interesting things that the author brings up is the initial point that "human failure" is very often a failure of design. As a software developer, I often want to blame the user for not understanding interfaces, or those that complain because they did not follow some instruction. After reading the chapter I have to admit that many of these so-called "obvious" GUIs were perhaps not so. An embarrassing example from the user side is that I once accidentally overwrote all of the data on my hard drive when I used a command line filesystem cloning tool that took the destination argument before the source, which is the reverse of most other linux utilities. While this was "human error" because I could have looked at the man page more carefully, the behavior was certainly catastrophic and unexpected.

A friend asked me about the value of taking CS164, as vanishingly few engineers will ever write a compiler, and I responded that its important to understand the underlying concepts, even if you don't need to implement one. I think that this is a good illustration of how conceptual models help people accomplish tasks and avoid errors. As your conceptual model is refined by learning about underlying principles, it becomes easier to make things work.

Richard Lo - Jan 30, 2008 10:25:12 pm

As designers think about the natural function of controls, it would seem most people would understand the natural orientation of the control or would be able to get a strong grasp of the general consensus of the natural orientation fairly easily. It is the idea of considering the natural orientation of the control that the designer must keep in mind.

The mapping was intriguing as the only problem would arise if you know that the device is capable of such functions but just do not know what the mapping of the controls is to operate that function. Often times if there is a complicated or complex mapping of few controls to many functions, I would guess that majority of the users are not even aware of many of the functions. This is fine, because the simple functions of the controls should be the basic and only necessary functions needed for the device, everything else is usually extraneous. Take a telephone for example. The simple functions are clear, dial the number and it will call. It is the more difficult functions, which are also extraneous, that require difficult and complex mappings, functions that most people probably aren't aware of and don't care for. So in a way, the complex mapping of few controls to many functions almost acts like blinders to the ignorant; it abstracts away the complex, extraneous functions of the device and preserves the simple, necessary functions.

Ilya Landa - Jan 30, 2008 10:57:07 pm

Reading this text made me recall many similar instances from my personal experience. About 2 years ago I worked as the Tech Support in the library, and on my first day the library head technician gave me a tour. When we got to the conference room, he pointed to a contraption sitting on the table and said: “And this is what we call a “Star Trek Phone”; no one knows how to operate it fully”. This “phone” had a central unit located in the middle of the table – a triangular base with buttons and speakers, but no handset. The handsets were located on smaller terminals throughout the table. At the time, I thought it was natural for an everyday device to have “bonus” functions – capabilities that can only be discovered over an evening with the manual. But now, after reading from Norman’s book, I’m thinking: why is that? Why should someone by a device without knowing all of its functionality. Or, more specifically, why should someone buy an everyday device without knowing completely what it does. When working with specialized equipment, operators need to learn all of its functions. In the ideal case, the same should apply to telephones or microwaves. Unfortunately, many designers believe that a person reads the entire 50-page manual before even plugging in a new coffee maker. Consumers, on the other hand, get easily intimidated by something that looks “like Hollywood’s idea of a spaceship control room.” So, in case of a new phone, users simply try to get to a point when they know how to dial a number correctly. After this lever of knowledge is reached, the manual goes into deep storage in a cabinet somewhere, and games like “What do you mean I called you, you called me!” ensue.

P.S. This was the most fun 23-pages reading assignment I’ve received lately.

Glen Wong - Jan 30, 2008 11:25:02 pm

I found this reading to be of particular interest. Something that really stuck out to me was how ridiculously bad some of the items he mentions are designed. It makes you wonder how such a product even makes it to market in the first place. I found the points he made to be valid in my own experience. His point that controls should imitate the physical effect of actuating the button especially resonated with me. However, in some of his examples I found it hard to imagine how the designer could effectively improve the interface. For example, with the car fader control. I agree that a left/right knob doesn't do a good job of conveying which way the sound will be affected. If the designer were to change it to a up/down knob, it would make operation more intuitive for the fader control, but since the knob is a shared control this could affect how intuitive the other operations would be. The alternative would be to add another control just for the fader, which would add unnecessary clutter to the design for a feature that is really not accessed that often. I think a point that the author did not emphasize, but that I feel is important, is making commonly accessed items prominent and easily accessible. It is inevitable that features will surpass the number of controls for certain devices and the best solution in this case is to make the base case uses quick and painless and make rarer uses as painless as possible.

Mike Ross - Jan 30, 2008 11:34:36 pm

I'm ashamed to admit the number of times I developed input for a program based soley upon which function I needed first. In my defense, these were generally toy programs intended for class or myself, but regardless of intent, this always resulted in a huge map of keyboard bindings or a list of buttons, and it left me with a result similar to the phone examples. Maybe the features were developed and implemented over time, maybe their arrangement made sense for the order in which they were developed, or maybe they make perfect sense if you memorize the entire feature set. Whichever, it's clearly the wrong way to design an interface, and it feels like some poor engineer was given a 12 digit pad first and told implement a list of features, when they should have taken the feature list and built the interface around that.

I was also slightly depressed to read about innovations crashing and burning on the market, just because they're the first or second model, thus preventing actual working products from seeing the light of day. It's depressing to see all the barriers to innovation, it almost makes early adoption sound like a noble cause.

Harendra Guturu - Jan 30, 2008 11:26:48 pm

In POET, I agree that there needs to be a nice midpoint between enough visibility and too much visibility for a object to be visible. But how do find this midpoint? The experiences of many users varies and some may find the "correct" amount of visibility to be overwhelming and others may about the same "correct" amount to be too little and restrictive.

