More Vista

Well, after using Vista a bit more on the new laptop, I wanted to comment on two new features that I really like from an interaction standpoint.

First is the subtle tweak to the file renaming process - before, when a user wanted to rename a file the whole filename (including extension) would highlight, and typing would wipe out both the existing name and the extension. For most renaming tasks the file extension isn't going to change, so this just caused needless keystrokes and a possible short-term memory issue. Now, only the file name is highlighted, so it saves time and makes renaming things a bit more pleasant.

Second is a pop-up preview the user sees when hovering over a window on the taskbar. As of yet, I can't wrap my head around how this will really increase my productivity, but I know I have a strong visceral attraction to it (garlic press round 2).

So far the interaction designer in me likes Vista, except for the horrid user access control. We'll see how my feelings evolve as I become more expert with it. =)

Is I Or Isn't I?

So I have this nice new laptop with Vista on it, and for the most part, I'm pretty happy with it. I like the new look and feel of the interface elements, and I'm finding some of the new features to be more usable than the XP stuff that I'm used to.

However, Vista is doing one thing that's blowing my mind. My default account is an administrator one, but many programs will not "run as administrator" unless I specifically open them that way via a right-click context menu at run-time. The issue here is that when the programs don't run "as administrator," they're sometimes denied write access, so they may work in unexpected ways (or not at all). What kills me about this is that I'm logged onto an administrator account -- all my programs should run in administrator mode automatically! It's like my computer doesn't trust me, and it's a very frustrating experience.

Emotionography

Well if you saw the post title and thought I was talking about the Big Bud track, too bad! Given our recent class discussion in HCI 470 and the continuing relevance of emotional design, I wanted to talk about my experience with a gift I recently received: a garlic press by Michael Graves!

Now I haven't used this product to press any garlic yet, so my reaction to it is so far completely visceral. When I first picked up the press, I noticed it had a definite heft to it; this sense of weight gave me the feeling that it could really... press some garlic (heh). The handles have a very pleasant texture - a kind of soft, grippy rubber. Aesthetically, I like the object; the actual pressing of the garlic involves spinning one of the handles in an arc, which creates a pleasing symmetry.

So without having actually used this thing, I'm already a bit infatuated with it. But, as Norman says in his book, visceral is only one layer of the emotional experience. When I actually do press some garlic, the behavioral component will become a second contributor to my emotional response. Will the press save me time and effort over peeling and dicing the garlic with a knife? Will it be easy to use and clean?

Together, the interaction and aesthetic characteristics have the potential to make this one of my favorite kitchen tools (although I doubt it will ever eclipse my dear knife). But even if the interaction is sub-par, my visceral response has earned the object some "credit," where I may be more forgiving of a troublesome behavioral layer.

Disaster Recovery

The topic today is kind of personally motivated for me, since my desktop PC just had a catastrophic hardware failure on Monday of project week... In all the free time I have with no computer to occupy myself, I've been giving the topic some serious thought.

I'll start out by saying the hardware failure happened when I was trying to install a new component, and I was very lucky that I took the time to back up my class files and personal documents before I cracked open my case. I'm finishing my projects downtown at the CTI labs, but if I hadn't backed up, I'd be up the creek something fierce, with only a few days to recreate what was essentially a quarter's worth of work.

I think this ties into HCI because we as designers are supposed to provide a safe environment where the user feels free to experiment, and can easily diagnose and recover from errors. Now although a hardware failure can be difficult/impossible for software to predict or deal with, I think there are still steps the system can take to assist in recovery from these types of errors. One thing might be some sort of timed, automated backup to external media. I know this capability exists now, but as far as I know it's something mostly employed by users that have a high awareness of data safety issues. Shouldn't we work to make this type of recovery ubiquitous?

Maybe computers could have some sort of built-in removable memory that could act like the extra house key you hide under a plant (try to work with me on that analogy, heh). Specify which files are important, and the system backs them up every night for you. Then when your motherboard melts, all you need to do is open the small panel inside your case and pop out the emergency memory (see, like grabbing the key from under the plant! I knew that analogy would come together...). Then you could go easily finish your class projects at another machine instead of committing seppuku over the shattered corpse of your computer.

