Scientific and Technical Communication: Theory, Practice and Policy

Introduction

Our daily perceptual tasks take place in a world in constant motion. We perceive objects with a moving head and eyes while traveling different speeds — walking, running, riding a bicycle, driving a car — even sitting — as the Earth rotates. We travel across various contours and types of land. Sunlight shifts with each season and passing day. Plants, animals and insects move in the wind. Over time our eyesight develops and deteriorates. We use instruments to correct our vision and view microscopic or distant objects. But many studies of vision fail to account for the dynamism of our perceptual environment. In fact, the way we understand graphic perception has changed little since the time of René Descartes (1596-1650).

Descartes postulated a "geometry" of sight. The lens of the human eye collects light rays and makes them converge to corresponding points on the retina. Retinal images excite the brain and cause it to move. Descartes claimed the movement of the brain provided the basis of sight — not graphic images in the back of the eye. The brain intervenes and interprets the graphic sensations we receive. After Descartes, psychologists fixed their research on determining the mechanisms the brain employed for processing graphic information. With few exceptions, philosophers and scientists since the seventeenth century have argued that perception results from mentally constructing images or representations out of sensory inputs. Contemporary theories of graphic perception also share the Cartesian legacy. However, these theories differ by making use of computer metaphors. Graphic perception, according to some versions of cognitive science, is a process by which humans "compute mental models of the world based on sensory inputs," "match recognition templates" or "measure input arrays against stored information."

A controversial response to Cartesian theory and its more modern forms came from the perceptual psychologist James J. Gibson (1904-1979). Gibson claimed that humans receive graphic information directly from the environment without a process of mental interpretation. The environment contains ecological information separate from us and specific to environmental sources. Additionally, objects in the environment possess functional properties of which animals take advantage. For example, humans perceive the properties of an apple as a potential source of food, liquid and even defense (if we throw it at a charging animal), but also perceive that apples do not make good building materials or a source of fuel. How did humans first know to eat apples and not use them as material for building shelter? According to Gibson, humans directly perceived the apple's functional properties — they did not decode and assign meaning to mental representations. In his final work (1986), Gibson stated, "Perceiving is the simplest and best kind of knowing" (263). Here, Gibson promoted the superior ability of humans — as biological organisms — to receive visual information in a moving, three-dimensional, natural environment.

Our ideas of how graphic perception works determines the strategies we use for designing graphic aids and graphic representations. In pursuing the Cartesian legacy, we need to develop strategies for graphically encoding information so it is accurately decoded by the viewer. In pursuing Gibson's theory, we need to develop ways to design graphics to realistically convey the information found in the moving, natural environment. From this starting point, we can begin to examine how graphic images represent nature and provide information in science and technology.

Representing Observations

Our daily perceptual tasks take place not only in nature, but in a world of our own making. For twenty or thirty thousand years of human life, we have made pictures. The achievement of picture making, like language, belongs to humans alone. Frequently the graphic images we create — drawings, photographs, maps, graphs, diagrams - represent objects — microorganisms, subatomic particles — and phenomena we cannot directly observe — voting patterns, world-wide population growth, and the rising price of oil. The world portrayed by many graphics is often flat and two-dimensional. Edward Tufte (1988 b) notes:

... the world portrayed by our information displays is caught up in the two-dimensional poverty of endless flatland of paper and video screen. Escaping this flatland is the major task of envisioning information — for all the interesting worlds (imaginary, human, physical, biological) we seek to understand are inevitably and happily multivariate worlds. Not flatlands. (62)

How best then to represent "multivariate" worlds when using graphics? Tufte suggests the "Methods for escaping flatland include layering and separation, micro/macro readings, contours, perspective, narratives, multiplying of images, use of color and dynamic graphics." (62-63)

The natural world experimental scientists observe is graphically presented as a flatland. Conventions governing graphical design and technical description reflect the role of scientists as incidental observers and recorders of natural occurrences — like courtroom stenographers. The use of third person and passive voice in scientific and technical writing is an example. Flat, abstract representations of information — statistical tables, bar graphs, pie charts — reinforce the notion of nature as solely an object of observation. Conventional two-dimensional graphic designs neglect the dynamic relationship between observer and observed. Tufte heralds a shift in practicing the art of graphic representation away from flatlands to high-resolution, three-dimensional representations. Through multiple images and perspectives, the dynamic interaction between observer and observed is more faithfully represented.

Cultural critics suggest that by virtue of growing up in the age of the rapid cut editing style of MTV, interactive film and video, and cheaper, more powerful computers, you process graphic images more rapidly than your parents. As you read this chapter and follow its practices, keep in mind how representations of nature in science, technology and society depend on our evolving awareness of the relation among graphic perceptions, minds and machines.