I also think fumbling with the occasional door may not necessarily be the result of faulty design, but due to the fact that we are being flooded with may different types of designs simultaneously, thus we may be used to one type of door and simple ignore the signs present for a different design since we figure we know how to work a door already. It is mentioned in the article itself that there are around 20000 everyday objects and that we don't actually learn how to use these objects due to the sheer amount of time required. Therefore, the information that we used to infer an object's usage may be significantly biased by prior experience such that no matter how clear the design clues are we ignore them.

Jonathan Chow - Jan 31, 2008 12:01:43 am

Two thoughts came to mind when reading POEM. The first is Norman's comment that it is seldom the user's fault when errors occur. He reasons that the cause of this is because poor designs are made that are not intuitive or are too complex for users to understand. This is similar to the comments made in the reading for last lecture in the Jones and Marsden paper. Many poor designs are the result of geeks thinking about the technology, and not necessarily the end-user.

The second thought was in regards to touch screen devices. I think that touch screen devices (the iPhone comes to mind) can make very complex actions simple, while adding the intuitive touching dimension. Since the screen can change to display different buttons, touch screen devices do not have the same problem as other things that might require one button to be used for multiple purposes. Touch screens a bit different that using a mouse too since you are actually using your fingers to press buttons, thus being more intuitive. Hopefully we will see more technology advance in this direction.

Reid Hironaga - Jan 31, 2008 12:29:24 am

Norman's POEM presents an interesting perspective on the importance of visibility and feedback in the design of interfaces. I strongly agree with his ideas that instruction manuals shouldn't be necessary for normal functionality of tools. However, he gives some weak arguments, suggesting the consumers have more power in their array of choices than they really do. "If people keep buying poorly designed products, manufacturers and designers will think they are doing the right thing and continue as usual." This statement assumes that purchases are based on function and usability, whereas in the real world people are often peer pressured by cultural trends to buy horribly designed products. In another extreme case, there can be an entire set of tools that are equally poorly designed; to not purchase one would harm the consumer more than the benefit of 'teaching the manufacturer a lesson'. Norman also states that "When simple things need pictures, labels, or instructions, the design has failed." I believe this is an overstatement. While with experiences, our intuition to meanings of symbols grows, from the standardized on/off computer power switches to red/green phones on cell phone control pads, the utter failure of intuition has no threshold. Regarding the modern age, when people will seemingly go out of their way to be hurt or inconvenienced just to obtain a lawsuit, it would be hard to argue that someone's intuition failed, whereas symbols and signs always provide some solid fragment of information.

Edward Chen - Jan 31, 2008 01:24:13 am

While reading POET, I couldn't help but laugh at the all the poor examples of design, thinking that nothing around me could be as ridiculous as that. However, upon closer examination of my surroundings and rethinking some of my annoyances over design, I realize that poor design is a rather easy thing to make. Like the reading said, a lot of designers have a certain model in their head that they apply to the design, but the user may not necessarily have the same model in the head when they try to operate with the design.

One thing that I found I could relate in the reading was the control for fade in/out on the car. I remember that many times when I want to fade to the front or back speakers in the car, I would first turn the knob to see what the effect was and then adjust accordingly. I almost did this everytime when I had to adjust the fade in/out because it's not exactly something I'd remember and the control is definitely not intuitive.

Johnny Tran - Jan 31, 2008 01:25:04 am

I could not agree more with the idea that technology is a trap in introducing more and more features into products. While it would seem that more advanced technology could go a long way in simplifying products (at the very least, by making them smaller or more portable), the trend seems to be for people to buy the most "advanced" products with the most features crammed into them.

The discussion about intuitive interfaces such as holes was very fascinating. While a seemingly simple design choice, they are highly effective in cluing users into sticking their fingers in the right holes. I wonder whether it is possible to find similarly-effective interfaces for other features--for instance, could phones be improved by adding holes to them, with increasing depths indicating higher digits?

Diane Ko - Jan 31, 2008 12:34:28 am

Norman writes in POET that for designs, you want to get as close as possible to a one-to-one ratio of functions to buttons. Looking at my own phone, it occurred to me how much that idea really comes into play. Certain buttons, have all their possible functions visible (a necessity for good design according to Norman) such as the 1 button with the voicemail symbol, the zero with the space symbol, the star with the symbol symbol, and then pound with shift and vibrate symbols. However, if I had never used a cell phone before it's likely that I wouldn't be able to figure out how to check my voicemail (the symbol resembling an upside down pair of binoculars doesn't really signify anything to do with voice messages), or how to add symbols (which on my phone is merely signified by '+'). In fact, for the longest time I didn't even know how to switch to number entry in texting with my phone because the method was so counter intuitive. Normally I would just tap through the different possible writing methods: T9 alphanumeric input, regular alpha input, and numeric input. Instead, this phone was originally set to regular alpha input. Switching to T9 requires a quick double tap but switching to uppercase regular alpha input requires 2 successive taps. Switching to numeric input actually requires that same button to instead be held down until a menu appears where you can select numeric input. Highly unfortunate since texting is useful for sending phone numbers. This feature was only recently discovered after looking through the manual.

Norman also mentions the importance of feedback in older technology and how it's becoming less and less available in the latest technology. Immediately I thought of the iPhone and all of its many features. The iPhone has become one of the most sought after cell phones of late and yet, ironically, doesn't even have numerical buttons. Typing on the iPhone can also be quite a pain considering that there is no tactile feedback for the letters and that people who don't have the daintiest of thumbs have a hard time pressing the right letter.

