Sketching and Storyboarding
From CS160 User Interfaces Fa06
Lecture on Sep 25, 2006
Readings
- Cognetics and the Locus of Attention. The Humane Interface. Chap 2. Raskin.
- Meanings, Modes, Monotony and Myths. The Humane Interface. Chap 3. Raskin.
Ramy Ghabrial - Sep 20, 2006 08:34:26 pm
Cognetics (chap. 2): Error messages are indeed pointless. Also, I agree that randomized confirmation messages are just obnoxious. I know a free program that randomly switches the placement of startup confirm/cancel messages in the hopes of tricking you into purchasing a "better" version. And windows xp's own "automatic updates have been installed, restart now/restart later" random popups never attract my locus of attention for long; at least in my case, they only seem to generate resentment. If you are going to try to capture the user's attention, it had better be worthwhile.
Modes (chap. 3): This chapter made me think of an interesting question. Does anybody actually use most function keys anymore? We have some function keys that are worth remembering and work the same in most situations (F1 for help, F5 to refresh, alt-F4 to close... why do you have to hold alt? Why not just F4?) but most of the rest just seem useless. I've been using computers for quite a while, but I think I've only ever used a quarter of the function keys in any manner. Do they have some hidden purpose I am unaware of?
Patti Bao - Sep 23, 2006 04:43:59 pm
Chapter 2: I never really thought about confirmation messages, which I suppose illustrates Raskin's point about habit formation, until recently, when I tried to delete a batch of photos using Flickr. I prefer their old organizing interface, so I usually use that, but this time I decided to try their new one (http://flickr.com/photos/organize/). After I pressed Delete, the first message that popped up was a little window that clarified my intent - and also displayed mini thumbnails of the photos I wanted to delete, which, just because they were there, made me stop and look them over for a moment. I pushed Delete again and got this message, "You really want to delete 27 photos? They will be gone forever, and cannot be recovered," so I pushed "Yes, go ahead". A third message appeared that said, "Are you really, really sure you want to delete 27 photos? They will be gone forever, and cannot ever be recovered," and after I pushed "Yes, go ahead" it finally disappeared and deleted the photos. This was my first time deleting photos in this interface, so I did actually pause and think about my actions. I am not sure whether I will still pause in the future (or if the habit has already developed!), but either way, I think the designers did try hard to break the habit by introducing rather unusual confirmation messages - not only are they more colloquially worded than normal, but they also flash you a visual (the mini thumbnails) and offer you three chances to abort the operation. According to Raskin, however, this is probably still not enough.
It's also interesting how Raskin critiques the desktop-based system. His example of the Apple OS starting up on the state last saved when it was shut off is a good one, and it's a feature I appreciate. However, I've recently switched from using Word to OmniOutliner, which surprisingly does the same thing - it opens on the spot where I was last typing before I saved it. In this case, I actually find it fairly confusing, and I'm not sure if it's because I want to see a blank slate, or if it's because I'm used to seeing a blank slate.
Chapter 3: Raskin makes a compelling argument against splitting interfaces into "beginner and expert subsystems", but I wouldn't go so far as to argue that all interfaces should start off as strictly monotonous ones. For instance, I think keyboard shortcuts are not quite as visible or easy to learn as the actual GUIs themselves, yet once they are learned, they can be extremely effective, which makes them something of an expert subsystem. Should designers not include keyboard shortcuts simply because they divide the user group into beginners and experts? What exactly would be a better alternative?
P.S. In response to Ramy, Apple laptops make full use of the function keys - each one is linked to basic controls like screen brightness, volume control, expose, and dashboard. And they can all be configured, so I guess it depends, but I actually use my function keys a lot.
Patrick Rodriguez - Sep 23, 2006 07:56:24 pm
Chapter 2: I wonder how the theory of a "Locus of Attention" meshes with the reality that we are multitasking more and more each day. Check this Time cover story from this year: The Multitasking Generation. Raskin says that multitasking only really works when everything except one activity is automatic. But take, for example, doing your homework while also listening to music, chatting with friends online, talking to your parents on the phone, eating snacks, etc. What tasks are automatic and what aren't? I'd say that more than one is, and even at the expense of productivity, sometimes it's better to multitask. To me, at least, activities can become more tolerable this way. So maybe, when it comes to UI, sometimes more is better than less, if the user is able to customize the workspace in a way that accomodates his favored way of working. Just a thought.
Chapter 3: Having written my post on Chapter 2 before reading Chapter 3, I appreciate that Raskin anticipated my thoughts about user customization. Key quote from page 49: "an interface that optimizes productivity is not necessarily an interface that optimizes subjective ratings." But I just can't agree with his presumption that the UI designer can be be perfect: "The central point of this issue is that if we are competent user interface designers and can make our interfaces nearly optimal, personalizations can only make the interface worse." Does one size fit all? That's a good question.
Scott Friedheim - Sep 23, 2006 08:35:57 pm
cHAPTER2&and3three Looking at users habits proved, in my opinion, to be a valuable method of analysis of user interface design. My experience in production environments has shown that doing things out of habit and having the tools run smoothly is very important. As Raskin pointed out, if users are always getting interrupted in their tasks by choices, options, popups, and warnings then the conscious mind need be alert effectively slowing down and frustrating the user. This idea of getting rid of unneeded options, and modes is key to simplifying a design. I personally frown upon and overlook products that don't have a clear intuitive user interface regardless of the features the product offers. If I can't do what I want with a product immediately then this means I have to fidget, which should only be done when you want to enable/configure a feature.
Raskin made another good point by stating if a user interface can be made more useful/productive by allowing a user to configure it then a bad job has been done in the design of the interface. This gets back to simplifying a design; if an option is used frequently by many users then just include it rather than requiring user configuration. This is what he calls the beginners vs. experts scenario. Of course depending on the application, if your designing a system, design it for beginners; experts, if they really are experts, will know how to get into the expert settings/mode/options (ie. these options needn't be visually available anywhere).
I really found interesting the idea of designing for emergencies and important decisions. It's a good design puzzle in figuring out what is to be done when a product demands attention and requires special notice. It's a puzzle because you must also consider the design trade offs in the products other functions, the users surroundings, etc.
Johnathan Hawley - Sep 23, 2006 08:13:28 pm
Chap 2 - I felt like I was in a psychology class while reading this chapter. Psychology classes have been some of my more favorite classes so I enjoyed this reading. A knowledge of psychology seems important for bridging the gap between machine and human. It was mentioned that the fact that we only have one locus of attention causes problems with designing a UI. So many user interfaces I have come in contact with throw so much stuff at you at once that it becomes confusing and frustrating. It would be a good idea to think of ways to methodically transition the user's attention from one part of the interface to the next while designing our application.
Chap 3 - I guess I have been using modal interfaces for so long I have a hard time visualizing how a given interface I am familiar with would look modeless. I have run into difficulties, like the author, with Word's adaptable interface. There are still things I would like to turn off that I haven't found yet. I do like the option of reversing something I've done rather than hitting cancel before it's too late. They said that noun-verb models are better and gave an example of what paint would be like if it were strictly noun-verb but I think I would like the way it is currently better.
Natalie Nguyen - Sep 23, 2006 10:23:52 pm
Chapter 2: I also feel that confirmation is mostly annoying and useless. In Windows, I always mindlessly accept, especially when it just alerts me of the number of items that I'm deleting; being able to look into the "recycling bin" and recovering items I didn't actually mean to delete is a much more useful feature. I also have not-so-fond memories of other CS classes which asked us to confirm (type "y") for every single additional file we wanted to submit, each time we wanted to submit. Mindlessly accepting confirmation messages is a habit that eventually inconvenienced me greatly; when I installed visual studio for the first time, I kept inadvertently accepting to open up the debugger when something crashed -- so not only did I have the frustration of watching a program crash, with all my work with it, but also the pain of waiting for the debugger hogging resources as it loaded up slowly.
Chapter 3: It is indeed arguable that a "more competent" designer would have alleviated the need for personalization, but I feel that if I am given a "suboptimal" design, I would like the opportunity to try to optimize the rest of it to cater to myself. In an example that probably reflects sadly on me, I will talk about the game World of Warcraft. It has a solid UI to begin with that is very easy to learn and use, compared to some other MMOs; however, as players progress, they almost universally wind up downloading player-created modifications to the UI. Is this proof of failure of the designers of the game? In my opinion, no; how could they expect to anticipate the needs of so many different players worldwide, each with their own play styles, goals, and preferences, for a game where being able to react with the right command at the right moment makes the difference between virtual life and death?
Tak Wong - Sep 24, 2006 12:43:11 am
Chapter 2: When I was reading about the cognitive conscious/unconsciousness, I thought a good UI should keep the users in the conscious state so they know and actually think about what they are clicking. An example is that I saw one of my coworker that mindlessly "clock in" on a timeclock program without thinking about what the pop up messages say. She was actually clocked in for the past 3 days and forgot to clock out all 3 days. So every morning she comes in and she would click "clock in", then popup says "you are already clocked in". She never thought about it (and don't think she even read the message) and clicked OK like the program is working fine. The thing is, the popup message for properly clocking in and the "you are already clocked in" message have different size and font color. I guess it became a habit of clicking OK to whatever she saw. I guess they don't really need to be in the unconscious state if there's only 1 purpose of the task.
Chapter 3: I think having expert shortcuts is a good interface. Imagine not having the ability to do Ctrl-C and Ctrl-V. Probably all of us in the class will be very annoyed. So probably keeping some of the common ones is not a bad idea. I think the author is pointing out that having too many ways to do 1 thing gets annoying (not having the one-to-one mapping), but perhaps having 2 ways to do 1 thing is not too confusing and it may be able to speed up some users since they can carry over some past experience.
Jason Shangkuan - Sep 24, 2006 10:46:22 am
Chapter 2:
I think that current interfaces do not keep users cognicent of the information the computer or device is relaying. As previous posters have given examples such as typing "y" for submitting CS project and homeworks or Windows messages that only provide the user with the option of selecting "Okay". In such instances when the user is only given one option or does not think it is necessary to understand the message, the user defaults to what the selection is given and take the chance or figure out what the reprecussions are afterwards. A good way to keep users aware and cognicent is to have the error message display: "Do you understand" and have a yes or no. If the user sees yes or no, then when "no" is selected more detail is provided. This helps the user backtrack to what happened instead of just skipping over the message and not trying to figure it out.
Chapter 3:
I find the concept of designing or having modeless interfaces a far-fetched idea. I think that trying to have modeless objects is a goal that we should strive to achieve in the sense that interfaces need to be simple and that modes should only be used when necessary. However, Raskin does mention that the advantage and disadvantages of modes have to be weighed, which does accomodate for designing for simplicity. I think that what would be most beneficial to design a simpler interface is not to try and design a modeless interface while weighing design advantages, but each feature that is designed should have its affordance measured as to what kind of mode or whether it should have a mode functionality.
Tabassum Khan - Sep 24, 2006 11:25:15 am
Chapter 2: I often find myself looking at a bruise on my arm or leg and not remembering when and where I got it. I guess this is where my locus of attention comes into the explanation. I might have been so absorbed in doing or thinking about X that I didnt even collect the fact that I accidently ran into a table and hurt my knee, only until later when I notice the bruise that I realize its existence on my body and begin to question it. If I dont get a bruise I would have no means to know that I had run into the table earlier. So my locus of attention came to this accident only when I saw the consequence of my action which would not have been possible otherwise. Similarly, if computer interfaces are made so that the effects of any actions are readily reversible rather than simply notifying users about the potential consequences of their actions, the users would still need some kind of indication that they have made an action with bad consequences and then they can simply reverse their actions if they so desire. Unfortunately this is not possible in my story. Moreover, in case of deleting files, how can a system know whether or not deleting a file can have bad consequence for the user or the user deleted a file he/she didnt mean to. For this reason, I think that it is not feasable to replace the annoying confirmation interface with reversible action interface, but a combination of both should be less annoying and more desirable.