Designing Pages

Upon picking up a text you respond immediately to how it looks. Within a matter of seconds you decide whether to pay attention to the words and graphics on the page. Your initial impression is not a reasoned evaluation, but a reflex based on your interests, experience and familiarity with different reading situations. A personnel manager looking at a pile of resumes may pull one out of the stack based on the kind of paper or font used, or the spacing of lines. You look at a glossy, popular magazine with certain expectations about the quality, design and imagination of the graphics and text. You likely have different expectations about the graphic design of a best-selling biography, a corporate annual report or a research article in a journal. Page design can intimidate, bore, intrigue or invite the reader's critical interest — and in so doing can persuade the reader to regard the subject in a certain way.

In each reading situation the appearance of the page — margin size, use of color, use of numbers and notation, font size, number of words and lines on each page, placement of graphics and so on — constitutes an important rhetorical tool. You may wish leave the reader with impressions concerning your professionalism, authority, and accessibility. Page design offers a reader important clues about your regard for, and knowledge of, a certain subject.

In designing a text, just as in writing a text, you must be aware of your audience's needs and expectations. If you work in a field in which new information is rapidly produced, document design is governed less by what information audiences want to know, than what information they need to know. In this instance the design of the abstract of a report, or a specific section of an article can be more important than the design of the full text.

The presence of graphics and graphic aids has become increasingly popular in the natural sciences and engineering. The placement and appearance of graphic aids on the page provides a direct link between your explanation of a subject and the representation you choose for it.

We will look at the following elements of page design:

using white space
using leading
using margins
using columns
using justification

Using White Space

About half of the page you are now reading consists of white space. White space is the blank part of the page used for margins, around graphic aids, among paragraphs, between lines in a paragraph and between words. You can manipulate white space to make a document easier or more difficult to read. The consistent use of margins and white space between graphic aids, paragraphs and lines makes a text easier to read. And conventional wisdom suggests the goal of page design is to make documents as attractive and easy to read as possible. But the goals of technical writers are not always straightforward. Page design is influenced by the availability and knowledge of technology, the economics of publishing, the conventions followed by specific publications and the writer's rhetorical choices. The widespread use of word processors and desktop publishing techniques, for example, presents opportunities for writers and publishers for the rhetorical use of white space. A writer can use less white space — smaller margins, less space between lines and paragraphs — to persuade the reader the content of a document or section is difficult, authoritative or even less significant — as is sometimes the case with footnotes. More white space suggests the audience is less knowledgeable and the subject less intimidating. Writers and publishers have many reasons for wanting a document, a section of a document or a subject to appear complex or simple; inaccessible or accessible.

Using Leading

The use of white space influences the ease and speed of reading of a text and is used to set off special elements of the text such as block quotations (longer than four lines) and graphic aids. The amount of white space used to frame special elements of a text is a product of the space between lines. The spacing between lines is also known as leading (pronounced ledding). Leading is determined by the kind of document you are writing and the specific elements in a text. Here are some rules of thumb in using leading with text and graphic design:

Leading between sections of a document is greater than leading between paragraphs;

Leading between single-spaced paragraphs is double spaced; one-and-a half-spaced or double spaced paragraphs require no extra leading. The first line of any paragraph is always indented;

Leading between graphics and block quotations in a single-spaced text would be double spaced; in one-and-a half- or double spaced texts, the leading is triple spaced.

Using Margins

Margins serve to frame the page. Pages on which the text extends to the edge of the page, or leaves erratic space between the end of lines looks unprofessional.

If the text is unbound, even margins on all four sides are appropriate. Although the size of margins can vary, they usually extend 11/4 to 11/2 inches on each side.

If the text is bound, each of the four margins are a different width. Of the top and bottom margins, the top margin is narrowest; the bottom margin the largest. In many cases the bottom margin is use for footnotes. Of each side margin the inner margin, next to the binding, is larger to prevent text from being hidden. Outside margins are narrowest, but some books and journals have begun to include summary statements or margin notes about the main text in outer margins. Large outer margins are also provided to encourage the reader's note taking.

Margins serve to put a document in an economic and rhetorical context. Publications are bound by cost as much as they are bound by consideration of the reader. The number and size of pages allotted to a journal for a given year by a publisher, budget — determined in part by the number of subscribers and the contributions of sponsoring organizations — and cost of mailing are reflected in the size of the journal's margins. The size of a journal's margins indicates the size of its budget, readership and impact on, or across, a profession.

As with the amount white space, the size of margins also indicates the technical level of the text. Larger margins make the text appear less technical and more accessible to a larger audience. Small margins indicate the text is designed for audience knowledgeable of the information and the format of the presentation. Again, as with white space, margins serve to achieve a specific rhetorical effect.

