Visual Information Design
From Cs160-sp08
Lecture on Apr 15, 2008
Readings
- Information Visualization. Readings in Information Visualization. Chap 1. Card, Mackinlay, Schneiderman.
Benjamin Lau - Apr 10, 2008 08:55:50 pm
I never thought of information visualization as a form of external cognition but I suppose it makes sense. Creating or analyzing a good diagram or graph or some other visual representation of abstract data effectively grounds the information in something more fundamentally basic and comprehensible to the human brain-- the tangible external world. The discipline of information visualization makes use of the mind's ability to map abstract notions of properties and similarity (among other things) to concrete things like locations and physical proximity. To start off, one could imagine a 2d map with Cartesian coordinates, that has several points representing buildings. If 2 points on the map are close to one another, the corresponding buildings must also be close. But the interesting thing is that this extends to much more abstract domains. For example we could imagine several flasks containing different amounts of oil and water. Let's say it was crucial to be able to distinguish quickly among the various flasks. We could do this by mapping each flask to a point on the map, with X for example possibly being the percentage of oil and Y being the percentage of water. Flasks then that contain similar percentages of oil and water will be closer together than the rest. The similarity between flasks due to relative proportion of oil and water had nothing to do with distance but the metaphor makes it possible to create a new visual representation that's much more amenable to perceptual inferences. Stuff like this is just one of many different kinds of techniques that visualization uses to transform data for external cognition, bringing the power of human perception to bear. (In particular this would be an information chromatography technique, there are other ways using temporal encoding, color, etc; as far as spatial representations go there are also more complicated ones like hyperbolic trees.)
Nir Ackner - Apr 12, 2008 12:58:19 pm
A lot of research seems to be heavily focused on visualization, with some very impressive results. This focus suggests that computer programs are highly effective at conveying information to visual learners. However, many people are auditory or kinesthetic learners, meaning that they gain insight from what they hear or do. It seems that much less focus is in these domains, and most of the research is trained on helping the visually impaired. As the technology for things like text-to-speech and touch-based interfaces develops, perhaps we can discover ways that the senses of hearing and touching could be used to aid in cognition as well.
Eric Cheung - Apr 12, 2008 03:13:16 pm
I understand the need for visualization as a way of representing data in an easy to understand way, but I found a lot of the examples in the reading were very poorly designed and impossible to understand without a lot of background information. I don't know whether it's the fact that some of the examples were quite old (~1995) and as such, the technology wasn't available at the time, but I found figures like 1.8 and 1.13 very difficult to understand. Also, I fail to see the need to devote 6 pages to how to make a good table. Granted, picking which values you include in the table can be very important, but I've probably seen many more examples of hard to read designs such as figures 1.31 and 1.32 than I have of bad table design.
Ilya Landa - Apr 12, 2008 07:44:06 pm
The paper was bit long, but, overall, bearable. I think the reading section of this class should be called “How Not to Design Interfaces.” And I don’t think it’s a bad thing – it is much easier to show bad design examples that to try to create a formula for a perfect design. These bad examples are also much more useful, and always more fun, to watch than reading a dry prose. I wish an above comment about lack of information about kinetic or auditory interactions. I don’t find that fact surprising or unfortunate at all. Vision is just the best medium for information transfer in humans. With even a simple screen, one can convey multiple pieces of information at once, while a smell, sound, or touch, can hardly convey more than one piece of data simultaneously. Plus, specifically for sounds, I don’t think that much research is needed – even without any background knowledge, humans can easily realize that the pleasant “Ding” is good, and the error buzzer is bad. Well, another hour or so spent, indifferently reading page after page. Honestly, I really hoped that the break in reading we all enjoyed over the last couple weeks would go on till the end of the semester.
Hsiu-Fan Wang - Apr 12, 2008 06:21:54 pm
I agree with Eric about the table design thing. I've always wondered if maybe tables ended up being better designed (on average) than other visualizations because tables are relatively simple looking so they survive only if they are truly visible, whereas a number of info visualization projects seem to capture attention based solely on their looks.
Another thought I had looking at the nomograph example is that I've found Excel is often more effective at visualizing things than the calculator. (I tested the small angle approximation once, trying to figure when it broke down by just putting an angle and its associated value on separate columns and then just scanning)
Also, a reference back to previous reading, the mentions of external memory remind me of the brainstorming, where naming and numbering ideas and placing things around the room help ideas percolate.