Gordon Mei - Jan 31, 2008 01:51:10 am

Among the considerations that arise in preventing poor design choices was the affordance of materials. The perceived uses of glass as an object to break, or of wood as a material on which to write and carve, were reasons why British Rail panel shelters of glass and wood were vandalized with breaking and graffiti, respectively. It plays to us psychologically, and essentially "asks for it" by being unwittingly inviting to these urges that the sight of those materials stir in us. While some of those who still would not succumb to these temptations have done so because of a decision to hold back, we still fall 'victim' to the affordance of materials, namely with more benign examples like bubble wrap or large balloons (both of which many of us find difficult to resist popping).

Likewise, it is affordance on which designers ought to rely to successfully signal whether to, say, push or pull a door without signs. Take the unlabeled, unexplained mysterious R button on a telephone system by British Telecom. The defense behind the R was that it was simply a button that had fallen into obsolescence, but had remained because its presence had no negative effects, much like the appendix organ in the human being. This is much like some of the older computer keyboard keys, such as the "Pause"/"Break" key, which (yes) does have some functions, but is not absolutely crucial to a vast majority of the users. Note that there are keyboards that have rid of some of these keys, suggesting that there are probably better ways to redistribute those niche functions among the active keys anyway.

Furthermore, some controls using natural mapping are fairly visually intuitive, particularly the seat adjustment control in a Mercedes-Benz car that used buttons in the shape of a car seat for reclining and shifting function If a passenger wanted to recline, they would spot the "back rest" seat button, and flick it in the direction in which the actual seat would move. It shows that even with the higher complexity of multiple directions possible in seat adjustments (recline back/forth, tilt seat, slide seat), the complexity of the design to the user does not necessarily have to follow.

Kai Man Jim - Jan 31, 2008 02:02:39 am

After reading the article, I understand how important is the design of a product to a user. A product can be very creative and useful, however, if a user doesn't know how to use its features, then this product is just a garbage to a user. Therefore, designers have to understand how to get into the way of how users think, and from there to make a simple design based on visual and easy concept to get the user's attention and link up their mind with your product. The camera example is a good one since it really get my attention about not using all the features from the camera. Or all I know is to press a button to take a picture, others new/creative way of shot is not my business because I don't know how to use it if it is not simple to me. So, I agree that design has to be based on simple concept and visual.

Andrew Wan - Jan 31, 2008 01:54:46 am

The anecdotal format of POET definitely helps illustrate examples of good and bad user interfaces. Visibility seems like an obvious requirement for a workable device, so it's interesting to see how frequently it's not thought of. Cost and aesthetics can play too important a role in design, resulting in inadequate interfaces. In other cases, device elements are used for too many purposes, hiding functionality.

As for the "paradox of technology", the conflict between complexity and usability seems to be one of the most fundamental challenges in design. It is interesting that "natural mappings" themselves change with time. As people become accustomed to using various devices, it sets a clear precedent for the design of future products. That said, it also becomes more difficult to resolve more basic problems with a type of interface. Given this challenge, creativity ultimately makes the most drastic advances in UI design.

Jonathan Wu Liu - Jan 31, 2008 02:02:25 am

I really liked the fact that Norman questioned everyday items. It's hard to find faults in those items when we use them everyday and we put them in a box in our mind. This skill is very useful in knowing ourselves which can translate into creating killer user interfaces. I also appreciated the Natural Mapping section. Motions in the user interface that imitate the actions that should take place seem to be the most intuitive. That is on a touch screen, when you scroll, you naturally flick down; or when you want to zoom in, you spread out your fingers. I know now to consciously consider natural mapping in creating any user interface. I also think today that Feedback is not used as often as it should. Many times the user is left wondering if a message was sent or if a setting was changed. No notification to the user puts the user at unease, and I often check to make sure that a setting has indeed been changed after I tell it to "Save Changes". I think Feedback also gives an element of personalization to the user, so that it feels like whatever service/product they are using just catered to their specific needs.

Fan Yang - Jan 31, 2008 04:02:36 am

I never noticed how many common every day objects are designed whether well or bad. POET brings up a lot of interesting points about how it shouldn't be hard or require thinking for somebody to operate simple objects. His comments about doors really hit home, I know there have been many times when I've pulled instead of pushed on a door and always felt stupid afterwards, but looking at what he considers well designed doors, I agree, I've never tried to pull on a door with a big flat bar across it. I think the most important thing to take out of the reading is that humans are designed to act certain ways when they see things or when things happen, so the design of an object should take this into account. Knobs are meant to be turned and buttons are meant to be pushed, and all of these should give feedback when they are operated so the user knows what the immediate consequence of his actions are.

Jun Kang Chin - Jan 31, 2008 07:00:11 am

Visibility seems to be the recurring theme throughout the chapter. It is visibility that gives users an idea of what mappings are possible with the device. I can't imagine using the complicated telephone system back in the days. The author is right that some functions that require arbitrary number weren't intuitive. That could be why we have more buttons on our business phones today, to provide the one-to-one functionality mapping as well as visibility.


On a different note, I found the introduction interesting. Norman quotes an example of ruffians always shattering newly replace glasses in the train yards, but the plywood replacements last because of material psychological connection. This reminds me of how I always thought stainless steel brushed phones has that 'German well-engineered' feel to it, while wood surfaces remind me of classical traditional designs.