Chapter 3: The author seems to be a little bias about the way in which keyboard shortcuts should be explained. I dont see why a person would interpret Ctrl+. as Control and + key, while not interpret Ctrl followed by a down arrow as pressing Control key and the down arrow key at the same time. I agree that memorizing keyboard shortcuts requires time; however, this time should be thought of as investment and not as time wasted from the task at hand. The reason is that once these shortcuts are memorized, they can help you save considerable amount of time and effort of switching from the keyboard to the mouse to execute simple commands.
Bowen Li - Sep 24, 2006 12:27:35 pm
Chapter 2
His thing is somehow oddly reminiscent of the talk of RAM vs. disk. Everthing is on disk (ie people have all the knowledge), but you only bring it into RAM (conciousness) when you need to.
For the pressing Y or N to continue issue, I've noticed that when installing Windows, Microsoft will have you hit different keys. Like F8 to continue, or L to delete partition, or something like that. It forces you to stop because hitting 'y' repeatedly does nothing.
I don't agree that it is obnoxious to prevent the user from developing a habit. If the person's job requires making decisions at almost every step of his work, then I say it is the designers job to BE obnoxious and make sure that person doesn't fall into a habit. From experience, I can say that it is far too easy to fall into a habit of repeated actions while at work. But luckily, my job didn't have many risks attached to it as far as affecting human life, etc.
I really like the idea in the Canon Cat of taking advantage of that 10 sec. delay. It shows that they did their homework to take advantage of people's psychological responses as well as their physical ones.
Chapter 3
I'm not sure why this guy is so anal on shortcut key notation. Most people would not interpret Ctrl+. as control and plus sign. In fact, I had more trouble figuring out his "better" system with the up and down arrows. Those aren't even standard characters, so how would you even type that? I think if there is confusion, then the shortcut key itself is too complicated.
Again, it seems like this guy is just pushing his luck. On the AutoCAD example, if you're using your mouse to do something, and you can't bother to notice what your cursor looks like (which is graphically right on top of the object you're manipulating) then you should not be blaming the designer.
Modes define the world. If we didn't want modes then my keyboard would have a million buttons, each one for a different function, and everything would be spelled out to the period. Again, I don't see how any semi-intelligent person can argue for non-modal systems. People are modal. All but the most basic single-celled organisms are modal. It is a completely simple metaphor to go from people sleeping and being awake to having an on/off switch on a machine. For systems where people use it often, I don't think it's too much to assume some responsibility on the user's part to keep track of the mode.
Indeed, I find that noun-verb makes the most sense, in most cases. Partly, I think because specifying the noun gives visual feedback (for example, highlighting a paragraph to copy specifies the noun) while verbs can't be visual. If you issued a 'copy' command first, what would that look like? It's interesting that this is quite contrary to HTML where the markup is specified, and then comes the content. Also note the difference between conventional and OOP syntax in common langues: perform_fuction(parameter) vs. object.perform_function In OOP the noun comes first, then the verb, whereas non-OOP has verb noun.
Kang Chen - Sep 24, 2006 12:22:39 pm
Chapter 2 The article pointed out several things that I've never really notice before like the cognitive conscious and cognitive unconscious example. Though my favorite part of the reading was the error messages and the attempt to randomize accept/reject response buttons. I agree with Raskin's point that users tend to form a habit that's hard to unlearn which ultimately leads to just pressing enter or Y for yes whenever error messages pop up. Most of the time, those messages are about minor related events or intended consequences such as "Are you sure you want to delete this item?" when the user did intend to do so. The suggestion of randomizing acceptance buttons was simply unhelpful and more than annoying. Just imagine having to read a couple sentences of instructions before you can perform a task repeated for a few dozen times.
Chapter 3 I disagree with Raskin's suggestion that there are "unlimited number of commands" which can help make an interface modeless. When I thought about the unlimited part, sure there could be unlimited commands, each of which has a single mapping to a single task to be performed. However, this approach greatly increases the number of commands the user has to memorize and thus increases the learning curve. Also, commands are usually no longer than 3 or 4 keystrokes so there is a finite number of combinations (some of which would be less intuitive such as Control + X for cut instead of Control + C because C was taken by copy).
Hiroki Terashima - Sep 24, 2006 01:30:18 pm
"We must master an ergonomics of the mind if we want to design interfaces that are likely to work well." "You cannot undo a habit by any single act of willpower; only a time-consuming training process can undo a habit." These are quotes that caught my attention while reading Chapter 2. They demonstrate the importance of understanding the human psyche and behavior. The notion of habit is especially important, because habits are easy to get but hard to replace. Habit is one of the reasons why people are reluctant to make a drastic change in UI in a short period of time.
Raskin suggests that we consider designing interfaces for individuals, not for a specific class of users, in Chapter 3. This way, we can come up with an interface that can easily be used the first time, and continue to be used in more expert-usage as you use it more often. This whole “expert” shortcuts is a nice affordance, but I think the challenge lies in determining which ones are for experts and which ones are not. I just noticed that my keyboard also has labels on the sides of the keys as in figure 3.9. It has a * on the ctrl key, and "* Save" under the S key, "* Find" under the F key, and so on. I just learned that "Ctrl-Y" is "Redo". But why is "Undo" "Ctrl-Z" and "Redo" "Ctrl-Y"? Why are they so far apart? The BART example was interesting once again.
Yimin Yao - Sep 24, 2006 04:49:35 pm
Chapter 2: As a neurobiology major, I cant help to think about all the theories on the biological basis of consciousness and memory. I found this reading rather enjoyable, and I agree that the limit of one Locus of Attention at any time can be very harzardous. I remember last semester when I was pondering on a problem as I walked to campus; even though I saw the red light at the intersection, my Locus of Attention was on the problem and it prevented me to respond appropiately to the red light; someone had to drag me out of the street because I was almost hit by a bus. This article also reminded me of the airline accident we talked about on the first day of class where the pilot entered the wrong airport for landing. I've always been doubtful about the claim that there was no confirmation message for such an important input; and now I think the pilot hitting the yes buttom to a confirmation msg due to habituation is more likely to be the case.
I've also noticed that some software designers have tried to avoid habituation by changing the wording and the answers of the question for confirmation windows. When i was installing programs lst week, I clicked on the opposite answer as my intention due to habituation, so I ended up having to read through a lot of the messages more carefully. I actually found that annoying because all i wanted was to move quickly through the installation process. Thus I think the interface designer must weight carefuly on the cruitiality of the decision and desirability of habituation for each confirmation message.
Chapter 3: One of the modal errors I usually make is with language input system. Often times, if I am chatting with different people in different languages, say English and Chinese, and I switch between the two using ctrl+space. So in Chinese mode, the gesture ctrl/space returns English input; while the same gesture returns the Chinese input in English mode. Although the mode is supposely indicated in the toolbar at the lower right corner, but my locus of attention is on the chat window and chat content, i never really think of looking over before what I typed is not in the language I meant to type.
Regarding the aim to build modeless interface. If it means to trade modal errors with many more gestures to memorize or many more keys to manage, I dont think it will improve the overall experience of using the interface; rather a more smart indicator of modes would reduce the frequency of errors would be acceptable to me.
Robert Taylor - Sep 24, 2006 05:53:21 pm
Chapter 2: I strongly agree with a need for users to not even be able to form habits. If they're doing something habitual and not even thinking about it, then the developer has probably added a needless step for the user. This step is just one more thing the user has to click through before they accomplish there task. Obviously I don't think we need to do away with popups completely, or anything of the sort; rather, when it is completely necessary to inform the user, perhaps only do it in a way that requires the user to read what's there. After all, if it's worth making a popup for, it must be very important and worth having the user read. Windows implements something like this for license agreements, such that you have to scroll down before you hit enter. Some might say this is too extreme for a simple error message, but I say it should either be something like this or nothing at all, to prevent habituation. At the very least, different keys for agreeing should be used; enter should not default to "Go Ahead", but rather "Go back". This prevents the user form doing anything stupid. If they "habituate" through a set of windows, hitting enter will only send them back, and perhaps they'll think a bit more about what they're rushing through.
Chapter 3: This reading brings up interesting points about the order in which things are done. However, I think details like these are more something that come naturally from designing and prototyping an interface, and do not necessarily require a whole lot of thought. Over-thinking in designing a process causes it to seem almost unnatural and inefficient. In almost any task, the best performances are done without even really thinking about what one is doing; it comes naturally, subconsciously. I like the author's syntax for shortcut keys as well; although obviously for generic commands, like Control Q to quit, it's not that necessary. Another, perhaps better syntax than the author's, at least for advanced users, is that which is sometimes seen in emacs: C-x-s is "hit x and then s while holding control". We'd also have C-x s for "hit x while holding control, then hit s without holding anything else".This brings up space with certain special keys such as space and -, but those typically aren't use as shortcut commands because of the confusion they cause. As a final note, one-button mouses are NOT a good idea. I'm not sure what the author is thinking here. I'd like to see his Appendix A.
Sean Carr - Sep 23, 2006 11:24:22 pm
Locus: This was a very informative and easy to understand reading about the locus of attention and aspects of the conscious and unconscious. The section about habit forming reminds me of so many incidents from my own life. The one part of this reading which I don't agree with and I wish Raskin had gone into more detail about is when he says that having many ways to do the same thing is bad. I don't agree. To me this means he doesn't approve of shortcut keys or right-click menus. As a "power user" I find this ways to access tasks much more efficient than others, such as navigating menus. However I do not think they should replace menus because they are much harder to learn and do not lend themselves to first time users. Even if the interface has icons for every task I want to do, using the mouse is slower than the keyboard and takes me away from my task of typing this post. Since people interact with computers in varying ways (mouse/keyboard) and have varying levels of experience it makes sense to have access methods tailored to these needs. I personally find interfaces which allow me to set my access methods to be the best. And of those I like ones which allow me to set multiple shortcut keys and mouse options for the same task because depending on exactly what I'm doing any of these various methods may be more efficient or easier to do. At the very least I need ways of accessing items from both the keyboard and mouse because I hate switching between the two.
Modes: This was a good look at the idea of modes. I never had really thought or heard much about modes using this particular definition but it seems to make sense. A problem I see is that there is no specification or widely accepted standard for what gestures do what. There are so many applications developed by so many different designers that it is inevitable that there will be differences in gestures for certain tasks. What if the interface you are designing has an action that is specific to that program or at least very rare so there is not standard for what gesture to use? You can look at all the interfaces you know of to make sure not to pick a gesture that is already taken, but what if another designer goes through the same process for a different rare action and comes up with the same gesture? Then a user using both interfaces will experience two different actions using the same gesture depending on the mode. If though this seems like it will always be possible to happen, having a set of standards for most actions would be very helpful and make many interfaces easier to use.
Randy Hilarbo - Sep 24, 2006 05:51:09 pm
Chapter 2 - It's interesting how this chapter gets your locus of attention into something that you haven't really thought about before. The way it engages me into their examples really prove its point. This article had me thinking that designing a system with an interface similar to existing interfaces may not always be a good thing. Similar interfaces may result to habituation, which may become a problem when it comes to important tasks (deleting a file, etc). I also find it helpful to keep in mind the idea of predicting where the user's locus of attention is (so you can make changes to the system elsewhere) in designing a good interface.
Chapter 3 - Thanks to this article, I was now able to turn off autoformatting in word. Users have different preferences and designers can not always create an interface that would generalize these preferences. So, i think that allowing users to customize their application is a good idea. But this option should be provided so that it's easy to undo/remember/see what you have done. I remember when my niece started messing with my keyboard while Microsoft Word was open. It triggers something on the application that I couldn't undo. I don't like reading the help manual so i just reinstalled the application. Anyway, I think that the noun-verb is very effective since this takes advantage of the user's locus of attention.
