And Visual Display

411

Chapter 6 Vision, Graphic Design, and Visual Display

pleDof Effective Visual Communication for Graphical User Interface Design ...... 425

We can only hint at the relevant material in this volume and

ems must therefore be to help us extract data that is required for Although the chapter deals with vision, certain characteristics a task from the mass of data available, and to suppress that which of the human sensorimotor system apply regardless of the modal- is superfluous or irrelevant. ity. One of these is the difference between the magnitude of a Butsimply providing a selective "filter" is not enough. Inevety physical stimulus and the magnitude of the psychological sensa- transaction with a computer, the user is making decisions. We tion of that stimulus. We must distinguish between visual intensity need to provide the relevant information for the task at hand, and and brightness, between wavelength and hue, between aural to provide it in the most articulate, succinct, andappropriate man- intensity and' loudness, and hetween frequency and pitch. ner possible. In doing so we exploit the human capacity for per- Research in psychophysics has concluded that psychological mag- ceiving structure and organization-in short, for understanding. nitude is related to physical stimulus by a power law that applies 'l%emain vehicle today for delivery of information from a com- across a wide variety of sensory phenomena, with exponents PUter toa human being is the visual chamel.This chapter deals with ranging from 0.33 for visual brightness to 3.5 for the subjective the visual representation, presentation, perception, and compre- strength of electric current applied to the finger (Stevens, 1961). hemion of information. It also deals with the discipline of graphic Thus we cannot rely blindly upon the physical properties of a design, which has as its mission the design of effective visual pre- visual or auditory display, but must instead consider carefully how sentations. Finally, it deals with methods and technologies for pro- particular colors or sounds will be received by a human observer. ducing effective visual presentations and delivering them to users. The science of psychophysics also contributes a number of other useful concepts (Van Cott and Kinkade, 19721, including FROM SENSATION TO PERCEPTION sensitivity range along a stimulus scale from minimum per- Effectivehuman -computer interaction requires the presentation ceivabie to maximum tolerable for a human observer information so that the eye and brain can see what the pre- relative discrimination sensitivity, measufing an observer's abiiity Senterintended to be seen. This is not always a trivial probiem to distinguish between two stimuli or to detect a change in one because the viewer does not perceive an image that conforms stiiuius, also known as the "just noticeable diierence"(JND) to the physical image displayed on a computer screen. ability to make absolute judgments of stimuii or of their The visual sensations reaching the eye are translated into per- magnitude along some scale, expressed as a number of 1956) cePbal experience by the brain through processes such as pat- distinguishable values (Miller, tern recognition. We therefore need to acquaint ourselves with Absolute judgments are typically much more difficult than rel- experience and perceive such phenomena as brightness, ative judgements. Thus, although we can distinguish several hun- cOnuast,flicker, motion, and color. dred different intensities of light or sound on a relative basis, we 412 Vision, Graphic Design, and Visual Display

can only reliably idenOfy on the order of a half dozen or a dozen one constrction--not as separate, independent, and free- different absolute stimuli arrayed along a single psychophysical floating elements. And all the objects areperceived as related dimension, such as brightness, hue, saturation, or value. Figure to each other--near: far: behind, a&ining, and so forth. ... 6.1 reproduces a summary of visual coding methods, and pro- Examples of perceived structure include the "perceptual con- vides some guidelines regarding our ability to discriminate dis- stancies" that result in buildings appearing to stand still, people plays that are coded using various individual stimulus dimen- remaining the same size and shape, and snow remaining white, sions. Multiple stimulus dimensions can often be combined to despite the fact that the retinal images may have radically differ- give us more discriminable stimuli. ent positions, shapes, and brightnesses depending upon the rici<,!r.gcocd.tiorls. .Another ccntra, coriccpt 153: emcrges is that of conrcn, now what is seen in one par[ of an inage is aflected Vuual coding methods The a!~orwn'mafenumber ollelels that h\ what an~earsin ad~o!ninp.?arts in s~ace(w::hin the currenr * - z z .. -. can he discriminated on an absolute basis under optimum con- scene) and in time (in the immediate recent past). More gener- : ditions (Sanders and McCormick 1993, p, 126) ally, what is seen is very often not what is presented-optical illusions are perhaps the most dramatic example of this. Illusions Alphanumeric, single numerals ...... 1O e are often the result of multiple possible structures and competing Alphanumeric, single leners ...... 26 interpretations, as in the familiar figure-ground phenomena or the Color (hue of surfaces) ...... 9 Neckar cube illusion (Figure 6.2). Thus, in structuringour screens, Color (hue, saturation, and brightness combinations) ... .24 we must not automatically assume that our intentions can be eas- preferable limit ...... 9 ily translated into effective presentation and communication-we s * Color(hue of lights) ...... 10 must work to achieve the desired resulb. preferable limit ...... 3 Geometric shapes ...... I5 ...... ,a lhe Neckm cub+an optical illusion. Notice how one stimulus * preferable limit .5 * Size of forms (such as squares) ...... 5 causing one optical sexsation can result in two dramatically "different perceptions of thefigure. preferable limit ...... 3 Brightness of lights ...... 3 * preferable limir ...... Z

The process of organizing sensations into perception also shares common elements independent of sensory modality (Lindsay and Norman, 1977, Chapter 1). Perhaps the most important of these is the brain's attempt to organize sensory messages info meaningful j%aftans and struchrres. It does this by..... applying rules that summa-

.. , -.", .. .. ,~ i , ri1.e irs model of the way the world is organized and tht way ii .- ... ,.,-. . '." a,. , .. : . . . , ' ,' ;*,* 4 should aoncar in var:ocs coiitexrs 'This allows us to ilucnra't.- ofts!i incomplete and sometimes inconsistent sensations into a plausible perception of the structure of the source of the sensations. Given this framework, the interested student canaccess the lit- Perception is an activeprocess. What is seen is not merely a erature to fill in other needed background-the basic anatomical passive response to visual input, but the result of hypothesis for- and physiological properties of the eye and of the neural path- mation and testing conditioned by our expectations. Lindsay and way to the brain, the dimensions of vision and the perception of Norman (1977) describe this as the combination of "data driven" brightness, contrast, fiicker, motion, and color. Particularly useful and "conceptually driven" processing, what computer scientists is a superb publication that presents an encyclopedic survey of might call "bottom up" and "top down" computation. perception and human performance (Boff, Kaufman, and What kinds of hypotheses are generated? Haberand LVilk'mon Thomas, 1986). Other useful sources in the psychology literature (1982, p. 25) suggest how we attempt to perceive structure in the are Thompson (19851, Chapters 2 and 3 of Lindsay and Norman images we see: (1977), and the book by Gregory (1966). The human hualsysrem is designed toproduce organized The human factors literature is complementary, focusing on the perception. Information consisting of a variety of such spatial optimum characteristics of specialized kinds of displays. Typical features assize, shape, distance, relativeposition, and tex- treatments are Chapter 3 of Van Cott and Kinkade (19721, Chapter ture is strucrured by the mind to represent visual scenes. 7 of Kantowitz and Sorkin (19831, andchapters 4 and 5 of Sanders These spatial features areperceived asproperties ofthings, and McCormick (1993). These consider, among other topics, the objects in the scene, and not merely as abstract lines or sur- appropriate design of alphanumeric displays, visual codes and faces: We do notperceive lines or unattached extents; wper- symbols, quantitative and qualitative displays of data, status indi- ceive objects. Allparts of each object areperceived together in cators, signal and warning lights, and representational displays. Vlslon, Graphic Design, and Visual Display 413