Hannah Hu - Jan 31, 2008 09:10:49 am

I find the glass/wood juxtaposition both interesting and amusing. It never occurred to me that even materials can affect user perception and expectations of the workings of a device. But then, sometimes when I read reviews about electronic devices, there would be an occasional gripe about a certain feel to a moving part - the battery compartment door feels like flimsy plastic, that shiny screen is prone to scratches (even if it is scratch-proof), etc.

Visibility is a definite must for complex devices, especially as more and more devices accumulate more features. I'm seeing more and more devices requiring instruction manuals that are novel-length (and not just because they have to accommodate multiple languages); and the trouble is that most designers are already technologically experienced. Given the vast technological gap between current generations and those who are in their 50s and older, which has widened as technology advance ever faster, there is more design-error potential. When designers in their 20s and 30s have to design something that should also appeal to people at least 20 years older, more user testing is needed. I can't even count the times that my mother keep calling me over to figure out what just happened to her databases (because she accidentally hit the wrong key). Frustrating indeed.

Roseanne Wincek - Jan 31, 2008 09:13:51 am

In Norman's The Psychology of Everyday Things, the author stresses that visibility, intuition, and feedback are extremely important in product design. I think that the author feels as though the best designs need no instructions at all. Natural design elements are intuitive and relate back to real-world conventions. The author used his new, European car as an example. Although it had 112 controls and he did need an instruction session with the dealer, he quickly mastered almost all of its functions. This is because the features of the car were designed to be extremely intuitive. For example, pulling the window button up to roll up a window in a car. This intuitive approach is bolstered by a one control-one function framework. Physical mapping greatly enhanced the ease of use of the car control system. In my car, the window controls are in the middle, between the two front seats, rather than on the door. My passengers are always groping for a control on the door, and then I open or close the window for them. By placing the control directly next to the object that it controls, the control is easier to learn and use.

The author also stressed that feedback, both for the user and for the designer, is extremely important. The user needs feedback to validate what they have done, and to know where the system stands. Likewise, the designer needs user feedback to find if the designed system is really usable for someone who is not an expert on it (since the designer is).

I also thought the U shaped technology cycle was really interesting. When we have new technology, the interface can be really difficult. Then we get good at the technology, and can automate a lot of the necessary features needed for use (like setting the frequency and antenna, in the radio example), making its use easier. However, then we get really good at that technology, and we add all kinds of features (which are really new technologies in and of themselves), and the system gets hard to use again.

Timothy Edgar - Jan 31, 2008 10:57:10 am

The POET article presented a lot of great examples of failures in interfaces. The other day myself, I pulled a door instead of pulling a door despite it said push in big letters since it had a vertical grip. Someone told me to push and I realized my mistake. Unconsciously I associated the vertical grip to pulling. The most interesting aspect to the article I found was the mapping comparison between the car and the telephone. Cars are designed rather differently, yet they are fairly easy to pick up (usually). Conversely, I rarely bother myself but with the basic features of a telephone. The idea of natural cues to explain things either intuitively or visually makes a lot of sense. Complexity can be reduced by relating to the natural intuition of people.

Robert Glickman - Jan 31, 2008 12:15:27 pm

This text was particularly interesting because it dissected many of the technologies that I take for granted and caused me to view everyday objects in a new light. The doors were particularly interesting because I see those glass doors everywhere and I have never consciously registered the associated cues (as is intended). The type of handle and the location of the handle are often very natural. The washing machine example also made me conscious of further design issues. I realized that there are so many technologies with which I ignore many of the features because their design is just so complicated. Some of the interfaces given just made me laugh and shake my head (the refrigerator, for example). The phone system also offered an interesting look at an everyday technology.

Andry Jong - Jan 31, 2008 11:01:46 am

For some reason, I get the feeling that, through this chapter, the only thing that Norman is saying is that each control in a device should only control one function; otherwise, put a display on the device.

While I do not have any other choice but to agree with Norman's example about doors and refrigerator, I do not think putting a controller that can go forward and backward in a car sound system is necessary. Let's consider the possibilities that can happen if, we put a scrolling controller (like a computer mouse ball) to our car sound system device. In the motion of the car, we would not even be able to make sure if our hands are moving left, forward, or forward-left. And what does the system recognize? Or is there any other possible shape that we can use to navigate forward and backward motion? Besides, if the designer had "[rotated] the control 90[degree] on the panel so that it moves vertically" (p.26), I believe it would have disturbed the drivers as they try to switch gears. It is true that good design is necessary for every device exist. However, from time to time, it is also necessary for the users to be willing to learn other type of control for (maybe) safety reason.

Cole Lodge - Jan 31, 2008 12:41:03 pm

This reading was quite refreshing, One thing I have noticed is that in todays world every one wants to be able to do more with each device. For example, where I work, we just had a meeting about all of the new functionality we are planning to include in the next release; what was noticeably missing was where all of these functions were going to be on the interface. This is were the contradictions arise, customers want more features and a simpler interface. Compromises have to be made; sadly, the simplicity is always what seems to suffer. All though this seems to be the way most companies are headed, it is also possible to simplify an interface to much; for example, the slide projector with a single button. The reason I found this article refreshing was because no one seems to think about where a feature will go when it is developed, instead it is thrown in wherever it will fit. This article on the other hand suggests that features should be developed around the interface. If the interface does not allow for a new feature, do not add it.

Lita Cho - Jan 31, 2008 12:48:06 pm

Norman presents very interesting ideas in chapter 1 of POET. I can relate to the frustrations people go through due to poor design. When I was trying to change my clock in my car due to date light savings, I couldn't figure out how without flipping through the manual. I realize how importance affordances are within a design, and I do agree that a good designer takes advantages of the affordances within an object. This also leads into the importance of visibility and making sure that the user can figure out how to use the device just by looking at it.

