Visual Information Design

From CS 160 User Interfaces Sp10

Revision as of 03:47, 16 April 2010 by Mattkc7 (Talk | contribs)
(diff) ← Older revision | Current revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Lecture on Mar 31, 2010


Contents

Readings

Mattkc7 - Mar 31, 2010 08:16:32 pm

Alexander Sydell - 3/27/2010 2:34:50

Most of the various examples and descriptions of visualizations in this reading seemed to contribute to the notion that the more data there is, the more a visualization can help a human understand and process the data. I would even venture to say that the increase is even exponential ("a picture is worth a thousand words" is a good example of this). For example, having pencil and paper when solving a multiplication problem helped somewhat. However, when a broker-dealer needed information about a thousand bonds the visualization in the example was incredibly helpful. Thus, it is probably wise to pay more attention to how an interface with large amounts of data presents it and spend less time on the subject in interfaces where there is relatively little data shown to the user.


Daniel Ritchie - 3/27/2010 10:49:35

I'm intrigued by the idea of multi-modal information visualization (or "sonification" and "tactilization," as the authors call it). Vision definitely provides the most bandwidth of all human senses, but why not augment it with the other available senses? Consider the so-called "retinal properties" of methods for data visualization. I would be very interested in seeing a study like MacEachren's from 1995 that examined the effectiveness of different "aural" and "tactile" properties. For instance, how effective are pitch, timbre, and volume at distinguishing the three different classes of data? How about tactile pliancy, texture, and heat? Perhaps with some solid data on their effectiveness, these properties might find their way into more real-world systems. I can see how such multi-modal infovis systems could also help vision-impaired individuals understand data.


Mohsen Rezaei - 3/28/2010 20:01:59

As discussed before data presentation and visualization is one of the things that researchers are still working on to better understand how users like to view raw data from experiments or any other data gathering events. Diagrams, whether active or static, and large-scale data monitoring are two types of data visualization that has been tested and worked well with users. This is important in User Interfaces design since after experiments we have seen that Cognitions is proportional to visualization, which by other means, good visualization can amplify the way users understand data and graphs. There are some steps into data visualization that should be carefully followed. Steps are taken as follows: carefully map the raw data from user to tables, then show them in a visual structure way that helps users understand the material and data well. So things that we need to be careful about in data presentations models are tables, graphs, and charts' design, structure and layout.


Tomomasa Terazaki - 3/29/2010 18:12:31

Like last reading, this time was about visualization but it was different this time. It was more about information visualization. Therefore, it concentrated more on how to make an information, data, or statistics into a visual form like graphs. I did not really enjoy reading this article because it was something that was obvious like making an effective graph so it makes sense. It is obvious that we need to make graphs so that the people looking at it would understand but it did not go into the details of how to make it because those are for the later chapters that talks about in more details (this is the first chapter so it talks about a lot of things in small details). However reading the article reminded me how much I like the periodic table of elements. When I first saw it in middle school, I remember saying, “what is this” but when my teacher started explaining why it was organized that way I was amazed by how much it made sense why all the elements were put in that weird format. At first view it is annoying but afterwards, I noticed that if it is not in that format it would be very difficult to memorize all the information like number of electrons outside or other things I had to memorize for my science class in middle school.


Michael Cao - 3/30/2010 10:28:28

Visual data is usually much easier to understand than just data written out in plain text, but it also depends on how that data is visually displayed. In the lecture about quantitative analysis, we looked at three different ways to represent data and try to determine which one worked best. These were the bar graph, the area graph, and the tree graph. Clearly the visual representation you choose affects how the user interprets the data. And while visual data usually makes data easier to understand, it sometimes might make it more confusing, and sometimes certain visual representations might be more advantageous.


Jason Wu - 3/30/2010 15:23:56

For me, one of the key insights of this reading is that "the power of the unaided mind is highly overrated." Humans are visual creatures, so it makes sense that finding methods to display information in a compact yet insightful way can help people amplify their cognition, thus allowing them to solve problems much more quickly and efficiently. The authors' examples of using pen-and-paper for multiplication, navigation charts, and nomographs brought to mind a type of external cognition aid that helped me quite a bit with boolean algebra in classes like CS61C: the Karnaugh map, which takes advantage of human pattern recognition abilities to reduce to amount of work necessary to simplify boolean algebra expressions. With such powerful computational tools at our disposal in the present and near future, it is an exciting time to be working on user interface design since their are many opportunities to apply information visualization principles to create interfaces that allow people working in a variety of different fields to work more effectively.