RAPHIC DESIGN FOR EFFECTIVE tent layout and formatting conventions applied with reference to ISUAL COMMUNICATION an appropriite grid. The resulting visual structure serves to guide the eye as the brain attempts to parse and interpret an image, and The literature of psychology, psychophysics, and human factors helps the user to navigate through the succession of displays and typically describes and quanfies the possibilities and limits of images presented by a system in the course of its use. human visual performance. The results tell us what cannot be Marcus develops hi recommendations for effective visual done, what cannot possibly work. Graphic design is the art of the communication in terms of three basic principles-organize, possible, the discipline of effective visual communication. Its economize, and communicate. He stresses that the application of is essential to the.design of effective displays and design principles to the medium of computer displays and inter- usable interfaces. We cannot here provide a course on graphic active systems is not a trivial process. The medium typically has design, but can introduce key concepts and provide reliable numerous special features and restrictions such as limited spatial po&e*s to the literature. extent and resolution. Thus there is a compelling need to edit Bowman (1968) provides a brilliant and systemtic introduc- the participation of skilled graphic designers in the design of tion to the visual language of effective graphic communication. screens and human-computer interfaces. He describes a vocabulary of form elements-point, line, shape, A comprehensive review of research results that inform effec- value, and texture; a grammar of spatial organization-plane, tive screen design is Tullis (1988). Videos that deal with graphic multiplane, and continuous; an idiom of uolumefricpnspective- design for effective visual communication include Easel Carp. parallel, angular, and oblique; and a syntax for phrasing the (1992 video) and Henry Dreyfuss Associates (1992 video). imaghreiationship, differentiation, and emphasis. Then he discusses the visualization process in terms of the three basic steps Typography at the Interface of concept, design, and production. He concludes with an exten- The most basic element of graphic design, as it affects each and sive design library of examples illustrating how to show every user interface, is lypography. The craft of typography begins whacnatural appearance, physical structure, and with the design of attractive and legible [errerfm;n a variety of organization of parts in relation to the whole typefaces orfon4 that may be either saifor sans saxand that are haw-physical movement, system of flow in relation to com- arranged in families encompassing variations in weight, such as reg- ponent parts, and process as a succession of related events ular and bold, and variations in slant, such as roman and italic haw mucb-physical size, numerical quantity, trend of Typography includes the careful arrangement of sequences of let- inaease or decrease, and division of its paris in terms of telforms on the page with the goal of enhancing readability. This is the whole done by controlling parameters of individual characters, such as where-natural area, environmental location, and position pointsizeand leltmpacing, of words, such as mdspacing, and of lies, such as line lagth, lading, andjut~fiationFinally, typog- with respect to other individual elements , i raphy encompasses Ee augmentation of raw text with simple i Dondi (1973) stresses that the process of composition 1s the graphic elements such as rules, leader lines, and logo^. These most crucial step m visual problem solving She illustrates with typographic parameters and possibilities are illustrated in Figure 6.3. i examoles combining what she defines asfhe basic elements of visual communication-the dot, the line, shape, direction, tone, color, texture, scale, dimension, and movement. The composi- Some typographic varialiom. 7bese are only a far, of the my tional technique she views as most important is that of contrast, : possibilities. and she devotes some effort to its explication. She also presents a variety of purposely polarized techniques for visual communi- A line of 10 point type, Times Roman serif font cation, for example, balance and instability, sfinmetry and asym- A line of 10 point type, Helvetica sans serif font. metry, regularity and irregularity, simplicity and complexity, and A line of 10 point type, Helvetica bold. unity and fragmentation. A line of 10 point type, Helvetica italic. 4 A third source of design ideas is Hofmann (1965), a catalog of A line of fixed-width type. challenging visual design examples organized around the themes Type with normal letter spacing. of the dot, the lie, confrontation, and letters and signs. These volumes may seem far removed from humamcomputer : T~wihcadasedkW interaction, but their insights can and must be applied to the design And with expanded letter spacing. of effective displays and usable interfaces. No designer has within Two lines with a the past fifteen years argued this more forcefully and more articu- ",shorter line length. * lately than Aaron Marcus (1983, included as a reading in Baecker * Two lines of type and Buxton, 1987). Good introductions to the discipline of graphic I: with increased leading. design and its application to interface design are Marcus (190), e included here as a reading, and a recent book, Marcus (192). He A line of type flush left. G presents examples that demonstrate the importance of careful rn A line of type flush right. selection and arrangement in using typography, signs and symbols, e s A line of type centered. charts and diagrams, color, and spatial and temporal arrangement. Echoing insights from psychology, he suggests the use of comic 414 Vision, Graphic Design, and Visual Display

The discipline of typography was always relevant to screen Color in Interfaces design, even in the days of single font fixed-width 24 x 80 char- Color is another key component of graphic design for effective acter displays (Marcus, 1982a). Bit-mapped displays on personal human-computer interaction. The appmpriite use of color is diffI. computers and workstations allow for richer possibilities, yet cult. Our reactions to color result from a complex set of physis restraint is required. The computer scientist's undisciplined logical, perceptual, and cognitive phenomena, which are translated delight in typography sometimes leads to "fontitis," the inclusion into principles and guidelines for the effective use of color in the of so many fonts and point sizes in a single display that the result third part of the reading by Marcus (1190), and also in the reading is visual chaos. As the eminent graphic designer Chuck Bigelow consisting of a short excerpt from a paper by Murch (1985), which cautioned at CHItGI '87, "What good is all this power if you lay in turn summarizes a longer treatment (Murch, 1984). Murch (1985) waste to the literate landscape?" is reprinted in its entirety in Baecker and Buxton (1987). Two other Rosen (1963) presents a catalog of classic typefaces whose readable introductions to the use of color in computer displays and grace and expressiveness put to shame the typical appearance of interfaces are Marcus (1982b) and Salomon (1190). our screens and printouts. Ruder (1967) presents a rich body of Both~urchand Marcus stress the need for restraint in the use evocative examples showing some of the uses and appearances of color, avoiding the "fruit salad" appearance resulting from the of typography within design. undisciplined use of color. The serious student of color will need Bigelow and Day (1983, included as a reading in Baecker and to review considerable material from the art and design literature, Buxton, 1987), introduces some of the issues of digital typography for example, Itten (1961), Albers (19631, Chevreul (1967), Yule including letterform design, representation of digital letterfarms, (1967), Birren (1969a, 1969b), Hunt (1979, and Wong (1987). letterform quality, legibility and readability of digital text at differ- Wyszecki and Stiles (1982) is a comprehensive treatment of color ent resolutions, and the problem of aliasing, in which high spatial science, including colorimetry, photomey, and color vision. frequencies in an original image are reproduced as local spatial The computer science literature has tended to focus on issues of frequencies, leading to the characteristic jaggies that are still often choosing the appropriate color space and of automating the map- seen in low-resolution digital images. Morris (1989) reviews the ping from one space to another (Joblove and Greenberg, 1978; history and key research issues of digital typography. More com- Meyer and Greenberg, 1980; Smith, 1978). Robertson and prehensive treatments may be found in Seybold (19841, which O'Callaghan (1986) deals with the use of color to display quantita- presents digital typesetting in the context of the entire typesetting tive information. In 1986, the journal ColorResearchand Application and printing process, and Rubinstein (1988), which pays particu- contained the proceedings of a meeting covering a broad range of lar attention to the technology of composition and layout. issues in both softcopy and hardcopy computer-generated color. W~thinthe literature of computer science, W~tten(1985) presents Stone, Cowan, and Beatty (1988; see also Stone, 1987 video; Berena the fundamentals of computer typography andsome solutions to the and Stone, 15% video) discusses the problem of mapping gamuts of difFcult problems of line breaks, hyphenation, justification, and page display monitor colors to gamuts of printer colors in order to enable makeup. Brown (1989) reviews and critiques two international stan- the faithful reproduction of screen images on paper. dards for the representation and exchange of structured documents, Two recent books are particularly valuable. Travis (1191) reviews the Office Document Architecture (ODA) and tbe Standard color display hardware, color vision, methods of color specification, Generalized Markup Language (SGML). Naiman (1191) is a compre- and the appropriate use,of color. Durrett (1987) discusses many of hens~etreatment of technical and perceptual issues in the use of the same issues, and then examines the use of color in applications grayscale for improved character presentation. Bibliographies and ranging from orocess control to cartoara~hvto education. ies&ch results indigital typography may be foundin-van Vliet Pictures, Symbols, Signs, and Icons (1986, 1988), Andre and Hersch (19891, and Naiman (1191). There is also much to be learned from the literature on the Human-computer interfaces increasingly incorporate images as legibility and readability of typography. Tinker (1963, 1965) and well as text. Arnheim (1969) states that images can function as Huey (1968) present some classic research on the legibility of pictures, symbols, and signs. type and on the psychology of reading. Hartley (1978, 1980); An image sows mere@as a sign to the exrent to which it Kolers, Wrolstad, and Bouma (1979, 1980); Pelker (1980); and stands for aparticular conrenr wirhout refiecling its charactais- Felker et al. (1981) are more modern collections of articles on the tics visual@.. . . Images are pictures to the mtto which they readability of documents and structured text. portray things located ai a lower level of absrracmers than they Generally, text on screens has been found to be less legible are hem-selves. They do heir work by grasping and renden'ng some relevant qualitiesshape, colot; mouemencofthe objects than text on paper Gould et al. (1987) surveys much of the rele- or actiuifiesthey depict. . . . An image acts as a symbol to the vant literature and presents the results of some experiments that exrent to which itporirays things which are at a higher level of fail to explain why this is so, but demonstrate some of the con- absrractms than is the symbol ifseIf . . . tributing factors. Gould et al. (1986) and Gould el al. (1987) show that reading speeds equivalent to that on paper may he obtained One trend in user interfaces, arising from advances in computer through the use of high-quality antialiased fonts displayed with graphics, is the incorporation of increasingly realistic three- dark characters on a light background on a high-resolution screen. dimensional portrayals (Greenberg, 1982). Mills (1982, 1985) sug- Vid.:os dealing with typography include Xerox (1985a video, gests that this by itself is not sufficient to make optimal use of the 1985b video), Russell (1987 video), Giannitrapani (1987 video), medium of computer graphics and argues that we need also pay and Small, Ishizaki, and Cooper (1994 video). attention to the insights into pictorial representation and communi- Vision, Graphic Design, and Visual Display 415

