About Face. Ch22 Review

Chapter 22. Menus

Menu has very long history from command line menu to today’s standardized menu. Very earlier command line menus are can be shown in sequential hierarchical menus, which is display the numbered menu in text to command lines. After select one number then also display the inner menu related to selected number menu. The menu has a disadvantage that it seems like a maze. Very difficult to see total menu hierarchy. However, Lotus 1-2-3 solve these problem, they displayed the total menu hierarchy. After GUI possible, the menus advanced to two types: drop-down menu and pop-up menu. Today’s menus has a meaning that it contains the whole possible actions.

File, edit, windows, and help are kinds of standard menu, and view, insert, settings, format, and tool are optional menus. Most recently used(MRU) is useful on file menu. In edit menu, selecting, cutting, pasting, and modifing something are usually represented. And multi-documented application, they need windows menu.

Furthermore, menu itself has some types: cascading memu, adaptive menu, and recently ribbon type. Cascading is similar to traditional hierarchy menu, but it’s visually more strong, but still suffered from difficult to understand especially novice users. Adaptive menus are show frequently used menu at first, and if user clicks the expand button at the end of the menu, then show whole menus to users. It also not easy to understand and makes users annoying. Recently introduced ribbon menus are combined menus with toolbars so that they visually strong but it needs time to used to.

Lastly, menus can be more strong when it combines with accelerators, checkmarks or enable and disable status when needed, icons, and mnemonics.

About Face. Ch21 Review

Chapter 21. Controls

Controls can be categorized 4 different types: imperative controls, selection controls, entry controls, and display controls.

  1. Imperative controls : command, action statements, corresponds to a verb in language.
  2. Selection controls : corresponds to a noun upon which to operate
  3. Entry controls : enable users to enter somethings
  4. Display controls : display and visualization of information

Buttons, butcons, hyperlinks, etc are kinds of imperative controls. Each has advantages and disadvantages. Button is the most familiar and obvious control. Users can be noticed that it’s a button and it will work. Another one of imperative controls is butcon. Butcon is a word composed with button and icon. Butcon is useful in toolbar. Butcon will be difficult novice user but for power and intermediate users, it’s very fast and strong tools. Nowaday, hyperlinks are very widely used, because of world wide web. Its visual affordance is very easy, click then move new page. We do not have to use this as a action command. Because the hyperlinks almost always used as a navigation.

Check boxes, flip-flop buttons, radio buttons, combutcons, list controls, and tree controls are kinds of selection control. Check boxes are have strong visual affordance that can be selected or deselected. The check boxes can be expanded as latching butcon. Flip-flop buttons have two states: on state and off state. However, usually it’s not easy because users does not know it has two states before clicking the button. Radio buttons are similar to check boxes but radio buttons can select only one in the group. So, if select somethings, then automatically deselected previous selected one. It can be expanded as a radio butcon. Combutcons are a composed word of combo, button, and icon. List controls has several text messages in the control and can select one or multiple lines of them. Do not use horizontal scroll in list controls, because it makes users very annoying. Combo boxes is also one of the select controls. Combo boxes can reduce spaces compared to list controls. But it’s hard to see at glance.

Most famous entry control is edit control. And additionally spinner, dials and sliders are also one of the entry controls. Careful consideration about data boundary is needed for edit control. Bounded edit controls must inform their boundaries to user gently. Spinner controls are variant case of edit controls in that it has two additional buttons at the end of the edit control. Increase the value or decrese the value. It has visually strong meaning and also controlable with mouse wheel. Dial and slider controls are also important although it’s not too frequently used. Edit controls has two error correction modes: active validation and passive validation.

Finally, tex controls, scrollbars and splitters are one of the display controls. Text controls can display text message without input. Scrollbars can display the total pages, current page number, and first sentence or thumbnails of indicated page. Lastly, splitters can divide the application controls into several group.

The above controls are not total controls, however, each of those controls are very frequently used and it’s important for standard program. If we do not use appropriately, users will be very confused and hard to use.

About Face. Ch7 Review

Chapter 7. From Requirements to Design: The Framework and Refinement

Chapter 6 and 7 introduced four major activities of design solution: developing scenarios, using those scenarios to define requirements, to define interaction framework using the requirements, and filling in the framework with ever-increasing amounts of design detail. And requirement definition which consists of developing scenarios and defining requirements is major subjects in previous chapter 6. On the other hands, in chapter 7, framework definition is the major subject, and the framework definition consists of the remain two process: to define framework using the requirements and filling the framework with design detail.

