Education, Communication & Information, Vol. 3, No. 3, November 2003
An Analysis of Horizontal and Vertical Device Design for School-related Teaching and Learning
WALTER M. STROUP & ANTHONY J. PETROSINO, The University of Texas at Austin, USA
This paper reviews and analyzes 12 horizontal and vertical design attributes of devices for school-related use. Horizontal design is the familiar all-things-to-all-people, ‘just-in-case’ design associated with desktop, laptop, and mainframe computing. Less familiar is the ‘vertical’ or ‘just-enough’ dimension of computational design, where device functionality is tightly coupled with specific needs in identifiable speciality markets. We believe that equitable access to key forms of learning functionality for all students and issues of total cost of ownership will provide the impetus for K–12 schooling to integrate horizontal and vertical technologies. We use two physically similar but functionally distinct portable handheld devices to illustrate the 12 design attributes. Representing horizontal design are Palm operating system-based handhelds and representing vertical design are graphing calculators.
ABSTRACT
Introduction Computing in business settings reveals at least two top-level dimensions of design and implementation. One dimension is the familiar all-things-to-allpeople, ‘just-in-case’ design associated with desktop, laptop, and mainframe computing. This form of computing constitutes a ‘horizontal’ dimension of use and design because these platforms—both the physical device and the operating system—support an extraordinary breadth of potential uses and kinds of interdevice compatibility. Less familiar is the ‘vertical’ or ‘just-enough’ dimension of computational design, where device functionality is tightly coupled with specific needs in identifiable specialty markets. These vertical devices have become a ubiquitous presISSN 1463-631X print; 1470-6725 online/03/030327-19 2003 Taylor & Francis Ltd DOI: 10.1080/1463631032000149665
328
W. M. Stroup & A. J. Petrosino
ence in settings like parcel delivery, inventory management, and some areas of medical practice. Most people have used such devices to sign for package deliveries, to scan product bar-codes, or to take blood pressure. However, few people have considered the potential implication of this kind of design for educational computing.This paper reviews and analyzes horizontal and vertical design attributes for devices in school-related use. Our goal is to help educators think in new ways about how we might provide important forms of computing functionality for all children. By ‘functionality’ we mean the ability to carry out various kinds of operations on data types, e.g. plot a set of data points for a science lab or scaffold a student’s engagement with learning to read by having a device speak out loud a word when the learner touches it in a sentence. Computing is, by nature, flexible. This means significant learning functionality can be implemented in many ways and across a range of kinds of hardware and software. With this article we hope to provide a framework for educators to think about important school-specific design issues. How, for example, might a district technology committee begin to think about whether to invest in a general-purpose device like a personal computer and then install specialized reading software on it, or invest in an elementary-focused reading device like a LeapPad? Does it make sense for schools to buy, as Maine and Henrico County, Virginia in the USA have, laptop or desktop technology as general tools to support teaching and learning? Would it be better to focus on more dedicated uses of technology that address specific needs and that are tailored from a hardware and software point of view to meet each of those needs? What are the tradeoffs and benefits? In considering these issues we do not believe there is a ‘one size fits all solution.’ Nor do we believe that either horizontal or vertical technologies, by themselves, are ‘the’ solution for schools. What we hope to do is to structure the analyses decision-makers might use to understand how horizontal and vertical design considerations are relevant for schoolbased technology implementation. Our sense is that it will be the thoughtful integration of these capabilities that will characterize school-based technology use in the future. We begin the article with a familiar example of a successful, largescale adoption and integration of horizontal and vertical capabilities. This example comes from a non-educational setting. We chose a non-educational example not because we think the design issues for business are exactly the same as those in education, but rather to highlight how the integration of horizontal and vertical design can make sense on its own terms, even when the financial resources might be present to pursue more costly alternatives. The sense is that even for these large multinational corporations, ones with the kinds of financial resources available to them for technology implementation that far exceed those of schools, it was the integration of horizontal and vertical design capabilities that ended up making the most sense. Accordingly, schools and districts with much more
Horizontal and Vertical Design
329
constrained per-person funds might, then, be able to learn from this integration of horizontal and vertical design. To move the conversation forward, issues of horizontal and vertical design should be situated relative to school-based practice. With this in mind, we introduce 12 horizontal and vertical design characteristics of relevance to K–12 schooling. After discussing these characteristics in general, we then illustrate them using two physically similar but functionally distinct devices. We conclude with some observations about where the conversation related to integrating horizontal and vertical design for schools may need to go next. Illustrating the Integration of Horizontal and Vertical Functionality The use of ‘vertical’ in the context of this paper is based on the business idea of a ‘vertical market,’ a particular industry or group of enterprises in which ‘similar products or services are developed and marketed using similar methods (and to whom goods and services can be sold)’ (TechTarget, 2002). Traditional vertical markets include insurance, medicine, real