cation that arise out of art, art history, design, and the psychology the roles, design, construction, and testing of icons for computer of visual perception (Arnheim, 1954; Gombrich, 1961). Mills shows systems and documentation. He stresses that the perceived mean- how man's background, history, and knowledge are embodied in ing of an icon depends upon the context in which it appears and "cognitive schemata" and his capacity for metaphorical thinking, the background, knowledge, and interests of the viewer. Thus, and how the choice of pictorial representation can facilitate hi icons are not automatically universal and do not necessarily work understanding of images and his problem-solving ability. The imag- in all cultures, but must be carefully designed and tailored for inative and appropriate choice of visual representation is thus a key international use. Horton presents a process for developing icons determinant of the success of a user interface. that reflects the principles of task-centered, iterative design we Another trend is that of incorporating in the interface icm have presented throughout this book. images representing system commands, objects, states, or results. Is the use of icons in interfaces a fundamental breakthrough Many icons function as pichlres. For example, a schematic pen may enabling new levels of ease of learning or ease of use, or a fad, 1 be used to indicate that the user may now paint, or a short thick eventually to disappear and be replaced with traditional textual line can indicate that lines about to be input will be drawn with a repraentations? Arguments on both sidesof this debate are pre- thick stroke. Some icons function as symbols, as for example, in sented in the reading by Baecker, Small, and Mander (1991). Both the use of an hourglass, clock, spinning globe, or "smiling Buddha" sides would likely agree that the meaning of an icon should be to indicate that the system is working and the user should be obvious to experienced users of a system, and be evocative and patient. A discussion of the development and testing of icons for self-evident to new users. Some icons fail to meet the former cri- use in the Xerox Star appears in the introduction to Case B. terion; many fail to meet the latter criterion. The reading suggests The design of icons is a demanding craft. Dreyfuss (1972) has that the use of animation could be a solution to the problem of catalogued the incredible variety of icons in his guide to some of icon comprehensibility and provides some evidence for this the 20,000 known international graphic symbols. The American hypothesis through a quasi-experiment comparing static to Institute of Graphic kts (AIGA) has analyzed the strengths and dynamic, animated icons. We shall return below to a further dis- weaknesses of passenger/pedest?in-oriented symbols in three discussion of the possible roles of animation at the interface. tinct dimensions: semantic, syntactic, andpragmatic(l981, p. 20): PRINCIPLES FOR EFFECTIVE Besemantic dimension r&s to the relationship ofa vhal VISUALIZATION image to a meaning. How well does this symbol represent the message? Do people fail to understand the message that the A special class of images used increasingly in computer-generated symbol denotes? Do peoplefrom various cultures misunder- displays are those portraying quantitative datai geographic data, stand tbis symboVDopeople ofvarious ages fail to under- and complex symbolic relationships. This is done by encoding stand this symbol? Is it dfficult to learn this symbol? Has this and interpreting the data and relationships in a chart, graph, map, symbol already been widely accebted? Does this symbol con- or diagram. The formulation and generation of intelligent varieties tain elements that are unrelated to the message? of these images can contribute significantly to the communicative Thesyntactic dimmuion refers to the relationship ofone and expressive power of computer displays. visual image to another. How does this symbol look? How No individual has presented this case more eloquently than mil do the parts of this symbol relate to each other? How mil Edward R. Tuftq (1983). His goal is graphical wellace, which does this symbol relate to other symbols? Is the construction of he defines (p. 13) as "the efficient communication of complex tbis symbol consistent in its use offigure/ground, solid/out- quantitative ideas": line, overlapping, transparency, orientation,format, scale, &cellence in statisticalgraphics consists of complex ideas ~010);and texture? Does this symbol use a hierarchy of recog- communicated with clarity, precision, and efficiency. nition? Are the most important elements recognizedfirst? Graphical displays should: Does this symbol seriously contradict existing standards or conventions? Is this symbol, and its elemen4 capable ofsys- .show the data tematic application for a variety of interrelated concepts? .induce the viewer to think about the substance rather than %?pragmatic dimension r$ers to the relationship ofa about me tho do lo^, graphic design, the technology of visual image to a user. Can aperson see the sign? Is tbis sym- graphic production, or something eke bol saiously affected by poor lighting conditions, oblique avoid disrorting what the data have lo say uiewing angles, and other visual "noise"?Doesthis symbol .present many numbers in a smallspace remain visible throughout the range of typical uiauing dis- make large data sets coherent tances? Is this symbol especially vulnerable to vandalism? Is encourage the eye to compare diiferentpieces ofdata this symbol d$ficult to reproduce? Gn tbir symbol he .rmal the data at several levek ofdetail, from a broad i enlaqed and reduced successfully? overuiew to thefine stnrcture I Every aspect of this discussion applies equally well to the .sew a reasonably clearpurpose: description, exploration, I design of icons for computer systems. The difficulty of the task is tabulation, or decoration I beclosely integrated with the statistical and verbal descrip- I illustrated in a reading included in Baecker and Buxton (1987) that provides excerpts from AIGA (1981) showing possible designs for tion of a data set. icons representing "information," "arriving flights," and "exit." "Graphics reveal data," says Tufte. He then presents in his Horton (1994) is an excellent recent comprehensive guide to beautiful book "a language for describing graphics and a practi- 416 Vision. Graphic Design, and Visual Display cal theory of data graphics." He concludes with humility and a and visualizing data are needed (Goldstein and Roth, 194). healthy skepticism (p. 191): Traditional methods of portraying data such as charts and graph Design is choice. 7he theory of tbe uinral display of quantita- are being augmented by new computer-based techniques for tive information consisd ofprinciples thatgenerare design information visualization. optim and thatguide choices among optim. 7hep'nciples The reading by Ahlberg and Shneiderman (1994; 1994 video), should not be applied rigidly or in apeetish spirit; they are for example, describes the use of dynamic query filters and not logically or mathematically ceriain; and it is better to uio- starfield displays to facilitate the process of visual information lare any principle than toplacegraceless or inelegant marks seeking. Dynamic query filters allow the rapid updating of search on paper. Mostprinciples of design should begreered with results as query parameters are adjusted with sliders and buttons. some skqticism, for word authority can dominate our uision, Starfield displays are two-dimensional scatterplots of sets of and we may come to see only tbrough the lenses of word results (for an early example of such a display, see the last scene authority rather than m'th our own qes. of Baecker, 1981 video). Zooming and panning over the displays, What is to besought in designs for the display ofinfoonnation and recomputing them rapidly as queries are reformulated, pow- is tbe charponrayal ofcom@zity. Not the complication ofthe erfully support human capabilities for intelligent browsing and simpk; ratber the &uk of he designer is to give visualaccess to the pattern recognition. subtle and tbe dficulcthat is. therevelation of the complex. Part VI of Shneiderman (1993) reviews other work on infor- Those wishing to learn about the art of "the revelation of the mation visualization carried out in Shneiderman's lab, including complex" will need to supplement the Tufte book with additional earlier papers on dynamic queries (Ahlberg, Williamson, and readings. His more recent work, Tufte (1990), is another beauti- Shneiderman, 1992; Shneiderman, Williamson, and Ahlberg, 1992 ful and insighthl compendium of heuristics and examples of video; Plaisant and Jain, 1994 video) and their application to effective information visualization. searching a real-estate data base (Williamson and Shneiderman, Another useful source to consult is Cleveland (19851, which 1992). Another technique is that of tree maps (Johnson and presents priinciples of graphical construction, a catalog of appro- Shneiderman, 1991; Turo, 1994 video), which displays large tree priate graphical methods, and a paradigm for graphical percep- structures by partitioning a rectangular display space into a col- tion. The use of a theory of visual perception and experimental lection of hierarchically nested rectangular bounding boxes. results on visual perception to guide the design of displays is a Viewers of large information visualizations need to perceive welcome addition to the literature. both local detail and global context (see Spence, 1983 video, for Three valuable books with a nice "how-to" flavor are Spear an early demonstration of this). Traditional computer graphic sys- (1969)). Schmid and Schmid (1979), and Schmid (1983). These tems allow users to pan and zoom over information spaces. three books, as well as the classic by Huff (1954) and the more Unfortunately, when one