Robert Held - Sep 24, 2006 07:13:23 pm
Chapter 2: Raskin makes a point that the the user should be given the ability to undo actions that could have been made by habit. This could include the ability to undelete a file or remove an effect in Photoshop. The feature certainly requires additional planning and coding on the part of the developer, but the benefits are worth the effort. One could essentially consider it a safeguard in case the interface, despite its well-intentioned design, fails to provide the user with clear notice of his or her potential actions. In fact, not including the ability to undo ones actions would be a major detriment to an application and likely lessen its value to the consumer. Most people want to be able to quickly use a program with the confidence that they can go back to an earlier state if need-be. Such a mindset is much more pleasing than constantly worrying about the far-reaching implications of one's actions.
Chapter 3: I especially enjoyed Raskin's point that separate "beginner" and "expert" interfaces for a single program are unnecessary. If a system is designed well from the beginning, then it should be easy to learn and just as powerful as an expert version. It seems like creating novice modes dumb-down programs and reduce the number of tools available to the user. If the tools are simply made more intuitive and labeled efficiently, the novice would then be able to progressively gain familiarity with the same interface. Then, once he/she is an expert, he/she would still use the same tools, but likely more quickly and with greater confidence.
Alex Wallisch - Sep 24, 2006 08:42:49 pm
Chapter 2
I have to say that I agree with Raskin in that being shown the previous task you were working on makes it easier to recall what you were doing. I have a firefox extension (SessionSaver) that saves all the tabs you had open when you closed firefox and opens them the next time you start it up, and I'd say that it's quite possibly the most useful extension on my system. However, there are also some applications for which I'm not so keen about them always remembering what I was up to. When I close my movie player, I generally intend for it to forget about whatever I was watching. It often makes me double take when I start it up days later, go to open whatever media file I wanted to view, and suddenly see the last movie I was watching start up wherever I left it off.
Imagine if every time you started up a shell, it began in the directory you were in when you last closed it, or if every time you started up a video game it brought you straight into the game rather than taking you to the main menu.
Chapter 3
This isn't the first time I've heard the distinction between noun-verb and verb-noun, but Raskin brings an interesting perspective to it. I was reminded when I read it of an article I'd read previously concerning the merits of standard imperative programming (i.e. C) vs. the merits of object oriented programming. By Raskin's claim, the latter would always be more intuitive, as it focuses on choosing the object that is going to be manipulated (the noun) and selecting a method of that object (the verb) rather than choosing the function first (the verb) and then passing arguments (the nouns). As far as I know, however, neither of these programming styles are definitively "better" than the other, which makes me wonder whether Raskin's claim can fully hold up. Raskin admits that there are some places - such as a pallette - where verb-noun makes more sense to the user, and I believe that these are more plentiful than he claims. When I use the command line, for example, it makes far more sense to me to say "cp source dest" than to try to put the source first.
Maksim Lirov - Sep 24, 2006 11:39:55 pm
Chapter 2: This chapter, focusing on the Locus of Attention, points out several properties of the human brain. The example of the "Are you sure you want to delete this?" message shows that it is necessary to consider whether a feature will be useful or annoying during the design. The warning message was put into Windows to prevent the user from deleting desirable files, but many users often answer the same way without thinking. I guess the users can't sue Microsoft for letting them delete desirable files however. After reading the several stories of pilots getting confused by the plane's warning system, I am now more aware that it is often a good idea to try to limit the pilot's locus of attention as much as possible. Finally, I do not quite agree with Raskin's critique of the desktop system. While it is useful to be returned to the last task that you were working on, oftentimes it is desired to go in another direction.
Chapter 3: I disagree with Raskin's assertion that if designers make their interfaces nearly optimal, personalizations can only make the interface worse. No two people are alike, and what is comfortable for me may not be comfortable for someone else. I think personalization allows people to customize the interface according to their habits. It is true that some people customize the interface to make it look good rather than be more useful, but I think in many cases personalization makes it easier to use the application for the user. For example, I personalize the Windows taskbar by placing the quick launch icons in an order that I am comfortable in using.
I think that Raskin's assertion that a modeless interface would be something to strive for is not totally true. Modes allow additional functionality to be added to the system. The text editor vi has many modes of operation, but with some practice many users achieve the same monotony that Raskin talks about.
By the way, is it just me or does Raskin's book feel like an advertisement for his Canon Cat?
Jason Lee - Sep 24, 2006 11:47:01 pm
Chapter 2: Cognetics and the Locus of Attention: The relationship between the locus of attention and habits in our daily lives is interesting to think about. Raskin brings up the point that "once you start a sequence that takes less than 1 or 2 seconds to complete, you will not be able to stop the sequence" and that you cannot interrupt sequences...unless the sequence becomes the locus of your attention." It is a simple fact that seems to be intuitive, but it says quite a lot about the nature of where we tend to pick up our bad habits when it comes to interfaces or how we make simple mistakes simply because we are not thinking about things and on auto-pilot. Because we are so accustomed to performing certain tasks, our minds simply lose focus as the task becomes like second-nature, allowing for imperfections in the way we accomplish things. Though we may concentrate intently on how the task is performed when first learning to become accustomed to some of the nuances or intricacies involved with it, as time passes, slight variations creep in to the way we perform the task, usually making it less efficient. In addition, performing repeated tasks causes it to cease being the locus of attention, which becomes hazardous when there is a small population of tasks within the repeated tasks that need to be treated differntly. Natalie Nguyen brings up the excellent point of confirming which files in our folders we wish to turn in for assignments. For folders that contain a lot of test files or extraneous files that do not need to be turned in, it is exceedingly easy to tune out our attention and type in N for all the files that do not need to be turned in, only to accidentally hit N again for a crucial file for the assignment, requiring me to abort the turn-in process and start all over again. The locus of attention seems to be a very tricky thing to manipulate or take advantage of, as what triggers it seems to differ from person to person. This became very obvious when conducting our group's contextual interviews and asking our interviewees to think out loud as they were performing their tasks, as what people gravitated towards differed from person to person.
Chapter 3: Meanings, Modes, Monotony, and Myth: Though interfaces would be simpler without modes, it is difficult to imagine how some programs would work without the use of modes. Perhaps this is a by-product of me being used to already established interfaces, but in some cases, it seems that modes are required to make efficient use of what the program has to offer (paint programs being an example cited in the text). It seems a bit too simplistic to say that the easiest solution is to get rid of modes altogether. Raskin states that "modes that vanish after a single use cause fewer errors than do those that persist". However, if you wished to perform a lot of a specific task that requires being in that mode, it would be very tedious to continually switch back to that mode after completing each task. For example, if you wanted to create many blue strokes in a paint program, it would be far too tedious to have to constantly switch back to blue after completing each individual stroke. In this case, it is much easier to switch modes to blue strokes and make as many strokes as you need. Perhaps the best solution in using modes is a compromise that allows users to decide if the intended task is to be used only once in the mode (after which it is immediately switched back to the default mode) or if the mode should persist. A good example of this compromise is the shift and caps-lock buttons. The shift is often used if only a single letter needs to be capatalized, such as at beginning of each sentence. However, there are times when all caps letters need to be used, and this instance, given the layout of the keyboard and the methods of typing, it is much simpler to use the caps-lock key.
Interesting tidbit about the locus of attention: a recent UC Irvine study found that the average worker can only focus on a task for about 11 minutes at a time before being interrupted by something, at which point the worker is unable to refocus for another 25 minutes.
Kimberly Lau - Sep 25, 2006 12:42:14 am
Chapter 2 Multitasking is a tricky process. I’ve always believed that in performing multiple tasks, one will execute neither task well, no matter how automatic either task is. For instance, take someone that is talking on the phone while driving. Initially, the public believed the cause for bad driving while on the phone to be that one hand was occupied with holding the phone instead of on the steering wheel. The natural solution was to develop handsfree headsets. However, people continually blame talking on the phone for bad driving whenever they encounter such a driver (even one on a headset). This is because the attention is split between multiple tasks. Raskin claims that multitasking works if a task is automatic, and that driving is an automatic task. I disagree, because even when everything that happens during a drive is predictable (and automatic), as in nobody darting onto the street, no cars cutting into your lane, or other sudden movements that call your conscious to brake immediately, etc, it takes a certain amount of concentration to drive safely. Likewise, it takes much focus to register a phone conversation.
Chapter 3 The shortcut commands people frequently use (Ctrl+) are such regular functions because people have been using these shortcuts since they were initially developed, and have become so accustomed to them that they automatically apply these same commands to other programs. To reapply these functions to new systems would be effective, since users already have these commands engrained in themselves and can easily adapt them with newer programs. As described in Chapter 2, people learn from repetition. After using these commands so repeatedly, they are bound to become habit.
Melissa Jiang - Sep 25, 2006 01:10:29 am
Chap 2: Locus of attention is indeed fascination. I remember a Cog sci professor of mine showed the whole class a video where we were told to watch two teams--one in black and one in white--and count how many times each team threw the ball and how many times each team bounced the ball. The task itself is pretty difficult because they are throwing and bouncing the balls rather fast. However, the most interesting thing is that while you are focusing on counting, a person in a gorilla costume walks right in the middle, pounds on their chest and then walks away. 3/4 of the whole class did not see the gorilla!
This article also reminded me of the typical saying regarding how humans only use 10% of their brain. Many have argued that we should be able to use 100% of our brain but honestly, why would anyone want to. That's just sheer exhaustion on your brain and the locus of attention can possibly just be an evolutionary advantage so that we'd function efficiently. Of course, some believe that if we are able to utilize 100% of our brain then we would have magical powers or ESP...but that's just all speculation (or is it...).
Chap 3: I actually do really like shortcuts and found that many times, it is easier simply doing commands with the keyboard than switching back and forth between the keyboard and mouse. However, I do realize that many people are different and therefore, they may take to a particular UI differently. Regardless of how good the UI is, I still think there will still be those users out there who are picky enough to personalize the interface to their likings. One size rarely fits all. In fact, if anyone knows of anything at all that fits everyone, I'm all ears.
Michael Moeng - Sep 25, 2006 01:08:00 am
Chapter 2: The point made by Raskin that I would most strongly disagree with is that developing habits is good for all interfaces, especially with driving. As with the deleting files example, people tend to skip repetitive steps, even when those steps include, say, stopping completely at a stop sign. He does qualify the statement by saying that habits developed from prolonged use of an interface should not "cause problems for the user." This is not quite the same as habits that "smooth the flow of [the users] work" -- especially in an example where inattention can be dangerous. For most user interfaces, especially commercial computer-based user interfaces, this completely applies. Users want to complete their tasks as quickly as possible with the least amount of effort, thats what computers are for, after all.
I'm also not sure about users only being able to consciously focus on one item at a time. A prime counter-example would be computer gamers. A goal of most games is to be able to simultaneously track many tasks at the same time--but these cannot be repetitive enough to slip into the unconsciousness otherwise the game would be boring--so they usually change, and these gamers must react to to different circumstances quickly. The tasks must mostly all be conscious, is playing the entire game one, single task under Raskin's definition, and some people can simply handle a "broader" task?
Chapter 3: I liked the menu displayed in page 58 of the text, where the most recent item of that group was shown, but the bar came out containing every item in a set order, so selecting the most recent as well as picking a different item were both made simple.
I agreed with most of what Raskin had to say about the noun-verb and verb-noun constructions--however, I appreciate the ability to modify items already placed verb-noun style, especially if I need to resize multiple objects, there's no simple way I can think of to do this noun-verb style.