Design framework means defining total structures of the users’ experiences. And there are three parts of the design framework: interaction framework, visual design framework, and industrial framework.

Firstly, defining the interaction framework is very first part of framework definition. Additionally, in order to define interaction framework, designers usually follow the 6 steps. Defining form factor, posture, and input methods is very first part of this step. And second step is to define functional and data elements. Step three is to determine functional group and hierarchy. Fouth step is sketching the interaction framework. The fifth is to construct key-path scenarios. Final step is ckecking design with validation scenarios. In fact these 6 steps are have several principles and important parts. From these 6 steps designers can determine structures of the design, but not specific detail.

Secondly, defining the visual design framework means the process of defining visual structures of the design framework. Visual design framework consists of two parts: developing visual language studies, and applying chosen visual style to the screen archetype.

Lastly, industrial design framework is lart part of design framework. These process also consists of several parts: collaborating with interaction designers about form factor and input method, developing rough prototype, and devleoping form language studies.

From those three processes of the design framework, designers are able to determine the overall structures of design framework, and also able to define the framework. These processes are not one time process, but interative and repeatalbe processes. Thus, the final design framework usually can be established by several refining processes, design validation, and usability test.

About Face. Ch6 Review

Chapter 6. The Foundations of Design: Scenarios and Requirements

This chapter shows the process that is bridging the research-design gap, and this process consists of four major activities: developing stories or scenarios, to define requirements from the scenarios, to define fundermental framework using these requiremenets, and filling details of the framework.

Scenario is a powerful tool for interactive design, because of its effectiveness of visual depiction. John Carroll’s concept of scenario-based design in his book, Making Use, is based on accomplishing users’ task. Carroll’s scenario-based design concept, however, has two major drawbacks: lack of understanding user, and lack of users’ goal ands motivations. The two major shortcomings are because it’s not based on personas. Thus, in order to overcome these drawbacks, we need to use persona-based scenarios. There are three types of persona-based scenarios: context scenarios, key-path scenarios, and validation scenarios. Each scenarios has their own purposes, and in this chapter will be concentrate on the context scenarios. Additionally, persona-based scenario is different from traditional use case in that persona-based scenarios focus not only functions of the system, but priorities of the functions and how to interact with users. On the other hands, use case equally concentrate on the functional requirements of the systems and their descriptions.

The definition of requirements are a little bit different from the usual requirements definition, which is just function or feature of the product. In this case, requirements has a meaning, users’ needs. In requirement definition phase, we need to distinguish the problem and solution. It means we have to know what is the problems, and then how to solve the problems. And requirement definition using personas and scenarios comprise five steps: creating problems and vision statement, brainstorming, identifying persona expectations, constructing context scenarios, and identifying requirements.

In first step, creating problems and vision statement, we have to know what directions these scenarios and requirements should be headed in. If we lack a clear product mandate, then we will considerably confuse. In this step, we need two statements. One is problem statements that defines the purpose of the design initiative, and the other is a vision statement which is inverse of problem statement.

Secondly, in brainstorming step, main purpose of this step is to eliminate as much preconception as possbile. So designer remains open-minded and flexible.

Thirdly, identifying persona expectations step is a step for formally recording persona’s mental model.

Fourth step is constructing context scenario. Context scenarios’ purpose is describing broad and relatively swallow context such as usage patterns, exhibits, and other considerations.

Finally, identifying requirements step is that analyze context scenarios and extract persona’s needs and requirements.

About Face. Ch14 Review

Chapter 14. Visual Interface Design

This chapter introduces principles of visual interface design. Although somtimes visual interface design seems visual art or graphic design, it is not entirely aesthetic related. The visual interface design relates with finding the best way of interaction between human and product. There are several principles of visual interface design.

At first, visual properties for group elements and hierarchical representation of data is one of the principles. In order to express group elements, applying similar and same visual pattern will be needed. And visually representing hierarchical importance is also a good way of grouping. Another way of grouping method is spatial grouping that represent elements based on usage frequencies, not based on similar functions. Spatial grouping method need for minimizing users’ movement. The squint test is a way to test the grouping and relationships of elements are effectively organized.

Secondly providing visual structure and flow at each level of organization is one of the principles of visual interface design. Alignment of elements and grid system are important to accomplish visual structure and flow. Dividing the pane vertically and horizontally and representing group elements to divided grid is the method of grid system. Grid system has several benefits for visual interface design such as usability, aesthetic appeal, and efficiency. Additionally, making a efficient logical path for human eye movement and arranging symmetrically and balancing elements are another important parts of visual structure.