is zoomed in to see detail, the context recent Monmonier (1990, include illustrations of what not to do, is not visible, and when one is ibomed out for purposes of ori- that is, how to lie with graphic presentation. Far more insidious, entation, the detail cannot be seen. A solution to this problem is however, is the well-intentioned but ill-conceived chartjunk to provide overview and detail displays in multiple windows with (Tufte's term) intended to enliven and entertain, but actually con- independent pan and zoom capabilities, yet this can iead to ceals or even distorts the data. Herdeg (19811, a classic catalog of divided visual focus, as yell as screen clutter or window overlap. lively diagrams now in its fourth edition, presents numerous Furnas (1986) proposed generalized fisheye vieus as another examples of imaginative and appropriate work interspersed with soiution to the problem of providing both context and detail in a chartjunk. Holmes (1984) is even more replete with chartjunk. The single information display. The key idea is to show "local" detail in student should look at these various books and think about what fuiI while displaying successively less detail for "global" informa- makes a visual presentation honest, communicative, and effective. tion at the periphery. This can be done in three ways, by graphi- 'Ibe advanced student can eventually progress to the monu- cally distorting the view, thereby shrinking the peripheral informa- mental works of Tukey and Bertin. Tukey (1977) focuses on the tion; by fiitering the view, so that less detail appears at the periph- exploratory analysis of data, and the use of novel forms of graphic ery; or by replacing details at the periphery with smaller represen- presentation to assist in the task. Bertin (1983) begins from a carto- tations, absmctions, and clusters. Examples of such approaches are graphic perspective, but enlarges this to a comprehensive explo- Sarkar and Brown (1992; Brown, 1993 video), which proposes fish- ration of methods of applying graphic presentation to the construc- eye displays of graphs; Rao and Card (1941, which investigates tion of complex yet communicative diagrams, networks, and maps. fisheye displays of tables; and Noik (1993) and Scbaffer et al. (1999, which study the fisheye display of hierarchically clustered INFORMATION VISUALIZATION networks and graphs. The latter paper also includes a recent SOFTWARE review of this literature and an experiment comparing the efficacy Charts, graphs, maps, and diagrams are examples of vehicles for of fisheye and full-zoom methods in a process control application. what is now known more generally as infwmation vinralizaiion Another area of recent research is work on intelligent auto- (McCormick, DeFanti, and Brown, 1987; Gershon et al., 1994; matic information visualization. A pioneeering work in this area Visualization Software, 1991 video; see also the proceedings of is Mackinlay's Ph.D. thesis (1986a, 1986b) exploring the auto- the new ACM and IEEE conferences on scientific visualization). mated design of graphical presentations of relational information. As computers are used to manage larger and larger sets of more This is accomplished using A1 techniques, graphic design expres- and more complex data, better means of manipulating, analyzing, siveness and effectivenesscriteria, anda composition algebra that Vision, Graphic Design, and Visual Display 417 composes a small set of primitive graphical languages. More More recently, Card, Robertson, and Mackinlay (1990, recent work in this area includes Roth et al. (19941, which uses Robertson, Mackiay, and Card (1991 video), and Robertson, automated design knowledge to assist both in the process of cre- Card, and Mackinlay (1993) apply human information-processing ating new designs and in the process of finding and customizing principles (see Chapter 9) to suggest that information work- relevant existing designs, and Myers, Goldstein, and Goldberg spaces be mediated and accessed through three-dimensional (13941, which shows how to create charts by demonstration. representations. A prototype information visualizer suggesting Videos documenting work in information visualization include how to facilitate information retrieval is presented in these Hill and Hollan (1392 video) and Tweedie et al. (1994 video). papers, and in more detail in two companion papers, Mackinlay, Finally, one kindof informationvisualizationthat desemesspe- Robertson, and Card (19911, which proposes 3D, fisheye-like cia1 mention is software virualization, the use of interactive com- views of linear information; and Robertson, Mackinlay, and Card puter graphics and the disciplines of graphic design, typography, (1991), which explores cone trees, animated3D visualizations of animation, and cinematography to portray the function, structure, hierarchical information. and behavior of algorithms and computer programs. Pioneering Gomez et al. (1994) is a recent review of issues in facilitating work in this area is represented by Sorting Out Sorting, a 30- 3D interaction. Brookshire Connor et al. (1992) explore the minute computer-animated portrayal of sorting algorithms design of three-dimensional widgets. One interesting example is (Baecker, 1981 video); Balsa, an elegant system for the interactive the "interactive shadow" widget, which provides valuable per- construction and display of such visualizations (Brown, 1988); and ceptual cues about the spatial relationship between objects and Baecker and Marcus (1988), which applies graphic design princi- also a direct manipulation interface to constrained transformation ples to improve the presentation of computer program source techniques (Herndon et al., 1992) More recentiy, Zeleznik et al. code. Recently, Eick and hi colleagues have developed lovely (1993) discuss the design of a toolkit for constructing 3D widgets. applications of high-resolution color data visualization to aid in the Further discussion of 3D interaction appears in Chapter 7. comprehension of large computer programs, networks, relational The most comprehensive vision of the potential of three- data bases, and other complex systems (Eick, Steffen,and Sumner, dimensional display and interaction is that of immSrsing users in 1992; Eick, Nelson, and Schmidt, 194; AT&T Bell Laboratories, an environment that is rendered and portrayed so vividly that it 1994; AT&T Bell laboratories, 1993 video). appears to be real. We shall return to this concept of virtual A comprehensive recent ovemiew of work in this field is Price, reality in Chapter 14. Baecker, and Small (1993). Other videos presenting work in soft- Animation ware visualization as well as programming environments with superior visualization capabilities include CCA (1983 video), Interactive computer systems should portray phenomena using Abelson (1985 video), Lieberman (1985 video), Xerox (1985~ dynamic displays as well as static images, so we require, in video), Tekvonix (1985 video), Eisenstadt and Brayshaw (1988 addition to the skills of the graphic designer, the skills of the video), ParcPlace Systems (1989 video), Carnegie Mellon cinematographer and animator. We can look to the language of University (1991 video), and Brown (1992 video). cinema for models of how our interfaces should behave. They should enable us to include, as global changes to the entire DISPLAY METHODS AND screen, the cut, the fade in (fade out), the dissolve, the wipe, TECHNOLOGIES the overlay Ysuperimposition), and the multiple exposure (image combination). Locally, within a region of the screen, Our final topic in this chapter examines a number of methods for interfaces should allow the pop on (pop ofn, the pull down defining and creating user interfaces and visual presentations- (puM up), theflip and the spin. Finally, they should allow, as three-dimensional displays, animation, structured editors, and either global or local phenomena, reverse video, color changes, graphical user interfaces (GU1s)--as well as the technology scrolling, panning, zooming in (zooming out), the close up, whereby such displays are produced and presented to users. and full motion, possibly through simulations of three- Three-Dimensional Display dimensional worlds and environments. With such effects, an Most interactive applications to date have used the computer to interface can more easily be made memorable and vivid, cap- display two-dimensional worlds. The dominant conceptual model tivating and enjoyable to use. is that of the "desktop metaphor," originally pioneered with the We have already presented one reading that illustrates this Star personal workstation which we analyzed in Case B. Yet the very well. Baecker, Small, and Mander (1991) employ animation field of three-dimensional computer graphics (Poley et al., 1990) to make icons livelier and more comprehensible. The addition of has vividly demonstrated the power of three-dimensional model- movement enabled viewers who were confused by several static ing, rendering, display, and interaction through compelling appli- icons to identify their roles quickly and accurately. cations to fields such as computer animation and computer-aided Baecker and Small (1991) is a systematic analysis of possible mechanical, industrial, and architectural design. roles of animation at the interface. Card, Robertson, and Why limit ourselves to two-dimensional metaphors inhuman- Mackinlay (19911, Robertson, Mackinlay, and Card (1991 video), computer interaction?An early attempt to use three-dimensional and Robertson, Card, and Mackinlay (1993) employ animation to representations for information visualization of complex semantic facilitate exploration and perception of large information spaces. networks was the MCC SemNet project (Pairchild, Poltrock, and Sukivirya and Foley (1990) demonstrate the automatic generation Furnas, 1988; Fairchild and Poltrock, 1987 video). of context-sensitive animated help. 418 Vision. Graphic Design, and Visual Display

