Billie J. Wahlstrom
"User-friendly" seems an easy enough concept to comprehend, but like so many lively notions, it slithers out of the field of vision when we turn the research microscope on it. Increasingly, the confusion caused by this term has created difficulties for those of us involved in teaching about and doing research on computer software, interactive video, and other educational technologies to which it has been applied. The difficulty of defining user-friendly results from the fact that software, like all messages, has two components--the content itself and the format in which the content is presented.1 So far, the concept of user-friendliness has been applied both to format and to content rather indiscriminately. We call a software application user-friendly if its format can be used easily by a computer novice. But we also call it user-friendly if the persona used as part of the software's content talks nicely to us once we get the program up and running. Content and format
are, of course, connected, but it is possible to separate what is said from its method of presentation. And it is this separation that is not yet clear in our writings.
Working toward a clear definition of user-friendlv is not just an exercise in nitpicking. It gets at some of the more important issues involving the creation of computer-assisted instruction (CAI) and other courseware; its use; and, more basically, the nature of the computer-mediated communication.
I suggest that user-friendly be used more accurately in the future to refer to the format and not the content of courseware. What we have been calling user friendly with respect to content, I would argue, belongs more appropriately to the discussion of narrative and point of view. The following criteria, all of which deal with format alone, would then be central to determining whether or not the product of an educational technology is user-friendly. First, a user-friendly format should allow a lesson to be easily accessible. Second, in user-friendly software, students should know where they are in the lesson or package at all times. Third, in user-friendly software, students should be able to enter and exit the lesson easily and quickly. Fourth, they should be able to exit and re-enter a program where they left off without difficulty and also without having to go through sections they have already covered. Finally, in userfriendly software there should be internal guidance aids to let users know where they are,
how to proceed, and how to get help. If these criteria are met, the product of educational technologies is user-friendly.
When it refers to a condition rather than to content, user-friendliness is useful in evaluating the format of any communication product. A book, for instance, can be user-friendly. If one is literate, the content of most books is easily accessible. If one has a book and is literate, one has only to open the book to get at its content. Unlike video tapes, video discs, or computer software, no additional equipment or training is necessary to access the content of book. Whether or not one can understand the content of any particular book and all issues about its quality are separate considerations beyond the scope of this discussion.
With respect to CAI, the issue of accessibility becomes more complex. To use CAI requires electronic equipment and an additional kind of literacy; yet, given these additional constraints on its accessibility, the relative user-friendliness of CAI can be measured by the ease with which the content of the software is reached. If one can get at the content--that is, run the program--without difficulty and without much prior knowledge of computers, then this CAI meets at least one criterion of user-friendliness.
To be user-friendly, a computer or videodisc program should also be designed so that users know--through maps and menus--where they are in the program at all times and can enter and exit it at any time or place they might wish. The content of software should be displayed in such a way that users know where they are in the lesson as well as what their options are for going on,
backing up, printing or saving their entries. Books are user-friendly in most of these respects. One can open a book quickly, get to the section one wants easily, close the book and open it at any point in the text with little problem. Page numbers and chapter headings provide readers with clues for orienting themselves in relation to the book's content. And, of course, one can jot notes in the margin and refer to them at a later time.
Although in some ways CAI programs cannot be as friendly as books, the example below from a user-friendly CAI program called WORDSWORK2 illustrates how special features can be designed to aid the user in getting at the content of the lesson. The two consecutive screens in Figure 1 have several of the user-friendly features, or orientation clues, mentioned above. The bottom lines of the screens indicate which choices students have. The program allows users to back up as well as go forward; in this way, the program facilitates rereading.
In WORDSWORK, there is also a help key, which explains the function keys, and, in addition, a key that displays a map (Figure 2) so that users can see exactly where they are in the program. Maps are important because they facilitate the process of reentering a program where a user left off.