Glen Wong - Apr 13, 2008 01:14:15 pm
I agree with what others mentioned about the long section on tables being somewhat overkill. Overall, I agree with the principles mentioned in the readings, but I feel that they didn't really say anything new. I already understand that external visualization aids help people to solve/visualize complexities that aren't possible without such aids. I already agree that tables are a good way to organize data. In fact, given the huge popularity of SQL, I would think pretty much everyone understands the importance of tables in the organization of data for visualization. I don't mean so disparage this reading, but the author does a lot of restating things I already take to be obvious especially in light of previous readings we've done in this class. The one thing I found very interesting was the chart on O-ring failures for rocket boosters at various temperatures. It is pretty ridiculous that the company that manufactured the boosters didn't try plotting failures versus temperature.
Khoa Phung - Apr 13, 2008 02:50:46 pm
This text was published in 1999 and I believe we have come a long way from back then. I believe that visualization has become very natural now and is widely used in almost all fields. The text mostly talked about improvements in computer applications and I believe they have neglected one important field. Computer games especially have not really improved their stories or ideas, but just made it more realistic and therefore, easier for the brain to process the vast amount of information that was displayed as text before. While the text was mainly focusing on reading charts, patterns for statistical analysis, games also now have developed that complexity, but managed to put it nicely in a way for gamers to easily adapt to and interpret that information. Examples are real-time strategy games where the user has to build out a headquarter, build units, upgrade technologies, watch enemies, manage repairs, manage power consumption, etc. It is important to be able to read information quickly and visualization has evolved tremendously. Now, we are moving towards 3D gaming experience using stereoscopy to enable the user to be able to process even more information even faster such as depth, angle, distance, and many more. Tables and graphs are still the very important in the 2D world, but as technology gets better, I believe that having a virtual office such as a holo deck can possibly become a reality.
Bo Niu - Apr 13, 2008 04:34:17 pm
Visualization is really something that can help the users to understand complex structure or ideas in a easy way. As many examples have been shown in the reading, just buy showing the data in a diagram or just lay out data in a logical way can reduce a lot time for the users to process the data. My favor example in the reading is how changing the lay out of the multiplication problem can reduce the calculation time greatly, since it's a very simple example yet the difference is so obvious. From this example, a very important concept of good visualization is also revealed, which is that a good visualization would be a simple graphical layout that present the idea in a simple logical way rather than just a pretty display.
Johnny Tran - Apr 14, 2008 01:36:20 am
I can see how the evolution of UI can actually be the evolution in visualization techniques. The jump from text to graphical interfaces parallels the jump from reading a text file of data to reading a graph on a spreadsheet. While a lot of UI designers might focus on how curved the buttons should be and whether the mouse cursor exhibits the requisite amount of eye-candy, the focus in UI design should really be on the way data is presented, first and foremost. Visualization is the key to doing this.
As the old cliche goes, "a picture is worth a thousand words." It amazes me how certain things would be difficult to describe in text, but would be easily understandable using a diagram. And the idea of allowing users to interact with the visual displays goes back to the idea of direct manipulation interfaces--a powerful idea.
In addition to the usual buttons and menus, perhaps UI widget sets should include standard visualization components, such as graphs, tables, and charts. It would prevent developers from redesigning the wheel, and perhaps lead to more software using efficient visualization aids.
Benjamin Sussman - Apr 14, 2008 03:29:55 am
An interesting read. I data visualization is an extremely important field which is only getting progressively more important as the internet continues to grow. Because of the simplicity (and thus beauty) of IP, the packets which bounce between routers are actually all that make up the internet, yet it is extremely important to both back-bone engineers (whose job it is to keep the spice flowing [whoops, I meant packets]) and front end designers who need to see how users are behaving on their websites to know the meta-information associated with those packets. Things like web analytics, where we need to visualize flow and hot-spots on the site are now essential because we don't have a direct communication with the vast majority of the customers whose visits alone are the new business model. Also, the geographic and adjacency graphs which can represent the internet are getting so complex that we will need complex mapping tools to assist maintenance and upgrades of the network.
In short, it's an extremely important field and one that deserves a lot of attention.
Jason Wu - Apr 14, 2008 01:18:08 pm
It seems like in an age where access to information becomes so much less restrictive than our ability to comprehend it, data visualization and its study will be fairly important, especially in the area of UI design. The way data is represented can make processing that information so much faster than the most straightforward way of reading it, since it allows you to do instant comparisons, or leave out things you don't need to pay attention to at the moment. As Khoa mentioned, a large amount has been evolving and changing in computer games which demand quick data access and feedback. It seems that outside of games, it's very possible to go beyond the simple graphs, and tables, and charts, and possibly move to something more interactive and dynamic, possibly with auditory and tactile elements. This field seems to be fairly unexplored, but demands will evolve data visualization soon enough.