The Structure and Editing of Graphic Electroluminescent displap, in which a thin film deposited Representations in layers on a glass plate glows when a voltage is applied The pioneering work of Ivan Sutherland (1963,1963 video) demon- Plma displays, in which neon gas sandwiched between stratedthat theunderlying mcture of displays is wen more impor- glass plates is ionized and thereby emits light tant than what appears on the screen. A number of videos present A number of books and survey papers (Refioglu, 1983; recent advances in exploring this concept. Pier (1989 video) deals Aldersey-Williams andGraff,1985; Perry et al., 1985; 'Iamas, 1986; with the processing of digital images. Bier (1989 video) considers Castellano, 1992; Werner, 1993) report that flat panel technologies "graphical search and replace" in synthetic pictures. Furnas (1990 are capturing an increasing percentage of the market for computer video) demonstrates a prototype system capable of reasoning about displays. Yet the only technology with significant success has been graphical interfaces. Pier and Bier (1591 video) presents a prototype LCDs, both passive and active matrix, because of their use in lap- system in which documents act as user interfaces in the sense that top computers. Meanwhile, the CRT has continued its many any document element can behave as a "button." Gleicber (1992 decades of improvement in brightness, luminous efficiency, color, video) demonstrates the use of constraints in a direct manipulation resolution, and cost. The interested reader may want to begin with graphical editor, and Karsenty (1993 video) explores ways in which Tannas (1985) and Castellano (19921, which present comprehen- the computer can assist the user in dealing with constraints. Finally, sive overviews of the technologies of visual displays, and with the so that users need not think about two representations, what is on monthly Information lhsplay Magazine (published by the Society the screen and what underlies it, Kim (1988 video) treats the pixel for Information Display). Snyder (1988) reviews both vision sci- on the screen as the sole representation, the state of the system. ence and disnlav hardware determinanb of imaee aualitv. . , "1. The Technology of Visual Displays One important realization of the past two decades has been of the significance of scale, that is, physical size of devices, in As interface designers, we need to understand the kinds of images enabling new kinds of computer use. First came personal com- that facilitate human-computer interaction, and also be aware of puters; which are now daily tools of hundreds of millions of indi- the technology capable of generating and displaying these images. viduals around the world. Next came laptop computm, which The process occurs in four steps: (1) understanding the funda- allow machines to be used free of the desk and the office. This mental characteristics of images as represented on computer dis- was followed by palmtop computm, sometimes known as per- plays; (2) understanding the kinds of changes to images that must sonal digitalassistants (PDAs), which reduce the size and weight be supported by the technology to enable interactive applications and increase the mobiiity of computational devices even further. to be effective; (3) understanding the basic technology and archi- Even more recently, researchers have begun to work on what has tecture of display systems; and (4) understanding how to apply become known as ubiquitow compuling (see Chapter 14 and this technology to the generation of appropriate imagery and the especially the reading by Weiser), in which highly portable tiny dynamic changes to the imagery that are required. badges and tabs as well as large wall-size displays, all of which The technology is advancing at breakneck speed (compare, for contain compute engines, are interconnected and communicating example, Sproull, 1986, to Baecker, 1979). Sproull(1980, which is in order to enable new kinds of applications. included as a reading in Baecker and Buxton (l987), is an overview of the technology and architecture of current visual displays. It Putting It All Together: GUls focuses on how technological pameters and architectural design Ultimately, typography aid layout, symbolism and color, dimen- of display update processors, frame buffer memories, and video sionality and animation, must be brought together in an interface generators can either facilitate or hinder the production of the dis- style for a particular program, system, or environment. The term plays, picture changes, and interactive techniques that interface graphical user interface, or GUI, is variously used to refer to the designers require. Tbese techniques include the incorporation of "look and feel" of an interface style, the common functionality color and antialiasing, smooth scrolling and dragging, rapid pop-up available to a user, and the underiying software that impiements menus and color table animation. Chapter 4, "Graphics Hardware", and supports user interactions. Employing the terminology intro- and Chapter 18, "Advanced Raster Graphics Architecture" of Foley duced in Chapter 5, we can state that the look and feel of a GUI et al. (1990) are more recent treatments of computer graphics hard- is typically implemented by a user interface toolkit sitting on top ware, with an emphasis in the latter chapter on the use of pipelined of a windowing system. and parallel architectures for computing in hardware relatively real- Recall our discussion of consistency in Chapter 2. Users would istic renderings of three-dimensional scenes. find it very difficult if every application and every computer sys- Of equal or often greater importance to the interface designer tem presented a different look and feel. The software industry is are the physical and pragmatic characteristics of the resulting trying to respond to this problem, Apple Computer (19921, for workstation and terminal. The last two decades have seen example, is a published set of human interface guidelines increasing interest in the use of flat panel displnys to overcome intended to encourage those developing software for the the fundamental problems of too much size, weight, and power Macintosh to provide consistent interfaces. Corresponding docu- consumption that characterize cathode ray tube (CRn terminals. ments from other dominant look-and-feels are IBM (1991), Work has concentrated on three technologies: Microsoft (19921, Open Software Foundation (1992), and Sun Liquid crystal displays (LCD$, in which a voltage reorients Microsystems (1990). molecules within a layer of liquid crystals sandwiched Marcus (19921 defines a "common user interface" as "a set of between glass plates so that light is reflected or absorbed rules used across a group of products that formally specifies the $ v~slon,Graphlc Deslgn, and Vlsual Display 419