Nathaniel Baldwin - 3/30/2010 18:13:37

This reading started out promising - the authors presented a clear argument for the usefulness and importance of presenting information in well-designed visual way. Somewhere around figure 1.16 - a truly hideous UI example, data visualization aside - I felt like the article devolved into a near-endless stream of terribly dry definitions (many of which it sounded like the authors were making up on the spot), lacking interesting real-world examples and just not providing me with much of interest. The "visualization is important" got repetitive, and I had to struggle through the rest of the article without feeling like I'd gained anything.


Matt Vaznaian - 3/30/2010 20:54:42

I know that I myself am a huge visual leaner. Visualization helps me understand and remember almost everything better. Cognitively I feel like I can correlate and analyze data better when visualization is used. The reading uses a lot of graphs and maps as examples different visualization aids and I agree that these are a great help in mapping data to a visual which can be easily processed.


David Zeng - 3/30/2010 22:11:18

While information visualization can significantly speed up cognition by increasing the working memory as well as as presenting a large amount of information, using it incorrectly can lead to large amounts of time wasted. For example, the chart on page 23 took me 2 minutes to understand what it could possibly meant. This was due to the incorrect ordinal relationship that was expressed. At the same time a chart or visual aid can unintentionally draw the attention away from the focal point of what the information is supposed to represent. For example, if you used a bar graph to show a histogram of student scores, most people would look at the raw data and notice things like the mode or the mean. However, if you wanted to demonstrate that there was a large variation in scores, that information wouldn't be noticed such as easily.


Daniel Nguyen - 3/30/2010 22:20:50

In the reading, I found the section section "How Visualization Amplifies Cognition" to be most interesting and useful. By referencing aspects of the Model Human Processor and in what ways visualization could optimize it, referencing working memory, search time, perception, the reading makes the technical benefits of visualization very obvious. Additionally, I feel that this reading approaches the concept of visual grouping of related information/objects much better than the previous reading did. Although these two readings accomplish the same thing overall, emphasizing the importance of grouping visually, this reading does a much better job of justifying why in more technical terms. I especially liked the mention of exposing patterns in the information in the reading, because I feel like this is something that can benefit both the user and the designer as opposed to just the user.


Annette Trujillo - 3/30/2010 22:38:40

The main thing that I got from reading this article that relates to user interfaces is that it reminded me of the issue of recall vs recognition. Information visualization is important because we want to make as much information as needed visible to the user, but without cluttering or providing too much information. It is easier for people to do tasks when they don't have to remember too much stuff, because our temporary information is limited. If we design our interfaces using appropriate information visualizations, this makes for a more user-friendly interface.


Bobbylee - 3/30/2010 23:07:13

I definitely agree that information visualization is getting more important in the world full of information. If you are trying to explain the information of a spatial object, it is better that you draw it out. I recall that when you have to design the structure of the a program, it is more recognizable to you graph the data structure on the paper or on the computer. Another thing that I have in mind is that, information visualization make me recall some interactive learning programs. For example, when one learns how to delete a node from a tree and something like that. It is easier to learn than just reading the instruction although it is best to have both.


Kathryn Skorpil - 3/30/2010 23:31:39

I noticed on some IQ exams that they have a portion where they give you a 3D shape on paper and ask you to select one that actually has a realistic rotation. I am sure some people have a very difficult time trying to visualize a 3D shape in their mind and rotating it, but I am not quite sure if this indicates someone being smart or not. Some people just have a better time visualizing things. Using some kind of program like Maya would certainly help people who can't visualize these types of things in their heads.


Jonathan Hirschberg - 3/31/2010 1:54:23

It is important that visualization techniques are used to simplify our mental computations by extending working memory into the external world. User interfaces should use these techniques to aid users’ computational tasks. These help for tasks where visualizing something facilitates completion of the task. Are there tasks where it is easier to think it through mentally than to visualize it? Would people use computers to solve these tasks? For example, procedural memory (like riding a bike) is not something you can simply visualize or describe. You have to just do it.


Sally Ahn - 3/31/2010 2:24:29

As they define "visualization" on page 6, the authors write, "the purpose of visualization is insight, not pictures." I think that the authors' point was that the goal of visualization is to facilitate understanding rather than create pretty pictures. However, this made me wonder about the possible relationship between the effectiveness of visualization and its aesthetics. A quick search on Google led me to a site that examples of "aesthetic" visualizations: http://pingmag.jp/2007/03/23/infosthetics-form-follows-data/ . At one point, it describes "information aesthetic infographics" as visualizations that "appeal both the mind and soul. While they positively stimulate our senses, in terms of engagement, involvement, and imagination, they are also optimized for the specific task of conveying complex data-driven concepts in intuitive and easily comprehensible ways." However, I think that some of these examples illustrate how this endeavor to be visually pleasing detracts from the usefulness of the visualization. And of course, "aesthetics" is much harder to define (as is the word "soul") and measure quantitatively. It would be interesting to explore how one can achieve an optimal "balance" of clarity and visual appeal.