estate, and banking. Naturally, the presence of vertical markets results in the development of products and services tailored to these industries. Software and hardware technologies have long targeted selected vertical markets. Parcel delivery services use relatively mature vertical technologies, with a suite of closely coupled forms of devices and software working together to address critical business requirements (Stepenson, 2002). What is interesting, from a design point of view, is how markets with mature, well-implemented vertical technologies continue to use both horizontal and vertical capabilities in cross-organization technology. Working right alongside well-developed vertical devices are horizontal, ‘just-incase’ technologies like standard desktop computers and popular commercial software packages (e.g. Microsoft Excel). The common use of overnight mail in the workplace illustrates this organization. The sender fills out an envelope with a tracking number bar-coded on the outside of the mailing container. Then a delivery person from the company picks up the envelope and scans the bar-code on the envelope. As the package travels to the shipping terminal, the airport, the destination city, and, ultimately, the intended address, a series of vertical devices scan information about the time, location, and person involved in the delivery of the package. At any time in this process, the sender, sitting at a networked desktop computer, can access the shipping company’s web site, submit a tracking number, and get immediate feedback regarding the package’s location. Is it still en route? Has it been delivered? Who signed for the package? This apparently seamless integration of vertical devices specific to the shipping business and common horizontal devices in the workplace exemplifies the power of horizontal and vertical devices working in concert. In the future, we anticipate schools will develop a similar integration
330
W. M. Stroup & A. J. Petrosino
of horizontal and vertical technologies. What may drive this integration is a commitment to equitable access to key forms of technology-supported learning for all students. Indeed movement in this direction has already begun. For example, the State of Texas in the USA has mandated access to graphing capabilities for all students1. Given the shortcomings in equitable access to computing—typically discussed under the name ‘digital divide’—schools are struggling to provide this sort of meaningful and ubiquitous access for all students. The need to provide universal access will continue to interact with issues of total cost of ownership and will push schools to explore solutions that go beyond simply providing a shared computer lab ‘down the hall,’ a mobile cart of laptops, or a small number of machines in the back of every classroom. We believe horizontal and vertical designs are complementary when used for school-related learning. We also believe that needs assessment and implementation require school-based educators and administrators to understand and to situate specific devices relative to these design attributes. As is true for any technology-related needs assessment and implementation in schools, the overarching goal has to be supporting the best possible fit between computing design and the kinds of functionality that matter for teaching and learning. The following horizontal and vertical design attributes are proposed as a way of organizing and situating issues of hardware and software design relative to patterns of implementation, support, and ongoing learning activity in schools. This article looks to clarify these dimensions as a first step in moving toward the kinds of successful integration of horizontal and vertical design now found in many non-educational settings. Horizontal and Vertical Design Attributes We present 12 design attributes, organized and analyzed in terms of horizontal and vertical dimensions. Although a specific device may tend more toward one of the dimensions, it may also have at least a few attributes that fit better with the other dimension. As is true for most analytic lenses, pushing toward clarity can, to varying degrees, oversimplify certain ‘gray areas.’ For example, our association of teachers with a vertical experience—an algebra teacher in her role as algebra teacher—is not intended to suggest that teachers don’t have horizontal (cross-context) experiences. Teachers do have these experiences, and these do need to be addressed in supporting the practical aspects of their work in schools (i.e. a school system might issue to each teacher a horizontal device like a laptop computer or personal data assistant to support work in a typical school day). Nonetheless, in her role as algebra teacher, the secondary teacher’s design needs may be seen as fitting better with vertical design. We’d also note that despite some references to secondary subject areas in what follows, a similar sense of verticality might
Horizontal and Vertical Design
331
FIG. 1. The teacher’s experience tends to be vertical and the student’s horizontal.
also hold for elementary teachers, even if their overall teaching responsibilities are more cross-disciplinary. The elementary teacher’s experience may be thought of as vertical because at any particular point in the day, he is typically focused on one kind of domain learning. Students in elementary settings, either in various groupings or individually, may be working on a range of different tasks. Yet despite this variability teachers might structure the settings so that each of these tasks is supported by the use of a particular ‘optimized’ device. Currently, most elementary teachers expect to use a range of different ‘specialized’ manipulatives for teaching math concepts. In a similar way, the use of optimized vertical technologies—a keyboard for learning about music and a graphics tablet for working on handwriting—might become a ubiquitous feature of a teacher’s experience working in elementary settings. Thus while we fully acknowledge the ways in which a push toward clarity in what follows may oversimplify some of the nuance found in real schools, we would maintain that the issues of vertical and horizontal design are useful in addressing practical issues of technology integration across the full spectrum of K–12 teaching and learning.