Max Preston - Apr 14, 2008 01:13:31 pm
Indeed, the idea of data visualization is absolutely critical in computer science for everything that requires human feedback or processing. This just goes back to how computer systems are the product of many layers of abstraction. Even though data consists of only 1s and 0s, which by themselves would be almost impossible to visualize as an application, they can be interpreted using set rules and complex frameworks to create more easily understood systems. For example, a text document in binary would be extremely unpractical for a person to read, and could take hours or days to decode by hand. Visualization of this kind of information into human-readable text is simply a vital aspect of computer systems. And this text can be further visualized into charts and images which can be useful for human understanding.
Michelle Au - Apr 14, 2008 03:20:32 pm
Card, Mackinlay and Schneiderman really emphasize the growing importance of information visualization and how critical a well-designed visualization is in our complex world. The example of the two diagrams depicting temperature and damage for the Space Shuttle Challenger illustrated just how much effect the right visual representation of data can have on our cognition. Having well-designed visualizations is even more important today. New developments in computer technology and graphics have given designers new tools and capabilities to further enhance the quality of visualizations. There's more potential for good visual representations with real-time processing and direct manipulation of the data. It would be interesting to see examples of the visualization research going on today utilizing these new advances in graphics and user interfaces.
Jonathan Chow - Apr 14, 2008 09:24:33 pm
I certainly agree with the others who mentioned the importance of Information Visualization right now. There is so much data out there that it is essential to find a good way to visually organize it so that we can actually perform an analysis of the data. As far as the article is concerned, I found it interesting to read about how they modeled perception through the actual anatomy of an eye. This is certainly a contrast to the other human information processing models we read about earlier. I also found the comment about distortion not being effective "when the features or patterns of use to the user are distorted in way harmful to the task." I think this is something that has been somewhat forgotten with the fancy new 3D rendering technology that we have now.
Gary Miguel - Apr 14, 2008 10:14:20 pm
I see we've arrived at the requisite part of every berkeley CS class where we spend time introducing a tangentially related topic and talk about it very briefly. The example of the two representations of the O-ring damage was a striking one, and a good way to start a book on visualization. The book does a good job of systematically enumerating the key processes involved in producing a good visualization, and the tasks such a visualization should support. As I was reading the parts about graphing calculators and physics problems, I had flashbacks to Math 53, when I would spend hours struggling to accurately draw a vector field or level sets of a 4 dimensional graph. I imagine it would have been a lot less work for me to write a graphing calculator, but then again CS projects always sound easier than they end up being.
The point is that automatic visualization software can make some tasks much faster, and in some cases, possible at all.
William tseng - Apr 14, 2008 10:49:19 pm
On a related subject to visualization, I thought I'd share this link with the class its a great example of how data when presented visually gives us more insight into the actual trend's and problems of the world.
http://www.ted.com/index.php/talks/view/id/92
In terms of the reading it was cool to see this reading tie back into concepts from previous readings. Especially those concepts relating to gestalt principles in terms of graphs and visual structures. Also I think one of the important things to take away from this reading is that when designing our visual interfaces we should pay attention to the loss or gain of information that occurs when we perform any of the 4 types of data transformations. As always which type of transformation your own program should use probably depends on the target user group and how well they would be able to process the data depending on how you present it.
Reid Hironaga - Apr 14, 2008 10:56:53 pm
I find the most interesting part of this article to be the wide variety of methods of conveying graphical data to a user. I also believe that there are many updates that could be made to this paper due to the advances in the field of graphics in the past few years. I am amazed by the possibilities in interactive interfaces, and visualization of data will definitely be impacted by the onslaught of new technologies. The transformation of raw data into a data table is definitely a nontrivial task to ensure that the users will safely interpret the data correctly. At the same time, the advent of artificial intelligence is making the need for humans to process data and the severity of decisions made by human interpretation less of an impact.
Alex Choy - Apr 15, 2008 12:34:08 am
I felt that the reading was interesting and gave many examples. Data visualization can be very important and especially in graphs, like in the shuttle launch example. I agree that by using the right graph and representation, much of the underlying information can be gathered and a better decision based on the information can be made. I also agree that the way that space is used is important. Not only can white space be treated as if it has metric structure, but it can also give a good balance between text/data and empty space for a more aesthetic feel. The way that data is presented is important and can reveal information in a way that text cannot. The growing importance of Visualization is good and can really cut down on the amount of text that we need to read.
Yunfei Zong - Apr 15, 2008 03:07:37 am
"There are right ways and wrong ways to show data; there are displays that reveal the truth and displays that do not." This quote strikes an important point in that visualizations can be used to conceal the truth, purposefully or not. We see this happening in cleverly produced pie charts, where illusions and tricks of the eye imply to viewers something completely opposite of what the diagram is actually showing. The reading gives many fine points on how not to design a visualization in this way, like the o-ring and the car manufacturer example, which are both completely ridiculous yet extremely serious problems at the same time.