Using Columns

Columns offer an economical format for including more words and graphics on a single page. As many reader's find shorter lines of text easier to read, using columns offer the unique opportunity to integrate an economic and accessible way to present information. The advantages of using columns are:

Columns allow more words to be fitted on a page;

Columns present shorter lines of text which can be easier to read than longer lines;

Columns allow more graphics to be fitted on a page. Graphics can be fit into single columns or can be spread across two or more columns;

Columns allow for the full integration of text and graphic. For example, single columns of text can be placed along side a graphic so that the information conveyed by the text can be reinforced by the graphic. This design can be particularly helpful in presenting technical reports, instructions or any step-by-step process.

On standard 8 1/2 by 11 inch paper two columns is the most common design. Single and multi-columns formats can also be used, however, columns need to be separated by white space, or vertical columns with white space on either side to prevent the reader's eyes from wandering across the page. As a rule of thumb, the white space between columns should be roughly a half-inch, although, as mentioned earlier, white space can be manipulated to achieve a desired rhetorical effect.

Using Justification

Justification refers to the arrangement of lines along the left and right hand margins of a page. Justified text is presented along a uniform margin. Unjustified text has uneven or "ragged" right hand margins. In almost all technical writing, the text is left justified except for paragraph and heading indentations.

Typewritten and word processed texts often have ragged right hand margins. Typeset and formal documents are usually justified on both the right and left margins. Unjustified, or ragged right, margins suggest a more personal type of communication and are used in internal (memos, personal reports) or personal (letters) documents. Justified text on both margins suggest a formal type of communication and are used in formal documents such as newsletters, journal articles, and books.

Both justified and unjustified text presents scientific and technical writers with several advantages and disadvantages. Depending on the computer software you use, right-justification can actually make a text more difficult to read. In order to achieve regular right margins, the software will push words out to the right margin, creating irregular spaces between words. The reader's eyes must adjust to irregular space between words and letters, making the text more difficult to read. Reader's can easily confuse breaks between words and sentences. To compensate for irregular spacing, words at the end of lines can be hyphenated, and many word processing programs include automatic hyphenation options. While hyphenation allows for even spacing, and is necessary when using columns, the effect can be distracting.

Choices of whether to justify, hyphenate or leave ragged right hand margins suggest a tension among economics, leaving the impression of formality, or of making the text easier to read. Justified text is necessary and more economical in the case of columns, but, in turn, is meant for specialized audiences, not the lay reader. Ragged right margins are easier to read, depending on their length, and help make complex information more accessible.

Designing and Structuring Text

The impression a reader gains about a document from its appearance begins from the ground up. Writers have learned the advantages of designing texts graphically and hierarchically through the use of typographical design, headings, lists and numbers. Typographical design refers to the use of type fonts, type size and styles of type such as boldface and italics. Headings are a word or brief phrase announcing the subject matter that will follow. Lists are an accessible arrangement of words, terms, phrases or descriptions on the page. Numbers are used in conjunction with headings and lists to further organize and emphasize the arrangement of topics.

As stylistic features of proposals, lab reports, memos, résumés and journal articles, typographical design, headings, lists and numbers allow readers to follow the order and emphasis of ideas being presented. The availability and ease of word-processing gives writers the opportunity to their design text to achieve a specific rhetorical effect. In this section, we will look at several aspects of the graphic elements of text design.

Using Typographical Design

Typographical design of the text provide the reader an opportunity to distinguish or gloss over the information being presented.

Using Fonts

A font (also called a type face or type font) is a collection of styles of one type used in printing a set of letters, punctuation marks and characters. Each font has a specific name and series of variations. The availability and ease of using word-processing programs allows you to make several choices with regard to fonts, but these choices are limited to the print elements, software and type of printer you use.

Fonts create different impressions about the formality, accessibility and simplicity — among others — of the content of the document. These impressions correspond with how easy or difficult a particular font makes a document to read.

The ornate structure of the Zapf Chancery font shows why its use is reserved for certain kinds of correspondence — invitations — and formal announcements and declarations. This and similar fonts and would not be used in most technical writing.

A New York font is an example of a serif font - serifs are the lines which extend from the tops and bottom of letters. Serif fonts make each letter distinctive and guide the reader's eyes across the page making longer documents somewhat easier to read.

An Helvetica font, is an example of a sans serif font — the letters have no horizontal lines and fewer distinguishing features. The "clean", vertical lines of sans serif fonts make them useful for shorter documents, numbers, notation, equations and captions for graphic aids.

When approaching technical documents — especially as a layperson — you can draw conclusions of the author's and printer's consideration of the reader. While various factors influence the selection of fonts, the examples above point out rules of thumb of the impressions fonts create. In technical documents deviating from these rules of thumb, you can begin to form a critical of what impression the author or printer wants to leave. A longer document, for instance, printed in a sans serif font keeps the reader at arm's length, or invites the reader to look only at certain sections.