In simple programs, users can be kept aware of their progress by a line at the top of the screen that tells where they are in the program: (Screen 5 of 10). In the WORDSWORK example, there is also a "quit" key allowing a user to leave the program without finishing it or without having to waste time by "paging" through it to the end.
p. 17 (Figure 1)
p. 18 (Figure 2)
A user-friendly format may also involve assigning special keys only one function and displaying only those choices immediately available to the user. For example, the function keys F5 and F6 are used elsewhere in WORDSWORK to allow users to print text or to examine their files. Where such functions are not appropriate, the keys are not displayed.
Maps and menus like those in Figure 2 also contribute to user-friendliness by helping users control software programs. Ideally, in CAI programs, complex menus should be nested. That is, there should be a main menu listing the major segments of the program and secondary menus for each segment listing specific contents. As users finish various segments, the main menu reflects that fact; completed sections are marked with an asterisk or in a different color. In Figure 2, screen A, users select the segment they want by using function keys (F3 and F4) displayed only with menus, and their choices are highlighted so that they can double check their selection before they press F1 to go on. Thus, a user's chance of ending up in the wrong section because of an error in moving the keys is minimized. In this program, orientation maps are kept as simple as possible, showing the user only the segments of the program he or she has already encountered in the menus. Segments of the program that the user has not yet reached are not displayed. The user's current location in the program is indicated, in this program, by red highlighting.
Because of their interconnectedness, it is not surprising that initial discussions of
user-friendliness conflated form and content. When we discuss the content of an educational message, we have moved beyond the discussion of user-friendliness. All authored texts--whether print narratives, film dramas, or technical discourse--have personae. Essentially, a persona is the implied or perceived author of a text or of parts of a text. In Figures 2 and 3, three different personae appear. In reality, two actual authors created the examples but chose to present their material in several different voices to determine which was most effective. An actual author (in the case of video, film, and technical discourse, often more than one person) creates a persona as a vehicle to convey the information of the text. In technical discourse, for instance, the implied author might well use a objective or reportorial persona, but behind this persona lies the real author. There is nothing dishonest about the difference between real and implied authorship. It is simply a function of discourse.3
The implied author of a CAI program, for example, can be any of a variety of possible personae. It is the current thinking among some CAI designers that a friendly persona--that is, non-arcane, supportive, patient--works best in educational software because it encourages the user to continue with the program. This may or may not be true, but it is important not to confuse the friendliness (or neutrality) of the persona of a program with a program's user- friendliness, or with the friendliness of the author, for that matter. The persona of the CAI might be charming, a clever electronic Lothario and the actual author a nebbish, but if
the format of the software is inaccessible to the average user, the program is still not user-friendly.
A number of personae may appear in a program. The menus and instructions may be written in a neutral style, for example, in which the author is attempting to be as unobtrusive as possible. The persona for this part of program may appear to be without personality. The body of same program however, may be presented through the words and characters of a less transparent persona. Figure 3 illustrates one of the two persona used in the WORDSWORK lesson designed to provide instruction on the narrative.4 The first persona, seen in Figure 2 is the neutral and objective persona found in menus and help screens throughout the lesson. The second, persona, illustrated in Figure 3, is the friendly character who "narrates" the lesson. Neither persona was intended to resemble the authors' personalities. Both personae were conceived of as being the appropriate presenters--that is, implied authors--of their material.
Evaluating the user-friendliness of a program should not be based on whether or not personae seem friendly. Recently, I worked with a CAI program from a major publishing house. The context of the program was interesting, the persona conveying the information was pleasant and easy to understand. The software package was anything but user-friendly, however. First, users of the program could not back up; consequently, when I accidentally advanced two screens instead of one, I could not get back to the screen I had missed. When I tried to do so, I discovered that I had to exit the program, go
p. 22 (Figure 3)
back to the beginning, and advance through every section I had already covered, screen by screen, until I reached my place again. Next, I discovered that I could not count on a consistent use of the function keys. The same function key had three different uses in this short program. Before long, I had pushed the wrong key and found myself having to start the program all over again just to get back to a screen I wanted to read. Lastly, it was impossible to exit the tutorial. To get out in a hurry, one could only turn off the computer. This was a user-hostile program, despite the fact that efforts had been made by the CAI designers to create a supportive persona.
