Milekic,S. (1997) Virtual Museums: How to make digital information
child-friendly?, in Museums
and the Web: Selected papers, Archives and Museum Informatics
Technological breakthroughs in the domain of computers and the
increasing use of the Internet have had a major impact on the
availability and dissemination of information. However, one large
section of the general population has hardly been affected by this
revolution. Currently there are no established means by which young
children, ages 2-5, can access the relevant parts of this vast
collection of information. The major problem is the lack of a
child-friendly interface environment which would allow children to
interact in a meaningful way with digitally presented information.
Although the problem was commonly recognized during the past decade
it was not considered acute until the recent explosion of information
availability. Using the model of an existing childfriendly
interface environment (KiddyFace, Milekic 1996), I argue in this
paper for the necessity of creating a specific milieu which will make
information accessible in an age-appropriate manner.
Introduction
I must preface this paper with the statement that I am neither an artist (officially) nor art historian or a museum curator. Most of the work described here was inspired by observations made during the period when I was involved in neuro-psychological rehabilitation of children with brain lesions. It was in this environment that I realized how critical child initiated action and the manipulation of objects are to the child's cognitive development and acquisition of knowledge about the world.
We find ourselves in the situation where information about works of art is increasingly available in digital form. I emphasize the difference between artworks which use the digital medium, and the use of the digital medium as a carrier of information related to a specific work of art. I do this because of the protective reaction I sometimes witnessed in museum curators when I presented them with a child-friendly digital environment. In this environment children have access to digital representations of works of art and, using a touch-sensitive screen for input, can manipulate them in a variety of ways. In short, they can draw a mustache on the Mona Lisa. Which is exactly what museums fought for centuries to prevent.
In this paper I will argue that making digitized information about artworks not only accessible but also manipulable by children has a great educational potential. For the first time in human history, it is possible - on a planetary scale - to enable a child to create a personal relationship with an art object by playing with, and investigating, the parameters that led to its creation.
The experience of forming a relationship with an art piece does not know age differences. It is never too early, never too late. However, for practical reasons, in this paper I will focus on how to make this experience possible for the part of the population which has been, because of inadequate interface, practically excluded from the "information revolution". These are children younger than 5, 6 and even 7 years of age for whom digital information is not only presented in an inadequate manner, but is often completely unavailable. This is even more of a paradox when one considers that this part of the population literally thrives on, and is most profoundly affected by, exposure to varied and rich sources of information.
The reasons for this state of affairs are numerous. Some of them are rooted in misconceptions about what computers are and how one interacts with them. The most prevalent idea of the computer is that of a large box emitting potentially deadly radiation, a keyboard with dozens of keys asking to be pressed which often has unpredictable effects, and a delicate mouse which requires precise eye-hand coordination to be usable at all. The use of the keyboard and the mouse as the primary input devices effectively defined which part of the population could have access to computers' information handling capabilities. By inertia this has continued to be so even when other ways of interacting with computers were readily available, both technologically and economically.
In order to define the optimal ways in which to
make digital information available to children we have to take a
closer look at the characteristics of children's spontaneous natural
activities. The most prominent activity in early childhood is that of
play. The need to examine the relationship between computer use in
early childhood and play has become evident only recently in the
literature (Henniger, 1994a, 1994b). Although there seems to be a
general consensus that computers can be powerful educational tools,
most of the design efforts have gone into designing software for
computers as they are now. I see this as the major conceptual
barrier to the development of genuinely child-friendly digital
environments.
What needs to be changed?
There are three major areas which need to be addressed in making digital information child-friendly. They can be loosely defined as changes in a) location, b) mode of interaction and c) content structure. Each of these changes will be briefly described in the following paragraphs.
Change in location. Although it seems trivial at first, change of location of objects is the first indicator of the psychological change of domain perception. Just moving computers from desks to the floor will not only make them more accessible to children but would also indicate that they are legitimate parts of their environment. Of course, modern computers would hardly survive this change, because they were not builtwith children in mind. Moving computers to the floor would also mean making them at least as child-resistant as any good toy. As simple as it is, change in location also implies a host of other changes in the design of child-friendly digital devices. First, the computers should lose all of their appendages and the cords that connect them. This means getting rid of the electrical cords, the keyboard and the mouse and making the image-displaying part self-standing (battery operated). Putting the display on the floor also means change in the orientation of the viewing surface from perpendicular (where the child had to look up) to a more physiological upward-facing angle.
Change in mode of interaction. The change in the mode of interaction is not only dictated by the change in location, but is necessitated by the inadequacy of the keyboard and the mouse as input devices for children. Both devices depend on possession of special kinds of knowledge and skills, not readily available to children. The use of the keyboard relies not only on competent writing (and typing) skills but also on knowledge of specific vocabulary and its use (for example, that typing "exit" will end the current game). This does not mean that in a child-friendly interface environment keys would be banned from existence, but only that their number, size and function would dramatically change.