Using Type Size

Type sizes for fonts are measured in points. Varying type size provides the reader with graphic cues for determining the order and relationship among topics in a document. If you notice the headings on this page, for example, larger type sizes correspond to section headings, and successively smaller heading sizes indicate subtopics and their relation to the subject of each section. The standard type size for most documents is 10 to 12 point. The following suggestions for type sizes for specific parts of the text lends a rough template for organizing your choices. Type sizes for:

running headers and footers, captions for graphic aids, index entries, references and footnotes range from 7 to 9 point;

titles or section headings range from 14 to 16 point;

text on transparencies and slides range from 16 to 24 point.

Of course, your choices of type sizes depends on what looks best with the kind of printer you use, and the type of document you prepare. Variation in type size also allows you to capture the reader's attention by emphasizing the importance of the subject;

Do Not Store at Temperatures Above 0 Degrees Centigrade!

or allows you to indicate an idea that is less important, assumed or inconsequential - at least for the author's purposes:

A 10% interest penalty applies for early withdrawal.

As we emphasized in using fonts, varying type sizes makes a document easier to read. You do not, for example, to print a lengthy section of text in all upper case letters because you eliminate the variation among letter sizes. Let the content of the document guide your choices of type sizes, do not let the cosmetic appearance of the text become an end in itself.

Using Styles of Type

Styles of type - bold face, capitalization, italics, outline, shadow and their combination - influence the meaning readers draw from a text through the use of emphasis. Graphic emphasis in a text is created most frequently through by using bold face capitalization, italics or a combination of these styles. Warnings in a set of instructions, formulas, and headings in a proposal, for example, are set apart by using bold face and/or capitalization.

In many journals, abstracts, key words and terms defined in a glossary are frequently italicized. Examples or quotations may also be set off from the text by using italics.

Your choices in using type and style variations depends on the capabilities of the typing or word processing programs you use, your familiarity with them and the availability of a laser printer. While dot matrix can reproduce most type and style variations, ink jet and laser printers offer greater contrast and choices in styles of type. As with all elements of typographical design, you must use them judiciously, keeping in mind your resources, rhetorical goals and the needs of your audience.

Titles

Titles for scientific and technical articles must indicate content (Mitchell 90). Key terms in the title enable librarians as well as abstract and index services to catalogue and excerpt your work. In other forms of scientific and technical communication - including memos and letters (in which subject lines act as a title) - brief, succinct titles introduce the audience to the subject and purpose of your document, presentation or seminar. Titles signal the rhetorical context the paper or presentation wants to achieve and who is the expected audience. For example, these titles for papers appear in an interdisciplinary academic journal:

The Essential Tension in Science and Democracy
The Qualitative Confirmation of Claims in Social Anthropology: An Application
Gene Talk in Sociobiology

Interdisciplinary contexts pose interesting insight into the choices face by scientific and technical communicators. In this instance, since the journal maintains a diverse readership, each writer must decide whether to situate their work in a given specialty, or to appeal to a wider audience. Each of the titles above not only gives a disciplinary locus for the research - social anthropology, sociobiology - but indicates the relationship between the subjects of the article - the incompatibility between science and democracy. Titles offer contextual cues depending on the audience's background. If the reader is familiar debates surrounding sociobiology and discussions about the use of terms from genetics in other disciplines, the title of the last paper has added meaning. The level of specialization alluded to in the title allows lay readers to determine if the article offers any hope of accessibility. The second of the four titles above is aimed at a very specific audience with specialized knowledge. If a lay reader wanted information on social anthropology, the title of the paper does not invite investigation. Most of you will write technical documents in-house or aimed at specific professional audiences. In so doing, you must use precise terms in the title to provide readers with a clear indication of content.

Although the title of an oral presentation or seminar is usually the same as the title of a paper, you may want to consider a more general title. In an open seminar or presentation, the title should appeal to a wider audience. A more general title invites the participation of experts outside your field and layerpsons whose insights are useful for the development of your ideas.

Here are some guidelines for developing titles:

Use key terms, words and concepts from your paper or presentation in your title. For example, the title "Questions Regarding Social Science" offers little of use for either a librarian indexing the article or the potential audience of the paper. Whatever questions or social science to which the writer refers should be specifically mentioned.

Determine the central claim, thesis or definition guiding the research of the presentation and modify it to form the title. For example, the title -"Solubility of Niobic Oxide and Niobium Dioxyfluoride in Nitric Acid-Hydrofluroic Acid Solution at 25 (degrees) C" - although esoteric to the lay reader, indicates a clearly defined research project and includes a number of keywords to aid in library cataloging and reference searches.

Approach rhetorical questions as titles with caution. Rhetorical questions beg for an unqualified answer from the audience even if you equivocate in your answer in the document or presentation. Rhetorical questions tend to predispose members of an audience to take a position they might not normally take. If members of the audience answer 'no' in response to title of the seminar" Does Global Warming Make a Difference in the Weather Patterns of Southwest Virginia?" the speaker, in arguing global warming does make a difference, faces defined opposition. The title "Changes in the Weather Patterns of Southwest Virginia Due To Global Warming" indicates the speaker's certainty in her claims and does not invite uninformed speculation.