And then third principle is using cohesive, consistent, and contextually appropriate imagery. These imagery like icons and illustrative elements will help users’ understanding about interfaces. These visual symbols have some advantages that can help recognition, especially navigation. Thus, visually different symbols have to behave differently, otherwise it will make user confuse. Furthermore, communicating using visual images and associating images with behavior can help users effectively more than just words.

The fourth principle of visual interface design is integrating style and function comprehensively and purposefully. Designers need to make products and brands stylistic harmony. The products is not just a functional objects, but it affects users’ behavior and experiences, and company’s brands. So, harmonizing user interface between the functions of the products and styles is very important.

Lastly, visual noise and clutter should avoid. In order to do this, designers have to maintain simplicity of user interface, and have to stay consistency of text and color in visual interfaces.

A lot of other principles of visual interface design including above five principles exist. And at the last part of this chapter, this book introduces the pros and cons of following standards of visual interfaces, and gives a criterion of violating standards.

About Face. Ch8 & Ch11 Review

Chapter 8. Synthesizing Good Design: Principles and Patterns

There are two major subjects in chapter 8: design principles and design patterns. Here is rough definition of the two subjects: Design principles are general guideline for better design, and design patterns are general solutions, which can be patternized, for solving common design problems.

Firstly, the design principles have several levels from very abstract levels, design values and conceptual principles, to practical levels, behavioral and interface-level principles. And then, the subcategories have also some distinguishable works. To sum up the core of each principles, the behavioral and interface-level principles are based on minimizing users’ cognitive, memory, visual, and physical works. And design values are principles that can be represented as ethical, purposeful, pragmatic, and elegant interactive design. These principles be base of designing process, so we have to keep in mind when designing.

Second part of chapter 8 is design patterns. Design pattern comes from architecture area, Christopher Alexander. However, the concept of the architectural design pattern can be applied to other areas such as interaction design. Usually we encounter several design problems during actual designing. And the solutions of the problems are oftenly similar to each other. So those similar solutions can be patternized. This is called design patterns. Because of the reason, design patterns can solve similar design problems very efficiently. The architectural design patterns do not exactly fit to interaction design pattern in that interaction design patterns deal with humans’ dynamic behaviors and activities.

 

Chapter 11. Eliminating Excise

Excise is that does not directly relate to users’ goal, but necessary to accomplish the goal. The purpose of chapter 11 is to introduce several ways of eliminating excises. There are some kinds of excises such as GUI excises, visual excises, and several messages like error, notifier, and confiromation. These whole excises usually stop users’ proceedings and their flow(In this context, the flow means absolute concentration about something, maybe it’s from the book, Flow). Thus, in order to prohibit users’ distraction from excises, we have to know what is excise, and how to avoid distraction from unnecessary excise. One of the significant excise is navigation. But distraction from navigation excises can be improved by some design properties, such as breathcrum, signspots, annotated scrollbar, and mapping between physical and mental images. Also, even though hierachical data structures are very helpful and well-matched with computer internal structures, the users’ view point are not familiar with hierarchical data. Alternatively monocline grouping is better than hierarchical representation.

AA4. What Happened to Remote Usability Testing? An Empirical Study of Three Methods

Summary

The purpose of this paper is to compare remote usability test methods with conventional lab-based usability test method. And in this paper, the definition of remote usability test is “the evaluators are seperated in space and/or time from users”, derived from Castillo et. al[2]. Because lab-based usability test method has several limitation such as space and time, remote usability test is needed in order to overcome those limitations. For example, open source software’s usability test is very hard when using lab-based usability test, because the developers and users are in all over the world sporadically.

From those reasons, the authors tried to examine and to compare the remote usability test methods. They categorized remote usability methods to three different types. The three types are Remote Synchronous Testing(RS), Remote Asynchronous Expert Testing(AE), and Remote Asynchronous User Testing(AU). RS is different from the lab-based usability test(LAB) in that the evaluators are spartially seperated from the subjects. However, the evaluators are seeing and communicating with the subjects in real-time. On the other hand, in the other two methods, AE and AU, the evaluators are not communicating with the subjects, and only on-line questionnaire is conducted by the subjects. The difference between AE and AU are that AE’s subjects are usability expert and AE’s subjects are ordinary users. Additionally, the subjects are required to execute 9 tasks of email client program, Mozilla Thunderbird 1.5.