Andrew Tran - Sep 25, 2006 01:28:32 am
Chapter 2: In my opinion i think this article should spend more time talking about pitfalls of interfaces relating to the cognitive conscious and cognitive unconsciousness rather than spending a lot of time on the human mind. But when it did related things such as automaticity and people using computers i found that quite fascinating. Never before have i ever thought about how i do things until now. I guess my cognitive unconsciousness about this topic switched into the consciousness and now my locus of attention is on understanding how i perform tasks that i do automatically. Another great example about automaticity other than Raskin's example of pressing the Y button is just walking or driving to a place. Sometimes i walk to class, and when i get there i don't even realize that i had actually walked, i just asked myself "how i get here?" Raskin mentioned that it was better to have the task the user last performed be the first thing he sees when he turns the computer back on. I disagree with this, what if the user just forgot to close the application? Then i would find it annoying to see the application on again when i turn my computer back on and have to close it.
Chapter 3: I believe a good interface is a simple interface. An application can have many complicated features and still have a good interface if those features that the user will perform can be used simply. I take for example again Photoshop. To me i think their toolbar system is ok, not great. It has too many buttons and other buttons within the buttons. Meaning if i want to use the healing brush tool i might have to click on the little arrow under the color replacement tool to get to it. It would be simpler if these were two different buttons, but the downside to that is people might complain about more buttons. A way to make people happy i believe is to have the toolbar customizable, users can add or delete tools however they want. Raskin said that having an unlimited number of commands help make interfaces modeless, but i dont agree with him. There is just too many features and I think the max keys a user should push for shortcuts is two, any more than that gets confusing. Back to the photoshop example, i don't know of anybody who knows all the shortcuts to this program. I mean there is just too many, with a limited amount users can memorize but with a lot no way. Some photoshop shortcuts require you to push buttons and mouse click at the same time, now thats what i call complicated and bad.
Qingyun Tang - Sep 25, 2006 02:01:28 am
Chapter 2: This chapter has mentioned many things that we do everyday but we have never thought about before. A lot of information is going through our mind every second; however, without thinking about it, we do not realize what is conscious in our mind and what is unconscious in our mind. I used to think unconscious mind takes some time to access and never realize that I could pull out things from unconscious mind so fast like getting the last letter of my first name. The author also mentioned about the persistency of working that leads to do things without thinking, like walking, typing, etc. They are just built-in in our mind; we never have to think about them when doing these activities. Same thing apply to user interface: if we could design a really friendly and unproblematic user interface, people will develop habits of using the system through that interface. Therefore, making an unproblematic user interface will help to avoid crucial operating problems and accidents.
Chapter 3: It is extremely important to let user of the specific interface to clearly know what mode they are in at the moment. I personally have encountered many situations that I don’t know what state I am currently at, thus unable to switch to the mode I want to be at. This will result a small mistake in general, while switching to the wrong mode; it could be very serious if pilot switches to a wrong mode thus results an accident. With a poor user interface design, everything would be more complicated. The author has shown some examples such as unable to set the clock in the radio of car. To let users intuitively know the functions of the buttons is very important. This way, users get familiar with the interface quickly and they will be more than happy to operate with the interface. If for some reason, the interface has to be compact and no extra space could be granted for some extra description, a manual should always be necessary to describe how to check the current mode and the functions in detail.
Tony Yu Tung Lai - Sep 24, 2006 02:15:55 pm
Chap 2:
It was neat to read about Raskin's point of view on warning/error messages. Raskin claimed that because our habitual behavior, the warning messages are virtually useless. Although I do agree with Raskin's point of view on habitual behavior, I believe that error messages are useful in case of accidents. As for me, I tend to click on the 'x' button on the browser when I only wanted to close the current tab, not all of them, and the warning messages helps in those cases. Also, I agree that warning messages are useless for a task that requires a lot of concentration. In fact, I find them extremely annoying, since users would choose to ignore them in those circumstances anyway.
Chap 3:
I am not sure if I agree with Raskin's opinion about customization. In the reading, he implied that customization is more of a burden than benefit to users. But I think that has to do more with the user's ability to customize than whether the customize feature is a mode. Even the best UI designer in the world can't design an interface that every single person would find it to be perfect. By having a a customize feature, users can make minor changes that will make a already great interface to be more suitable to him/her. If the user happen to make changes that turns the UI to be a diaster, that's what the 'set to default' button is
Suthee Chaidaroon - Sep 25, 2006 02:48:31 am
CH2 There is no way to prevent users from forming a behavior like pressing "y" for yes. Although the author suggests the designers to develop some helper functions like Undo, it is still not sufficient enough for some application like auto-pilot. I think "Undo" is probably the most feasible solution to prevent user mistakenly deletes an important data. The new MAC OSX also develop an action history (Time-Machine) in which users can go back (undo) to their previous action.
I like an example of a "Cannon Cat". I have never realized about taking an advantage of a human psychology and its behavior. As I can see in a Cannon Cat example, the developers worked really hard on both machines and human as a result of a very comfortable and fast application.
CH3 I think allowing users to customize the interface is a good idea but troublesome. It is true that the interface designers cannot generalize all users' preferences into one universal interface. Customizable interface is really confusing and complex. It would be difficult for users to memorize all their own-defined shortcut keys. Why don't we just make them memorize the predefined one? That is why interface designers need to study on users' behaviors and tried to create interface that satisfy the majority of users as suggested in CH2. At the end, I think providing multiple modes is not forbidden. There are plenty techniques to minimize errors and confusion.
David Eitan Poll - Sep 25, 2006 02:41:19 am
Chapter 2: I found the analysis of human multitasking in this chapter interesting. I had once heard someone claim that under every circumstance, people are less efficient when attempting to multitask. However, this chapter claims that this is true for conscious processes, but not those that are automatic. This allows us to design interfaces that take advantage of automatic processes common to most people. This would allow users to maintain productivity while gaining the advantages that the application has to offer. At the very least, it keeps the interface from being an encumberance.
Chapter 3: The presentation of state and data is extremely important in design, as implied in this chapter. It's very easy to produce ambiguities by choosing the wrong control to represent a concept. One important point was that the "part of speech" of the represented value is a good indicator of the type of visual cues necessary to portray the state accurately to the user. Choices, adjectives, verbs, etc., can all be represented by controls that people have come to understand and expect through experience. For example, it makes far more intuitive sense to change properties through radio controls or checkboxes than using toggle buttons, since we're used to buttons being representative of actions (verbs).
Michael Mai - Sep 25, 2006 04:48:51 am
Chapter 2
This chapter was well written and enjoyable to read. Understanding attention and brain psychology develops a basic understanding of how to maintain the user’s focus on the interface. I feel that unless you have a fresh mindset everyday, there is really no way to avoid getting into habits and routines to achieve the same daily goal. Compare it to drawing a line between two dots. There are so many ways to draw lines to connect them, but why do a different line everyday when it’s much simpler to just do the same one each time? The message I get from the chapter is that you should think about the habits that the users might have while working with your interface and to see if you can implement ways to steer them in a general helpful focused direction.
Chapter 3
In reference to the GIDs, my G15 Logitech keyboard has the same notations for the pressing keys as described in the article. I believe that allowing people to personalize interfaces only adds distractions and fluff. It takes away from the goal of the interface while allowing for distractions. Also, it requires extra time to code these features. Finally, in terms of the noun and verb argument, I feel most comfortable with the noun-verb method because its easier to remember what you wanted to do with an object that’s highlighted compared to having the action ready and looking for the object.
Rayhan Lal - Sep 25, 2006 05:48:38 am
Chapter 2: Multitasking is not operating serially on several things (the common use of the term), it is doing it all in parallel. (I’d like to see anyone type an essay with one hand, while chatting with friends with the other, simultaneously thinking about both). The reason we can talk on a cell phone and drive is because we’ve switched one of those two to the unconscious. That is why when one is performing any task where life-and-death hang in the balance, at least one operator must be conscious of the process. I love how the single locus of attention was proved, and it warrants repetition. If we could process multiple streams of information at the same time, then we could be drawn into two opposite or discordant directions. Identity, a unified consciousness, is vital to our survival lest we be pushed and pulled by the myriad voices in our head. (As an aside, one problem with allowing users to restore files is the use of disk space. I would be interested if anyone had a solution that would allow for undoing a delete while keeping disk consumption low.)
Chapter 3: As for the key entry syntax, I think that they all can potentially work, but you need to tell the user what the -, + or arrow means. I HATE very modal interfaces <cough> vi <cough>. Consider programming, I think most of us spend more time than necessary coding trivial things because there are so many ways to do it, and there are some expressions that are just convenient syntactic sugar for others. I also agree with Raskin that “the user will, usually, not be a knowledgeable interface designer”. I always end up turning off customized menus in Word because I know I will eventually need to do something I have never tried before. Tools not visible from the onset become even less visible if items are laid out based on use.
Yang Wang - Sep 25, 2006 05:45:08 am
Chapter 2:
Talk about habit, and doing things unconciously. When I was working during summer at a computer company, I take the same route as if I am going to berkeley, and the only difference will be that I will be getting off at a couple exist before the berkeley exit. I only did this for 10 weeks. But soon after I finished my internship, I found myself having trouble getting to berkeley. I have to conciously remind myself to get off at the berkeley exit. Otherwise, if i wasn't paying attention when I am driving, I will find myself getting off at the work exit and go straight to work.
Chapter 3:
Nothing much to talk about, i agree on author's points, especially on the part of differentiating between expert and beginners. Just imagine if my parents and I have to use the same interface, it will be either too hard for one party or just boring for the other.
Roland Carlos - Sep 25, 2006 09:44:49 am
Chapter 2:One of my favorite points in Chapter 2 was Raskin's statement that an action can be both conscious and unconscious (such as the traffic light) based on the situation.
In any case, the chapter seems to reinforce the fact that the main obstacle that interfaces have to face are the natural habits of humans. Even the most perfectly well-thought interfaces can be taken down by the habitual actions of its users (such as the mentioned deletion situation). However, it seems the only way to handle these habits is through extreme measures, such as the "type the tenth word in reverse" scheme mentioned in the reading.
Therefore, we have to use the locus of attention to support our interface. Locus of attention is an interesting concept and is something we wouldn't normally think about until we're presented with it, as in this case. I bet you weren't thinking about the concept of a locus of attention until you read about it, but now that you did, you knew it was always there, just in your unconscious self. It also seems like it is a game of distraction with the users to support our interface so that the user may not become aware of some of the flaws in the design.
Chapter 3: As I said in my comment about Chapter 2, modes are a great idea in theory, but they will be derailed by a person's natural habits. It makes sense though, if we're not paying attention to the mode we're in, then it'll fall out of our conscious and then we won't know again until we make a couple mistakes (usually).
I think the best solution lies more in more distinct symbols for a mode being on or off (like the light for Caps Lock) rather than in forcing the user to learn multiple different commands for a certain function based on the mode it is in. Anything that forces the user to learn more/do more seems like it would be a drawback from the product. But then again, people adapt all the time, so it might not be a lot to ask a user to learn two different commands in the name of avoiding future mistakes.
Edward Karuna - Sep 25, 2006 09:56:21 am
Chapter 2: The author makes a telling point in the fact that we *have* built to ergonomics, that we understand our physical limitations, but it is often true on the opposite side that we do not build for cognetics, the ergonomics of the mind. It is because of this, I think, that when programs try to exploit the habits we form, such as several free programs that switch the positions of the continue and buy now buttons, that it seems so insidious. It is interesting that later on, the author goes and details the work of a magician, whose actions are required to exploit out tendency towards singular points of attention, and which most people don't find insidious. Perhaps it is sort of the stated intent behind the activity that makes the difference. After all, we watch magicians in order to be fooled, while we start up computer programs to use the program.
Chapter 3: Frankly, I find the author's newly invented style of notation to be tedious and confusing to read. Perhaps it's a product of growing up in the 'information age' as it were, or perhaps the very habit-forming tendencies that the author describes in the previous chapter, however, I have no trouble at all reading the initial notation presented. Also, in reference to the first comment's question about the Fkeys, I use them in all the OSes I use, in Linux to switch between virtual terminals using screen, to summon both applications and virtual desktops in Windows, and in Mac OSX for a plethora of things, not the least of which is the amazingly useful Exposé family of functions.