Headings, Subheadings and Numbering

Headings and subheadings serve much the purpose of titles within a document. A concise heading or subheading, formed by using key words, concepts, or the central thesis of a section, previews the subject matter of part of a longer text. Headings and subheadings provide an internal index for larger texts, allow transitions among subjects, and give the reader a chance to pause and determine which sections she needs or wants to read. In certain instances, a heading or subheading substitutes for introductory sentences that preview the subject matter of a section of the text. However, in many types of technical writing, a specific journal requirements, or in-house publication practices will govern your choice and use of headings. Scientific journals, for example, have a template in which the headings - 'literature review', 'methods', 'discussion' - define a specific and expected order in the presentation of information. You need first to become aware of the requirements of professional or in-house organizations to determine how the choice you have in how your document is structured.

You can adopt the guidelines for composing titles for headings and subheadings, but here are additional suggestions:

Make headings and subheadings as specific as possible. For example - Problems in the Relation Between Science and Democracy - lends the reader only a vague notion of what this section of the paper will entail. The heading can be improved by adding detail - "Four Problems in the Relation Between Science and Democracy" - and improved further by including greater specificity - "Four Problems in the Relation Between the Conduct of Experimental Science and Democratic Decision Making." Although the last heading is longer - and some publication have limits on the amount of characters in a title or heading - you must weigh the benefits of clarity against the need for brevity.

Approach noun strings with caution. A noun string (this applies to adjectives and adverbs as well) is a phrase in which a series of nouns are used to modify the last noun in the series. In the overblown phrase " postoperative recuperation program procedure indicator sheet" all of the words in the phrase modify the word 'sheet'. However, properly used noun strings offer the writer shorthand expressions for complex concepts or names. For example, the phrase "active noise control system" is a more direct way of saying " a noise control system that is active." Still, even short noun strings can be perplexing. It is not necessary to use the noun string "on-line coronary stimulator device" when "heart pump" would do just as well. When using noun strings balance your desire to use verbal shorthand with an understanding of the knowledge level of the audience and the requirement that technical information be accessible and useful.

Indicate a process by using a verbal form (-ing). The phrasing of headings and subheadings indicates how you want the audience to use the information. In longer technical documents, you may instruct the reader to complete several tasks. To that end you will incorporate many forms of technical writing - definitions, process descriptions and performance instructions - into a larger report or manual. Use headings and subheadings to distinguish the tasks you set for the reader. While one section of a computer user's manual, for example, presents a glossary, another offers instructions for trouble shooting common problems. Headings and subheadings can indicate parts of a process or set of instructions the reader must follow - Converting files to and from MS-DOS or ProDOS format - or - Entering Numerical Data on a Spreadsheet.

Numbering

The use of corresponding numbers with headings allows you outline and organize and a text hierarchically. This chapter serves as an obvious example, The text is organized by major headings which in turn are subdivided. Presented in an outline form, readers can go directly to the part of the text in which they are interested and determine the extent to which an author covers the topic. You can reinforce the distinctions between topics by adding a numbering system. For example:

1. Scientific and Technical Writing Style

1.1 Passive Voice
1.2 Pronoun Reference

1.2.1 Voice and Personal Pronouns

1.3 Vagueness
1.4 Faulty Parallelism

2. Collaborative Writing

The extensive use of numbers, headings and subheading in a text can assume a life of its own and become Byzantine. A rough rule of thumb would be not to go beyond three numbers - in the example above not having a subheading numbered 1.2.1.1. Readers can easily lose track what the numbering system indicates and how the subjects of the headings and subheadings relate.

Establishing Order

Writers can use formatting and style options (explained in pervious sections) to show the relationship among major and supplementary topics. Instead of a numbering system , you can use capitalization, different font sizes and styles, boldface, italic, underlining and indentation. The arrangement, style and indentation of headings in this chapter is an obvious example. Hierarchical levels among headings and subheadings are established by the designated and consistent use of formats and indentation:

Scientific and Technical Writing Style

Passive Voice Pronoun Reference

Voice and Personal Pronouns Vagueness

Faulty Parallelism

Collaborative Writing

Generally, limit to three the number of different format and style options you employ for heading and subheadings. Readers can quickly lose track of the hierarchy of subjects you want to establish if a number of styles and formats are used.

Lists and Boxes

In order to emphasize or make information more accessible you can separate it from the text by employing lists and boxes. Lists can make long sentences more accessible or help emphasize the importance, for instance, of materials needed and steps to be taken to perform a task in a set of instructions. And a list can serve as a self-diagnostic - allowing the reader to check and see if they have followed the right steps in the right order.

Generally lists are either bulleted or numbered. Bullets are symbols commonly found in résumés - that draw attention to listed items. Bulleted lists draw attention to the equivalent relation among the items. Numbered lists, however, suggest the sequence, priority and number of items presented is significant.