presentationof dataandfunctions, theuser'sinteraction, and Albers, J. (1963, 1975). Interaction of Color. Yale University Press. the logical organization and behavior of information," 2nd Suggests Aldersey-Williams, H., and Graff, G. (1985). The Flat Panel defining and documenting the common user interface in a "user Challenge. High Technology, 3945. interface standards manual." Smith (1988) compares the use of American Instituteof~raphic (1981). Symbols Signs.me standards versus guidelines in designing user interface software. System ofPassenger/Pedesttian Oriated Symbols Dewlopedfor the Despite the shared goal of interface cOnSistenc/-, there US.Department of Transpotlation. Hasting House. zemany differentGUIsinusetoda~.Thereadingb~Marcus(1992) J,, and He;sch, R.

Castellano, J. (1992). Handbook of Display Technology. Greenberg, D. (1982). An Overview of Computer Graphics. In Academic Press. Greenberg, D., Marcus, A., Schmidt, A., and Goiter, V. (Eds.), The Chevreul, M. (1967). 7hePrinciples of Harmony and Contrasr of Computerlmage: Applications of Computer Graphics, Addison- Colors and Their Application to the Arts Van Nostrand Reinhold. Wesley, 7-75. Cleveland, W. (1985). 7he Elements of Graphing Dab. Wadsworth Gregory, R. (1966). Eye and Brain: lhe Psychology of Seeing. World Advanced Books. Color Research and Application (1986). University Library. Supplement to Volume 11. John Wdey & Sons. Haber, R., and Wilkinson, L. (1982). Perceptual Components of Dickson, G., DeSanctis, G., and McBride, D. (1986). Understanding Computer Displays. IEEE Computer Graphics and Applications the Effeaiveness of Compter Graphics for Decision Support: A 2(3), 23-35. Cumulative Experimental Approach. Communicarim ofrhe ACM Hartley, J. (1978). Designing Instnrctional Text. Kogan Page 29(l), 4M7. Limited. Dondk, D. (1973). A Primer of ViwalLireracy. MIT Press. Hartley, J. (Ed.) (1980). ThePsychology of Written Communication. Dreyiuss, H. (1972). SymbolSourcebooP An Aurboriratiue Guide to Kogan Page Limited. Internaliaal Graphic Symbols. Van Nostsand Reinhold. Helander, M. (Ed.) (1988) Handbook of Human-Computer Dunett, J. (Ed.) (1987). Color and the Computer Academic Press. Interaction. North-Holland. Eick, S., Nelson, M., and Schmidt, J. (1994). Graphical Analysis of Herdeg, W (1981). Diagrams: lhe Graphic Vwalization ofAbstracr Computer Log Files. Communications of rbeACM37(12), 5C-56. Data, Pourth Expanded Edition. Graphics Press Corp. Eick, S., Steffen, J., and Sumner, E. (1992). Seesoft (TMbA Tool Hemdon, K., Zeleznik, R., Robbins, D., Brookshire Connor, D., for Visualiiing Lime Oriented Software Statistics. IEEE Transactions Snibbe, S., and van Dam, A. (1992). Interactive Shadows. Proc. MsT on Sofrware Engineering 18(11), 957-168. 92, ACM, 16. Pairchild, K., Poltrock, S., and Purnas, G. (1988). SemNet: Three- Hofmam, A. (1965). Graphic &gn Manual. Van Nostsand R6inhold. Dimensional Graphic Representations of Iarge Knowledge Bases. In Holmes, N. (1984). A Designer's Guide to Creattng Cham and R. Guirdon (Ed.), Cognitiue Science and I& Applications ro Human- Diagram. Watson-Guptill Publishing Company. Computer Inreracrion. lawrence Edbaum Associates. Horton, W. (19%). me Icon Book: VisualSymbolsfor Computer Felker, D. (Ed.) (1980). Document Design: A Revlevleuof rbeRelwanr Systanr andDocumentation. John Wiley & Sons. Research. American Institutes for Research. Huey, E. (1968). lhe Pvchology and Pedagogy ofReading. MIT Press. Felker, D., Pickering, P., Charrow, V., Holland, V., and Redish, J. Huff, D. (1954). How to Lie with Statistics. W.W. Norton. (1981). Guidelinesfor Document Designers. American Institutes for Research. Hunt, R. (1975). ReReproducrion of Color, Third Edition. John Wiley & Sons. i ! ' Poley, J., van Dam, A., Feiner, S., and Hughes, J. (1990). Computer Graphics: Principles and Practice. Second Edition. Addison-Wesley. IBM (1991). IBM Systems Application Architecture: Common [iser I Access: Advanced Interface Design. I Furnas, G. (1986). benerallzed Plsheye Views. Proceedings of CHI j '86 ACM, 16-23, Itten, J. (1961). The Arf of Color. Van Nostsand Reinhold. ! Gershon, N., Busllell, C., Mackinlay, J., Ruh, W., Spoerri, A,, and Johlove, G., and Greenberg, D. (1978). Color Spaces for Computer Tesler, J. (1994). lnformation Visualization: The Next Frontier. Proc. Graphics. Computer Graphics 12(3), 20-25. SIGGR4PH '94, ACM, 485-486. Johnson, B., and Shneiderman, B. (1991). Treemaps: A Space-Filling i Approach to the Visualization of Hierarchical Information I Goldstein, J., and Roth, S. (1994). Using Aggregation and Dynamic i Queries for Exploring Large Data Sets. Proc. CHI 94, ACM, 23-29, Stmctures. Proc. 2nd Intemtio~lIEEE Visualization Confmce, 284-291. Gombrich, E. (1961). Art and Illusion: A Study in the Psychology of Picrorial Representah, Second Edition. Pantheon Books. Kantowitz, B., and Sorkin, R. (1983). Human Factors: Wdmtanding People-Systa Relarionrhips. John Wiley & Sons. Gomez, J., Carey, R., Fields, T., van Dam, A,, and Venolia, D. (1994). Why Is 3-D Interaction So Hard and What Can We Really Kolers, P., Wrolstad, M., and Bouma, H. (1979) Processing of Visible Do about It? PIOC. SIGGRAPH '94, ACM, 492-493, Language, Vol. 1. Plenum Press. Gould, J., Alfaro, I.,Barnes, V., Finn, R., Grischkowsky, N., and Kolers, P., Wrolstad, M., and Bouma, H. (1980). Processing of Virible Minuto, A. (1987). Why Reading Is Slower from CRT Displays than Language, Vol. 2. Plenum Press. from Paper: Attempts to Isolate a Single Variable Explanation. Lindsay, P., and Norman, D. (1977). Human Information Human Facrors 290, 269299. Processing: An Introduction to Psychology, Second Edition. Gould, J., Alfaro, I.,Pinn, R., Haupt, B., and Minuto, A. (1986). Academic Press. Reading from CRT Displays Can Be as Fast as Reading from Paper. Mackinlay, J. (1986a). Automated Design of Graphical Presentations. BMResearch Report RC-12083. Stanford University Department of Computer Sciences Report No. Gould, J., Alfaro, L., Finn, R., Haupt, B., Minuto, A,, and Salaun,J. STAN-CS-86-1138. (1987) Why Reading WsSlower from CRT Displays than from Paper. Proc. CHltGI '87, ACM, 7-11. Vision. Graphlc Design, and Visual Display 421 I Mackinlay, J. (1986b). Automating the Design of Graphical Open Software Foundation (1992). OSF/1MotifSlyleGuide. presentations of Relational Information. ACM Transactions on Prentice Hall. Grnphics 5(2), April 1986, 11@141. Perry, T.,Wallich, P., Apt, C., and Baldauf, D. (1985). Computer Makinlay, J., Robertson, G., and Card, S. (1991). The Perspective Display: New Choices, New Tradeoffs. IEEE Spectrum, July, 52-73. Wall: Detail and Context Smoothly Integrated. Proc. (28 91,175179. Price, B., Baecker, R., and Small, I. (1993). A Principled Taxonomy Marcus, A. (1982a). Designing the Face of an Interface. IEEE of Software Visualization. journal of VialLangwgesand Computer Graphics and Applications 2(1), 2329. Computing 40,211-266. Marcus, A. (1982b). Color: A Tool for Computer Graphics Rao, R., and Card, S. (1994). The Table Lens: Merging Graphical Communication. In Greenberg, D., Marcus, A,, Schmidt, A., and and Symbolic Representations in an Interactive PocustContext Gorter, V., (EdsJThe Computer Image: Applications of Computer Visualhation for Tabular Information. Proc. CHI 'a% ACM, 318-322, Graphics, Addison-Wesley, 76-90. 481-482. Marcus, A. (1983). Graphic Design for Computer Graphics. IEEE Refioglu, H. (Ed.) (1983). Electronic Displays. IEEE Press. Computer Graphics and Applications 3(4), 6370. Robemon, G., Card, S., and Mackinlay, J. (1993). Information & Marcus, A. (1990). Primcipies of EffectiveVisual Communication Visualization Using 3D Interactive Animation. Communications of for Graphical User Interface Design. VnixWorld, August 1990, the ACM 36(4), 57-71. 107-111; September 1990, 121-124; and October 1990, 135-138 Robertson, G., Mackinlay, J., and Card, S. (1991). Cone Trees: (revised and reformatted). Animated 3D Visualizations of Hierarchical Information. Proc. CHI O Marcus, A. (1992). A Comparison of Graphical User Interfaces. In PI, ACM, 189-194. Marcus, A. Graphic Design for Electronic Dccumenls and liser Robertson, P., and O'Callaghan, J. (1980. The Generation of Color Interfacq ACM Press (Excerpt included 137-159, 187-188). Sequences for Univariate and Bivariate Mapping. IEEE Cornpuler Marcus, A,, Smilonich, N., and Thompson, L. (1995). The Cross-GU Graphics and Applications 6(2), 24-32. Handbookfor Mulliplatfon liser Interface Design. Addison-Wesley. Rosen, B. (1963). Type and Typography: Re Designer's Type Book. McCormick, B., DeFanti, T., and Brown, M. (1987). Special Issue on Van Nostrand Reinhold. Visualization in Scientific Computing. Computer Graphics 21(6). Roth, S., Kolojejchick, J., Mattis, J., and Goldstein, J. (1994). Meyer, G., and Greenberg, D. (1980). Perceptual Color Spaces for Interactive Graphic Design Using Automatic Presentation Computer Graphics. Computer Graphics 14(3), 2%261. Knowledge. Proc. CHI '94,ACM, 112-117,476. Miaosoft (1992). The Windows Interface: An Application Design Rubinstein, R. (1988). Digital Typography: An Introduction to T@ Guide. Microsoft Press. and Compositionfor Computer Syskm Design. Addison-Wesley. I Miller, G. (1956). The Magical Number Seven, Plus or Minus Two: Ruder, E. (1967). Typography: A Manual of Design. Hastings i Some Linuts on Our Capacity for Processing Information. House Press. Psychological Reuiau 63,81-97. Salomon, G. (1990). New Uses for Color. In laurel, B. (Ed.) (1990), 1 Mils, M. (1982). Cognitive Schema and the Design of Graphics The Art of Human-Computer Interface Design, Addison-Wesley, Displays. Proc. Graphics Interface '82, 3-12. ' 269-278. Mils, M. (1985). Image Synthesis: Optical Identity or Pictorial Samuleson, P. (4992). Updating the Copyright Look and Feel Communication? Proc. Graphics Interface '85, 303308. lawsuits. Communications offhe ACM 35(9), 25-31. Monmonier, M. (1991). How to Lie wid? Maps. The University of Samuleson, P. (1993). The Ups and Downs of Look and Feel. Chicago Press. Communicationc ofd?eACM36(4), 29-35. Morris, R. (1989). Rendering Digital Type: A Historical and Economic Sanders, M., and McCormick, E. (1993). Human Factors in View of Technology. The Cvmputerjourmrl32(6), 524-532. Engineering and Design, Seventh Edition. McGraw-Nil. Murch, G. (1984). The Effective Use of Color, Part 1: Physiological Sarkar, M., and Brown, M. (1992). Graphical Fisheye Views of Principles. Teknipes 7(4), 1516; Part 2: Perceptual Principles. Graphs. Proc. CHI 92, ACM, 8591. Tebniques 8(1), 4-2 Pan 3: Cognitive Principles. Tekniques 8(2), Schaffer,D., Zuo, Z., Greenberg, S., Bartram, L., Dill, J., Dubs, S., 25-31, Tektronix Corporation. and Roseman, M. (forthcoming). Navigating Hierarchically Clustered @ Murch, G. (1985). Color Graphics--Blessing or Ballyhoo? Networks through Fisheye and Full-Zoom Methods. ACM Computer Graphics Forum 4, North-Holland, 127-135. (Excerpt TransacNons on Information Sysfmv included 133135.) Schmid, C. (1983). Statistical Graphics: DesignPrinciples and Myers, B., Goldstein, J., and Goldberg, M. (1994). Creating Charts Practice. John Wiey & Sons. - by Demonstration. Proc. CfU 94, ACM, 106-111, 475. Schmid, C., and Schmid, S. (1979). Handbook ofGraphic Naiman, A. (1991). The Use of Grayscale for Improved Character Presentation,Second Edition. John Wiley & Sons. Presentation. Ph.D. Thesis, Department of Computer Science, Seybold, J. (1984). The World ofDigita1 Typetting Seybold University of Toronto, March. Publications, Inc. Noik, E. (1993). Exploring large Hyperdocuments: Fisheye Views of Shneiderman, B. (1991). Protecting Rights in User Interface Designs. Nested Networks. Proc. ACM Conference on Hyperiexr and ACM SIGCHIBulletin 22(2), 1M9. Hypermedia, 192-195. I j 422 Vision. Graphlc Design, and V~sualDlsplay