Darren Kwong - 3/31/2010 9:25:57

This reading relates to previous concepts and readings. One of the key things in visual information design appears to be the reduction of articulatory distances. Also, an appropriate human information processor model can be developed and applied to the improvement of information visualization.


Weizhi Li - 3/31/2010 10:22:58

The reading does very a good job of systematically enumerating the key visualization processes. I think the concept of visualization is what makes computer friendly to the user and the data easy to understand. The reading provides a lot of UI design examples. I think it’s especially helpful for our project which requires us to display information in a mobile device which has a representatively small screen. Visualization is very important to UI design and should be considered very carefully during our project.


Wei Wu - 3/31/2010 10:32:27

Figure 1.15, which details the cyclic process of a knowledge crystallization task, is very similar to human-computer interaction diagram in the Direct Manipulation reading by Hutchins, Hollan, and Norman. However, the former puts the process more in the context of the human actions of gathering information, analyzing it, etc. Unlike Hutchins' diagram, Card's figure illustrates the phases of accomplishing a task that does not necessarily involve interaction with a computer. In fact, Card's article suggests that computers and visual representation can only help in analyzing abstract data to reach an accurate conclusion, that computers make it faster and easier for humans to reach those conclusions. This seems more optimistic than Hutchins, who talks about how human-computer interaction can either expand or close the "distance" between a user's goals and a computer's operations -- that is, some interfaces may make accomplishing human goals much more difficult than otherwise.

While visual representations from computers can dramatically aid in the interpretation of data, as Card argues, it is important to remember Hutchins' discussion of the semantic and articulatory distances created through HCI. Card presents different methods of visual presentation but offers little guidelines on what methods are more suitable for certain types of data. It is the form that this mapping of data to representation takes that actually determines whether the representation can provide greater insight to accomplishing a person's task.


Charlie Hsu - 3/31/2010 11:10:59

I felt that the six ways visualizations can improve cognition were the most important part of this reading, and provided inspiration for thought on how to design information visualization such that we maximize cognition improvement. Increased mental resources and enhanced recognition of patterns might come from us parsing and filtering data to only give the most relevant data in a well-organized manner. Reduced search might come from us automating the search of data via queries to a database. Designing good visualizations that afford inferences through the information, can be manipulated, and can highlight points of interest are also good ways to amplify cognition.

Mapping data structures to visualizations is where these principles come into play, and the reading shows that we need to keep many things in mind when designing information visualization, from Gestalt principles for graphic design to human perceptions: thinking about focus and peripheral perception.


Linsey Hansen - 3/31/2010 11:55:45

The main part I remember from the reading is the 6 outlined elements for knowledge crystallization, which is something that I found pretty useful. I mostly felt that a lot of the reading was about what you should do when designing a visualization for something, but it didn't always specify how to do it. I mean, it did point out individual strengths of certain visualizations, but those were strengths specific to that object, and couldn't exactly be applied to every visualization in existence. Anyways, the elements and their examples were nice, since they helped break down the thought process for coming up with a visual representation of some sort of information, and I feel like that short summary is a lot easier to remember then all of the varied examples the rest of the reading had.


Arpad Kovacs - 3/31/2010 12:05:37

The goal of visualization is to amplify cognition by providing external aids that leverage human perceptual abilities, which decreases search time, enhances pattern recognition, and reduces the load on working memory and processing resources. Upon viewing a visualization, the observer takes raw data, transforms it into data tables, and maps this to visual structures, which create a direct manipulation interface by encoding the data in a manipulable medium. This effectively reduces both semantic and articulatory gulfs of execution and evaluation in the process of knowledge crystallization, which involves gathering information, encoding it into a schema, problem-solving, and packaging the results into an output.

I think that it is useful for visualizations to rely on recognition, rather than recall, and thus diminish the need to pull in information from long-term memory. Thus the visualization can raise the cost-knowledge characteristic function to increase viewers' foraging efficiently by finding more information per unit of time. However, as the chapter mentions, badly designed visualizations can actually reduce cognition by misleading the viewer. For example, 3D pie charts often distort the perceived area of each sector due to perspective effects, and could lead to false insights and misguided decision-making.