Here are some conventions for assembling lists:

Present all items in parallel structure. Parallelism is using similar phrasing to reflect similar processes, ideas or situations. For example, on my "Things to do Today" list I might include:

walk to the store
jog around the campus
ride my bike to the campground
drive to my parent's house
fish on the banks of the New River

Although parallelism is more a stylistic issue than a grammatical one (technically, it is not wrong to break parallel form) - it aids comprehension in that it supplies the reader a predictable pattern;

Design bulleted lists vertically with aligned left margins;

Design numbered lists either vertically or horizontally depending on the amount of room you have and given publication standards.

Boxes also separate and emphasize material. Different software programs will supply you with options regarding the design of boxes including; dotted, hairline, single, thick and double lines, preset borders (none, box or shadow) and shading. Boxed material can include:

formulas, equations and logical notation; key words; anecdotal material (there are a number of examples in this book); document and section headings.

Some documents contain many types of box material. You must remain consistent in using box formats. For example, you might use a double lined box to set off anecdotal material, and a single lined box to separate key words. Use boxes and shading consistently and sparingly.

Viewing and Designing Graphics

When using graphics the technical communicator gives a clear and effective pictorial representation appropriate to a communicative situation. Graphic and graphic representations should never be used to distort information or mislead an audience. While graphic and graphic information may occur in many forms - tables, charts, graphs, slides, computer simulations - their relevance depends on a presentational format. A presentational format may be understood as both the physical appearance of the graphic information, and how it is planned, organized, and displayed within a given context. A bar graph, for example, may allow the viewer an accurate representation of the annual rate of inflation in particular state, but the graph lacks clarity unless compared to the context in which it appears. The bar graph may be presented as an example in a book on designing graphic aids, as a transparency used in a presentation before a governor's economic task force, or as the centerpiece in a story in a local newspaper. Although the information of the graph stays the same, the graphics do not speak for themselves - they need interpretation whether in the text or as a part of an oral presentation.

According to many current models of perception, when you design a graph you encode information into shapes, symbols and colors. The success or failure of a graph depends on how well the viewer is able to decode the information you present. To understand how to format and graph data, you must first come to understand how people decode and perceive information. Designing graphs requires knowledge and manipulation of a viewer's perceptual skills. The aim of most graphs is to convey great amounts of information in a small space. But the ability to comprehend quantitative information depends on designing the graph to make the best use of a reader's perceptual skills. According to Cleveland and McGill's (1985) and Cochran et al.'s (1989) research, "People do no judge angle, area or shading as accurately as they can judge position along a common scale." (25) In a pie graph, for example, viewers "must spend time and effort to compare the differences in the angle and area of each slice, as well as cope with shading." (27) In representing quantitative information in a pie graph then, the designer often demands the viewer take the time and effort to sift through other forms of perceptual information. Based on experimentation, Cleveland (1985, 254) lists from most to least accurate the perception of graphical presentations:

1. Position along a common scale
2. Positions on identical but nonaligned scales
3. Length, direction, angle
4. Area
5. Volume, curvature
6. Shading, color saturation, density

In designing graphics, keep in mind that your viewer is more adept at perceiving the position of points on a common scale, than in perceiving the differences among the shades you use to distinguish plot points. Accordingly, the more perceptual demands the designer makes on the viewer, the less likely the relevant part of the display will be seen and understood.

To aid in the viewer's scanning and search of information from graphics, Cochran et al., offer three general recommendations: (27)

Label points and ideas clearly by using captions to identify what is being shown;

Integrate text and graphics. Do not make the viewer flip back and forth between the explanation of the graphic and the graphic itself;

Remove extraneous graphic information (sometimes referred to as chartjunk).

To help you in deciding whether to use a graphic, which graphic to use, and how to draft graphics, consider these questions and guidelines:

Do you need a graphic? In many professions and academic disciplines, graphics are staples for presenting information. Graphics are effective for emphasizing important points and information. But an endless procession of graphics diminishes the significance of the ideas and is as mind numbing as turgid prose. Do not use graphics indiscriminately. Make sure the graphic offers the best way to provide information. Since the purpose of graphics are to convey large amounts of information in a small space, consider the efficiency of using a graphic opposed to a prose explanation.

Draft a graphic as you would draft a paper. Use the conventions governing the design of specific graphics to determine which type is most appropriate for the information you want to convey. Do not settle for the first template you select. See how effectively the information is present in different forms. Experiment with the design to achieve the least cluttered, most accessible graphic possible. In weeding out clutter, Tufte (1983) suggests the graphic designer maximize the amount of information presented and minimize the amount of ink used. If a simple line graph, for example, is crowded with tick marks, grid lines and multiple colors, the viewer will lose the significance of the information presented.