332
W. M. Stroup & A. J. Petrosino
1. Designed for Whom. As shown in Fig. 1, we associate the teacher’s experience with a vertical dimension of school activity and associate the student’s experience with a horizontal dimension. This is intended to capture and organize important aspects of what kinds of devices are designed for whom. Horizontal. A student’s experience varies considerably as she moves from class to class, studies different subjects, organizes homework, schedules personal activities, keeps track of phone numbers, and maintains a ‘to do’ list. Vertical. The classroom teacher focuses primarily on teaching a domain area in a specific group context. An English teacher teaches the domain area of English to a group of students who are intended to develop English-related insights and abilities. An elementary teacher, at a particular point in the day, may be focusing on cursive writing using graphic tablets. In each case the teacher may want sets of devices that can be ‘counted on’ to implement certain kinds of functionality in a consistent way for all her students. 2. Focus of Functionality Horizontal and vertical designs have clear implications for what the devices do (e.g. add or spell check). Horizontal. Addressing the varied experience of the student, the device should support a wide range of current and possible functionality. This is what we mean by ‘just-in-case’ design. Vertical. The teacher wants certain core functionality to be available to all the students in her class nearly all the time. Because the experience needs to be uniform and predictable, flexibility is not highlighted and may actually detract or distract from intended classroom use. This is what we mean by ‘just-enough’ design. 3. Physical Movement Between Contexts Contexts in school are usually associated with physical place. Moving from doing math to doing English requires physical movement between classes or between locations within an elementary classroom. Horizontal. A horizontal device will move with the student in a relatively unobtrusive way. Close physical proximity also highlights a sense of intimacy, customization, and personal expression (like clothing and cell phones). Vertical. Since functionality is often associated with a specific context, such as a math classroom or math area in an elementary classroom, physical portability is less salient. Characteristics like unobtrusiveness and
Horizontal and Vertical Design
333
personalization may still matter, but for different reasons. Unobtrusiveness may matter relative to not interrupting important lines of sight in the classroom or being able to fit on small desktop surfaces. Personalization may be associated with preserving individual identity during a given class. Students put their names on the tops of pieces of paper or, in an electronic context, are identified in a network-based simulation. 4. Interdevice Communication Teaching and learning are based on communication, so communication capabilities matter in technology design for schools. Horizontal. Many horizontal devices are closely associated with individual use. Consequently, communication capabilities may have evolved as a natural extension of individual needs. Point-to-point communication with physically proximate peers is the emphasized form of interdevice communication (similar to students passing paper notes to one another, Palm operating system (POS) devices allow for peer-to-peer communication). Vertical. Group interactivity is highlighted in a classroom. Typically, a range of group sizes and forms of interaction are present, so a device must be designed for whole-class, collaborative group, peer-to-peer, and individual use. These more flexible and powerful forms of communication require support capabilities associated with a true network. Much of this truly ‘group-oriented’ network design is relatively new, as opposed to the common use of networks to support peer-to-peer communication such as e-mail. 5. Operating System Complexity or ‘Weight’ Horizontal and vertical designs have implications for the level of complexity, or what is sometimes called the ‘weight,’ of a device. Generally, increased complexity is associated with higher cost due to the increased cost of components, software implementation, maintenance, and training. Horizontal. In order to satisfy the urge to be all-things-to-all-people, horizontal devices are required to be inclusive in supporting a wide range of possible cross-context needs. For example, in order to be able to print a document in a range of possible contexts, the device must ‘know’ about a wide range of printers. Supporting this inclusiveness increases the complexity or ‘weight’ of the operating system. Vertical. Operating system complexity is generally minimized to support a narrow set of functionalities. Moreover, as devices become integrated into classroom networks, much of the complexity can be offloaded to the network. For example, if a student wants to print a particular graph, the
334
W. M. Stroup & A. J. Petrosino
device itself doesn’t need to know how to print; it can just send the graph to a desktop computer that has the necessary print drivers. 6. Content Domain Interaction The nature of a device alone does not determine the nature of the learning experience (Papert, 1990). However, a technological resonance can exist between device and learning trajectory. Some devices more readily support certain kinds of explorations (the ‘right tool’ for the job). This idea helps us understand the relationships between device design and implications for domain-specific and interdisciplinary investigation. Horizontal. Due to its cross-context capabilities, horizontal devices more readily support interdisciplinary work by students. Statistical analyses and written discussions of results are likely to be part of what students might do on a single device. Additionally, a device that travels with a student outside the school building is more likely to support integrating in-school and out-of-school activities. Vertical. Although nothing about the design of vertical devices precludes the possibility of interdisciplinary explorations in a classroom, the device itself may not readily support this kind of cross-domain integration due to its specific focus. 7. Software Extensibility Horizontal devices allow the nature of the software experience for the user to be changed more easily. This user-driven flexibility is what we mean by ‘extensibility.’ There are two levels in analyzing extensibility: (a) the application level, and (b) the programming/ scripting level. Horizontal. Nearly all horizontal devices allow the user to add or delete application software. Moreover, horizontal devices tend to support programming or scripting at various level of programming expertise. This means that users can ‘author’ the nature of their experience from a simple level like moving icons around, to writing macros that function within or between applications, to—at the expert end of the spectrum—writing code in professional-level programming environments (e.g. Java, C, C ⫹ ⫹ , etc.). Vertical. Generally, it is hard to change the functionality of a vertical device both because access to an authoring layer is often limited and because the authoring itself is reserved for experts only. 8. Physicality of the Device Practical experience suggests that there is no such thing as a universal physical interface. The more inclusive the range of kinds of capabilities, the more compromises made in physical design.