The reading also addressed the importance of feedback and how nowadays devices don't focus on it anymore. Telephones use to make a tone to tell the user that he/she pressed the right button. Now, cellphones are more interested in adding more features than focusing on feedback. Feedback can be useful to the user to know if they are using the device correctly, and help the learning curve.

I realize now that designing is not an easy task, and a good design goes completely unnoticed. I didn't know how much work when into a simple telephone, yet it is very easy to use.

Max Preston - Jan 31, 2008 01:14:28 pm

There is one thing that I feel as though the author failed to touch on. Specifically, there does exist a sort of advantage to being forced to figure out how to use unintuitive devices. Really, the entire trying-to-figure-out-how-to-make-this-impossible-device-work process makes it easier to use similar devices in the future, does it not? Depending on what the device is, one can begin to understand the basic principles behind the workings of the device and different routines and general tactics for getting similar devices to work. I think this is a sort of general knowledge and problem-solving skill that can be useful in many, many different kinds of circumstances. For example, how would you know how to make a user-friendly interface without having first experienced what it's like to struggle with a completely nonsensical one? Knowing what not to do can make it much easier to know what to do. I admit that it would be convenient if all objects made perfect sense and were simple to use, but I think that they are some drawbacks as well.

Tam La - Jan 31, 2008 01:21:35 pm

In Chapter it interesting how the author start out by talking about people habit of turning on the wrong stove burner and how plastic wrapped packages that are impossible to open. Also how washing machine and dryers have become too confusing to use. For me personally, I have had experience problem with all the things that listed above. It interesting how they are mostly common products that we use through out our daily life but we still has problems operating them sometime. I never can figure out how to use a washing machine when it not the model or the brand name that my family have. We have a digital panel control washing machine so when I try to use the older model with the turning knob. I never can figure out where to turn it. I would never think how things as small as a user interface between you and the machine could affect it full function. Without the proper user interface a machine can not fully function for the reason it being build.

William tseng - Jan 31, 2008 01:37:21 pm

I find it interesting to look at the points brought up in the article with a currently popular application such as the iphone. While I can agree with him in that in circumstances such as entering an unfamiliar building you might not know which way the door swings, I think there is certainly more slack in designing the visibility of functionality on the iphone, specifically the functions afforded by the multitouch display. There is no visible button which tells you to zoom and even if there was it would be redundant with the double-tap gesture used to zome on the Iphone. Similarly because of the "sleek interface of the iphone" without buttons and large clues to give away how it works it is also hard to form a good conceptual model just by looking at it. If you were to see an iphone while it is turned off, you would have little inclination which side is up down, whether it should be held horizontal or vertical etc. I do have to admit that what the iphone does well is it does a good job of mapping functionality. I.e. the accelerometers detecting if the phone is horizontal or vertical and changing the display mode accordingly. This is a good example of how the movements cause results to happen with very little user effort.

JessicaFitzgerald - Jan 31, 2008 01:54:36 pm

I thought the author's discussion about the paradox of technology was both interesting and true. Technology is intended to make our lives simpler, but the more technology advances the devices we use become more complicated and much harder for the average person to use, so they don't get the full use out of them because they don't want to spend the time reading the directions to figure out how the device works. Instead they just use the functions of the device they are familiar with and need to use. The new technology isn't even used, so what is the point of having it in the first place? To me this shows how important design is. Design needs to be simple enough so that the average user will be able to understand how to use it easily. If not, then it is almost a waste of time to design it in the first place, because then only a few people will take advantage of the full functionality of the design.

Also, I found his description of the refrigerator cooling dial quite funny. Why bother putting an adjustment dial on the refrigerator if it is too confusing for anyone to use, and then also to include a completely wrong schematic design is pretty hilarious. Its like the designers were like "hmm.. how can we mess with the users to confuse them as much as possible" - and it worked!

Nir Ackner - Jan 31, 2008 01:56:16 pm

Norman discusses the new phones in his building and how "nobody ever thought about trying the phones in advance". Usually, when making a purchase, the primary factors considered are the features the product offers -- not how easily those features can be used. As such, there seems to be little economic incentive for designers to put the effort of iterated design in, when it won't affect the number of units sold. On the other hand, if users become more sensitive to the usability of products at purchase-time, or the cost of technical support begins to rise due to poor design, companies will begin to look more carefully at this important topic.

Another interesting point is Norman's discussion of the ideal design being one control per feature. This seems to be an argument toward many different mobile devices rather than a swiss-army device, an important topic in our previous reading. Norman, while discussing a mental map, fails to mention the critical concept of grouping: putting related features together in menus or areas of the device to help guide the user.

Pavel Borokhov - Jan 31, 2008 01:44:55 pm

In general, I mostly agree with what Norman is saying. All too often, we are presented wth doors that have pull bars but are meant to be pushed, and with cell phones on which getting to a specific place in the interface takes many more button pushes than one would expect. However, I would take issue with some of the "complexities" that he brings up. One such instance is the instructions for recalling a phone call that is in the process of ringing at another desk. The instructions in question say, specifically, "...to answer any ringing extension, dial ringing extension number, listen for busy tone. Dial 8 to connect to incoming call." Personally, I do not find anything confusing about this (perhaps the concept of extension numbers was not widespread at the time of writing?) – you simply dial the extension of the phone which is ringing, and then press 8 to pick up the call once the busy tone is heard (to be honest, I'd be more concerned about other people being able to intercept my calls without any input on my part). And while I do agree that the "hold" functionality he mentioned on numerous "new" phone models was harder to perform than it should be, I did not find it to be so inaccessible as to make it unusable. Of course, placing a single "hold" button would make much more sense, and most modern phones do in fact have such a setup for the functionality.