Jessica Cen - 3/31/2010 12:19:33

I agree with the authors when they state that information visualization is “a new upward step in the old game of using the resources of the external world to increase our ability to think.” In addition, I believe that an example of how information visualization is making us smart is a good and intuitive computer interface. If a computer interface is effective in its graphics and it provides intuitive ways (such as direct manipulation) for the user to complete tasks, then it benefits the user’s cognition. However, if the interface misuses graphics and confuses the user, then the information visualization is not effective.


Chris Wood - 3/31/2010 12:24:45

What we see surely has one of the biggest impacts on how we think, and certain visual cues are necessary for instilling specific trains of thought. How information is presented on a page has profound impacts on whether or not a user is able to understand the presented information and use it in a constructive way. I agree with the article in the idea that computers have the ability to present information in such an efficient way that it prompts cognitive growth and makes people literally smarter and more creative. Clever visualization schemes can reveal patterns in data, making people more perceptive. It is interesting because as we learn how to better represent information to amplify the potential of our minds, we get better at representing information to amplify the potential our minds. I agree with these qualitative ideas, but the quantitative stuff about measuring eye movement and the properties of a retina only served to cause me immense amounts of grief.


Dan Lynch - 3/31/2010 12:27:16

"To the thinking soul images serve as if they were contents of perception ... that is why the soul never thinks without an image. This quote from Aristotle describes how humans process information; simply, thought is not possible without images. Humans harness an innate ability to process visual information: visualization simply provides high rates of transmission of information and ideas between humans. For example, the human visual system can interpret a 3D image even if it is on a screen or paper, often providing a spatial understanding of the object of reference. This is often much faster than processing a 2D image and coming up with a 3D understanding. In other words, the amount of data that can be transmitted to the human brain is far increased with 3D renderings as opposed to 2D renderings with supplementary data.


Long Do - 3/31/2010 12:56:32

While it is interesting that visual information helps increase cognition, at what point does it cause a decrease in cognition. The ability for visual information to help us better understand requires some mental processes to understand what is being presented to us and with enough of these information, the mental processes might take up more than it would save. Being selective of the information, just like in the example with the O-rings for the rockets, is more important than the information itself. Because we have to be selective, we have to have some idea of what kind of information we are looking for before we see a pattern. We may also not be able to find small patterns. Computers are better at finding patterns and connecting coincidences than humans in this sense and so I think the need for visual information to determine patterns might be on the decline.


Vidya Ramesh - 3/31/2010 13:03:48

Information visualization deals with using visual representations of data to aid in the acquisition of knowledge. This is very important in the task of knowledge crystallization, which is when a person gathers information, constructs a visual representation of it, and in the process understands the information while packaging it for other communicative or active tasks. The authors point out this is extremely useful when looking for or detecting patterns within the data. Most of the visualizations produced do the former so they are identified to be visual knowledge tools. Other important characteristics of visualizations is the fact that they allow inferences to be drawn and reduce the search for data by visually relating information. This compacts the information into a small space and allows for hierarchical search.


Joe Cadena - 3/31/2010 13:10:59

Quoting the authors, "The purpose of visualization is insight, not pictures," I have come to realize the importance of the effectiveness of a view. As we develop an application, we want users to be able to interpret our design, make decisions based on the visual prompts, and receive the intended feedback. Amplifying the user's cognition through the use of accurate information visualization ensures our program is an effective tool.


Richard Lan - 3/31/2010 13:27:15

The big idea of information visualization is taking large collections of data and arranging them so that useful inferences can be drawn just from looking at it. A simple example of information visualization is a graph, which can demonstrate positive or negative correlation between a set of data points. More complex examples are computer-based, and oftentimes interactive, allowing the human user to manipulate the visualization by adjusting variables and other parameters. Information visualization is highly useful because it helps humans think by presenting data in a coherent way. The value of the visualization then lies in how effectively the information is presented, and how clearly the relations among the data can be seen. Visual aids increase a person's working memory, allowing the user to gain high level insight rather than getting stuck trying to understand the disparate data points. Visualization can be applied to physical phenomena in scientific settings, but is most useful when applied on more abstract data, such as stock prices and movie data. Taking raw data and changing it into a visualization involves applying data transformations to the data, first to organize the information, and then to map it to a visual structure. A good mapping highlights the relationships within the data, and do not introduce false relationships.


Kyle Conroy - 3/31/2010 13:34:29

External cognition is a powerful idea presented in the reading and can be utilized to interact on a deeper level with users. While many programs, such as calculators and spreadsheets, simply show the result of an operation, it may be useful in certain situations to explain a process to the user. Instead of simply showing text based steps, we can write out the process a user would take if he or she were using a visual aide, and therefore partaking in external cognition. Wolfram Alpha takes this approach when showing users the step by step guide to certain problems, such as integrals, that the search engine can calculate. However, it is also important to remember the limits of external cognition. Richard Feynman tells the tale of trying to explain certain math tricks to a man who relied on using an abacus for calculations. The man did not understand numbers, only methods of bead movement on the abacus. Thus, external cognition should be used a crutch, it which a user does not understand the underlying concepts, and only knows the process by which to get results.