Horizontal and Vertical Design
335
Horizontal. In order to function across contexts, horizontal devices must make tradeoffs related to, for example, keyboard availability and size, screen size, battery life vs. weight, and screen resolution. Vertical. A natural consequence of a vertical device’s functional focus is that the physicality of the device can be optimized for domain-related learning. A visual arts device can have a large and colorful touch screen, a language arts device a keyboard and headphones for text-to-speech learner feedback, and a science device a range of ports and integrated data collection tools. 9. Upgrade Path Among the hidden aspects of device ownership is the need for and ability to upgrade device functionality. Upgrades are distinct from software extensibility in that an upgrade is usually a required change in baseline functionality that often includes changing the nature of the interaction of the operating system with the hardware platform. Horizontal. As user needs change or functionalities are added to a horizontal device, the core capabilities of the device itself have to be upgraded. Showing pictures on a device designed to be a daily planner requires upgrades in the device. Ever-expanding user feature requirements and the just-in-case commitments implicit in horizontal design combine to drive a need for regular upgrades. These upgrades can vary from simple software updates to actually replacing the device. More frequent upgrades tend to result in higher costs for horizontal devices. Vertical. Because vertical functionality is closely tied to a specific core use, there is little to compel an upgrade to the device unless the core use changes. The usefulness of graphing functions in introductory algebra is likely to remain constant for the foreseeable future. Unless new kinds of functionality become compelling in the context of learning algebra, the need to upgrade is minimal. Teachers often will forego adding all the latest ‘bells and whistles’ in order to retain a familiar and robust core of meaningful functionality. 10. Learning to Use There are two levels in learning to use a device: learning to use the operating system and learning to use application software. Horizontal. For a horizontal device the operating system is both visible and designed to provide a consistent user experience. The user knows where to find the home screen or desktop, applications are launched in a consistent manner, and menus function in similar ways. Learning to use application software is often scaffolded by learning to use the operating system.
336
W. M. Stroup & A. J. Petrosino
Vertical. The operating system of a vertical device is often invisible. When turned on, the device jumps to the context-specific functionality. Learning to use application software is typically not scaffolded by working with a visible operating system but by the domain of use. The spell checker should appear near the thesaurus; statistics plots should be near function plots. 11. Evaluation The generally personal nature of using horizontal devices and the generally localized use of vertical devices have implications for how the devices are rated or assessed. Horizontal. Consistent with a horizontal device being literally and figuratively ‘close’ to the user, criteria for evaluating the device often will highlight personal satisfaction along with the frequency and quality of ongoing use. Helping to organize the student’s day might be viewed as being as valuable as helping the student to compute the standard deviation of a data-set. Entertainment capability is also often considered. Overall, the user ‘liking’ the device matters. Vertical. Two criteria tend to be highlighted in evaluating vertical devices. First, the use of the tool has to be engaging as a classroom activity. Second, there tends to be more of a focus on showing student improvement in learning subject matter—do quiz grades or high-stakes test scores improve? Curricular results tend to take priority over personal affinity. 12. Data Storage and Archiving In the school context, issues of storing or archiving student data are significant. Horizontal. Data archiving tends to be more important for a horizontal device because the user has entered a significant amount of unique or personal information into the device. This storage or backup of student data centers on iterating peer-to-peer exchange. This means students can back up their data to a desktop computer or to another device. Vertical. Students will often have relatively little data on a particular device and, because devices are shared, data tends to be purged frequently. Any significant data created can be collected or archived using a network. In some instances, it may be collected in a peer-to-peer way. Table I summarizes the design attributes for horizontal and vertical devices. To help clarify these attributes we will discuss two devices that have established a presence in schools. Two Handheld Examples Comparing the horizontal and vertical dimensions of design and use reveals tradeoffs of capability, total cost of ownership, and ease of implementation.