In those settings, the authors executed the 4 methods of usability tests, one is LAB, and the other threes are RS, AE, and AU. And the result statistical data indicates that LAB and RS has similar performance so that RS can overcome the geographical limitation of LAB. And AE and AU has bad performance for usability test. The authors led the results from several comparisons with broad criterions.

Analysis

I choose this paper because of interesting of remote usability testing. When we are working at company,  we usually encountered some issues that need to do small and fast usability test. It means that we want to know some small usability issues, but the work or job isn’t large enough for requesting usability lab. I also experienced such situations. At that time, the only thing I can do was to search user’s reaction(such as complains) about the issue on the internet.

Anyway, it seems that the authors designed the experiments very carefully with unbiased and effective manners. Also they analyzed the result data very thoroughly with several standards and also compared with previous results. The conclusion is clear that RS is also very effective compared with LAB. And AE and AU have some drawbacks in that they cannot find some critical usability issues. So if we cannot use lab-based usability test because of space and the other limitations, then RS can be good alternative way of usability test.

And during the reading, I have a question about the asynchronous methods that if they tested additional asynchronous method which record video and sound(think-aloud method), then the asynchronous result can be better than in the paper.

References

1. Morten S. Andreason, Henrik V. Nielsen, Simon O. Schroder and Jan Stage, What Happened to Remote Usability Testing? An Empirical Study of Three Methods. Proceedings of the SIGCHI conference on Human factors in computing systems 2007

2. Castillo, J.C., Hartson, H.R. and Hix, D. Remote usability evaluation: Can users report their own critical incidents? Proceedings of CHI 1998, ACM Press (1998), 253-254

AA3. Demonstrating the Feasibility of Using Forearm Electromyography for Muscle-Computer Interfaces

Summary

The purpose of this paper is to use human muscle motion as computer input. Normally, existing input devices need physical transducers in order to receive human input, but in this paper, tried to develop muscle-computer interface(muCI) which is directly receiving human muscle input to computer. It is basically hard to capture human muscle movement, but the authors was able to do these jobs by using electromyography(EMG).

At first, the authors tried to examine how EMG on forearm is enough to detect muscle movement and the error rate is smaller enough. And then, they found that EMG could be used as a input signals, and could get sample data. Based on these test, broader test was examined. In those whole tests, they tried to figure out 4 kinds of human finger movements by using EMG. The 4 finger movements were position, tap, pressure, and lift.

Each of muscle movements could be classified and captured with acceptable accuracy using EMG. To list up the results, the position movement was 71% accuracy with 9% standard deviation(sd), pressure movement was 76% accuracy with 6.1% sd, tap was 87% with 6.9% sd, and lift movement was 76% accuracy with sd 6.1% in a single-sample classification.  These results were significantly good compared to previous researches, which had 20% accuracy.

Analysis

First of all, forearm muscle input device seems very creative and interesting. Because the forearm muscle moves our fingers, so if we perfectly classify forearm muscle movements, then it means we also classify our finger movements. So it will have very broader application areas such as keyboard without physically keyboard, wearable computers, and game input devices.

However, this paper seems that also has some limitations of classification of finger movements. Only 4 kinds of figner movements were classified. I think there can be more general or specified finger actions. And the interval of input signal is 250 ms. It means that the input signal was occured only 4 times in a second. This frequency does not see continuous signal.

Although muCIs seems to have above limitations, the improvement of the accuracy was very impressive and even have very broader application areas.

Reference

T. Scott Saponas, Desney S. Tan, Dan Morris, Ravin Balakrishnan (2010). Demonstrating the Feasibility of Using Forearm Electromyography for Muscle-Computer Interfaces. CHI 2008, 515-524

AA2. Comparative Evaluation on Tree Visualization Methods

Title

A Comparative Evaluation on Tree Visualization Methods for Hierarchical Structures with Large Fan-outs

Summary

In this paper, Hyunjoo Song et al. did usability test and analyzed the result about tree controls which is used widely at several areas(Tree control is usually used to represent hierarchical information) The authors did the test using three kinds of tree controls.

First one is a traditional tree control(TRD) that represent each child nodes with a connection line to parent node. TRD has a weakness that if there are many child nodes, then it will be hard to see and acquiring information from the control at a glance.