Katy Tsai - Apr 15, 2008 09:39:21 am
I think if anything, the idea of visual information design really brings in all the concepts we've learned this semester to emphasis how important visualizing information is in translating different types of information. I liked how the basic multiplication example really emphasized how we need to visualize certain things, even the most basic ideas, because we can only hold so much in our working memory. It was interesting reading all the examples because the majority of them didn't display any extremely complex ideas, but being able to create a visual representation, like in the knowledge crystallization example, allowed for a more organized and comprehensive thought process.
Harendra Guturu - Apr 15, 2008 10:26:18 am
I agree with the reading in that visualizing aids can help making a task a lot easier. I think this goes back to the idea of recognition over recall. The ability to use the outside paper as extra ram acts as a way of just looking a the paper and "recognizing" what task you are doing and what needs to be done next. I also agree with Eric and Hsui-fan in that it is not necessary to always complete the model of visualization by taking raw data and mapping them into a data table and then mapping the data table into a visual structure. Some times data is far more effective in a table than in a fancy visual structure that may be hard to understand if its not a commonly seen object.
JessicaFitzgerald - Apr 15, 2008 10:42:39 am
This reading on Visual Information Design reminded me of one of my high school classes where we learned about learning styles and how people learn. This reading talks about how we all learn visually, which seems to suggest that everyone learns or sees things better this way. I am sure that some people might find other ways of absorbing the information helpful. However, in general to understand something a visual is almost always the best and easiest way to understand something. That way you get to see the big picture, and as the reading points out what is really going on. It also depends on the way the visual is presented. If it is not presented in a clear way, then you might miss some information that is quite critical such as the shuttle launch. I feel that by paying attention to create more correctly representative and clear visuals, the more users will be able to understand the meaning of and use these visuals effectively.
Diane Ko - Apr 15, 2008 11:09:25 am
A lot of this reading relates to the Information Visualization class I'm taking this semester. Specifically, all the visualizations of data in 3D. A lot of the bar charts that appear in 3D tend to be very hard to compare data values. Also, the 3D simulated workspaces look more complicated than necessary. While I do think there is something to be said about placing objects in 3D in that there's more use of the visual space (since an object lying down in 3D could take up say 50 pixels whereas if shown flat could take up 500 pixels), it just ends more often than not looking very cluttered and hard to navigate. I personally wouldn't want to a good portion of my computer's processing to be placed on rendering all the 3D stuff when it could be used to make something I'm running finish faster.
Jonathan Wu Liu - Apr 15, 2008 11:33:11 am
I definitely appreciate the concept of visual information design. Just having more information displayed in the same amount of space is an art, and I see how complicated and comprehensive ideas can be made easier to understand with the correct visualizations. These concepts are applicable to any application; for example, we run into the same problem in our application when we want to display the most amount of data in that small 2.5 inch screen. Though the table information was overkill, giving us a step by step method as to how to create a visualization can be helpful when we want to know what is the most important information to display. And as technology advances, we will have access to more complicated visualizations (b/c of faster processors, etc.) and create more creative displays of information.
Yang Wang - Apr 15, 2008 02:03:00 pm
I have always thought visualization is what make computer so useful to every area. During some earlier classes such as the cs61a, we have learned how visualization evolves in ages. This reading enhanced this idea once more to show how important the visualization is. Especially for our project which requires us to display information in limited space to users. How to efficiently use this space to display information clearly as well as achieve visual pleasure is defiantly an art.
Video gaming is another good example of how important visualization is. Many games, without graphic, are just math problems. Strategy games, for example, simply take calculation of opposite units and come out with a result of conflict. However, once those numbers are masked with graphic units, everything changed. In our perception we would treat them as a war of some kind. I guess that was a bit off topic, but anyway. Visualization as the main mean for human perception to take information is very important to the interface design and should be considered very carefully during our project.
Bruno Mehech - Apr 15, 2008 02:16:32 pm
This reading reminds me of some of the examples of search interfaces that tr to visualize search results Marti Hearst showed us. As I remember these interfaces were failures in visualizing search results. This is partially explained in this paper by the discussion of abstract and scientific data visualization. Apparently no one has figured out how to visualize search results in a way that makes it easier for users to find what he is looking for, but the reading takes an optimistic in that there is always a way to visualize some data.