Owen Lin - 3/31/2010 14:52:41

I think that designing visualizations play a very important part in interfaces, especially on handheld devices with limited display sizes. The advantage of visualizations is that a good visualization can represent a lot of data in a small area, so we want to be able to make use of visualizations to optimize the efficiency of our interface. We have already made decisions on our iPhone app for singers that lend more toward a 'visual' interface. For example, originally we planned to have our application be able to open sheet music files and display the music in the traditional format of staff bars, notes, clefs, and key signatures, but due to the limited size of the iPhone screen, we switched to the Rock-Band style lines because we feel that it conveys much of the same information and allows for much less clutter. I feel that visualizations play an important role in human interface design.


Conor McLaughlin - 3/31/2010 15:04:23

Card's citation of Norman's explanation of the limitation of human ability, and its need for external aids to extend intelligence, shows all these readings are starting to come together into one coherent vision. The need for good visual representation of data, and its ability to communicate the actual pertinent information of the data, such as the Challenger booster damage, is incredibly important when designing a user interface that represents large amounts of data or statistics or even the state of the system. I think the article does a good job in understanding the actual difficulty of interpreting or evaluating the important results of data, however, and I appreciate the systematic process it attempts to formulate to first UNDERSTAND the data, with data tables, etc, before actually going about implementing a visual representation. I actually felt the system described reflected the product design cycle, where one needs to plan, test the initial designs, and completely understand the system before beginning implementation. The design cycle is here being applied to the representation of data, but when thinking of an actual application, at a higher level that is all any application is attempting to accomplish.


Boaz Avital - 3/31/2010 15:13:38

I actually rather like the distinction between what the author calls scientific visualization and information visualization. The actual name "scientific" visualization may be more of a historical artifact of the visualization of physically-based data than anything, and it may not apply nowadays when a perfectly valid physically-based visualization is "Best places to get coffee", something few people would consider scientific. Aside from that, I think it can be incredibly important to distinguish visualizations of data based in the physical world and visualizations of abstract information. It's possible that people have expectations when viewing physical data, they may expect the visualization to map in an obvious way to the physical source from which it was construed. It may be easier to understand and analyze it that way, which is the goal of any good visualization. Meanwhile, visualizations based on abstract data has different considerations. It could be that people are used to viewing certain data in certain ways (eg stock data in line graphs). However this field could be more cognitively flexible for people in terms of what ind of visualization they can easily understand. I also find the broader idea of "perceptualization" interesting. Adding extra dimensions to the way we view data is tricky but has the potential to be very useful if the use is correct and well thought out.


Hugh Oh - 3/31/2010 15:15:07

The article talks about the relationship between data and the space it occupies. Something very interesting I took from the article is the conclusion that empty space can be used as a container. In addition, empty space can convey useful information when complimented with actual data. In conclusion, empty space is not an inefficient use of space but actually very helpful.


Brandon Liu - 3/31/2010 15:20:32

One interesting visualization is the design of the BART map: by default, the map is different from other transit maps since it uses physical location rather than pure schematics for its layout. It's somewhere between the all-the-same-angles map shown in class (Like on the London Underground) and the aactual physical map: for example, the maps show that trains run north through Berkeley when they actually run north and then west. My guess for why the map was laid out this way was sot that users weren't misled by the relative distance of stations. At the ends of the system, stops are very far away, while on Market street they're very close: a pure schematic diagram would make SFO airport seem much closer then it actually is. The visual artifacts specific to the BART map aid us computationally since we can estimate the time it takes to travel to a certain destination, without being cluttered with all of the changes in direction trains physically make.


Geoffrey Wing - 3/31/2010 15:22:40

Many people consider themselves "visual learners" and this reading affirms the importance of the visual interface. Knowledge crystallization is an important concept to consider when designing an interface. If users can understand the data/output they can correctly draw conclusions and accomplish their tasks. Whenever I have a problem, I often find myself writing things down - I often outline papers or write pseduocode before I actual set out on the actual assignment. As with the multiplication example, writing things down definitely helps my mental processes.

The way we are shown data is becoming increasingly important as we have more and more data to manage. As I learned in another class, humans are adept to pattern recognition so presenting data in the clearest and easiest way for us to process. This reading is especially helpful for my group's project, as we will be creating visualizations for attendance data.