I think that feedback is a really key element of any interface which is completely crucial to its success. I know that I am personally a feedback nut and absolutely love to have multiple feedback mechanisms available to me to make sure that I am aware of what is going on. This is another example of where multimodality is not only useful, but almost essential, as a way to guarantee redundancy in the feedback/output mechanism and provide a higher degree of certainty that the user was able to acknowledge and correctly interpret the feedback mechanism. As a subset of this and specifically for software applications, the software needs to remain as responsive as possible during all times. For example, if the we order a web browser to visit a certain webpage, the browser normally does two things immediately: provides some sort of indication (visual, aural, or both) that the request is being carried out and unblocks the interface to allow the user to perform additional actions. Yet it is worth noting that most file browsers do not use such an analogy. Say we are using a file browser to view files on network-mounted volumes. Let's say we are using Windows Explorer to browse the files on a machine located in Tanzania with many files and folders (the Mac OS X Finder, too, suffers from this problem). When we double-click a folder on the machine to view its contents, Explorer might freeze while trying to render the file list of the folder because the folder contains 1 million files. The interface will be completely blocked and no other actions will be possible until this action completes. Given the speed of the connection used by the Tanzanian computer, as well as the large number of files, this might never actually happen. Thus, a proper feedback mechanism would be to, first and foremost, inform the user that the file list is being retrieved so that they know that their request was accepted. But an important second part would be to do this action in the background so that the user may do other things or even completely cancel the operation, since he's looking for a folder with just 10 items, which would open immediately. This analogy, of course, would not apply to operations which would sensically require the interface to be (maybe partially) blocked, such as when an error dialog box pertaining to some window is displayed and we want to make sure that it is noticed by the user.

Jeffrey Wang - Jan 31, 2008 02:10:41 pm

I think that many of the reasons that appliaces are so confusing to the human mind is that human mind was not evolved to use such complex technology (not that we cannot learn to do so, but that the brain was not selected to do so). To clarify, one can easly handle a CD player that has a simple play, stop, pause button. However, with the addition or recording, file folders, and shuffling option, this makes the CD player a lot harder to operate. This can be seen as a social thing that we must learn through society; because older individuals are often incompetent at even the "well-designed objects" described by the articfle. However, because of the pressure for companies to create complex operating systems with a simple interface, in order to attract a greater customer base, they are forced to add all these additional (and may i say, superfluous) functions, while tryign to keep the interface simple for the "exquisitely" tailored mind to organized. One comment was that the feature had to be "visible" - in the old days telephones had visible hold buttons. However, with all the additional features that are added, some must be taken away because individuals still want a "Simple" interface. Thus, there is conflict between the features that the individuals want in an interface. (I also really liked how the telephone designer compared building a phone to natural selection -- in the genome, if genes are already there, and have no negative effects, selection will not select it out.)

Although the article talks about complications of, for instance, telephones that do not have a "hold" function, this is hard to enforce internationally. (Why is a hold function even necessary?) Even if non-literate symbols were used, symbols in each country are different. Thus the interface is culturally dependent.

This is not to say that the interface can be modified through psychological experiments. Of course, psychology is a science and through this science, technology can be greatly advanced so that individuals will understand its use better. (I.E. how to make computers easier to understand for elderly individuals who are not accustomed to such technology)

Jeremy Syn - Jan 31, 2008 01:24:45 pm

About the new telephone system mentioned in the reading, I was wondering why they would try to create a new telephone system, especially one that didn't work too well. The current telephone system works perfectly fine and everyone is familiar with it. Instead of creating a new one, they should have tried to perhaps build upon it and make it even more convenient or useful. In the case of the refridgerator, I must agree that I hate when a device has knobs that are dependent of each other. I hate wasting time trying to set the multiple knobs so that everything seems perfect.

Brian Trong Tran - Jan 31, 2008 01:55:03 pm

The beginning of the chapter COMPLETELY reminded me of those "How many Berkeley students does it take to ...?" jokes. I really enjoyed all the examples that made me feel slightly less incompetent when not using a door correctly or messing up on holding calls. The chapter also listed some pretty important design topics that all engineers should take into account such as mapping and feedback. I think mapping is very crucial because users will otherwise not develop a connection between their actions and the functions of a given product. I liked that discussion very much and feel that mapping is very crucial in the development of any software product. I disagree with the part on how more features results into more complexity to the point that the product is unusable. I think features have the benefit of lessening complexity if designed properly.

Ravi Dharawat - Jan 31, 2008 02:00:30 pm

It is interesting to consider the balance between making aesthetically pleasing devices and functional devices, and how these things can be at odds. It is also interesting to consider the balance between hiding increased complexity from users and giving the users more information so as to better use the device. This puts in many words what visibility is. Decreased visibility in the name of efficiency or aesthetics has long been the bane of progress. Consider the programmer who squeezes what easily could have been four lines of C into one. Sometimes it is useful, because it allows more code to be seen at once, useful for debugging. But it definitely makes single line statements more difficult to read.

Raymond Planthold - Jan 31, 2008 02:42:57 pm

I'm pretty sure office-type phones are designed by people who hate other people. My father recently got a new fax machine (well, used, but new for him) and in the process of setting it up, I discovered that it lacked a speaker phone function. Well, it actually had one, only the button was labeled "Hook" -- as in "take the phone off the hook." I'm sure it made perfect sense to somebody.