Cole Lodge - Apr 15, 2008 02:15:03 pm
Overall I found that this reading seemed enjoyable. This is probably because this is the first reading in a while. At first I was having trouble understanding what was meant by visualization, but after seeing the examples given, I seemed to have a much better idea of what it meant. I found this a little funny; I was only able to understand visualization after I saw examples. One example that really helped me understand was the multiplication example; it seemed so straight forward and just made sense. I can see this idea being used in my groups prototype to help improve our interface; specifically, creating a better visual of a task on our calendar will help the user understand what is going on.
Chris Myers - Apr 15, 2008 02:21:04 pm
It is interesting to see how Xerox was trying to get ahead of its time with desktop visualization. Their ideas for a 2D interface are everywhere today, but the 3 dimension never really took off. I suppose the difficulty of presenting more data with simulated depth of field is that it can easily become overwhelming for the user, even if it is simplified. For example, the paper notes that the distorted information on the sides of cube cause the information to become invariant. The information is then no help to the user. It just makes the display impressive looking but useless.
Siyu Song - Apr 15, 2008 02:16:35 pm
I liked the concept that visual information can help increase your active working memory base such as in the first example with doing multiplication in your head versus doing multiplication with a pen and paper. The next few examples with visualization of navigational charts and the graphic of the ozone layer were interesting in that it was able to take the overhead of processing data out of the process of analyzing this information. I also found interesting the Larkin and Simon study that was referenced on how visualization can amplify cognition. Using things like location recognition and recognizing geometric elements. This reminded me of some of the new visualization features on Digg.com, some of this was was such as having news items pop up as dots and then names of people who dugg the article linked to the dots and it was a 'swarm' like interface. I stared at it for a while, but I don't know how useful it is in terms of picking out and reading articles.
Brandon Lewis - Apr 15, 2008 02:28:40 pm
This is a very interesting topic. I learned multiplication and long division so long ago that I never realized the tabular method had to be invented by someone. Because of their basic nature, they have a kind of cultural significance that seems to transcend the pie chart or the scatter plot. They seem permanent and unchangable, but of course they are nothing of the sort. The idea of creating new visual aids that are nearly as useful is quite exciting. Whoever invented the multiplication method has achieved a kind of immortality.
Also, novel visual aids reminds me of why I have been interested in computer science in the first place. Creating new ways of understanding and conceptualizing complex phenomenon, essentially increasing our collective intelligence, has been one of my ambitions. At the same time, there seems to be a lot of information in this text that is of practical value for designing user interfaces. I liked this reading quite a lot.
Tam La - Apr 15, 2008 02:47:57 pm
It is remarkable just how much a visual representation of data can enhance the absorption and analysis of that data, and this reading make that point very clear. In addition, the reading demonstrates how different ways of presenting the data can either mask or uncover patterns in the data. In retrospect, such a property of graphical data representation seems obvious, but it didn't strike me until after I did the reading.
Jiahan Jiang - Apr 15, 2008 02:41:58 pm
This was a very interesting reading; it pointed out such normal phenomena that I have never bothered to consider. It is interesting how we use external visual representations to express and aid our cognitive activities. I enjoyed the section on perception; it shows how important visual representations can be and how intricate it is when things seem straightforward.
Kai Man Jim - Apr 15, 2008 02:51:12 pm
This reading is very long, but it brings a good signal on how important a visual information to us. It reminds me from my cog. sci. class that visual can be captured in our memory for long time. A good image can catch an observer's attention and that image will be store into our memory. I think this concept is very useful for us when doing a design, we have to know how to make our visual image interesting, and don't let our users feel boring.
Since it is very hard to have a standard on which is good and which is bad on designing visual information. I would suggest that the reading should have both good and bad examples to let the reading compare. And the examples in the reading should be more updated.
Jeffrey Wang - Apr 15, 2008 02:57:47 pm
While most of the examples of visualization are from a decade ago, the foundations of concepts still hold true today. Information visualization can be seen as a form of external cognition. One particular information I found helpful was how diagrams help in 3 ways: (1) By grouping together information that is used together, large amounts of search were avoid (2) By using location to group information about a single element, the need to match symbolic labels were avoiding (3) In addition, the visual representation automatically supported a large number of perceptual inferences that were extremely easy for humans. This is specifically useful for website design. Most of my programming are dedicated to website programming, where visualization is largely different from desktop application. Good visualization require a good grouping of text, as explained in the article.
Zhihui Zhang - Apr 15, 2008 01:28:41 pm
Although lengthy, the reading provides some pretty good insights on visualizing data. As they mentioned, there are good ways and bad ways to present data. In particular, I can think of the case of how to present information on filling a syringe. Obviously you can display numerics on the capacity or even rate of flow into the syringe. But a better way would be to simply present a visual model of a syringe being filled.