TL/DR summary: Disaster recovery is this is a topic I didn't think about much until I had to deal with it myself. And as a user-centered designer, I'd like to help other people not to have to think about it either. What about you all? Has anyone seen a cool technology or method for helping people get back on their feet after a computerboom?

Credit or Debit?

At CVS today I bought some stuff (frozen pizza and coke cause it's homework time tonight), and as usual, I used my Visa check card to pay. Now even though this isn't an "actual" credit card, I earn rewards if I run it as credit instead of debit (no idea why, but I like rewards).

So as usual, I swiped the card and punched "Credit" to start the process, and then got the annoying "Enter your PIN" prompt. But I don't have to enter my PIN... because I want to run it as credit. The (secret) method to continue running the card as credit is to hit cancel or clear on this screen... and the cancel or clear button is someplace different in every store. Plus, how horrible is it to force the user to mentally re-map the cancel button to "Move forward in the process please?"

So my question is, is there some kind of credit card with a pin that I'm just ignorant of? Or is this just an oversight by the manufacturers of these terminals, who are not updating the task flow to accomodate users with cards like mine?

Going back to edit in this final note: I've seen less experienced card terminal users become absolutely flustered when this screen comes up. For a novice user that may have trouble adapting to unexpected steps in the task flow, this interruption is catastrophic -- especially when the instructions on how to complete the intended operation are unclear.

My Favorite Menu, Ever

Gonna drop down from the usual high-concept ramblings here to discuss a usability gaffe that's got to be my all-time favorite. I came across this soon after moving to Chicago, when trying to find a channel guide for my Comcast cable. Let me just say, when I saw the prompt "Select your neighborhood...", I was looking for something like "Lincoln Park." You know.



After I finished laughing, I actually checked my documentation to see if these were some labels I was supposed to know or something... nope. I decided to go with the second option because I didn't see any rebuilding going on outside my window. The really frightening thing is that one screen before this, I was asked to enter my address to log in to the Comcast system or whatever. This, to me, is a classic example of what About Face (omg v3!) preaches: Don't make the user enter data that the system can figure out.

I'd be interested to hear if anyone else has good stories about entering unneeded information. If we get a few comments going it'll help me take my mind off of mental models for a week!

Superstar Cred

Ok this post is a little esoteric but hey, that's how I get down sometimes. Have you seen the commercials for the new T-Mobile Sidekick? The one that's designed by D-Wade?

Yeah, I've got multiple issues with this.

1. What is D-Wade's actual contribution to this phone? The commercials, I'm pretty sure, say he "designed" it. Now I'm sure #3 is quite intelligent, but somehow I have trouble seeing him pondering a user flow diagram in Visio on the plane between series. But ok, the weblink above is more forgiving - it says he "played an integral part in the design" of the Sidekick. So did he pick the colors? Are gold and cream D-Wade's signature colors or something? Did he choose the control layout? Well, his model is laid out exactly the same as the normal ones. Or, and this is the scenario I find most likely, did he get paid some money to be in a commercial holding a Sidekick which some actual designers colored to appeal to the NBA-hip-hop-success lifestyle demographic and to look good next to D-Wade in a white suit?

2. Isn't the claim that D-Wade designed this phone interesting commentary on the general public's perception of designers? Does this perception of design as needing no specialized knowledge speak to the troubling tendency of users to blame themselves for errors (cause you know, it's not like designers should be going out of their way to make it easy or anything)? Or am I reading too much into this?

3. If normal people freak out for a phone "designed" by D-Wade, would actual designers freak out for a [whatever] designed by some superstar designer? I guess this already happens somewhat in the consumer products sphere but I'm curious if someone in the software universe would have that kind of pull (Other than Apple of course).

It Ain't All Good

Travis' comment on my last post got me to thinking, maybe I came off too complimentary about Starbucks! Although I really like the name-asking, and as he pointed out, the fast and easy credit processing, there's definitely some aspects of Starbucks' interaction design I find troubling.