Horizontal and Vertical Design
337
TABLE I. Summary of horizontal and vertical design attributes. Design attribute 1. Designed for whom 2. Focus of functionality 3. Physical movement between contexts 4. Interdevice communication 5. Operating system complexity or ‘weight’ 6. Content domain interaction 7. Software extensibility Application level Programming/scripting level 8. Physicality of the device 9. Upgrade path 10. Learning to use OS level Application level 11. Evaluation 12. Data storage and archiving
Horizontal
Vertical
Student Just-in-Case Portable across physical/use contexts Peer-to-Peer/near neighbor topology Heavy
Teacher Just-Enough Fixed within use (e.g, math classes) Networked/flexible group topologies Light
Interdisciplinary by nature Easy Many and easy to do Available at various levels of expertise Compromises/Tradeoffs Often to ‘keep up,’ expensive
Domain focused Hard Fixed Experts only
Visible and consistent Scaffolded by OS and varied Personal satisfaction and use Iterating peer exchange (resident)
Invisible Scaffolded by the domain/context Contextually engaging and domain results Data collected using network (not resident)
Optimized Infrequent, relatively stable
We will illustrate these respective design attributes using two related technologies—the graphing calculator (vertical) and the personal data assistant (PDA) based on the Palm operating system (horizontal). We chose the graphing calculator and the Palm operating system (POS) device for two reasons. First, each device has an established or a growing installed base in schools. A range of handheld computing devices seems to be an important new component in school-based technology use. The second reason relates more directly to the focus of this article. While a graphing calculator and a POS device are visually similar—both are handheld and have a similar amount of screen real estate—their designs are very different. A POS device is more of a horizontal learning tool and the graphing calculator is more of a vertical learning tool. We wanted to choose devices that had surface similarities to underscore the sense in which design features include much more than just the physicality of the device. A similar point might have been made by discussing how a horizontal device can be limited to working in a vertical way—e.g. a desktop or laptop computer constrained to run only graphing software. This too would have provided a looks-alike-but-works-differently contrast. Our concern, however, was that we didn’t want the article to be seen as just another piece on all the different ways one might use, or limit, a laptop or desktop
338
W. M. Stroup & A. J. Petrosino
personal computer. Instead we want to explore design principles that we believe are ‘larger’ than any particular kind of device. Already we feel that too much of the conversation about school technology is collapsed to being about purchasing a very specific kind of device (e.g. what kind of laptop computer) or kind of suite of software (e.g. what kind of integrated ‘office’ application) and not enough about general design principles aimed at all students gaining sustained access to powerful, meaningful computing functionality. We want the reader to think through whether, for example, certain kinds of use patterns and/or tradeoffs of cost might argue for devices that embody—from the ground up—either vertical or horizontal design, and not whether one kind of device can be ‘turned into’ another.2 Vertical Example: the graphing calculator The graphing calculator inherits a vertical legacy from scientific calculators (Texas Instruments, 2002b). Professional scientists and engineers use scientific calculators as tools. The graphing calculator retains a focus on math and science, but is now optimized for the kinds of functionality needed at various levels of math and science learning. Indeed this specialization has made the graphing calculator not one device, but a number of different devices targeting specific areas and grade levels (e.g. more fraction-specific functionality for late elementary/early middle-school grades; more calculus-related functionality for the early university level). Relative to the 12 attributes discussed earlier, sets of graphing calculators are typically brought to the classroom by the math or science teacher. The focus of the functionality is on a predictable and uniform set of features that relate to learning mathematical or scientific concepts. Because of this tight domain focus, the devices themselves tend to stay with the teacher from class to class and are typically purchased by schools. Students may be assigned a particular device during class but it is expected that out over the course of a day the same device may be used by a number of different students. Communication was not initially an issue. But now that teachers are increasingly interested in installing mini-applications, passing out data-sets to analyze, or collecting a range of student-generated artifacts (such as graphs, particular functions, or programs), some manufacturers are developing networks to support forms of group-oriented interdevice communication. Artifacts and electronic gestures (e.g. pressing an up/down arrow key to ‘vote’ on a particular question) can be shared across all the devices. Some of these networks support various forms of role-playing and participation in real-time simulations (Wilensky & Stroup, 1999). Operating systems on vertical devices are relatively simple3. Only those capabilities required to implement the just-enough functionality of the device are active—system-level functionality like printing or data backup is offloaded to other devices. Data or print jobs can be uploaded to a computer or, in the near future, sent through a classroom network to a desktop computer for archiving or printing. The printer drivers themselves
Horizontal and Vertical Design
339
are not on the calculator. To support the domain-related core functionality associated with a graphing calculator, the operating system doesn’t need to recognize lots of different kinds of printers, disk drives, and input devices. This makes the operating system very ‘light’ in terms of complexity, and all but invisible to the user. In general, the operating systems on vertical devices are more robust than those of horizontal devices. Focused devices like graphing calculators don’t tend to ‘crash’ or ‘hang’ as often as horizontal devices such as standard laptop computers. In addition, the transparency of the OS makes the learning curve short and the related implementation and maintenance costs small. Although uploadable applications are becoming increasingly available for these devices, acquiring different core functionality is still closely associated with simply purchasing a different device. For example, users may buy device A to learn about fractions and device B for learning calculus. The physicality of a graphing calculator is integrated with the core functionality. As part of the optimization associated with vertical design, numeric keyboards are built into math devices, QWERTY keyboards are built into text-oriented devices, pen-based input is built into graphics-oriented devices, and a piano-like input is built into music devices. In school-based use, graphing calculators tend to be upgraded rarely because the core functionality associated with students learning algebra does not change rapidly. Learning how to use the devices is closely associated with domain-related competence. This means understanding something about calculus plays a major role in structuring a user’s ability to interact with a device having calculus-related functionality. Evaluation of a vertical device is tied to how well it does the task for which it was designed. Graphing calculators are expected to be useful for learning mathematical and scientific concepts and are evaluated in those terms. Consistent with this domain focus, the largest professional organization of mathematics educators in the USA, the National Council of Teachers of Mathematics (1998), specifically noted that graphing calculator use ‘has been shown to enhance cognitive gains in areas that include number sense, conceptual development, and visualization’. Similar domainspecific evaluative statements have been made by other content-specific professional organizations and in relation to reports such as the National Science Standards developed by the National Research Council (1996). Because vertical devices are typically shared, storing and archiving user data, especially personal data and functionality, is less central. A great deal of mathematics can be explored without the user ever having to retain individual data. Indeed, many teachers begin each class with students clearing the memory of the graphing calculators.