On a more personal note, for my last project at work i was really surprised to find well users responded to a visual timeline of when things are scheduled to occur rather then simply a table values.
Randy Pang - Apr 15, 2008 02:05:38 pm
Although I think Information Visualization is one of the most fascinating and useful problems today, I can't say I was exactly thrilled with this reading. I felt that most of the visualizations did not visualize their data effectively (granted they are from 1999, which while more modern than most of our readings, is literally eons behind actual modern information visualization), and that the reading was relatively dry (yet another term-list. Also, 74 megabytes? Seriously?). This is not to say that all modern information visualization is great (for example, the 3d graphs in the new Microsoft Office suite), but there are a lot of interesting things going on that I would have found much more useful and interesting then this reading (for example, what are people's thoughts on iPhone centric interfaces, the coming tabbed Facebook re-design, useful visualizations for desktop (e.g. Konpose for OS X, all the silly Compiz effects on Linux), widget data layout, visual search engines like searchme.com and managedq.com, completeness of data vs. simplicity and ease of grasping it, etc. etc.). I personally found the 10 minutes Marti Hearst spent showing off FLAMENCO was far more useful then this reading (I also recently discovered clusty.com, which is a commercial implementation of this kind of hierarchical categorization). I just would have hoped that an article on information visualization would have realized that extremely large bodies of text do not make for good data representation.
Jeff Bowman - Apr 15, 2008 02:58:49 pm
While the in-depth foray into data tables proved to be more of a low-level reference than I was prepared to understand, I very much enjoyed the section detailing precognitive attributes and "retinal properties".
I was however most impressed by Tufte's representation of the data behind the Challenger explosion, and how presenting the data correctly could literally have saved lives. The importance of presenting information well was not lost on previous lectures' discussion of pilot negligence versus bad interface design. The presentation of enlightening data in this format made me think of Al Gore's film An Inconvenient Truth, for which his Powerpoint-delivered slides provided the most fascination--along with the realization that how data is described can drastically alter the way it is handled by a user. Likewise, the YouTube movie ["Did you Know?" http://www.youtube.com/watch?v=ljbI-363A2Q] and its sequel "Did You Know 2.0" illustrates the difference in perception of mostly the same data, in one [textual] format versus another [graphical] format.
Michael So - Apr 15, 2008 03:02:27 pm
I thought that idea that visual aid in cognition was not at all surprising. In fact, I think that most people already know this but they probably just never thought about it or realized it. The example about doing a multiplication problem in one's head versus on paper was a good illustrative example of how visuals can amplify cognition. However, a visual does not necessarily amplify cognition. It can confuse, or deceive as well. Therefore I guess there's a need to study on principles of good visualization. This is good to consider when designing good interfaces, or improving interfaces because good interfaces should amplify cognition.
Hannah Hu - Apr 15, 2008 03:03:25 pm
To state the obvious: We rather look at pretty pictures than read a bunch of text. Pictures are pretty. Very pretty.
Gerard Sunga - Apr 15, 2008 02:29:21 pm
The reading was fairly interesting, and it was nice to see a heavy amount of visual examples. The entire paper was fairly interesting, especially the part that discusses visualization as amplifying cognition. From our lectures, we can see this is clearly the case (especially when looking at Gestalt's principles and the human predilection to follow these principles). Another interesting concept is the fact that the author advocates a high degree of visualization. This seems counter-productive for power users and will hinder users from doing things more efficiently, as seen in the KLM analysis of "better looking" interfaces during lectures.
Edward Chen - Apr 15, 2008 03:02:25 pm
The reading was interested in that it illustrated how visualization was so widely prevalent in our day-to-day lives. Even the simple example about writing out the multiplication down to speed up the process shows how common it is to those who don't even notice it. Nowadays, with increased bandwidth and visualization software, the amount of visualization data is more widespread on the web, using all kinds of visualization techniques to convey an idea or highlight an important point from data. The graphing calculator was another example that stood out to me because I remember looking at various equations and not being able to visualize how they looked if graphed, yet when it was graphed on a graphing calculator, how the equation worked all made sense.
I find that visualizations will become even more prevalent in the future as our fast-paced society demands information even faster and the many of the fastest ways to comprehend raw data is via visualizations that highlight important aspects of the data and compress various kinds of data into a cohesive visualization.
Andry Jong - Apr 15, 2008 02:57:25 pm
Visual aids help people learn things more easily. Designers can convey many different ideas to others visually. I do not find this idea extremely new. Besides this reading was developed long time ago, where there might be a lot of development already took place after it was published. A shorter reading will help too. However, this reading is still pretty exciting to know. Relating to this class, it is always good to have a good Visual Information Design while thinking about User Interface. Mostly when we are designing a User Interface for mobile applications, it will be very useful to replace all buttons with descriptions on them with visual representations of things. It will save a lot of space and, in motion, this representations will help the user understand the meaning of the feature faster than trying to read descriptions.