Specifically, I'm freaked out by the way Starbucks casually lets their implementation model take priority over the customer's mental model (you can get a bit more detail on implementation vs mental models here if you like) of drink ordering.

I'm uncool enough to remember the first time I went to Starbucks, so let me describe the evolution of my coffee-ordering mental model. My first time, I saw they had a chai latte, which I knew I liked. I ordered it at the register, and then moved back a few steps, waiting there until I heard "Grande chai" shouted from somewhere to my right. Ok, drinks don't come out at the register. My zero-experience mental model of Starbucks was copied from fast-food restaurants, so it wasn't much of a leap to ordering at the register and picking up my drink at the bar (Even though I've seen some Starbucks where the big coffee making thing kind of blocks the view of the counter part).

No sweat, I've got it down. Order at register, pick up drink at the bar. This worked out until the first time I ordered a frappucino, stood at the bar, and oh! here comes my drink at the register. I remember this caused some trouble because it was very busy, and I didn't quite hear the cashier saying "frappucino" (low expectancy for input from that channel!). I left Starbucks that day secure in my new expertise: "Cold drinks come out at the register, hot drinks come out at the bar."

This one worked out until I saw someone order a regular coffee and it came out at the register - thus revealing the naivete of my mental model. You could write all these errors off as novice user mistakes, but at this point I'd been going to Starbucks for a pretty reasonable amount of time. With no signage or cues built into the system, it's really difficult to make the leap from a working mental model to the actual implementation model (read: from novice to expert).

I have two hypotheses about the effects of this situation:

1. Most people do not have a deep understanding of Starbucks' implementation model. I could be wrong about this - maybe I'm just coffee-retarded or something. But I'd be willing to bet that most people that succeed in ordering at Starbucks just have a robust working model of the things they most frequently order (this is currently the case for me).

2. This is a barrier to experimentation and new customers, because people are afraid to fail in front of others. Especially in a place like Starbucks, which, as Aaron pointed out to me a bit ago, sells a whole attitude of urban "coolness." I know intelligent, rational people that love specialty coffees but are afraid to go to Starbucks because they feel like the ordering system is too complicated and they don't want to mess it up. Now you might argue Starbucks is successful, so those customers (the ones not cool enough to figure out the implementation model) just aren't the target market. I guess we should be glad Starbucks doesn't "target" everyone then, or they might have enough money to buy the whole solar system.

Proving I Love to Say "Context"

Preface: You get some kind of some random thoughts this week, but I know they're significant because I originally wrote them in a moleskine notebook, and supposedly that's what Hemingway used to write in before he snapped and shot himself.

The topic this week is how the effects of different user contexts can be observed and designed for. Let me start with an example: a waiter serving customers at a restaurant. For the waiter, many customers will come and go throughout the night, and he will repeat the elemental actions of his service cycle over and over again with different subjects.

On the other side of this interaction are the customers. From their viewpoint, the restaurant is only one stop on their itinerary for the evening, and they only interact with one waiter.

Although in this case, no technological system is really designed to support this interaction, the waiter certainly takes advantage of his context to re-use effective strategies on subsequent customers. This gets more relevant to interaction designers when the system complexity increases - for an example, we can look at everyone's favorite coffeehouse.

At Starbucks, one context is a barista who deals with people all day. The overlapping context encompasses the legion of individuals who need their grande double shot no whip skim chocolate latte right now. It's not too hard to imagine how the interaction between these two can quickly degrade. But here, we have the advantage of designing a system to place in between the two contexts to help the interaction run more smoothly.

One of the things I like best about Starbucks' approach to the customer interaction is that their computer system requires (maybe? they ask me most of the time but not always) a name to associate with the order, which results in a self-disclosure from the customer. This helps the customer build a relationship with the barista, and the barista to see the customer as a unique person; both of these elements increase the likelihood of a pleasant interaction. And this is one of the things that makes me happy to be studying interaction design; the thought that subtle design decisions can positively influence people's experience without them even being aware of it... =)

All that being said -- what do you think? Have you seen some examples that bridge a contextual gap between participants? What about systems that fail at this?