Saba Khalilnaji - 3/31/2010 15:45:39

Although diagrams and navigation charts are great for relaying information to the view; skewing and scales can often distort data toward another conclusion. However, visualization is more effecting means of giving the viewers insight in their decision making process. But visualization leads to the problem of perception. If you have a diagram highlighted or color coded in certain regions to provide better visualization, each view may have their own perception of the colors and therefore come to different conclusions. It is important to use conventions with such things, for example weather maps use specific colors for high/low pressure. Designers can bring attention to certain regions by having it stand out, especially the user has to repeatedly process the data. The example given was that a driver will notice a car changing lanes more than a car merely driving straight because it is "new information"


Aneesh Goel - 3/31/2010 16:13:51

Knowing how to visually present information to users has obvious value when designing user interfaces; well-displayed statistical information, as the article demonstrates, will lead to conclusions that poorly organized data will cover up and ignore, as in the space shuttle O-Ring demonstration.

However, there's another use for designers that stands out; the discussion of organization of information and especially processes is valuable in the design phase itself, when trying to predict user interactions and workflow. Well-designed representations of workflow will allow a quick grasp of what the critical tasks are, what ordering is salient, and what should be grouped but not necessarily ordered; poor representations will devolve into a list of tasks, possibly a nested list of subtasks, and provide little else to guide how those tasks are implemented. Just as the designer needs to learn to provide a visual extension of memory to the user, they need to provide one for themselves as they plan and begin their work, and even later on as they enter into lo-fi prototyping.


Yu Li - 3/31/2010 16:18:29

Information visualization is when large-scale collections of non-numerical information is visually represented. For users, it is much easier to see something than imagining it in their mind. This is why information visualization is essential in iPhone development. For example, compare giving a user a list of locations versus giving them a map. With one glance the user can see how the different locations relate to each other from looking at the map, while with the list, they have to read the list and then relate the data by themselves.


Esther Cho - 3/31/2010 16:24:41

I think this article is indirectly addressing one of the usability heuristic, "Recognition rather than Recall." By allowing information to be visible for the user, the user's performance on a system is increased while the mental workload is decreased. Although the recognition rather than recall heuristic is important in design, the article illustrates a great point that just because data is visible doesn't necessarily make it useful. It's how the data is presented that would be most useful to the user.


Vinson Chuong - 3/31/2010 16:31:13

Information visualization seems to be all about searching for encodings that reduce the gulf of evaluation for getting the desired information out of a set of data. The examples in the book seem to hint that patterns are most easily observed when the data is encoded using a metaphor that relates it to attributes of physical objects in the real world, which is reminiscent of our discussion of the directness of interfaces.

We can further apply the concepts discussed in this reading to more than just illuminating data sets. We can apply them to our user interfaces. What's the best way to encode a set of choices or tasks that a user can choose from? Are there better alternatives to the typical widgets we use?


Calvin Lin - 3/31/2010 16:31:30

When we use information visualization as the basis for thinking, interpretation, computation, etc, we are essentially placing ourselves in the context of another mindset, or a different “world” where certain operations and data representations are changed. The chapter describes this as visual external artifacts that aid and enhance cognitive abilities. But I don’t see it as aiding our thinking. When we use visualization to perform some action, we become constrained and dependent on the scope and limitations of the representation. Little, if anything, carries over from each method. For example, doing 15 * 12 in your head is a very different process than using paper. On paper, you do partial computations [(15 * 2), (15 * 1)], write them in the right columns, and then do addition. Having much experience reading a 3D graph doesn’t directly transfer to solving the same problem with the data in a table.

When we rely on data visualization, the problem becomes less about how good one’s cognitive abilities are and more about how good the design and representation is. I suppose this may be more desirable because we can put much effort into optimizing a design so that cognitive effort is minimized, and therefore a wide range of users of varying skill can solve the same problem. A meteorologist probably can interpret a bunch of numbers and data and predict the weather, but most people can probably look at the maps they use on TV and interpret where it’s going to rain, where it’s humid, and so forth.


Wei Yeh - 3/31/2010 16:38:28

The way information is visualized greatly affects our ability to interpret data. Take for example two apps that helps you figure out which BART train to catch: bart.gov's QuickPlanner and The BART Widget (http://worrydream.com/bartwidget) for Mac OS X. Given a starting station and an ending station, the way the two apps display train times are very different. QuickPlanner shows train times one after another, in chronological order. One has to build a mental model in his head to figure out when the immediately next train is leaving and the time intervals between trains. This causes unnecessary cognitive load. The BART Widget, on the other hand, nicely lays out bars representing each train, one above the other, in an infinite graph over time. This makes cognition much easier, since one can visually determine how soon a train departs and time intervals between trains. Additionally, with routes requiring transfers, the BART Widget's bar lengths correspond to the actually trip times for each trip segment, whereas QuickPlanner evenly divides all segments.