Timothy Edgar - Apr 15, 2008 03:06:51 pm
I'd also agree with Nir about that visualization is very interesting, however a multi-sensory approach doesn't seem to be often looked at. The first examples explains multiplication taking less time with pen and paper. However, this also is a tactile experience, which I wonder the relationship between being something people are doing versus something visualized. I can't help but to think of my all the modeling I do in Mechanical Engineering and the attempts to convey such large data sets with visualization techniques. Often times, the simple graphs and charts are flawed in that they imply different relationships that the intentions.
Gordon Mei - Apr 15, 2008 03:20:10 pm
The readings on information visualization highlight the importance that good diagrams can play in conveying trends and the bigger picture much faster and more sensibly than any dense table of data values can. This corresponds to the visual nature of the external world that the human mind finds more natural. For example, there was a class of mine where multiple pin-ups of room temperature findings were presented, but the only one that captured the attention of the classmates in the audience was the diagram with color gradients showing varying degrees of room heat at different regions of the space.
Likewise, a national elections map color coded down to the counties level told me more about the direction of the voting populace than any of the pages and pages of tabulated percentages in the major newspaper print (telling that a candidate is stronger in coastlines and large cities rather than being five or ten percent ahead here or there). Or satellite imagery of the world at night indicates volumes about energy use based on lights dotting the coastlines and major cities across the globe, more effectively than bar diagrams.
As the readings state, visualizations amplify cognition by saving the person memory and processing resources, and cuts the need to sift through overwhelming amounts of information (pattern detection, perceptual inference, perceptual attention for monitoring, and encoding info in something manipulable). In other words, it's a more human way of learning and understanding, and more in tune with the way we interface with the world.
Jeremy Syn - Apr 15, 2008 03:14:31 pm
I've always known that doing arithmetic in your head was always much more difficult and time consuming than doing it on paper. However I really never thought about the reasons why this would be. Of course I knew that hard, concrete substances, helps us visualize and do the formulas. Its easier to visualize on paper in front of you rather than trying to do it mentally. So of course we should make our designs incorporate visual information since it wouild help the user to reduce the amount of time to complete a task rather than trying to do it from memory, mentally. Visual information would really help in trying to get a user to learn a complex task or operate something with a complex structure. Visualization will always come in handy when working with an interface.
Brian Trong Tran - Apr 15, 2008 03:06:18 pm
Visualization is very helpful in displaying large amounts of data. I think everyone can agree that using a graph can be more helpful than a basic text file in terms of organization when it comes to mass display of information. The 3D work space is really interesting because it allows me to work in it as if I am actually working on my own desktop. I think the article spent too much time talking on tables. I understand that they are important, but I don't think that much detail was necessary to illustrate the point.
Ravi Dharawat - Apr 15, 2008 03:14:59 pm
I agree with Edward, trends seem to point to a growing importance for visual information as individuals are required to process more data at faster speeds. Visual information representation is great for this because it allows us to use the real-world as a sort of ram to supplement the L2 cache that is our short term memory. Holding more data allows us to process faster. But I wonder about the upper limits. It is still very difficult to process a large number of points of data, as organization then becomes difficult, and the references to those datapoints need to be managed. I wonder if using sound and touch might be able to "parallize" human thinking, as it were.
Daniel Markovich - Apr 15, 2008 03:10:32 pm
... Still Downloading .... Still Downloading .....Finally.
In all honestly the article was interesting but hard to read. It was just too long for having a project due the same day. The article does a great job of presenting examples of different visual aids, and the visual aids for the examples are helpful towards understanding the material.
David Jacobs - Apr 15, 2008 03:07:59 pm
Typically whenever I see a reading related to information visualization, I can't help but think to share Hans Rosling's presentation using Gapminder. It seems however, that William has beaten me to the punch in this area. As awesome as Gapminder is, I have to mention that its unclear sometimes whether Rosling's popularity is a function of his visualization quality (which I think most people can agree is great) or rather his stage presence. In order for a visualization to be effective, it should be deliver the same experience whether or not someone is around to talk you through the interesting points. I wonder if Gapminder users could find the same wonderful trends that Rosling points out without prompting.
Maxwell Pretzlav - Apr 15, 2008 03:23:50 pm
I found this article interesting to a certain extent, but mostly tedious. I am very familiar with graphic design and have read several of Edward Tufte's books, so few of these concepts were new to me. A few of the examples shown were interesting, such as the fascinating graphic depicting the world's tide system. Overall, however, I found this reading spend 25 pages repeating a message that seems clear to most people: pictures are more useful than words, and well-designed pictures are best.