Rude Museum People?

Last week I had the opportunity to visit the Body Worlds exhibit at the Museum of Science & Industry. My friend (who is a grad student in Computer Science) and I decided to skip the accompanying audio tour and just wander through on our own, but we quickly noticed how popular the audio tour was.

(note: the audio tour used handsets a lot like these [link has sound])

After we'd been walking around for a bit, we noticed an interesting phenomenon. As we looked at displays, people with the audio handsets would kind of float into our personal space, sometimes going so far as to bump into us without realizing it! Not an "Oh, I'm sorry" bump -- these were "I'm totally engrossed in my own world and I don't even notice hitting you" bumps.

So the easy assumption is that these people are just inconsiderate, right? Well, interaction designers know better! My theory on this, which (as with all things HCI) I was quite happy to share with my CS friend, is that these awkward encounters are simply artifacts of poor design.

Specifically, the audio handsets were dominating users' selective attention to the point where they lost track of their spatial surroundings. My old HCI 450 book lists four factors that determine which sensory channels are attended: effort, expectancy, value, and salience.

In this audio tour example, the effort required to get full meaning from the audio tour is probably high. This is especially true since audio feedback relies more heavily on working memory than does written text. The expectancy of the audio channel is high, since the user presumably triggers the handset when they arrive at a new display. The value is also high, since the user is presumably learning a lot of interesting information about the display. And finally, the book indicates that auditory stimuli are particularly salient. All four factors are so high in this situation that the users slip into inattentional blindness and begin to perform actions that they normally would not (bumping into people without noticing).

In this case, the problem isn't too critical - we just chuckle at the "rude" person and step out of their way. But it's not hard to imagine situations where selective attention can cause more serious errors. My two-cent advice here is to remember that we have the power to shape the users' selective attention via our design choices, and that may be just as important as graphics, menu design, or anything else.

The First Post is About Bars

One of the universal questions you'll hear from strangers at a bar is, of course, "What do you do?" And if you reply, "I'm a student," you might as well be ready to field the follow-up. "Oh, what are you studying?"

So if you, like me, are a graduate student in Human-Computer Interaction, and you actually choose to tell the person that, you will likely need to act fast to stop their eyebrows from rising right off their head. If you're not talking to a technology person (and computer nerds don't even hang out at bars, right?), HCI apparently sounds like you're serving the evil robot masters in their quest to rule the world. So, you knowingly smile and launch into your explanation:

1. Try Wikipedia's definition of HCI: "the study of interaction between people and computers." Didn't you ever learn not to use the word in its own definition?

2. Describe a job title in the HCI field, like interaction design: "Interaction design defines the behavior of a system in response to its users over time." It's starting to look like broad descriptions of our field are not a quick way into that cute stranger's pants.

3. Make a generalization: "We figure out how to make things easier for people to use." This one is a lot simpler, but in my experience, people start thinking "How the hell are you getting a master's degree in that?" I had one person say to me "So... you're getting a master's in... troubleshooting?"

I find this interesting because I think it really highlights the non-designer's paradigm of technology: if a system is difficult to use it must be their fault. So an advanced degree in making things easier is a mile away from most folks' context. Much less the concept of conducting extensive user research, using findings from cognitive science to design iterative prototypes, and following up with rounds of usability testing.

(warning, actual content: I think it's important for us to keep this paradigm in mind when conducting any user-centered activities - design, research, testing, whatever. We as designers don't (shouldn't) share this viewpoint, and as always, this makes it harder for us to step out of our own context and into the users'.)

4. Focus on a specific HCI activity: "Well, I do a lot of user research during the design phase of things." Excellent, now the person thinks you're a marketing kid. Unless they're actually in marketing themselves, in which case they probably think you're a weird marketing kid because you called it 'user' research.

5. Try what I tend to say nowadays: "Yeah, I'm studying psychology."