340
W. M. Stroup & A. J. Petrosino
Horizontal Example: the POS PDA While POS-based devices are used as examples of horizontal design, an interesting aspect of the history of POS-based systems is that they began as vertical devices and much of the early acceptance was closely related to this tight vertical focus. POS devices replaced the ubiquitous daily planners that professionals tend to lug around all day to keep track of work-related appointments, contact information, notes, and so on. The devices were referred to as personal data assistants (PDAs) and the just-enough functionality associated with replacing the daily planner was built into the device. Had the design of these devices retained this tight focus on a particular vertical market—working professionals using daily planners—the PDA might have remained, for our purposes, a prototypical vertical device. The devices retain much of this core functionality, but they have evolved into all-things-to-all-people, just-in-case devices. The size and complexity of the operating systems have grown considerably. Consistent with this movement away from the early vertical focus, the companies marketing these devices no longer refer to them as ‘personal data assistants.’ Relative to school-based use, POS devices are purchased primarily for design attributes that we have discussed as horizontal. The target users of the POS devices are students. Students move from class to class and/or from subject to subject during a typical school day. They also keep track of assignments and other personal information. Rather than carry a number of distinct devices through the day, it is reasonable to trade off optimal designs for a more inclusive design4. Consistent with the flexibility associated with horizontal design, a range of kinds of domain-related functionality can be added to a single POS device and then moved from class to class. The related need to optimize for portability during a school day requires that the POS device be physically small. As a natural extension of personal communication, POS devices emphasize peer-to-peer communication. For between-class or out-of-school interaction this communication may be ideal. Reports on POS-related projects point specifically to their usefulness in communicating homework assignments and notes among teachers and parents (Crawford & Vahey, 2002, p. 27). In the group setting of the classroom, however, peer-to-peer communication sets an upper limit on both the number and kinds of interactions supported. In addition, because all of the communication is initiated locally, with the student, efforts to orchestrate whole-group trajectories or forms of activity with groups larger than pairs are more challenging. Sharing information is also challenging because without network hardware and software functionality, students must cascade out the day’s activity much as they might pass out stacks of paper worksheets. In addition, any given pair of students can choose to pursue their own agenda independent of, or even at odds with, what the teacher may intend. POS devices can be a source of distraction. In a recent project, 41% of teachers reported ‘inappropriate uses’ of POS devices such as ‘inappropriate games,
Horizontal and Vertical Design
341
privacy issues, cheating, or distraction during class time’ (Crawford & Vahey, 2002, p. 30). Students being able to use a wide range of kinds of functionality, when and how they want, is a strength and a weakness of horizontal design. Moving well beyond the initial form of ‘synching’ with a desktop computer, the current generation of POS devices are able to interact with a extensive range of kinds of printers, keyboards, data probes, MP3 players, digital cameras, projection systems, and forms of removable storage. The wide range of hardware and software supported by the POS allows students to readily break down barriers between domains. Writing, for example, can be integrated into the activities of a mathematics classroom. Photography can be part of science classes. Teachers whose students have used POS devices have described ‘easier transitions across activities and contexts (data collection, analysis, reporting; in the field, in the classroom)’ (Crawford & Vahey, 2002, p. 19). New functionality can be added to a POS device, and it can be programmed in a range of ways. As is true in general with horizontal design, adding possibilities in terms of either new versions of software or new kinds of hardware interactivity tend to drive more frequent upgrades. This, in turn, can add significantly to the total cost of maintaining POS devices. Additionally, flexibility of support requires the operating system to be more visible—there has to be a place for newly installed possibilities to appear—and complex. The operating system, then, becomes a substantial presence in the user’s experience. For a device that may embody significant design compromises in order to attain a high level of portability, issues associated with interacting with the operating system can be noteworthy. For example, in one evaluation study (Crawford & Vahey, 2002, p. 24), 40% of teachers reported students having either ‘some’ or ‘a lot’ of difficulty using the ‘graffiti’ handwriting recognition feature of the POS. A device with a built-in keyboard would not need this handwriting recognition feature. On the positive side, a visible OS usually requires that a consistent set of interface standards be implemented (e.g. application icons appearing on the home screen), which helps to support the user’s interaction with a wide range of software applications. Increased flexibility and complexity in an OS have consequences in terms of an increased initial cost of learning to use the devices and these costs need to be factored into budgets for school-based implementation. Reflecting a sense that using a POS device is to be a personal, individual kind of experience, the evaluation of POS devices tends to focus less on domain-related learning (e.g. results with ‘number sense’ as was mentioned earlier for the graphing calculator) and more on personal activity, enjoyment, and integration with out-of-classroom life. As a student-focused device, self-reported student use patterns become particularly salient. These self-reports clearly show that the value of the device is identified with significant forms of out-of-domain functionality. The top