What is your rhetorical purpose? Consider your goals and the contexts in which you are using graphics. Are you:

proposing (minimal, moderate, radical) changes in beliefs, ideas or actions?
introducing a new idea?
creating consensus with the community?
establishing your credibility or authority?
refuting preconceptions or common held beliefs?

Most single interactions do not result in achieving your convincing the audience of your views. In every community hearing, employee meeting and journal publication, members of the audience possesses many different moods, ideas and goals. Obviously you cannot address all of these factors. But you can establish the foundation for a second hearing or a sustained dialogue by connecting you rhetorical purpose to the audience's concerns.

Who is your audience? Nonspecialists are more familiar and comfortable with certain graphs - pie charts and pictographs for example - than others. These same graphics may not be appropriate for presenting information to specialists. Of course, the reverse is also true, sophisticated and unfamiliar graphs will alienate nonspecialist audiences. In drafting a graphic consider your role and the audience's role within the setting of the presentation. Groups of people will have various beliefs, values, interests and expectations, but you can make general, educated assumptions about an audience's views as a touchstone for your design choices.

Number the graphic. Give clear titles and labels.

Explain the graphic. Tell the reader what to look for and what it means.

Use color selectively. As software and printer technology continue to develop and become more affordable, you will have more opportunities to use color in graphic design. With the ease of using color comes the temptation for overuse. In using color consider the following guidelines:

Technical journals, in certain instances, do not accept color graphics because they are difficult to reproduce. Check the journal's publication guidelines;

Reproducing color adds to a publication's costs. If you are working, for example on a corporate report or manual with a production budget, you must weigh budgetary constrains against allure of color graphics;

Use of color can confuse audiences. We do not normally assign order or priority to colors. It makes no sense to say green is greater than blue. Complex graphics force viewers to memorize and keep in mind color keys, while trying to gather information from the graphic. The lack of a visual hierarchy among colors can frustrate viewers. A possible solution is using one color and establishing a hierarchy through shades of the color. Darker shades suggest greater emphasis, lighter shades suggest less emphasis. Darker colors and shades make items appear larger and closer.

Consider your rhetorical purpose, audience and information you present in choosing color. Bright colors grab an audience's attention, but can appear juvenile and inappropriate for serious or tragic information. Muted colors appear more dignified, but can underplay innovative or new findings and information.

Give the graphic a good deal of white space. While text and graphic are integrated, keep a clear border around the graphic.

Tables

Tables display numeric and quantitative data arranged in columns and rows. Tables benefit reader by organizing and classifying data. For small sets of data with exact numerical values, tables provide an alternative to using multiple graphs. In texts, tables present large amounts of information in much less space than words and paragraphs allow. The organization and classification of information in columns and rows allows the reader to see to relationship among numerical values. In oral presentations, tables can reinforce the speaker's points in an efficient, self-explanatory way.

Remember to check the recommendations provided in the style guides of professional societies and journals concerning abbreviations, the size and number of columns preferred, systems of units, unit conversions, decimal places to be retained and methods for "rounding off' decimal values. The recommendations we give here should be taken as general guidelines, each discipline and profession adheres to a unique set of standards.

Tables usually appear in the text in boxes set off by white space. Guidelines for constructing tables follow:

Number and give each table an informative title. Center the number and title above the data. Do not use a generic label such as "Findings" or "Experimental Results"; include the number of the table with a title including necessary and specific details for the reader's understanding. For instance a table titled "Table 4. Sample of New York Social Register Marriages, 1981 through 1985" is more effective than "Table 4. Test Results." Always include the duration of the study.

Arrange the table so that it can be read vertically, not horizontally. In so doing:

limit the table to a single page;
group items in the left-most column in logical order: alphabetical, chronological, topical;
skip lines;
label each column and row;
align numbers and decimals in a column;
limited decimals to two places.

Use standard units of measure and symbols as determined by the style guide of the discipline and journal for which you are writing. Generally, these same standards apply to documentation procedures used in oral presentations

Indicate the units of measure you are using in the title of the table. For example, "Table 3. Post-World War II Funding of Science in the United States (in constant 1992 dollars)." Also indicate in specific columns or row headings if you are using different units of measure within the table. For instance, it is easier for the reader to understand the distance of a particular marathon as;

26 miles, 385 yards and 15 feet rather than 133,170 feet.

However, in most instances if you give data in a table using a one unit of measure - feet, miles - do not shift to another unit of measure - meters, kilometers. Also, some units of measure have more than one name. Although both technically correct, hertz, as a unit of measure, is preferred over the unit cycle per second by the Institute of Electrical and Electronics Engineers because of the use of cycle alone as a unit of frequency.

Use footnotes for any items - column headings, sources of data - requiring explanation or taken from other sources. Use letters instead of numbers to footnote numerical data. Footnote form is also set by the standards of the professional society, discipline and journal for which you are writing.

Integrate tables into the text whenever possible.

Interpret the data for the reader. Tables present data, they do not interpret it. You must show and tell the significance of the data.