Siu Pang Chu - Sep 25, 2006 10:03:28 am
Chapter 2: Ergonomics mean studies of the people-machine relationships. A good ergonomic design can provide safe, healthy and comfortable environments for human being. Another factor for a good design is Cognetic limitations which can be described as an interaction between knowledge-driven processes and sensory processes. The interface should not require the user to have an unreasonable ability.
In the article, it also mentions about habits that are subconscious, efficient and parallel. So, a good interface should provide a habit-building function. When you perform a task repeatedly, it tends to become easier to do. When we repeated the same process, practice a sport, it will. Also, avoid interrupting it by conscious activity
Chapter 3: In the chapter, the author mention the word monotony. It means a design which is lack of variation. "Gesture g always results in action a." The author believe that both beginners and expert users of the same system require the same interface. That is so true, let say firefox. I have been using firefox for many years. I don't see how I use its UI diffrerently in those years. The button I will click are still that a few buttoms, location bookmark , new tab..... So there is no need for the interface to be too customizabile.
David Hoffman - Sep 25, 2006 10:59:17 am
Chapter 2: Raskin’s most potent point in chapter two refers to habituation. The goal of designing a user interface is to make it highly succeptible to habit formation in such a way that tasks which initially require attention become automatic. He does site one counterexample, being that the confirmation box after a delete become part of the delete process and no longer serves to capture attention from the user. When a computer is asked to shut down it gives several options such as restart, or shutdown, or sleep. However, it puts the default option as whatever the last one was selected. So there have been many times when earlier in the day I had restarted and then when I intended to shutdown later, I accidentally restarted the computer by just clicking the ok button. Raskin also makes the point that people get absorbed in things and do not always acknowledge the problems and aids being offered to them. The take home message of this chapter is that you need to make sure that the processes that should be trivial are easy to learn habits, and things which require full attention are at the locus of attention.
Chapter 3: The main point of this chapter is that users come to expect certain things from an application. Such expectations include that their time devoted to learning certain habits in the program will be honored later and these habits will continue to be useful. This chapter also discusses the problem with modes, in that mainly the mode can cause user error after the user stops thinking about it and also, that setting a mode can be distracting. At the same time, there must be some affordance for allowing users to perform more actions than there are physically buttons. Raskin seems to shut down almost all the ideas for correcting these problems. He doesn’t like elaborate keyboard shortcuts, or giant menu structures or modes. However, his main point is well taken that there is a problem with having a user perform different actions with differing commands, and how to design the commands. What Raskin argues is the worse interface, is when the available resources are diluted with many repetive ways of doing the same thing. Its up to the designer to determine the most effective means of control and then implement it and make the users learn the way. In Microsoft word one example is that there have to be ten different ways of increasing the font size. Instead of being a blessing and making many people happy, everyone resents that Microsoft has wasted resources on making the software more complicated than it needs to be.
Suneet Shah - Sep 25, 2006 10:59:48 am
Chapter 2: I distinctly remember a time when I was about 5, my father was backing up his entire computer to tape. There was a message on the screen that said "cancel this backup". For some reason, I felt the dire need to click on the button stop. Why? Becuase I wanted to see if it would say "Are you sure you want to stop?". Guess what. It didn't. If you've ever used tapes before, you can understand how annoying it is to start a 12 hour tape backup all over again. This was one critical situation where a warning was not included but could have benefitted greatly from one. However there are often times where warnings can become tedious. For example, when trying to delete a single file. Even if you delete it, if you made a mistake, you can always recover it from the Recycle bin. Users do tend to ignore warnings and mindlessly press accept. However there are other certain circumstances where I feel warnings are necessary: e.g. when emptying the recycling bin, when doing something that can not be undone, etc.
Chapter 3: I really agree with the author's point that good interface design should easily lend it self to habitualization. Just think of some of the interfaces that we have built habits to using. E.g., driving. The standard interface of a steering wheel, pedals to stop and go, etc. This is something that we have habitualized. Also think of the keyboard and mouse we all used to type now. This is also a standard interface that lends itself to habitualization. However I do not think vice versa is true. Just because one can become acclimated to a certain design, and form habits, does not make it a great design! But great designs should lead to habitualization. I did not understand the author's obsession with ctrl + keyboard shortcuts. His new notation was more difficult to read, and most users would interpret his "poor notation example" properly. I also don't see how we can strive for modeless interfaces. Other people brought up the point, and the author does too, that there are an unlimited number of commands which can help make modeless interfaces. Perhaps I misread his point, but it seems like the tradeoff for simplicity here is huge and unfavorable towards "modal" interfaces.
Anirudh Vemprala - Sep 25, 2006 11:32:11 am
Chapter 2: Through this chapter, Raskin lays out the psychological and cognitive definitions for cognetics, the cognitive conscious and unconscious as well as the locus of attention. He defines the former as the feature or object that we're actively thinking about. He then proceeds to discuss the benefits and draw backs of habituation and simultaneity in interface design. One of the most enjoyable parts of reading this chapter were all the interface examples he used to illustrative his points on psychology. The one on the Canon Cat was particularly interesting and so was the one relating the persistence of critical information for a duration longer than that of a user's memory decay period.
Chapter 3: In this chapter, Raskin discusses modes, user preferences, monotony and the expert-beginner myth. I had a lot of trouble accepting Raskin's belief that user customizations were a bad thing. I believe its important that users should be able to personalize their work environment (going off metaphors like desktops, we all have different ones). Customizations not only allow users to be more productive (the macro feature in Word, say) but it also allows companies to gauge how to innovate their products. Podcasting, e.g. was born out of users preferences and customizations. Raskin's problem of missing toolbars could easily be solved if the program had a "restore defaults" option.
Utsav Shah - Sep 25, 2006 11:13:28 am
Chapter 2: I had to read this chapter with all my attention to get the message behind it since I don't prefer reading psychological/cognitive scientific articles. It was interesting nonetheless as it taught me things which I've never heard before such as cognitive consciousness, cognitive unconsciousness and locus of attention. I see how an user interface which makes user unconscious or loses his locus of attention can be a bad one. For example, using gaim(chat client) on unix machines in soda loses one's locus of attention in no time, it has no flashing the borders of the window system to inform you about the message you're receiving.
Chapter 3: This was more fun chapter to read. The author talks about modes and uneasiness of using keyboard shortcuts. I personally like using shortcuts because I think it's much more faster than pointing and clicking, especially on a laptop. In fact, I HAVE to use keyboard shortcuts/function keys to control the sound, brightness, eject a CD, etc. on my laptop because it offers no other external buttons to do those tasks. I like the point he made about Verb-Noun vs. Noun-Verb constructions. I've used both and after readint this chapter, I also agree that Noun-Verb is the ideal way. Lastly, he talks about visibility and affordances using BART example which was a good recap of the same example we discussed during class.
Julius Cheng - Sep 25, 2006 11:32:54 am
Chapter 2: I'm very glad to finally able to hear about actual cognitive theories on human attention and habituation. I feel that the other readings mostly give empirical information about human behavior but offer no explanation for it. Whether correct or not, Raskin's application of locus of attention and the contrast between the conscious and the unconscious (cognitive models that are not his own) to interactions with user interfaces offers us a way to provide reasoning for the observations we make ourselves during our user interviews. Merely knowing empirical observations about humans, like "people tend to do this" without applying a cognitive psychological model to it is, to me, a really unsatisfying approach to study user interfaces.
Chapter 3: Whew! That was a long and difficult reading. I had a hard time understanding what a mode was, and why we should seek to eliminate them. Raskin says that "I believe an interface that is both modeless and, insofar as possible, monotonous [...] would be extraordinarily pleasant to use)", but this seems pretty impossible for an moderately complex interface. Would attempting to simplify the interface by elimating modes make it even more non-intuitive and hard to use? I'm pretty confused, so I don't have much else to say about the reading. I hope we cover it in class.
Joe Hart - Sep 25, 2006 11:48:35 am
Chapter2: Raskin brings up some very interesting concepts I had always thought an annoyance but never really directly thought about. Error messages, pop-ups, and etc are all concepts that have been around since the beginning of the desktop metaphor. They may have once had a role in informing the user about the interpretation of the task they are making, but now they seem to be more of an annoyance than information. Through the advent of pop-up advertising, error messages that are too vague to interpret, and pointless confirmationdialog boxes with one option, I know that I just click through them without reading or caring. The whole interface needs a new concept before these issues will really go away.
Chapter3: I somewhat disagree with Raskin on the expert/beginner concept. One of the main reasons I am a user of linux is that the interface is highly customizeable. However, simple commands that have become commonplace should be included as a default config. I know that I have a habit of pressing ctrl-c to copy and ctrl-v to paste. If these shortcuts are not present in an application I am completely lost, slowed down, and struggling to interface with the environment. So a balance must be struck between standard expected tasks and an interface that allows a user to make small adjustments to better suit their needs.
Leo Chen - Sep 25, 2006 12:13:53 pm
Chapter 2: I wonder if our generation has expanded the Locus of attention. I mean, even though, we aren't supposed to we drive and talk on our cells. In this case which one is the automatic action? Driving? Driving requires active awareness. So I guess to make that example fit the Raskin's comments, our locus while driving and talking would be constantly shifting between those two tasks.
Chapter 3:
I'm not a fan of his notation for shortcuts. First of all, he uses non standard characters for his notation. I mean up arrow, down arrow.. People may interpret that as pressing those keys on the keyboard. The + notation has been around long enough that people can decipher what it means. The existance of "expert shortcuts" is quite handy. Shortcuts aren't necessary for everyday work, they just make things easier. Tasks may still be accomplished throught the GUI.
Huangnankun - Sep 25, 2006 12:15:54 pm
Chap 2:Raskin breaks down the human cognitive processes into 2 types, conscious and unconcious.He talks about the difference of the 2 and how our locus of attention and habits affects the formation of the 2 cognitive processes. He then talks about the role of these processes in terms of computer interface design. The example of the error message is a good one. Raskin makes the argument that error messages lose their purpose because users will soon make this an unconcious process and develope a "stock response" to the problem. I do not totally agree with this. Even though confirmation messages are common, they still serve as a reminder to the user that something irreversible is about to happen. Together with visual cues such as the big red warning sign (common on many confirmation diaglogues.) will alert the user of that.
Chap 3:Raskin talks about interfaces and how splitting up a user interface is a bad idea. I do not really agree with this. Adobe photoshop is a good example, a person can for example navigate photoshop by clicking on the icons on the toolbar. But as users get better, the toolbar becomes absolete as using shortcut keys is much faster and more efficient. So the "beginner" interface actually helps the user to learn the "expert" interface through well-designed built in tooltips which tells the user of the hotkeys.
Heung Tai - Sep 25, 2006 12:28:41 pm
Chapter 2: To understand how the brain works is not easy. However, interface designers should really consider the transition of each screen of an interface that doesn't surprise users. Most of the time, the interface is not advertisement, it doesn't need to set the mood of the users. The interface should function what the users expect and provides feedback as much as it can. Able to follow what users expect (unconsciously) is a way to make the interface more user friendly.
Chapter 3: The idea that separates users by beginners and experts is really bad. It doesn't clearly tell what is the difference between the two. A good interface should separate users by what tasks they want to accomplish. For example, dreamweaver has different mode for normal mode and layer mode. This logical separation can reduce confusion and allows the users to focus on more specific tasks. If designers really want to provide further customization, they should provide an interface for saving the settings in View->Toolbar which is in many softwares. A beginner would not bother to change the view and I believe only expert would really need to customize it for speeding up their tasks.