Feedback is definitely something that gets overlooked often. On the newer AC Transit buses, the controls for opening the rear doors (in addition to being physically behind someone who is standing right at the door) offer no feedback when pressed, and the door itself takes a rather long time to start opening. This leads to confusion for new riders, who may think that the door is broken or that they didn't press it right the first time.

On the other hand, even when feedback is given, so many people still seem convinced that more pressing == faster response. Take elevators, for example. How many people do you see walk up to the call button, clearly lit, and press it again? Then, when the elevator is taking too long for their tastes, they begin to mash the button. Some of them probably only do the first out of habit, and the second out of frustration, but there are those who actually think it has an impact.

Jesse Albini - Jan 31, 2008 02:39:48 pm

I really enjoyed this reading. Many of the designs in cars, are things that I have spent a lot of time complaining about/praising. For example, I have an old style BMW radio (quite similar to the one pictured in the reading) that has the strangest way to change the bass/treble. There is a slot that you can put your finger into, and you slide it in the direction that you want to change it. It took me nearly half a year to figure out how to change it, and it's still hard to use even knowing how it works.

The idea that I found most useful was the idea that reducing the number of buttons and controls doesn't make something easier to use. In fact, it makes certain things much less intuitive. The idea that a low learning-curve is more important than spatial economy is something to remember.

Scott Crawford - Jan 30, 2008 11:54:35 pm

The concept of natural mappings is a seemingly obvious one, but once spelled out adds a lot of clarity to defining the innate usability of 'everyday things.' An apt example that the article made me think of is door locks on a car. The 'old fashion' analogy of having that small rod pop up and down mimics what you naturally think of when something is unlocked or locked (respectively). If it's sticking out of the door, it can be thought of as being 'out of the way' and when it's nearer to flush with the door it can be thought of as being in the door and so 'in the way'. In the operation of the door, something being 'in the way' naturally implies that it obstructs the usual operation of the door (that is, the ability to open/close) whereas something 'out of the way' should allow it to function as usual. Despite this natural analogy, the designers of the Toyota Rav-4 (or at least the model from a few years ago that I have) have corrupted this analogy and made a very unnatural system. Set into the door frame is the ability to lock/unlock all the doors in the car with a single press of a button (which hinges back and forth - pressable in 2 directions). However, the button instead of being labeled with 'lock' and 'unlock' has it's two pressable directions distinguished by having a bump on one end and an indentation on the other. It would appear that they are trying to appeal to the natural analogy of the 'old fashion' locking system (when down is locked and up is unlocked), but no matter what, I always find myself puzzling about whether I should be pushing down on the bump to lock the car, or pushing down on the bump to achieve the state it's 'label' indicates (a raised position, or an unlocked car) - and vice versa for the indentation. This is a failing of the mapping in that the relation of bump/indentation to locked/unlocked is not readily apparent, and so should be replaced with the actual words lock/unlock for greater clarity. In other words, this particular system cannot easily appeal to a natural physical mapping, and so should in lieu of that use words to convey it's function (since labeling one end with 'lock' and the other with 'unlock' would be perfectly clear to anyone who understands English).

Megan Marquardt - Jan 31, 2008 02:55:10 pm

One thing that was very apparent in this reading is the fact that elegance does not directly correlate to usefulness. Such as his explanation of his friends' experiences with elegant doors and elegant projectors, the simplification does sometimes over-simplify such that an easy task is turned into a very difficult one. It seems that anytime a certain, simple function needs a consultation with the user's manual, then the interface with which the user is presented is much too complex or ambiguous. This is indeed very surprising, because it would seem the designers of these products didn't conduct user studies, and failed to iterate through the design cycle we've been talking about in lecture.

Jason Wu - Jan 31, 2008 03:02:02 pm

There were a lot of interesting ideas he has in this paper, discussing the faults of everyday objects that I have always taken for granted. However, I feel like in some cases he takes idiot proofing too far. For example in his anecdote about his friend being trapped between two rows of doors, I feel like this is definitely the user's fault rather than the door design.

I did like his attempt to come up with the function to control ratio, but I feel like that's not always the case. I believe you can still create a rich, full featured product with minimal number of well designed, slick controls. I believe the iPod scroll wheel is a great example of this.

However, he does approach the idea of design with a very critical eye, which is a great trait to have for that job. He comes up with a lot of great notions to think about such as visibility, and feedback, allowing the user to pick up and use the products without reading an extensive manual beforehand.

Zhihui Zhang - Jan 31, 2008 02:41:49 pm

Norman gives several examples of design choices where gains in aesthetics were made at a sacrifice in usability. I've personally had many moments of embarrassment when I've tried to pull a push door or even attempting to open a full size window that looks identical to a sliding door besides it. Particularly, I think it's important that designers consider the ease of use for the various features they try to implement; especially whether a user will be able to intuitive use the product without reading the manual (which most people will forgo). At the same time, it shouldn't be the case that aesthetics have to be sacrificed. For example, in the case where British Rail replaced the glass panel shelters with plywood - which in my opinion is not as visually pleasing - in an effort to discourage vandalism, an alternative might've been to use clear plastic. Neither plywood nor plastic are for breaking, but plastic also allows for visibility.