342
W. M. Stroup & A. J. Petrosino
two ‘favorite activities’ of students using POS devices were playing games (64%) and ‘organizing, planning, and scheduling’ (20%) (Crawford & Vahey, 2002, p. 35). This view is consistent with teachers’ assessment of student use. Second only to promoting the ‘integration of computing’ into a wide range of instructional activity, teachers listed advancing ‘autonomous learning’ and supporting ‘students’ organization’ as the second and third most significant benefits of using handheld POS devices (Crawford & Vahey, 2002, p. 11). Personal affinity and organizational utility seem to matter much more in the evaluation of horizontal devices than they do for vertical devices. With so much more information and capability related to personal use resident on the device, issues related to backing up and ‘synching’ data are more significant with horizontal than vertical devices. With POS devices teachers are encouraged to ‘assign desktop computers to students’ or assign specific computers ‘in the classroom, and/or in the computer lab for HotSynching [backing up the POS data to a desktop].’ Then the teachers need to install the POS-related software and ‘set up accounts on desktop computers under the Palm ID numbers or student names for HotSynching’ (Crawford & Vahey, 2002, p. 38). In many ways a POS device is designed to be an extension of the capabilities of an existing desktop computer and, relative to school-based use, this has significant implications in terms of students needing regular access to desktop computers for data storage and archiving. The need for some number of desktop computers has to be factored into the costs of implementing the use of POS devices in schools. Learning from the Present, Looking to the Future Horizontal and vertical devices already are a presence in schools. Whether it is the significant number of AlphaSmart keyboards or graphing calculators as exemplars of vertical devices, or the full gamut of personal computers as exemplars of horizontal devices, these devices are there because they speak to some educational need. The presence of both kinds of devices in schools suggests there is a ‘story’ to be told about how horizontal and vertical devices can be part of an integrated approach to educational computing. What we have attempted to do with this paper is to take a first step toward weaving a whole cloth out of the strands of possibilities represented by the presence of these devices. The 12 design attributes are offered as a way of organizing and situating issues of hardware and software design relative to patterns of implementation, support, and ongoing learning activity in schools. Rather than purchasing software and hardware designed primarily for business settings and then representing these business solutions as a ‘standards-based’ approach to school-based computing (International Society for Technology Education [ISTE], 2003a; Microsoft Corporation, 2003), the integration of horizontal and vertical design stands to support an
Horizontal and Vertical Design
343
approach to technology implementation that would grow more organically and sympathetically out of reflective school-based practice (e.g. Bell, 2001; Petrosino & Dickinson, 2003). For this to happen, the voices of educators with competency and interest in domain-based learning—visual arts, music, dance, literature, reading, history, science and mathematics—need to be heard over the din of those who would narrowly define technological ‘literacy’ in terms of surface familiarity with a particular integrated suite of office applications and/or a set of low-level skills like inserting a CD-ROM or being able to ‘use keyboards and other common input and output devices’ (ISTE, 2003b). Again, the overarching goal in these efforts has to be supporting the best possible fit between computing design and the kinds of functionality that matter for teaching and learning. Keeping costs under control and ensuring equitable access to important forms of functionality by all students should help push this integration forward. Successful models of integrating horizontal and vertical design do exist outside education and we can learn from these efforts. A range of different kinds of integration, customized to particular teaching and learning settings, are likely to typify the future of this development in education. By coming together as vertical markets, it seems likely that some number of commercial partners would be willing to work with school-focused educators in designing ‘ground up’ education solutions. This would help significantly in making ubiquitous computing a reality in our schools. While this paper has focused principally on clarifying and contrasting the affordances of horizontal and vertical design, future efforts need to speak more directly to what this integration might look like in schools and how we might begin to get there. Even with this work ahead of us, we still believe the analytical framework presented in this article can provide near-term assistance to educators charged with thinking about what kinds of devices might best advance their educational goals and vision. Overall, when it comes to school-based computing, more may not always be better and optimization may not always be optimal. Instead, the integration of the best of horizontal and vertical design may be a more promising and equitable way forward. Acknowledgements We would like to acknowledge the support of both the Learning Technology Center (Paul Resta, Director) and the Center for Science and Mathematics Education (James Barufaldi, Director) at The University of Texas at Austin. This work was also supported by the following: National Science Foundation grants (14656-S1 Amendment 4) entitled Challenges to Projects: VaNTH K–12 Partners in Education, (09093) entitled CAREER: Learning Entropy and Energy Project, (126227) entitled Integrated Simulation and Modeling Environment, and a grant from the Department of