The use of the mouse suffers from similar shortcomings. Not only is it inherently abstract (what I do here moves something there, but not always…) but it also involves fine visual-motor coordination. The surface area of the 'folders' displayed on the monitor I am currently using is approximately 1/4 square inch and it is within this range that the child has to coordinate the movement with the 'click' (sometimes even 'double-click') in order to make something happen. The size of the interface elements is not the only problem. One could easily increase the size of typical 'buttons' on the screen and use another input device (like the touch-sensitive screen), but the problem of interaction still remains. The adult-designed ubiquitous 'desktop' metaphor with its files and folders, and subfolders and 'windows' (on the desktop?!) is hardly a typical child's handy metaphor. The necessity for changing not only the input devices but also of the way the digital information is structured is the topic of the next section.
Change in content structure. The change in content structure does not mean change in content per se, but rather change in the way the content is organized and presented to the child. It is blatantly obvious that to an illiterate person (or a child) all the 'folders' on my computer monitor look pretty much the same. Thus, in a child-friendly digital environment the indicators of content should be clearly distinct visually and represent familiar aspects of the child's experience. However, this is the most superficial change necessary. There are other aspects of children's activity that call for more radical changes. These are a) making the information (content) manipulable, and b) making the content structure compatible with the child's social environment.
a) The idea that making the information experientially accessible to the child leads to more efficient knowledge transfer has been around for some time but has not been consistently implemented in the area of child/computer interface design1. Thus objects should be made manipulable in the way which makes sense to the child and provides feedback which can compensate for the unavoidable impoverishment of sensory input in comparison to equivalent realworld manipulations. Several ways of achieving this goal will be discussed in the section describing the KiddyFace environment.
b) Making the content structure compatible with the child's social environment means that it readily supports the social interactions a child is likely to engage in: interactions (playing) with peers and interactions with adults (educators, parents). Originally, computers (as the 'personal' part of 'PC' implies) inhibited interaction and collaboration between individuals. It is only recently, with the popularization of the Internet (which was made possible by a friendlier interface design) that the importance of supporting collaboration is being discovered again, and that different software products supporting it are boasting three-digit return rates. Although it is possible to imagine similar solutions for children's computer environments, it is important to realize that a child-friendly environment has to support social interactions more concretely, and respond appropriately to simultaneous and possibly divergent inputs on the same physical unit.
What is child-friendly?
In this section I will propose some of the concrete steps necessary to make digital information child-friendly. Most of the described solutions were implemented in an interface environment (KiddyFace) designed to assess various cognitive functions of young children. I will discuss the practical solutions following the same order of presentation as in the preceding section.
Location and design. As described above, a simple change in location from the desktop to the floor, leads to, dramatic changes in computer2 design. The computer loses its recognizable parts like the keyboard and the mouse, and is reduced to the display unit. Ideally, the display itself should be compact, mobile, and with an upward-facing touch-sensitive viewing surface (see fig. 1). It should also be rugged, scratch-resistant and use a built-in power source.
The unit should also have ample storage capacity (a large hard drive) and a way to quickly access, modify and update stored information (via infra-red communication ports). The paradox is that units with the above characteristics are already commercially produced but have never found their way to the floor of a child-friendly environment! They are typically used in ATM machines, for finding merchandise in modern superstores, or for finding locations in shopping malls or airports. Some are even used in museums but, as far as I know, with very limited possibilities of interaction.
Child-friendly modes of interaction. The mode of interaction with the display unit in the KiddyFace environment does not pose additional cognitive overload like the use of the keyboard or the mouse. It is physically supported through two input devices: the touch-sensitive screen, and the voice recognition mechanism (through the built-in microphone and cheap, commercially available software).
User input through the touchscreen was selected for a variety of reasons. Pointing to and touching an item are the most natural ways of indicating its selection, and require no training even in very young children. Touchscreens are very durable, have no moving parts and require almost no maintenance. Since they are superimposed over the viewing surface, they demand no additional space. A host of studies on adults (summarized eloquently by Sears and Schneiderman, 1991) indicated that touchscreens were the fastest pointing devices. However, if used for the selection of very small targets (less than 10 mm in diameter), they were also the ones with the highest percentage of error rates. The results were partially caused by the low resolution of older touchscreens and the returning of multiple pixel locations by the touchscreen hardware. In the past several years both the increased resolution of touchscreens and the software-implemented strategies for stabilizing the touch location have reduced touchscreen error rates and brouight them in line with those of the mouse. It is worth noting that even with the older touchscreens there was no difference in error rates between the touchscreen and the mouse in conditions where larger selection targets were used (Ostroff and Schneiderman, 1988). In a pilot study conducted at the Hampshire College Cognitive Development Lab we have found that even children as young as 2 years find the use of a touch-sensitive screen intuitive and easy. Furthermore, their performance on a simple visual mapping task was quite good, possibly as the result of the decrease in cognitive load associated with the interface (Milekic and Rattermann, 1996).
The change in the input device from the mouse to the touchscreen is not in itself sufficient to provide a child-friendly environment. A number of the typical actions one can perform, as well as the graphical "look and feel", had to be redesigned. Some of these features will be briefly described.