Jiahan Jiang - Jan 31, 2008 03:07:05 pm

This is a very interesting and enlightening article; a lot of the ideas and issues discussed is extremely helpful; the real-life examples given in the article is just hilarious. It reminds me of the websites that feature creative designs or re-designs of everyday objects. It's one thing to have creative, aesthetically pleasing designs, it's a whole other story to make sure they actually "make sense." I truly enjoyed the reading and especially loved the discussion on "visibility."

Daniel Markovich - Jan 31, 2008 02:56:00 pm

I felt one of the most important topics in Norman's POET was relational/natural mapping. After reading the section I realized how much we take everyday designs for granted, and that there is a deep psychological principal behind designing a product that requires little or no thought to use efficiently. Although it is not strongly stated in the article, throughout the reading Norman proposes that to design a product that not only provides the correct functions but is convenient and easy for the common and first time user to operate requires a deep understanding of the user, and how they will interact with the product. He reinforces that designer today are more pre-occupied with a products variable features than with the user-product relationship and that this is the pitfall of designers and their products.

Jeff Bowman - Jan 31, 2008 03:13:05 pm

One of the things that I thought about when reading about the conceptual models was whether they were actually necessary. To some degree, I believe that while conceptual models are necessary for a learning curve, they can be abstracted out such that users learn the system fresh without relating it to an existing system, and don't need to know how the system functions in order to use the system. However, I also realized how quickly that system breaks down: People don't think of computer windows as sets of lit pixels, or as rectangles of light, but rather as physical things they can open, close, and resize. Web pages, files, links, and many other facets of computer systems are common nouns in English, to which we apply common verbs. Finally, I failed to realize that regardless of whether the conceptual model is communicated well to the user, the user will likely concoct one anyway to match patterns of input to patterns of output, reinforcing how useful it is to create a clear conceptual model in the first place.

Randy Pang - Jan 31, 2008 02:37:18 pm

Regarding Max Preston's post, I don't feel that one should ever purposely design systems that unintuitive, nor do I feel that using such interfaces are helpful (I mean, why suffer through bad UI's just so you can be better equipped to suffer through future ones? Why not just solve the problem at it's root?). The assumption here is that all nonsensical devices are nonsensical in similar ways, which I don't believe to be true. Just because I struggle and learn the ins and outs of one UI, doesn't mean I can transfer any of that knowledge to a seperate one. Even among good UI's, there is a flawed notion of transitive familiarity. For example, recently I was playing around with my friends HTC Mogul (a smart phone that runs Windows CE), and I initially struggled immensly. Partially this was because of certain UI decisions, but more prominently, it was because I had already become accustomed to the iPhone's UI, which is immensely different. Learning what to and not to do on the iPhone actually hampered my ability to interact with my friends Mogul (because I was trying to treat it similarly). After my friend showed me how to interact with Mogul, the UI actually turned out to be fairly nice and usable, just different, and I probably would have had a better first experience had I had zero experience using the iPhone previously. This is why, as Norman suggests, we need to make things like visibility, mapping, and feedback our foremost concerns, because they will always lead to a good user experience, regardless of our assumptions of the user (provided he's not blind, nonsensical, and numb, of course).

Siyu Song - Jan 31, 2008 03:22:20 pm

I thought the main idea of this article was very similar to the Palm article about simplifying features to make the most important functions of an object work easily and without the need for explanation. Specifically, the anecdote about the women who complained new phones in their offices lacking a hold function because it was not readily apparent, seemed like exactly the situation Palm was trying to avoid. This article however is more universal than the Palm article by saying all everyday objects should be simplified, not just handheld organizers. I liked how this article gave very clear guidelines that well designed products should follow.

Zhou Li - Jan 31, 2008 03:32:11 pm

The key points for a good interface design presented by Norman are: visibility of the controls and easy-to-understand instructions, natural mapping of the controls and the resulting operations, and immediate feedback of interacting with the controls. But there are always trap-offs between usability and simple elegant design. Space and size constraints can also make the interface design more challenging. So a great product that also looks good really requires the interface designer to think thoroughly from a user’s point of view and test the product with potential user groups to get their feedback before delivering the final version. Even a simple everything objects such as the door can cause frustration for users if visible operating clues are not present. With new technologies coming out every day, it is tempting to incorporate them into devices we use as soon as possible, but the main point to consider is: are those new features really worth the extra complexities added to the design? If the new features require me to read the manual for half an hour to figure out, then I probably would rarely use them. So there is no value of added complexities due to those features. And if the features are useful then they should be made easy to use via the interface.

Daniel Gallagher - Feb 03, 2008 08:36:18 pm

Norman's POEM touches on many of the common frustrations with "should just work" objects, those things in everyday life that are conceptually simple and yet still trip up users because of poor design. I agree with Ben Sussman that he could have gone into some of the trickier aspects of design where there are several workable designs and the success of one over others is more difficult to figure. Also there was little in this chapter about design iteration or refining design besides a quote from a designer about it taking 6 tries to get something right. However, POEM struck me as less technical (if that is the right word) and more of an overview of various broad design concepts. Like Design Concepts for Dummies. I found it quite readable and informative, but most of all the excellent examples drove home the concepts. I laughed at some of the ridiculousness in office telephones because I've seen that first hand. Last summer my cubicle was adjacent to the IT guy in charge of setting up the phone system for the company and he spent literally two weeks on the phone with various phone and electricians trying to get the building connected and then the many features of the system working. The incomprehensibility of using the features on the phone eventually forced him to write directions for every function, with the executives still dropping in to have him personally set up their conference calls. Bad design cost a lot of money and wasted time in the end and gave this guy grey hair.



[add comment]
Personal tools