Graphs

Graphs represent the relationship among two or set of data by the use of dots, lines, or bars usually plotted on an X and Y axis. In this section we will discuss guidelines for designing scatter graphs, line graphs and bar graphs.

Scatter Graphs

Scatter graphs (also known as "scatterplots") use single, unconnected dots, plotted on a proportionate X and Y axis to represent the correlation between the rate of change over time of given variables. Scatter graphs are also useful in analytical or exploratory data analysis. The pattern of dots indicates to the viewer the relationship among variables. Random scattering generally indicates no correlation among variables. Dots along a diagonal (either imagined or drawn) in which the left endpoint is lower on the graph than the right endpoint indicate a positive correlation among variables; dots along a diagonal in which the left endpoint is higher on the graph than the right endpoint indicate a negative or declining correlation among variables. You can use shading and grid lines to distinguish groups and a correlation among data.

Scatter graphs provide an interesting example of when less is more in designing effective graphics. In designing scatter plots consider starting and extending the frame of the graphic (the X, Y axis) in relation to the range of the plotted points. You will notice viewers can easily determine the ranges and trends the plotted points represent.

Line Graphs

Line graphs are used to plot changes in quantity or value over time. Line graphs are also time-series plots - the horizontal axis (X axis) frequently represents a certain amount of time - 12 hours, 12 months, 10 years - and the vertical axis (Y axis) a certain quantity - the number of houses sold in Virginia during a year, or the failure rate of small business over the last 5 years. The rising and falling of the lines within the coordinate field express particular trends. One of the most common line graphs we encounter on a daily basis shows the movement of the value of stocks on the New York Stock exchange. The movement of the line indicates the overall change in value of the stock. However, one of the problems with time-series plots is that the simple passage of time does not offer any causal explanation. To know, for instance, that stocks rose 4 points in a period from 11 A.M. to 2 P.M. tells us little until we know what drives the trend. If two blue chip corporations announced their merger at 10 A.M., the viewer could determine if the announcement had an immediate effect on the market, as well as determine the net effect. To provide a causal explanation for time-series plots, you need to; 1) bring more variables in the graphic design, and 2) explain the graphic thoroughly.

Consider the following conventions for designing line graphs:

Use the horizontal axis to represent time; use the vertical axis to represent quantity;

Keep the horizontal and vertical axis proportional, include variables or units of measure at regular intervals along the axis;

Limit the number of lines within the coordinate (X-Y) plane ; if lines cross consider using no more than three;

Label lines clearly and always provide an explanation;

Designate each line by color when necessary. Although many lines graphs use colors and corresponding keys designate each line, keep in mind the limitations of a publication to reproduce color images, the number of lines in the graph, and the difficulty reader's have in easily determining color contrast;

Use grid lines when necessary to clarify the relation among lines and variables in the coordinate field. Grid lines serve as helpful guides in initially plotting data, but consider eliminating them in the final version of the graph. If you include grid lines, keep them in the background so as not to compete with the primary data. Keep mind the ratio of information represented to ink used.

Line graphs have many variations, but two of the most common are stratum graphs and ratio graphs. As with all lines graphs, stratum graphs show change in quantity over time. Stratum graphs are distinct in partitioning each section separated by the lines. Sections bounded by lines are typically distinguished by combinations of cross-hatching, vertical lines, dots and color. The advantage of stratum graphs is in furnishing clear contrasts among sets of data. However, designers of stratum graphs frequently display a proclivity for chartjunk. Tufte (1983, 107-112) notes the interaction of cross-hatching and vertical lines with the physiological tremor of the eye - moiré effects - in which the graphic design produces the disturbing appearance of movement and vibration. Computer graphics programs make it easier to design partitions in stratum graphs which generate moiré effects. Tufte suggests replacing cross-hatching with varying shades of gray, and/or labeling partitions with words.

A ratio graph represents a percentage of change in a quantity over time, rather than the real number of change. Although the horizontal (X) and vertical (Y) axis are proportional, ratio graphs are distinguished from other lines graphs by compressing the scale on the Y axis. Ratio graphs are used to compare large quantities with much smaller quantities. For example, you many want to compare the defense department budget to the budget of a small, privately owned engineering firm that deals with government military contracts. You could not graph the true relation between the two institutions given the disparity of the values - even if the percentage change in the respect budgets was equivalent over a given year. The line tracing the small firm's budget would appear flat compared to the much larger budget of the defense department. To accommodate the radical difference, compress the scale of the Y axis.

Formatting, Designing, and Using Graphics
Chapter 15: Part 1

David M. Toomey and James H. Collier

Book Contents

Chapter 15: Part 1

Formatting, Designing, and Using Graphics Chapter 15: Part 1

David M. Toomey and James H. Collier

Book Contents

Chapter 15: Part 1

Formatting, Designing, and Using Graphics
Chapter 15: Part 1