Sung Yi - Sep 25, 2006 12:35:26 pm
Chapter 2:
I think the cognitive consciousness and cognitive unconciousness can be compared with the relationship between DRAM and hard disk drive in a computer. Also, I think that habitualization is not always bad because it allows some space for the users to multitask -- if the UIs that the users interact with is intuitive and easy enough, cognitive unconciousness can come along to help the users to multitask. For example, when driving a car (as mentioned in the lecture), the pedals, audio buttons, air conditioner buttons, volume controls etc are pretty intuitive so that the driver can manipulate them (unconsiouness) while focusing on driving (conciousness). Another example would be when a user carelessly press enter when warning or error messages come up without reading them. If the message is more attention-grabbing, the problem will be solved.
Chapter 3:
For the lock example, where toggle is confusing for users, I think the problem could have been solved with better labels such as "the data is locked/unlocked now" or "press this to lock/unlock the data." Clarity rules over conciseness.
Antonis Mannaris - Sep 25, 2006 12:18:40 pm
My high school math teacher had the most annoying demand. Our proofs needed to comply to a strict format on the paper, and the answer clearly indicated. While this made my work "pretty" it significantly slowed me down, especially in the beginning. I had a habit of writing out my thoughts immediately on paper, wherever. I guess this would be an example of what the article describes as a "bad habit". Further into the semester, I started enjoying the new way we were writing out our proofs. It made everything much clearer, and I had gotten so used to it that it no longer slowed me down. So my habits changed! The reason I mention this is that the Raskin chapter has a great point. In the beginning, a new interface will seem strange and awkward to a user no matter how well it is designed. If however getting used to the interface is fast and easy, then the designer has succeeded. Another important aspect the article mentions is perception and memory. System designers should never assume that once they provide an instruction, the user will later on obey it. It is more likely that the user will forget and act in a way that seems natural to them. For example, to save a document I will first try Ctl-S, which may be the command to close the program.
Regarding modal interfaces I agree that a good interface should require no personalization. Previous postings disagreed with this saying that some level of personalization is good. But what if for some reason I need to use a computer other than my own. If I am so adapt to my own settings, I will be seriously impared in using someone else's settings or even the default. What I do not completely agree on is the idea that dividing an interface between beginner and expert is bad. I think that as long as the two do not intervene, then such division is acceptable. By not intervening I mean that whatever you can do in the beginner interface, you can do in the expert interface in exacly the same way. Essentially, the expert interface simply has more options, not different options.
Ming Huang - Sep 25, 2006 12:21:30 pm
Chapter 2:
The theme of habit developing when using an interface translates into the observation that it is more successful to "plagarize" existing working interfaces, so that the habits developed by old interfaces the user have used can be successfully (and automatically) applied to the new interface without too much mental effort. The new interface thus reduces the cognitive resource the user needs to engage in its operation, because it only require cognitive unconcious acts.
Despite the user confirmation critisism I still believe for certain kinds of users they can be really helpful. Expert users of critical functions, such as system administrators, are trained to take every error dialog seriously. It does not mean that they don't make mistakes (I am looking at you, AOL), but considering that there is little alternative to alert users of actions that are not reversable due to technical reasons, a confirmation is the most effective.
Raskin also mentions that the interface should do a better job to direct users to the last task in progress when they finish the current task. Systems nowadays have the option of returning users to the desktop if there are no other windows/tasks running, in case the usr wants to start new tasks, or the last window the user was working on.
Chapter 3:
I don't quite agree with Raskin's skeptisism on toggles, I believe checkboxes are sometimes more useful than radio buttons. Consider a menu of more than 2 items that you have to select from, say from a list of components in the car that needs to be repaired, in a customer service system in a body shop. When the items needs exceeds one and the outcome of toggling an item has obvious consequences (checked - needs reqairing, cleared - leave alone), a group of check boxes labeled with their items will do well with the task, while you cannot select more than one with a radio buttom group.
A large cause of user errors, besides that the interface is modal, is that the interface lacks standard. Every systems chooses their own way of deciding functions of specific key presses, that a Windows user gets unexpected results when he is using a Mac. Sofware like those open source projects in Windows place and draw their buttons and dialogs inconsistently that it destroys the look and feel (in other words, long-developed user habits) of the system and thus make the interface less efficient and more prone to errors.
Vahe Oughourlian - Sep 25, 2006 11:19:57 am
Raskin - Chapter 2
I appreciate two points that Raskin brings up at the beginning of this chapter: the level of detail one requires in the study of an idea, and the commitment of one to his or her metaphor. It's a tricky line to define, but he puts it very plainly; we need not know the physical details of the brain to study its reactions through empirical data collected. Of course, some of the details may aid in understanding, but, at first glance, we should not let our lack of knowledge of the brain keep us from trying to understand the mental processes involved in users utilizing interfaces. The attachment to the metaphor is something designers tend to do, and it hurts because they often use the metaphor as a basis for their design, which could end up hurting them since the metaphor may be flawed.
Raskin - Chapter 3
While I agree that modes are made confusing by a lack of information regarding the modes, modes are part of what make computers so useful and how devices with limited input methods (i.e. - cellular phones) are able to be as versatile as possible. One suggestion about indicating modes (that I have seen in video games, specifically Chrono Cross and Chrono Trigger, but never in commercial software) is to indicate the mode by changing the color of the user interface (and, on occasion, the tint of the screen). That way, the color is always in the user's locus of attention; there is no need for the user to move their attention from what they are looking at, since what they are looking at is indicating the mode.
The point about radio buttons is also well taken. I believe the lack of radio buttons is partly due to the "of course" factor that programmers often associate with the default option or mode in their program. Options using radio buttons, however, may end up getting too long for one particular option. However, at least this idea would end up correcting the error where, when a toggle is clicked, then clicked again, the mode does not return to the default mode, but gets "stuck" in the toggle mode because the programmer forgot to detect the uncheck and set the default back.
Eric Vacca - Sep 25, 2006 11:32:04 am
Raskin begins with several mind experiments diffentiating between the conscious and unconscious mind. Using this idea in combination with the idea that humans have a singular notion of self, and that (most) see themselves as the same person no matter the task, he build a cognitive approach to user interface design. Many of the examples he brings up (automation, absorbtion etc), are things we have all experienced, and from there he builds definitions (Modes, gestures, chunking), and illucidates many of the underlying problems with user interfaces.
I found this book very well written, and lays out, even for a lay person, what designers should be thinking about when making interfaces. I especially found the later sections of chapter 3 that gave tips to make menial computer tasks automated using quasimodal interfaces. The main concern is how can you strike the correct balance between designing an interface for the novice who is learning, and one fo the expert who needs to get stuff done. Is it possible to do both without sacrificing the other? Whenever i run into modes called "beginner" and "expert" i'm almost always choose the expert mode to start because i do not want to learn functionality after i have already learned all the functions.
Yen Pai - Sep 25, 2006 12:38:01 pm
Ch2: The most interesting thing in this piece was the focus on the cognitive processes and the consequences for UI features that we take for granted like confirmation and error messages. The differentiation between habitual actions and actions driven by the locus-of-attention is very well taken and explains quite a bit about why traditional implementations of error and confirmation messages are ineffective. Another possible solution for confirmation messages is to only use them when an action appears to be out of bounds with the user's regular usage pattern, and thus breaking the habit; however, this comes at a cost of making the interface less consistent, possibly increasing user confusion.
Ch3: The topic of modes and monotony is rich and in the text one can find and think of many, many examples of how modal interface design plagues us. For me a particularly annoying trait of many software applications has always been the desire to allow a user to overcustomize. Examples of overcustomizability are widespread but the Linux desktop comes to mind because on sites such as Slashdot (http://www.slashdot.com), a common lament is the lack of Linux adoption as a desktop OS. In particular, I find the implementation of many Linux desktop packages guilty of this tendency, as well as being guilty of presuming a pronounced expert/beginner gap. Linux on the desktop will never succeed unless it reduces the modes created by an overemphasis on user customizability and more on productivity. Users have enough to worry about (learning another OS UI) than to have to deal with different "themes" that drastically change the interface. With Ubuntu's distribution of Linux, you see some of this happening.
Tom McClure - Sep 25, 2006 12:56:11 pm
Chapter 2
Fascinating topic. I take issue with the assertion that offering multiple ways to accomplish something causes problems. To the contrary, some ways of doing things are intuitive to some, and some ways are more intuitive to others. By offering both, the user that is uncomfortable doing one or the other will not be slowed down. Whichever one is naturally picked will become the habitual way for that user to do that task, and they will not thereafter be consciously distracted by the availability of the other method (they may never even discover its existence).
Simon Tan - Sep 25, 2006 12:16:33 pm
Chapter 2
So the author is recommending that, instead of having confirmation messages to signal if something dangerous is about to happen, interfaces should simply make everything reversible so that such messages are never needed. (2.3.3) The example was in the context of deleting files, so my mind first turned to Apple's upcoming "Time Machine" feature, or Microsoft's Shadow Copy feature already existent in Windows Server 2003. However, both these solutions require an extra hard drive to be attached in order to function, which makes me feel that it is a lot of trouble and cost to have this confirmation-free environment set up. Is it worth it to be able to restore files one at a time, or would it be easier to just do regular backups as we are all told to do? Also, would such a policy of no-confirmation work well in cases when malware is needing Admin-level privileges to do bad things to your system in upcoming Windows Vista?
I agree that confirmation messages fall mostly to habit, though. I have accidentally lost something through my Recycle Bin because I just never think when I empty it. I would rather it show me a preview of every file I'm about to permanently lose, as a more visual confirmation (in addition to being a confirmation that is always different each time I try the task of emptying the Bin).
Chapter 3
I like the "radio buttons instead of a check box" concept the author introduced. However, such an interface can still go bad if there are options that aren't listed but can happen if a user picks particular options.
After reading this, I am convinced that modal interfaces are typically a bad thing, even if they do reduce the button count. On a keyboard however, a modal interface is almost required; otherwise, there would just be too many buttons to fit plausibly on a board. Where the threshold is between simplicity and usability is still fuzzy for me.
I find it very humourous that the author included the BART ticket-dispensing machine as an example of a bad interface. I use them all the time, and never experienced the interface he described - rather, I believe that the BART authority read this author's book because the way they are right now is very close to his description of how they should be.
Siyan Wang - Sep 25, 2006 12:41:51 pm
Cognetics: It seems to me that one can never avoid the automaticity induced by confirmations. Even with his very draconian confirmation method, I think I would eventually habituate myself to look to the very end to find which number word to spell backwards, and then count towards it. Even though this may be more of a pain than simply typing Y, I don't think it would make me stop and think about it. Thus the fine line seems like it would be very hard to walk, and that, like Raskin said, actions should either be reversible or not possible. However, this is a bit harder for the interface developpers.
Meanings, Modes: Once again, this book touches on many common sense ideas that seem to be just absent from the design of many interfaces. It seems obvious that modes and a constant changing thereof would be a source of many errors from the user, but many applications still adopt this method. Perhaps they have many different functions and thus would like to ease the burden on the user to remember all those buttons, and thus change the modes to allow the same button to do many different functions. However, in the end this would most likely confuse the user.
Michael Udaltsov - Sep 25, 2006 12:58:38 pm
Chapter 2 - I agree with the statement that user habits should serve as the basis of interface design. This will help prevent critical mistakes, such as automatically pressing Yes to delete files without verifying the confirmation message. One of the suggested solutions is to allow users to undo erroneous commands. A lot of applications use it already (Word, Photoshop, etc), however, some critical applications don't - for example OS file management and database systems. Most of the time when you delete a file, it goes to the 'Recycle Bin' or 'Trash.' Once it's deleted from there, it's very difficult to get it back, unless there is a backup system behind it that stores the deleted data. Same is true for a lot of database systems. Deleting records, or dropping entire tables can often be done with a simple statement, but recovering without a backup is not possible. Because of this most database systems rely on continuous backups that require a lot of work to properly restore and synchronize with the rest of the system. A better solution would be to integrate the backup and allow undo functionality for deletion in both of these cases. Sure, there are additional concerns with performance and disk space if deletion doesn't actually happen, but that's another concern. Disk space has grown greatly in the past few years, and shouldn't be a problem. Performance may suffer in databases, but if they are modified to ignore records that are used for backup only, it shouldn't be a problem.