Brian Taylor - Apr 15, 2008 03:26:41 pm
It was really interesting to see how various ways of presenting data can so drastically alter the user's perception of what the plot is meaning to convey. I had heard in the past that businessmen excelled at manipulating plots so that increases i revenue looked good and losses looked not that bad (mostly by axis manipulation), but I had never realized that one could perhaps present the data in a whole other form or model that would potentially destroy all thoughts of correlation. In particular, the example of the shuttlecraft and the o-ring failure chart exemplified this.
Overall, though, visualization is a great tool that helps reduce the cognitive efforts required by the user to accomplish some task. I found the evolution of tools from paper to slide rules to calculators to graphing calculators a particularly interesting topic, although I tend to wonder at what point is the machine doing everything for us, instead of just assisting us.
Zhou Li - Apr 15, 2008 03:14:07 pm
We encounter the usage of information visualization in our everyday life, especially when computer generated data visualizations are getting easier to produce. Even without the justification of this reading, well designed graphs generated by excel or 3D models produced by matlab obviously help us to see the relationship among plain data points much better. The reading explains the reason behind this enhancement in more detail and in a scientific way. information Visualization acts like a visual aid to our cognition process. Partly because it allows us to extend our working memory by holding partial results outside the mind. However, not all visualizations are helpful, some them might even convey the wrong idea. The example of the O-ring damage in the reading illustrates this point. Choosing the right type of visualization requires a good understanding and insight of the information at hand.
Robert Glickman - Apr 15, 2008 03:22:28 pm
This reading was very interesting and particularly hits home for me, because I am very much a person who relies on charts, graphs, and visual representations of data to understand them well. When I recall information, I usually recall where it was on a page that I read, or if there is an associated chart, what the chart looked like. One quick comment I must make is that this article definitely needed some better editing as I was constantly seeing words and especially homophones misused, a pet peeve of mine. This was a slow read for me because I have such a fascination with visual representation of information, that I would stop at every chart and scrutinize every detail, dissecting the meaning. But it certainly was a fast read as well, since it didn't seem as slow as it was. I especially liked the parts which incorporated biology, including the discussion of the eye, different parts of direct and peripheral vision, and the blind spot.
Scott Crawford - Apr 15, 2008 03:29:16 pm
Visualization is a very very cool thing to develop. Embellishments can either get in the way, or provide a great extra touch to a product. The difference between over-embellishing and making things look cool but still be effective is a line that is difficult the thread. Personally, I find spartan visualizations very effective.
Henry Su - Apr 15, 2008 03:22:26 pm
I definitely agree that visual information is a key aspect of communication. Many people, in fact, learn visually. That is why they say a picture is worth a thousand words. Visual information, such as diagrams, tables, graphs, and layouts can convey more information than a page of words can. Perhaps even more importantly, the information conveyed by visuals is often easier to interpret and absorb than textual information. The user can just glance at a visual, and see exactly what is going on. For this reason, it is important to design visuals in a way that communicates information as effectively as possible.
Andrew Wan - Apr 15, 2008 03:22:23 pm
Given the general similarity with which people interpret visual information, it comes as no surprise that certain types of visual models are essentially standard. I found the reading's discussion about using visualizations as way of extending cognition most interesting. Having an extra space for thought, be it visual, auditory, or other, provides an enormous amount of use to an individual. It seems interesting that while 3d graphics have been commonplace for a while now, few applications use them to actually extend human faculties. That said, there are any number of good reasons for this, least of which being the limitations of 2d-screens and input. For now, 3d visualizations remain most useful for data representation, but new technologies are already extending what people can do in a given environment. Case in point: geolocation on cellular phones, which provides a way of organizing physical information that would otherwise be difficult to keep track of.
Daniel Gallagher - Apr 19, 2008 11:25:19 am
This reading was quite long but interesting. It made explicit many of the concepts I've implicitly believed about visualization of data. Extending human long-term memory through putting down data (in books for example) is probably obvious to everyone, but I feel like people do not think as much about using external aids for things like multiplication- they just do it and move on. The multiplication visual aid allows users with only the most basic of times-table knowledge to compute large problems, and it makes me exited to consider whether there aren't many other visualization techniques that could similarly enhance the human mind people have yet to create (digital or otherwise). Another thing that I found interesting was the description of human vision in the second half of the document. The image on p25 of a photograph distorted to show the visual field at a point in time was especially nifty- I knew humans only really see what's in front of them clearly, but now I've been paying attention to just how narrow that really is.