Mikhail Shashkov - 3/31/2010 16:59:23

All of these examples and approaches seem very focused on data visualization for scientific work rather than anything we would do in this class. Nonetheless, a decent read, although long. By the way, why don't they have small screen with stuff in keyboards? Like the Nintendo DS... Seems like another place to put more external aids.


Angela Juang - 3/31/2010 17:00:15

This reading talks about the difference between interactive visualizations and non-interactive visualizations, and seems to emphasize that interactive diagrams are more effective than non-interactive ones. However, how do you tell if a diagram is interactive or not? Whether or not a person finds any particular visualization to be interactive depends on their own individual experiences and knowledge. For example, a person who knows music may find a diagram containing musical notation "interactive", while someone who does not know how to read music will not be stimulated by it. While this is a very blatant example, more subtle differences can make a difference in how people interact with diagrams in a similar way. How should we judge how interactive a diagram is?


Eric Fung - 3/31/2010 17:01:46

I think it is important to highlight emphatically that there are multiple ways of displaying the same data. In the example of the ship navigator, the navigator can choose from multiple kinds of maps (Lambert and Mercator projections), depending on the task. The Mercator projection is ideal for calculating straight lines between smaller areas, but the Lambert projection more accurately represents straight lines between longer distances, due to the curvature of the earth. And both projections do not totally accurately reflect the proportions of the earth; a globe would do that. But a globe is impractical for a navigator's purposes.

The point is, there are sacrifices to be made in order to reap the benefits of using any visual representation at all. Much of the process of selecting which representation is concerned with what your goals are: detecting overarching patterns? reducing the time it takes to search for information? an easier way to view changes in a large data set? a combination of these?


Andrew Finch - 3/31/2010 17:25:23

As the article states, visualizations play a key role in conveying information and making it more interpretable. Most data sets appear fairly meaningless when viewed as a simple table of data, but when transformed into an image, graph, or diagram, they suddenly become much more meaningful. One thing this article does not cover very thoroughly is how important it is to choose the correct form of visualization for a given data set. In many cases, a substantial amount of planning and design must go into this selection process. Information represented with an inappropriate visualization looks just as meaningless as it did when it was just a table of information.


Bryan Trinh - 3/31/2010 17:26:04

Recognition rather than recall is the guiding design principle of this paper. It provides a myriad of examples that show that by allowing the user to recognize tasks rather than recall them the cognitive ability of the user is much higher. This paper makes clear that by making your thoughts visual one can amplify cognition. This is not only true in groups but also in individual tasks. By drawing things out, we expand our working memory to a larger pool of information.

Although the paper does not present a list of hard fast rules for design as previous papers have, it provides a philosophical grounding for understanding different visualization concepts.


Jeffrey Bair - 3/31/2010 17:27:37

A UI design oftentimes is a success or a failure just from the information visualization. Displaying information in an easy to read system and being able to interact with this interface can make or break a design. Navigation charts are a great interactive interface whereas diagrams can display a great deal of information or lack thereof. In our own project we have to make sure that the information that we present to the users is useful and complete. Having an incomplete amount of information shows the user that we have a lack of information about our subject and thus an incomplete product. Being able to visually see the data that we present is an important aspect of the project and we have to keep in mind that organization of our data can help in creating our product. For example, for our skater app we have the rating system which we keep a standard for and we can arrange the data in that order. Being able to create visual information allows both the user to enjoy the product and use the product more effectively.


Spencer Fang - 3/31/2010 17:28:08

The ideas in this chapter can be seen in many modern file managers. The OSX Finder allows users to attach color labels to files, which make them stand out from one another. This feature increases Perceptual Monitoring because it makes files stand out by their appearance, and is especially true if a user chooses to sort the files by label, which will group together files with the same label.

Nautilus from GNOME is interesting for its spatial mode feature (also found in Finder from older versions of Mac OS). It opens each directory in a new minimal window, which displays only its contents (no navigation toolbar). The idea is to have the user think of each directory as an independent entity, instead of thinking of the filesystem as the data inside of a file browser. With this feature enabled, the window position, window size, and the arrangement of each directory's contents are remembered to complete the metaphor. Now the user can for example remember that school related directories always open windows on the top left of the screen, while music related windows always open on the bottom right. This improves Enhanced Recognition of Patterns and Reduced Search through the spatial arrangement of windows and their contents.