Bryce Lee - Sep 25, 2006 12:52:07 pm
Chapter 2: Jeff Raskin's focus on the locus of human thinking and the transfer of thought is essential for designing programs which are compatible for human beings. The power and ability of computers in the past has led to a development based on features and capabilities - the human ability to harness these tools was assumed. A perfect example of this is the advances in multi processing. The modern computer allows multiple tasks to run in parallel, while a human can only keep one of these processes conscious. The computer may be able to context switch within quickly, but the the equivalent human operation leads to a break in productivity and thus actually hurts the user work flow.
Chapter 3: The issue with modes arises from the need to have some sense of state. In this chapter, Jeff Raskin actually seems to contradict himself. Earlier, in chapter 2, he praises the ability of a computer to resume with all settings the same as when it was left to preserve the conscious locus. However, such setup would also entail a certain amount of state be remembered, which are yet further modes which can be easily forgotten and error-prone. However, I do agree with his example about the remote controls - in a more common scenario, the inverted settings on game controllers makes it difficult to control, due to learned habits which force a certain set of opposite actions.
Dexter Lau - Sep 25, 2006 01:08:13 pm
Ch2: One of the main points of the reading is the locus of attention where the user is busy concentrating on one thing, and not another. This can be easily analogous to a web designer wanting to place a table and text in a certain spot (e.g. drag and dropping it to where they want it). Their locus of attention is on just that, making the page and not fiddling around with the buttons and going through tedius dialogue boxes in order to complete their various steps. This problem gets a bit more hectic as functionality increases and going back and forth between the nitty-gritty and what's supposed to be in your locus of attention causes a loss of time and trying to get back in the train of thought of your original work.
Ch3: This chapter deals with the cognitive link between the user and their interface. One example they use is the flight buttons that have dynamic relabeling abilities. This means that depending on what type of mode the pilot is in, the buttons would change accordingly. This seems great in that there are a lot of automation in the types of buttons used and the number of buttons that need to be in the cockpit. The big caveat however is knowing where the buttons are. In the example, a pilot would have trouble getting to their manual control and it would be disasterous if they had to go through several screens just get back to the right keypad they need to hit the "manual flight" button.
Andrew Hao - Sep 25, 2006 01:22:53 pm
Chapter 2 Raskin deals with the cognitive concious and unconscious, and makes the observation that their distinctive properties (that the conscious deals with the new, the novel, the immediate and the important while the unconscious deals with the repetitive and the expected) have applications to interfaces. His claim that I do not have control over my locus of attention makes me uncomfortable. Whether I like it or not, my attention is out of my control. Interface designers can take advantage of this by allowing those events in the locus to be taken over by the unconscious.
Chapter 3 From experience, the semantics used to indicate progress or state is incredibly important. Oftentimes I will be uncertain whether a certain link on a webpage is an indicator of status or is an action indicator. I am unsure of what 'mode' I am in, whether my actions will bring about information or if my action will bring a change of state.
Eric Yoon - Sep 25, 2006 06:30:14 pm
Chapter 2. It's an intriguing approach, to think of one goal of interface design as the effort to create a habit in the user, such that a certain repetitive action can be easily and quickly executed, almost by the unconscious mind. That is a very powerful concept that applies to some extremely complicated software tools. I would imagine one essential part of such an approach is to create shortcuts for various activities, make those activities capable of being committed to memory, and to create a smooth means of chaining the corresponding commands together. This is expressed no better than in the hands of a graphic designer who has mastered Photoshop -- I have seen it, and it isamazing to watch. From the observer's standpoint, it is itself an art form, to watch the designer apply countless alterations, from texture to shadow to light and transparency, to the same image in the blink of an eye. Such systems allow the users to think about what's more important -- in the designer's case, the aesthetic appeal of each iteration -- rather than get lost in the interface.
Chapter 3. I don't think I agree with the author that having separate modes for "experts" and "beginners" is a bad idea. I use tons of different kinds of software every year, and a great many of them have very complex operations that, I am sure, are extremely entertaining for advanced users. But many times, I want to use the program for just one simple operation. I don't want to have to search all around trying to figure out how to do that -- I just want one big button to press. Hence some software programs have "Wizard" modes that allow you to walk through easy steps to specific goals. But it's very hard for me to imagine that all tools could have interfaces that are open to both experts and novices alike -- in some interfaces, a novice wants just one button, but how could any expert tolerate that?
Jonathan Yen - Sep 30, 2006 03:04:57 am
Chapter 2: I think the idea behind cognetics and locus of attention is very interesting. I wonder how exactly scientific studies could be conducted that would evaluate the amount of time spent in a certain area of attention. Raskin makes some very good observations about the conscious and the unconscious, and it seems like further investigation into how unconscious processes work could potentially help user interface designers figure out how to make their interfaces easier to learn how to use.
Raskin's discussion about the single locus of attention gives a lot of insight into how to consider the development of user interfaces. In particular, the fact that we have just one locus of attention enables designers to enhance the user interface by focusing on one specific aspect of the interface that the user is going to interact with, and thereby exploit this limitation. If this was taken into consideration in designing interfaces, it would likely make things much quicker and seemingly more convenient for the user, provided that the interface can keep the user in that particular area of attention.
Chapter 3: Personally, I don't really feel that modes are significant problem in user interfaces. I think the problems with modes vary from among different users. Though people may be bound to make errors in using modes, I think it isn't difficult to learn to adjust. Furthermore, it seems rather natural to make errors from time to time, so long as it doesn't occur frequently. The examples that Raskin gives are anecdotal, and it doesn't necessarily mean that modes are bad for all users.
Raskin's discussion of the beginner-expert dichotomy towards the end of the reading is quite thought-provoking. I definitely agree that it is awkward to separate users between beginners and experts, even though I often believe that it is best to design an interface that is tailored to beginners' and experts' needs accordingly. Like Raskin observes, there aren't two separate poles, but rather different levels of knowledge of the different features of the system. I think this is a much better way to understand the evident differences and working habits among different users.
Chen Chang - Oct 01, 2006 08:12:24 pm
Raskin Chapter 2: Comment 1 - The reading stated that "the ideal human interface would reduce the interface component of a user's work to benign habituation as many of the problems that make products difficult and unpleasant to use are caused by human-machine design that fails to take into account the helpful and injurous properties of habit formation". Habit formation allowed me to think about confirmation steps in the current era, I personally think they should be rid of because I figure the user's mind is already made up when making a decision and often times will just mindlessly accept a popup dialog box without much further ado. For example, when it comes to software installations and you get to the screen where you must check the box at the bottom to agree to the terms and conditions of usage before allowing you to proceed to the next step, who seriously takes time to read all that?
Raskin Chapter 2: Comment 2 - Thanks to this article, I have now found somebody else who has trouble with the automatic "features" in MS Word. I completely understand Raskin's troubles when he gets frustrated that pressing command-T yielded a paragraph format rather than the intedned unitalicize. From my personal background, I've always turned off autoformatting and spell check in word because I absolutely hate the green and red lines that come up under each misspelled word or possibly grammar mistake and would much rather run the check at the end of my document manually. After all, its not Christmas so I say save the green and red for later because it changes my locus of attention and distracts my thought process as I'm thinking of what to type in an essay or lab report or etc.
Raskin Chapter 3: Comment 1 - I stand by the author in his BART example of it being a horrible interface as even those familiar with machines get baffled at the unnatural sequence of operations that they require. The slots and buttons are too widely separated and not labeled well and the lcd display is too high up so its hard for shorter people to see, but most importantly there is lag in response time after the user pushes a button so the user may habitually press it again and it again until you realize you way overdid the selection and have to start over again. My first BART experience was most frustrating of all because I heard the PA announcer say "Fremont train in 2 minutes" so I figured that to be plenty of time to pay for my ticket and hop aboard the train. Well I was stuck at the machine for a good five minutes and consequently had to wait for the next train. My problem arose in using a credit card for payment as the machine defaults to a $20 bart ticket and I only want one for $3.85 as thats the one way fare to Fremont station. It took forever to get the buttons to respond to lower the cost down to $3.85 as each button press only decrements one dollar so thats about 19 button presses including the one for decrementing a nickel each time.
Raskin Chapter 3: Comment 2 - I agree with what most others have said about function keys on a computer, the main question is do you ever use all of them? I think especially with desktop computer keyboards, many of the F1-F12 keys can be eliminated because I myself don't even know what they do (and guess that most others are in the same boat). I figure you can keep F1 for help and F5 for refresh and of course F4 to use in conjunction with alt to close a program, but thats about it in my experience. As for laptops, I can understand the use for all the function keys because they must have them to adjust brightness and contrast settings as well as optional wifi on/off and standby/hibernate controls. I think that shortcuts are useful, ie alt+f4 and ctrl+z and ctrl+x and ctrl+c and ctrl+v simplify life especially in those pesky programs where you don't get the program title bar to select the copy and paste functions from the drop down file menu (so knowing the shortcut in advance can save you).
Aleksandr (Sasha) Ashpis - Oct 02, 2006 10:48:19 am
Chapter 2
1) The article makes a good point that throughout history a model has always been made to compare to human thinking, to which this model was always disproved when the technology advanced, but with this logic, eventually the model will be so complex that it will match humans. If it were to match humans one day, then that day plus x days would make the machines smarter than us. The only significant variable that we cannot calculate is; how far away is the day machines match humans, it might be on the same scale as solving chess. The article mentions a comparison to the brain’s conscious and unconscious to a system of pointers, if this is true, “how much space does a pointer take up” and what is the hard drive space of our brain.
2) This part on habit forming rings loud and clear in my head. I have been a nail bitter since as long as I can remember, and although I don’t like the habit, I cannot stop it, because every time I consciously think about it, I can stop, but soon as I relax I realize that I’m biting again. There is no quick fix, but a time consuming process. According to the book there are no true multitaskers, except when the other tasks are so well learned they can be run on autopilot, thus convincing me that any person that a “good multitasker” is just someone that has taken a lot of time to autopilot a lot of daily activities. If humans only have one locus of attention, and we can switch back and forth, then why do our context switches loose information. Meaning u switch from one locus of attention to another and then back, but you don’t have all the information in your conscious mind anymore, can this trait be improved or automated so no information is lost, so that humans can context switch like good operating systems?
Chapter 3
1) Who hasn’t confused instructions or directions, but the key stroking example provided in the chapter is not a complete solution either. The problem arises with the fact to a relatively short combination of keys one must decode the formula, and if not understood read a paragraph plus that explains in detail what the one line cryptic means. Although the author makes good points he violates his own rule in the example of resetting parameter RAM, where after describing in depth the new approach of how to get rid of ambiguities in commands, he adds his own shortcut of grouping up arrows, without noting which keys, together to symbolize keys being released. Honestly was confusing at first, then I figured it out by reading the paragraph description that followed, I don’t think there needs to be a paragraph to describe such a simple operation. In addition, what if English is not a first language for users, the point of a diagram of drawing is that it removes almost all cultural tendencies from a language, but now the diagram is so confusing that it brings back the very thing that it is designed to get rid off, sounds a bit silly to me.