Selection of an object is carried out by touching it. There is no traditional highlighting of selected object (necessary for the mouse input) because of the existing haptic feedback. However, since all of the objects have defined 'anchor' points which are used for 'dragging' action, there is often a small movement of the selected object as it aligns its anchor point with the touch point. Selected objects can be dragged by moving the finger across the screen, and 'dropped' by lifting the finger. If objects are dropped over the appropriate slot (which usually corresponds in size and shape to the object being dropped), there is a suitable form of visual and acoustic feedback. In addition to the traditional mouse-supported actions, there are three more types of interaction supported in the KiddyFace environment: pointing to, throwing and pushing.
Pointing to consists of simple pointing (and touching) of the desired location. This action will cause the selected object (the one last touched) or the fitting one (with the corresponding size and shape) to move on its own to the chosen location.
The throwing action is executed when dragging of an object across the screen exceeds certain speed (which corresponds more or less to the speed of the natural throwing motion). In this case the 'thrown' object will continue to move in the same direction even when the finger is lifted off the screen.
The pushing action follows pretty much the same pattern of events as a reallife action. When one side of the object is approached and touched with a moving finger the object starts moving in the same direction in front of the finger. The motion is terminated when the finger is lifted up from the screen or when it stops moving.
The voice recognition mode of interaction is still in an experimental stage. The desired interaction would be to allow the child to perform simple actions and navigation using oral commands like "go", "no", "more", "new", etc. It remains to be seen whether the current hardware and software can cope with these demands in a satisfactory manner.
Child-friendly content structure. Since the focus of this paper is how to make digitally encoded information child-friendly, it is assumed that most of this information will be of a graphical nature. Although most of this information (for example, paintings in a museum) was not meant to be manipulated, it is necessary to allow the child to play an active and creative role in interacting with it. Again, it is important to remember that the child is not actually manipulating the original works of art, but rather the digitally encoded information about them. In the case of digital reproductions of graphical works of art it is possible to allow the child to play with all the parameters that the artist manipulated in creating the piece of art: color, space, proportions, etc. It is only in this way that the development of the most intimate and complex relations with the work of art is possible. For the time being, the KiddyFace environment allows the manipulation of simple parameters, such as size, orientation and proportions. I must emphasize here that although this approach is child-friendly, it is by no means restricted to children. Aesthetically 'illiterate' adults are in the same position in regard to their relationship with works of art as inexperienced children. The development of such interactive learning environments for art education has been described elsewhere (Milekic and Driscoll, 1996).
Manipulability of works of art is not restricted only to twodimensional works. Currently there are several tools which allow smooth manipulation and exploration of three-dimensional objects (like Apple's QuickTime VR player). Although not as successful, there are some attempts to achieve the same level of manipulability in the realm of classical music (Beethoven's 5th).
Conclusion
In order to allow children to share the vast
amounts of digitally encoded and stored museum information, it is
necessary to create specific childfriendly environments. This
entails both the psychological and the physical redesign of computers
as tools for manipulation of digital information. The anticipated
changes include changes in the location and shape of computers as
well as major changes in the types of input devices (touchscreen,
voice) used. In order for these changes to be effective it is also
necessary to develop the software which supports them. Shifting the
emphasis from observation to active exploration and direct
manipulation allows interaction with digital information to occur in
an age-appropriate manner.
References:
Henniger, M.L. (1994) Software for the Early Childhood Classroom: What Should It Look Like?, Journal of Computing in Childhood Education, 5(2), 167-175
Henniger, M.L. (1994) Computers and Preschool Children's Play: Are They Compatible?, Journal of Computing in Childhood Education, 5(3/4), 231239
Mikropoulos, T.A., Kossivaki, P., Katsikis, A., Svranides, C. (1994) Computers in Preschool Education: An Interactive Environment, Journal of Computing in Childhood Education, 5(3/4), 339-351
Milekic, S., Driscoll, R. (1996) Interactive Learning Environments in Art Education, unpublished article
Milekic, S., Rattermann, M.J. (1996) The Effect of Computer Pointing Device on Cognitive Performance in Young Children, unpublished research data from a pilot study
Papert, S. (1980) Mindstorms: Children, Computers and Powerful Ideas, Basic Books, Inc., New York
Sears, A., Shneiderman, B., (1991) High precision touchscreens: design strategies and comparisons with a mouse, International Journal of Man-Machine Studies 34, 593-613
Endnotes
1 Of course, the notable exception is Seymor Papert's LOGO and the use of a programmable 'turtle'. In his "Mindstorms" Papert also mentions principles similar to the ones exposed in this article. However, the major focus of Papert's approach was in the area of mathematics and was not generalized to the presentation of digitally encoded information of other types.
2I use the term 'computer' for the lack of a better one. One definitely needs a new term for the device whose main function is not any more to compute, but to store and represent the information.
Word 6.0 for Windows version, 130
KB