Peter So - 3/31/2010 17:33:53

Great read. I like to take notes in the form of mind maps and this reading validates concepts that support why mind mapping may be a better way to take notes because of the similar nodal structure of thoughts in the brain. In mind maps, a central idea has related ideas branching out to not only organize thoughts spatially but provide a hierarchy of ideas as well. Its a creative way to use space to better remember similar ideas.


Richard Mar - 3/31/2010 17:35:55

Information visualization is something that isn't always done well in games. Take radar displays or map overviews: in a lot of games, units and structures tend to be represented as colored dots, with green representing the player's possessions and red (or a team color) for enemies. The problem with this approach is that the player only knows something is at the location of the dot, and not much else. Using extra colors forces the player to spend extra time looking at the display, and can lead to confusion. Unfortunately, due to size constraints for mini-maps, this cannot be avoided.


Long Chen - 3/31/2010 17:38:12

I am a strong believer in visualization of mental operations. The first example about multiplication really hit home for me. When I do any double digit multiplication I visualize it just as if I had pencil and paper in my head. I "write" down the numbers in their proper places and "record" the numbers as if they were written down. This way I can devote my mental capacity to the next multiplication. This visualization has been the way I do math since elementary school, and I haven't even though about how other people might go about it.

Visualization to show others how people think is also a powerful teaching tool. I find it particularly helpful whenever one of my TAs realize to explain his mental process by drawing webs or diagrams on the board demonstrating the step-by-step thought process. This visualization of the compartments of thoughts floating in the brain helps teach not just the material but also strategies to tackle the problems.

Although information visualization has not been a subject of study until the 80s, I believe the concept has been around for generations. It is such a natural step to take when describing one's thought only in words becomes too difficult to express explicitly. Now, many of the major innovations extending from this class of user interface is providing tools to freely and simply visually express one's self. Any new hit software has visual tools that have evolved drastically from the prompt-based computing of the past. One reason the business world primarily uses PowerPoint and Excel for their main functions is the simplicity of the graphs and charts available for them to display the data in an intuitive manner.


Alexis He - 3/31/2010 17:45:18

Information visualization seems like a very smart way to represent and manipulate data for visual learners. But what about auditory, kinesthetic, and other types of learners? Wiki describes visual thinkers as 30% of the population. And it seems from this article that they are very well represented in research, but there may be even more innovative techniques for conveying information in these other areas.


Raymond Lee - 3/31/2010 17:48:04

External cognition brings to mind the design principles of recognition vs. recall and keeping necessary information in view when a user is performing a task. The designer cannot trust the user to have full concentration to remember all portions of a task, so it is helpful to have some record of the task as it is being completed for the user's reference. An example of this is adobe photoshop's History panel, which displays all actions the user has taken so the user doesn't have to remember exactly what stroke or filter he used in the past.


Jordan Klink - 3/31/2010 17:49:23

Like last reading, I once again enjoyed learning about the more aesthetic side of design. This reading in particular, though, had much more specific focus on organizing data. It brought to my attention that one should not simply collect data and throw it at the reader for him/her to manually interpret. In fact, just the opposite should be done: careful consideration should go into how the data will be shown to a user, as with a large amount of data, one can easily be overwhelmed. I wonder, though, just how much time should be spent into organizing data rather than collecting it? Perhaps that is a very important question to consider when researching.


Wilson Chau - 3/31/2010 17:53:24

This reading really reminded me of something I learned in psychology which was that a lot of how we interpret something has to do with the way it is presented to us. This reading helped to remind me of some of those principles and also dove a litter deeper than what I had learned before.


Richard Heng - 3/31/2010 17:55:10

The reading mentions that scientific data can be visualized using physically based information and information visualization can be visualized based on abstract information. It mention that there is no obvious spatial mapping for the information visualization, but one that is efficient can be created. Perhaps it might be useful sometimes to pretend that scientific data has no physical mapping, and try to find an efficient mapping. Better, less obvious visualizations might be found this way.


Victoria Chiu - 3/31/2010 17:55:54

Diagrams are helpful when we read it correctly. The same set of data can be interpreted to very different results looking differently. There are also a lot of different diagrams to present the same set of data. We can use marks to signify things. By using line and point marks, we can create topological structure such as graphs and trees.

Divya Banesh - 3/31/2010 19:46:33

I think one excellent form of visualization that exemplifies the visualization techniques from the paper is sign language (esp. American Sign Language). Since deaf/dumb people can not use speech, they have to do everything with their hands and many of their world and letters visualize the actual objects. For example, the sign for the letter "z" is to write out the letter 'z' in mid-air. Another example, is that the sign for "hill" is to draw a mountainous figure with your hands. Visualization is useful in many parts of peoples' daily lives.



[add comment]
Personal tools