2) As clearly demonstrated, modes are a classic catch 22; modes are there for the simplicity of using a machine, but occasionally cause hard ache. Much research has gone into proving certain aspects of modes and human machine interaction, but the only real solution to remove all ambiguities is to follow the number 1 way to minimize mode errors, as stated by Norman (1983), Do not have modes. This again presents a problem of its own, how will we limit our machines if modes are gone. My approach to finding a solution is tiered system of modes, meaning make computer or any personal machine use like any other new activity humans learn, have lessons/levels. Starting in the beginning and moving up or down if deemed necessary, put the user in charge of what level they think they are on. To further explain what I mean by levels, for new user which would be the lowest level, have no modes and simply anything and everything to basic options and operations, because a new user does a very limited set of things. As the user gets better and more acquitted, expose more complexity while still have good help and example menus. The help and example would be the difference then what is currently done by Windows, where key system folders are hidden, and is presumed a beginner would not be able to turn on the option of seeing hidden folders. This situation is not always true, because a curious person will type it in to Google and problem solved, but once hidden folders are visible no instructions or further restraint is there to prevent a user from doing damage to the OS or any other application inadvertently. But like everything else in life, there is no black and white, its always shades of gray, and thus computers/machines have to strive to adhere to the ergonomics approach, of providing ease of use and comfort to 95% of its users, and not concentrate on the extremes.
CharlesLeung - Oct 02, 2006 11:56:35 am
ch 2) I think the discussion about habits and how if the pedals on the car were reversed then the user would most likely have problems driving more than a few blocks. Right after reading this passage I thought of my recent experience with someone macintosh computer. On the surface, it seems like just like another computer, keyboard mouse and everything. However, after a few minutes I realized that there was no right mouse button and that I could not just follow my old habits. I guess you could say that using that computer was not in my cognitive unconscious.
ch 2) The passage about confirming intent was also very interesting because it brought up the fact that if a user has to write down too many things, then they will start just using some recycled stock answers. I kind of hate to admit this but with my software firewall that's installed on my PC at home, I sometimes get lazy and use the stock answer of "Permit" even though I don't really carefully look at what's going on. This is especially the case if I'm in the middle of playing some sort of comptuer game.
ch 3) The example of the Fluke scopemeter made the think that additional buttons, lights, and other I/O things actually help to confuse the user. If we think of an iPod which has a pretty good UI, which is in fact much better than other mp3 players I've tried, there are very few buttons, screens, and other things in which to confuse a user. This may just be a tradeoff between the difficulty of learning a new UI and how quickly you want to do things.
ch 3) It seems that although yes the BART machines are very horrible, that they shouldn't really make that big of a deal in one's life. If we look at most BART riders, they probably take BART very regularly (like going to work) and the process in which they buy tickets has migrated into their cognitive unconscious. So even though the system is very confusing, I don't think that the BART machines would pose too large a problem for customers because using the machines would simply become habit over time.
Jae Chang - Oct 02, 2006 12:45:52 pm
Chapter 2: Comment 1
Raskin divided human cognitive processes into 2 categories, cognitive consciousness and cognitive unconsciousness. I think, in the terms of computer, the cognitive consciousness can be considered as main memory such as DRAM, and the cognitive unconsciousness can be considered as secondary memory such as Hard Drive; secondary memory stores information permanently, and main memory stores information when information is process. Also, the singularity of the locus of the attention, only one process can be performed at a time, is similar to multitasking OS with single CPU. If the UI is very intuitive enough so that connecting between consciousness and unconsciousness happens so fast, users will be very comfortable to perform multitask.
Chapter 2: Comment 2
Understanding human brain is not a simple task and cannot be achieved shortly. However, understanding human brain is very crucial to design human interface. As the error message example, randomized confirmation interface can avoid users to press enter as their habits. The habit formation can be the two sided sword; it provides consistence interfaces so that users can work comfortable, but the consistence interfaces may lead users ignore importance message by their careless habit. Other may think that users are responsible for the careless confirmation, but I believe that UI should have a way to capture users’ attentions for error message by providing comfortable environment at the same time.
Chapter 3: Comment 1
Raskin mentioned about separating the UIs for beginners and experts. I think this is a very good idea as long as the UI's default is set for the beginners' usage and other features are customizable. The UIs for experts does not necessarily mean it's harder to use. For example, the computer experts still should feel more comfortable using Windows XP or Mac OSX, rather than using a command line prompt. The only thing, I assume, that experts feel uncomfortable using a beginners' UI would be that it is tediously detailed and of step-by-step. If these could be eliminated by setting shortcuts, hot-keys, and customizing buttons, the separating the UIs for different users would be no problem.
Chapter 3: Comment 2
When I started to use Emacs at the first time, it was too hard to use because Emacs does not have many intuitive interfaces so that beginner can use the editor comfortably; beginners expect universal user interfaces. I agree that Emacs are powerful and faster to edit files for experts. However, the lack of expected user interfaces makes beginners to take a lot of time to get used to it even if Emacs supports customizable commends. I believe separating the UIs for beginners and experts is crucial and will provide comfortable environment and expandability at the same time.
Charles Lee - Oct 02, 2006 12:48:24 pm
Chapter 2: Excessive confirmation is not only useless and annoying, but potentially dangerous, as it numbs the user to all other confirmations. It eventually becomes second-nature for the users to just click "OK", even when they are actually making a mistake. More than one confirmation for the same action leads to these automated actions spectacularly; I know several people who just press "enter" twice immediately, when a confirmation is expected.
Chapter 2: A comical topic is always a yes/no dialog instead of a confirmation. There are countless screenshots online of silly Windows or Macintosh error messages somewhat equivalent to "Something you do not want to happen, is about to. Please press yes to continue, cancel to continue, or no to pop up the same dialog so that you can choose one of the above, and then continue."
Chapter 3: Expert shortcuts are important to users as well. While it is not a good idea to overwhelm the user with too many features to process, it is also not a good idea to provide too few features to do a task properly. A nice in-between medium is to have extra features quickly available to users, but not directly visible. The important part is that these tasks are able to be completed quickly, so shortcuts are perfect.
Chapter 3: Although the "Ctrl+X" format might have been harsher to users a long time ago, they are now so generally accepted that it would be a waste of user learning to not implement. What was once possibly strange is now mainstream, and eliminating these shortcuts would actually hurt users, since many expect them to be there. Standardization would help users more than multiple new ways to do the same things. That way, users will not be caught attempting to use a feature that did not exist, or worse, inducing the wrong feature with an previously memorized keystroke.
Keenahn Jung - Oct 02, 2006 01:03:53 pm
Chapter 2: I especially hate the requirement of "agreeing to the terms" of the EULA for some piece of software you're installing. This is such a ridiculous step in an installation process. How many people, having purchased a piece of software, and in the process of installing it, decided after reading the EULA, that they no longer wish to do so??? This step, consisting usually of checking a box and hitting "ok," sometimes requiring you to scroll through the EULA, is required for installation, but is not automatic. Thus, this is highly annoying and definitely just makes me skip reading any EULA. This is definitely dangerous, as the company could potentially put anything in there, and I would have to agree if I wanted to install.
Error messages can be useful, but only if they provide the right amount of information. Users do not wish to be inundated with cryptic messages about memory locations and other things. They want to be alerted to things they can do to fix the problem. I often find myself googling error codes or error messages to find fixes. This could be a useful feature if this was built it. Something like "error xyz has occurred, click here to google it!"
Chapter 3: Again we have the issue of how much information is too much information? I personally like to twiddle knobs and move sliders as much as possible, but not everyone is like me. Mysql server has something on the order of 500 adjustable fields in its config files. I am happy to spend a day adjusting these to maximize performance, but then again normal users will probably not be digging through config files. This is why the invention of "right click" is so great. When you're using a GUI and you wish to dig deeper, right clicking on something is a great way to expose functionality that is normally hidden.
When you create a new version of a piece of software, customers want it to look different enough that they feel that it is worth upgrading, but they don't want it to be so different that they have to retrain their employees on how to use it. Thus, I believe that software shortcuts should be backwards compatible. Take for example the vi text editor. Most people would find this system quite archaic and strange, yet I spent all summer using it because this was the preferred text editor of the people at my workplace and had been for years. Of course, there are improvements such as vim, which support more notepad-like or ms word-like functionality, but keep all the old shortcuts and functionality intact. I think this is a good example of an upgraded version of a piece of ancient software that allows legacy users to keep doing what they've been doing for years, while also allowing the new generation of users to gradually learn it.
Cheng-Lun Yang - Dec 07, 2006 03:59:20 pm
Ch.2
1.Lots of the user interfaces this days have poor designs. Just as last night when I was submitting my homework for this class, the prompt asked me whether X file is what you want to submit. I clicked yes. Then the second one pops up. I clicked yes again. After a couple of this prompts, I got lazy and started to type yes to all the prompts. It is an example of me unconsciously answering the prompt.
2.I think randomize accept/reject buttons is a good idea. I have seen a few such idea applied in some software such as winzip. However, if the interface randomly mix up the response buttons, there should be a fall back option in case the user clicks yes when he intended to click no. This will let the user shift to the conscious mind and think twice about his choice.
Ch.3
1.The modeless model suggested by having unlimited number of commands doesn’t sound like a good idea to me. The more commands there is, the harder it is for users to get used to the system. For example, after 4 years at Cal and hundreds of hours using Solaris systems for projects, I still have hard time remembering all the commands. One reason that Microsoft software is so widely accepted is because there are fewer commands to memorize, which is easier for people to pick up.
2.The author’s point on separation of beginner and expert interface is definitely true. Throughout the course, we learned that user interface should be easy to use for all people from different target group. Having separate interface according to the expertise on the subject is contradicting to the theory of user interface.
Robin Franco - Dec 15, 2006 12:00:56 pm
Cognetics 1: The human brain has evolved over human existence to be able to process massive amounts of data. As much as we believe we are conscience and intelligent creatures, there are still limitations to the processing power of our brains. To account for this, our brains have also created a way of dealing with the flood of information, namely, ignorance. Our brain selectively choses what information to ignore, and at times fills it in with its own expectations. This reminds me of two pictures I once saw. The object of the task was to see what was different from the two pictures. It showed an airplane with passengers getting on. For the life of me, I was unable to determine the difference. I was looking at the passengers' hats, shows, luggage, everything. But it turns out that the difference was in the plane: it was missing one of it's engines. My brain selectively chose to fill in that missing data with its own information. It took the easier route of looking for more commons differences, which was a fallacy.
Cognetics 2: The concept of having a “Are you sure?” dialog box seems to have slowly evolved to being worthless. Currently, Windows Explorer pops up a warning at certain times when either sending passwords over an unencrypted connection or whenever a site tries to run ActiveX controls. At first, I could see how a user would read the warnings and selectively chose yes or no. But the dialog box now appears to often. The user just wants to see the website so is therefore accustomed to clicking anything that will allow them to see the site. So when they are actually face-to-face with a situation where it would be best to reject the website, they no longer pay heed to the warning and choose to accept. So this negates the whole reason for the warning in the first place.
Modes 1: Whenever I think about an interface that seems a little too “clunky”, I try to see how it would work in the real-world. And the concept of modes is one that I can never find a real comparison in our regular world. By regular world, I mean nature. Since our brains involved to be efficient in a natural environment, one would expect that our user interfaces would take heed to this but it's not the case. One example is modes. In nature, a rock is a rock. It doesn't enter various modes. This allows us to quickly be able to use a rock for whichever purpose without having to do much thinking. In a dangerous situation where I must protect myself, if i have a rock, I can quickly react and use it as a defense. But even a gun might not be so useful. Is it child-safety on? Is the bullet inside? Is it cocked? All these questions could enter my mind, slow down my response, and lead to complications. Therefore, I think modes in general should be used sparingly.
Modes 2: One mode that I find rather annoying is the caps-lock mode. More often than not, I end up accidentally pressing it since it's so close to the other useful keys. It is very rare that I ever have to enter a string of capital letters. Obviously, I see the usefulness of the Shift key, and since it has to be held down during its use, it is difficult to mess up. On the other hand, sometimes the caps-lock key is on without one noticing, which leads to problems such as the inability to correctly type in a password. This obsession with modes is something I think designers have to improve upon.