Shneiderman, B. (Ed.) (1993). Sparks of Innovation in Human- Williamson, C., and Shneiderman, B. (1992). The Dynamic Home Computer Interaction Ablex. Finder: Evaluating Dynamic Queries in a Real-Estate Information Smith, A. (1978). Color Gamut Transform Pairs. Computer Graphics Ex~lorationSystem. 92, AcMz 338-346. l2Q, 12-19. Witten, I. (1985). Elements of Computer Typography. Inte~tional Smith, S. (1988). Standards Versus Guidelines for Designing User Journal of Man-Macbine Shldies 23,623687. Interface Sohware. In Helander (1988),877-889. Wong, W (1987). Principles ofcolor Design Van Nostrand Snyder, H. (1988). Image Quality. In Helander (19881,437474, Reinhold. -~ ~ Spear, M. (1969). Practical Charling Techniques. McGraw-Hill. Wyszecki, G., and Stiles, W (1982). Color Science: Concepts and Melbods, Quanlitative Data and Formulae, Second Edition. John Sproull, R. (1986). Frame-BufferDisplay Architectures. In Annual Wiley & Sons. Review of Computer Science, Annual Reviews, 1H6. Yule, J. (1967). Principles of Color Reproduction. John Wdey & Sons. Stevens, S. (1961). The Psychophysics of Sensory Function. In Rosenblith, W. (Ed.), Sensory Communication, MIT, 1-34. Zeleznik, R., Herndon, K., Robbi, D., Huang, N., Meyer, T., Parker, N., and Hughes, J. (1993). An Interactive 3D Toolkit for Stone, M., Cowan, W., and Beatty, J. (1988). Color Gamut Mapping Constructing 3D Widgets. Computer Graphics 270),81-84. and the Printing of Digital Color Images. ACM Transactions on Graphics 7(4) 21p29.2. .- vi&os,"if ' " '7 , ,,., . ,.:, .."" ., . 'ae3 Sun~virya,P., and Foley, j. (1990). Coupl~nga LI Framework w.tri . ... . , . .. .. ,.. .+ ..<. .d ~utomaticGenerationof context-~ensitiv&imated Help. Plot. Ahelson, H. (1985). Boxei: Applications of a Personal Computing li7S 90,ACM, 152-166. Environment. ACMSIGGRAPH Video Rm'm (SGVR) 18. Sun Microsystems (1990). Open Look Graphical User Interface Ahlberg, C., and Shneiderman, B. (1994). Visual Information Application Style Guidelines. Addison-Wesley. Seeking Using the FilmFinder. ACM SGVR 97. Sutherland, 1. (1963). Sketchpad: A Man-Machine Graphical AT&T Bell laboratories (1993). High Interaction Data Communication System. AFIPS Conference Proceedings 23, 329-346. Visualization-Using Seesoh to Visualize Program Change History. Tannas, L. (1985). Flat-PanelDisplap and CRTs Van Nostrand ACMSGVR 88. Reinhold. Baecker, R. (1981). Sorting Oul Sorting. 30-minute computer-ani- Tannas, L. (1986). ElectroluminescenceCatches the Public Eye. IEEE mated teaching film. Distributed by Morgan Kaufmann Publishers. Spectnrm 23(10), 3742. Excerpted in ACM SGVR 1. Thompson, P. (1985). Visual Perception: An Intelligent System with Beretta, G., and Stone, M. (1990). Color Selection. ACM SGVR 55. Limited Bandwidth. In A. Monk, (Ed.), Fundamenlals of Human- Bier, E. (1989). MatchTool: Interactive Graphical Search and Computerlnteraction, Academic Press. Substitute. ACMSGVR 48. Tinker, M. (1963). Itgibili~ofPrint. Iowa State University Press. Brown, M. (1992). An Introduction to ZEUS. ACM SGVR 77. Tinker, M. (1965). Bases for Effective Reading. University of Brown, M. (1993). Browsing Graphs Using a Pisheye View. Minnesota Press. ACM SGVR 88. Travis, D. (1991). Effective Color Displays: Tbeoy and Practice. Carnegie Mellon ~niversi~(1991). GENIE: Developing and Academic Press. Assessing State-of-the-ArtIntegrated Programming Environments. Tufte, E. (1983). Tbe Irwl L)lsplay ofQuantitalive Information. ACM SGVR 65. Graphics Press. CCA (1983). Program Visualization. ACMSGVR 13. Tuhe, E. (1990). Envisioning Information Graphics Press. Easel Corp. (1992). Graphic Design of an Executive Information Tukey, J. (1977). Erporato y Data Analysis. Addison-Wesley. System. ACM SGVR 78. Tullis, T. (1981). An Evaluation of Alphanumeric, Graphics, and Eisenstadt, M., and Brayshaw, M. (1988). Visualizing the Execution Color llformation Displays. Human Factors 23(5), 541-550. of Prolog Programs. ACM SGVR 58. Tullis, T. (1983). The Formatting of Alphanumeric Displays: A Fairchild, K., and Poltrock, S. (1987). SemNet 2.1. ACM SGVR 27. Review and Analysis. Human Factors 25(6), 657-682. Furnas, G. (1990). Graphical Reasoning for Graphical Interfaces. Tullis, T. (1988). Saeen Design. In Helander (1988), 377411. ACM SGVR 56. Van Cott, H., and Kinkade, R. (Eds.) (1972). Human Engineering Giannitrapani, 1.(1987). Three Dimensional Font Design Guide to Equipment Design, Revised Edition. American InstiNtes ACM SGVR 34. for Research. Gleicher, M. (1992). Briar-A Constraint-based Drawing Program. Van Vliet, J. (Ed.) (1986). Tert Processing and Document ACM SGVR 77. Manipulation. Cambridge University Press. Henry Dreyhss Associates (1992). Computer Interface Design 1992. Van niet, J. (Ed.) (1988). Document Manipulation and Typography. ACM SGVR 78. Cambridge University Press. Hill, W., and Hollan, J. (1992). Pointing and Visualization. Werner, K. (1993). The Plat Panel's Future. IEEESpectmm ACM SGVR 77. 30(11), 18-26. Vision, Graphic Design, and Visual Display 423