All new educational technologies have a struggle with persona, though we seldom think much about this issue. Because its use precedes that of CAI in time, educational television (ETV) provides interesting insights to this problem. ETV is frequently awful, an interminable procession of talking heads providing information often better adapted for presentation in another medium. Because ETV was first developed when the "bullet theory" of communication was current, educators gave little thought to personae. They believed that to create an effective educational message all one had to do was generate a well-formed "bullet" of information and shoot it. All viewers would be hit with the content of the lesson. Consequently, ETV was conceived as "no-frills" television. The result were ETV personae who appeared earnest and objective in the extreme and who looked the camera right in the "eye." The effect is not very engaging, although information, can be conveyed this way.
CAI has gone in the opposite direction, having viewed the issue of personae more
flexibly. In its attempt to be engaging, to draw on the dramatic traditions of commercial television and to make use of the devices originated in arcade games, CAI is frequently dominated by distinctive personae. "Representing Numbers by Letters" is one such program.5 Designed in 1984 for community college students who have not taken algebra, this program features Mr. Mathyou. Mr. Mathyou is an animated figure who wears sunglasses and who speaks to users in cartoon-style thought balloons appearing above his head. Programs also have been designed to convey information as if the computer were a person addressing the user.6 Figure 3, screen A, illustrates the humanized computer as persona. Figure 3, screen B, indicates how personifying the computer introduces a curious blurring of the distinctions between humans and machines as well a blurring of historical eras.
CAI designers have attempted to create interactive programs that invite use, but it is precisely in our attempts to deal with personae that we have complicated research on user- friendliness. Although there are some very good reasons why the format of programs must be user-friendly, it is unclear whether or not they should have personae that seem friendly to the users. Many people have indicated that they get tired of cute programs if they have to use them frequently. Some users prefer a neutral persona, partly because of the time it takes to read supportive comments when users could be getting on with work. Other users have indicated that they respond well to hostility--in varying degrees--in learning situations. Research that I have begun with my colleague Jack Jobst at MTU is
beginning to examine the relative effectiveness of personae.(7) Our experiences have made us wonder whether not a hostile persona, in a user-friendly program, might not be best for some audiences and some applications. Maybe users like to be berated a bit when they know that they have been careless or have failed to master something they are, in fact, able to master ("By now even you should have learned to use the semicolon!").
Research into user-friendliness has been slowed, in part, by our own efforts to discuss the subject. Our discussions often jump from how the menus are organized to how students can save their words and to how patient the programs are. Until we are clear about whether we are considering format or content, we are going to continue to make confusing statements that do not contribute to further understanding of CAI. Imagine, for instance, an analysis of a novel that moved willy-nilly from the nature of the main character to the display of chapter headings in the table of contents and then back for comments on the fact that the neatest thing about the novel is that you can close it up and open it again at will. We have become more sophisticated in our literary discussions. We know that the author creates the whole text; that the characters, themselves, have nothing to do with the titles of chapters; and that the nature of "bookness" means there are format requirements associated with novels only some of which can be manipulated by the author.
More research needs to be conducted into both the content and format issues of educational technology. I teach, use, and study CAI and other computer-driven educational technologies and feel it is important for us develop a methodology for analyzing format and conventions in computer-mediated communications. I am also interested in learning more about the nature of human/computer interactions, especially how human beings best assimilate mediated information.
The form and content of novels have developed over many years, and it has taken even longer for criticism to learn to deal with these concerns. It will be a long time before we will have such well-conceived forms for computer-mediated messages, and it will be still longer before our theory and criticism catch up with developments in this field. We are not, however, aiding in the development of new knowledge in our field if we do not expand our theoretical understanding. Until we decide what is a matter of form and what a matter of content, manage to keep the author and the persona separate in our discussion, and do more research, we can only hope to muddy the waters and postpone the day when all the silt settles and we can view computer-mediated messages clearly.