Second control was made to overcome above weakness that is a tree control using list interface(LST). If the number of child nodes increase, LST partially shows child nodes, and provide “more” button at the bottom of the control. This method also has advantage that it does not need to make a bunch of lines between a parent and children node. If users click the “more” button, then scroll bar was created and by using this, users can see the whole child nodes. Additionally we can save spaces. However, it makes ours do additional clicking and scrolling.

The last one represents child nodes using multi column(MLC). We can see the whole children at a glance. It seems to overcome the second tree control’s weaknesses.

The authors tested usability about three important tasks: browsing, revisit, and topology understanding. And for doing the test accurately, they used two data sets: one has large amount of children, and the other has medium amount of children. They analyzed 5 parameters from the test: total time to complete the task, number of mistakes to find proper answer, the length of panning, correctness of the task, and time to expand child nodes.

The results of the usability test are that  MLC generally shows significant improvements and LST seems to more effective than TRD, but sometimes TRD is better. Furthermore, MLC got most positive assessment from user interview. In conclusion, MLC is very effective compared to TRD and LST.

Analysis

I select this paper, because I want to know how to make tree control effective. During my past work experiences, I made a lot of tools for game users and developers. And tree control was used very frequently at that time, because it can represent hierarchical structure very easily. Anyway I got an another improved way of hierarchical data representation.

During reading this paper, I have some questions; how many children in tree control is most effective, and list-based tree control can be more effective if we do not need to use scroll bar, such as automatical scrolling when mouse is on the “more” button.

And this usability test has a focus group, undergraduate students who are familiar with computer. It means that the results of this paper cannot apply to persons who do not familiar with computer.

Reference

Hyunjoo Song, Bohyung Kim, Bongshin Lee, and Jinwook Seo (2010). A Comparative Evaluation on Tree Visualization Methods for Hierarchical Structures with Large Fan-outs. Proceedings of the 28th international conference on Human factors in computing systems, 223-232

AA1. Enhancing Web Page Readability for Non-native Readers

[Summary]

Before starting to summarize this paper, it will be helpful to show related presentation movie.

Click here to go to see the movie.

The purpose of this paper is to improve non-native reader’s readability on web page. Chen-Hsiang Yu et al., the authors, categorized the readability as reading comprehension, reading speed, and user satisfaction. And they found that most influential areas of readability on web page are sentence separation and sentence spacing by iterative study. After that, in order to improve sentence separation and sentence spacing, they had tested 3 different formats, conventional web page which is called Standard Format(SF), improved methods that is called Visual-Syntactic Text Format(VSTF), and Jenga Format(JF) developed by the authors.

The reason that the authors tried to develop new formatting method is that although VSTF shows significantly improved readability, the VSTF has some problems when the formatting method used at web-based page. For example, VSTF splits a sentence with several phrases and clauses, and then shows the hierarchy of the sentence using the splitted phrases and clauses. Thus, VSTF has an advantage that we can see gramatical structure at a glance, however it also has a weakness that VSTF makes web page height longer than before and the authors said that this point can lessen the readability.

(This link is a example of VSTF. I found this from google.)

On the other hand, Jenga Formatting(JF) has several advantages compared to VSTF. JF does not increase the hieght of web page. And JF displays only a main sentence with normal font size when the remain sentences are displayed with small font size. It makes possible to read main sentence at a look. Furthermore JF reduces the stress of eye movement by positioning next sentence at the direct bottom of one sentence’s end position. (It’s the reason of the name, Jenga.)

By doing these methods, they did formal usability test about reading comprehension, reading speed, and user satisfaction. And as a result, they presented that VSTF and JF enhance the readability more than SF. Moreover though there are no significant differences in statistical data between VSTF and JF, but usually participants’ replies about JF are more positive than VSTF.(FYI, they developed JF firefox extension add-on program.)

You can download the add-on program here.

[Analysis]

I found some small problems during reading this paper. Because the authors’ target users are non-native readers, so they had to test several group of non-native readers. But their participants are all Asian. In order to argue that their target is non-native readers, they had to test more general groups with European, Russian, South American, etc. It could make some skewness to the result. I think another possible problem is that reading comprehension can be decreased when using JF. This point is also pointed out by test participants, and I also experienced. It is because the main sentence is not always first sentence.

Jenga formatting, however, is very impressive and interesting, and it can be evolved with natural language processing, synonyms connections, etc.

[Reference]

Chen-Hsiang Yo and Robert C. Miller (2010). Enhancing Web Page Readability for Non-native Readers. Proceedings of the 28th international conference on Human factors in computing systems, 2523-2532

Follow

Get every new post delivered to your Inbox.