Karsenty, S. (1993). Inferring Graphical Constraints with Rockit. Spence, R. (1983). Office of the Professional. ACMSGVR 13. ACM SGVR 89. Stone, M. (1987). Color Selection Tool. ACM SGVR 33. Kim, S. (1988). Vtewpoint. Toward a Computer for Visual Thinkers. Sutherland, I. (1963). Sketchpad. ACMSGVR 13. Originally filmed in ACM SGVR 58. 1963, published in 1983. Lieberman, H. (1985). ZStep: A Stepper for LISP-based Visual Tekuonix (1985). Magpie. ACM SGVR 19. Inspection. ACM SGVR 19. Turo, D. (1994). Hierarchical Visualization with Treemaps: Making ParcPlace Systems (1989). A Navigator for UNIX. ACMSGVR46. Sense of Pro Basketball Data. ACM SGVR 97. Pier, K. (1989). Digital Darkroom. ACM SGVR 48. Tweedie, L., Spence, R., Williams, D., and Bhogal, R. (1994). The Pier, K., and Bier, E. (1991). Documents as User Interfaces. Atuibute Explorer. ACM SGVR 97. ACM SGVR 63. Visualization Software (1991). Special Issue on Visualization Plabant, C., and Jain, V. (1994). Dynamaps: Dynamic Queries on a Software. ACM SGVR 70. Health Statistics Atlas. ACM SGVR 97. Xerox (1985a). Human Interface Aspects of Typefounder. Robertson, G., Mackinlay, J., and Card, S. (1991). Information ACM SGVR 18. Visualization Using 3D Interactive Animation. ACM SGW 63. Xerox (1985b). Multilingual Typing. ACMSGVR 18. Russell, A. (1987). MIT Visible Language Workshop. ACM SGVR 33. Xerox (1985~).Cedar Programming Environment. ACM SGVR 19. Shneiderman, B., Williamson, C., and Ahlberg, C. (1992). Dynamic Queries: Database Searching by Direct Manipulation. ACM SGVR 77. Small, D., Ishizaki, S., and Cooper, M. (19%). Typographic Space. AmSGVR 97.