344
W. M. Stroup & A. J. Petrosino
Education (DOE P342A000111) entitled Inventing New Strategies for Integrating Technology in Teacher Education (Project INSITE). The views expressed in this paper are those of the authors and do not necessarily reflect the views of the funding agencies. A final special thanks goes to Jennifer Cook for her editorial assistance. Correspondence: Dr Walter M. Stroup, The University of Texas at Austin, Science & Math Education, 1 University Station D5705, Austin, TX 78712-0382, USA; e-mail:
[email protected] NOTES 1. For example, for Algebra I and II the Texas Education Agency’s (2003) discussion of ‘Curriculum Connections for Mathematics’ states: Tools for algebraic thinking. Techniques for working with functions and equations are essential in understanding underlying relationships. Students use a variety of representations (concrete, numerical, algorithmic, graphical), tools, and technology, including, but not limited to, powerful and accessible hand-held calculators and computers with graphing capabilities and model mathematical situations to solve meaningful problems. And this is supported by a policy statement from the Commissioner of Education to Superintendents requiring access to this kind of capability (TEA, 2002). 2. This observation is intended to apply as much to turning a horizontal device like a laptop into a vertical device, as it is to trying to turn a vertical device like a graphing calculator into a more horizontal device by attaching an alphanumeric keyboard to the data port (Texas Instruments, 2002a). 3. Or the OS is a constrained version of a more complex operating system. 4. These compromises often do have noticeable drawbacks. For example, not having a keyboard for text input has been shown to dramatically reduce the usefulness of POS devices for writing (Crawford & Vahey, 2002, p. 24). REFERENCES BELL, L. (Ed.) (2001) Preparing tomorrow’s teachers to use technology: perspectives of the leaders of twelve national education associations, Contemporary Issues in Technology and Teacher Education [Online Serial], 1(4), available at http://www.citejournal.org/vol1/ iss4/currentissues/general/article1.htm. CRAWFORD, V. & VAHEY, P. (2002) Palm Education Pioneers Program, March 2002 evaluation report (Menlo Park, CA, SRI International), also online at http:// www.palmgrants.sri.com. INTERNATIONAL SOCIETY FOR TECHNOLOGY EDUCATION (2003a) National Educational Technology Standards (Eugene, OR, ISTE), also online at http://cnets.iste.org/index.shtml. ISTE (2003b) Profiles for Technology Literate Students (Eugene, OR, ISTE), also online at http://cnets.iste.org/students/s_profile-35.html. MICROSOFT CORPORATION (2003) Microsoft supports leading education organizations to help schools assess 21st century skills (Seattle, WA, Microsoft Corp. Press Pass), also online at http://www.microsoft.com/presspass/press/2003/jun03/06-30p21kentpr.asp. NATIONAL COUNCIL OF TEACHERS OF MATHEMATICS (1998) Calculators and the Education of Youth (Arlington, VA, National Council of Teachers of Mathematics), also online at http://www.nctm.org/about/position_statements/position_statement_01.htm. NATIONAL RESEARCH COUNCIL (1996) National Science Education Standards (Washington, DC, National Academy Press), also online at http://www.nap.edu/books/0309053269/ html/index.html.
Horizontal and Vertical Design
345
PAPERT, S. (1990) A Critique of Technocentrism in Thinking about the School of the Future. MIT Media Lab epistemology and learning memo no. 2 (Cambridge, MA, MIT Media Lab). PETROSINO, A.J. & DICKINSON, G. (2003) Integrating technology with meaningful content and faculty research: the UTeach Natural Sciences Program, Contemporary Issues in Technology and Teacher Education [Online Serial], 3(1), available at http:// www.citejournal.org/vol3/iss1/general/article7.cfm. STEPENSON, W. (2002) Tracking a wireless trailblazer, Wireless Business and Technology, 2, pp. 30–34. TECHTARGET (2002) Vertical Market, online at http://searchwebservices.techtarget.com/ sDefinition/0,,sid26_gci213286,00.htm. TEXAS EDUCATION AGENCY (2002) Calculators and the Texas Assessment of Knowledge and Skills (TAKS) Mathematics and Science Assessments at Grades 9, 10, and 11, online at http://www.tea.state.tx.us/taa/studass022502.htm. TEXAS EDUCATION AGENCY (2003) Curriculum Connections for Mathematics, online at http://www.tea.state.tx.us/technology/techapp/instruct/hscc/cc111.htm. TEXAS INSTRUMENTS (2002a) TI Keyboard, online at http://education.ti.com/us/global/news/ release/news34.html. TEXAS INSTRUMENTS (2002b) Electronic Hand-held Calculator, online at http://www.ti.com/ corp/docs/company/history/calc.shtml. WILENSKY, U. & STROUP, W.M. (1999) Learning through participatory simulations: networkbased design for systems learning in classrooms, in: C.M. HOADLEY & J. ROSCHELLE (Eds) Proceedings of the Computer Support for Collaborative Learning (CSCL) 1999 Conference, pp. 667–676 (Palo Alto, CA, Stanford University or Mahwah, NJ, Lawrence Erlbaum Associates), also online at http://www.ccl.sesp.northwestern.edu/papers/partsims/cscl/ or http://kn.cilt.org/cscl99/A80/A80.HTM. WILENSKY, U. & STROUP, W.M. (2000) Networked gridlock: students enacting complex dynamic phenomena with the HubNet Architecture, in: B. FISHMAN & S. O’CONNOR-DIVELBISS (Eds) Fourth International Conference of the Learning Sciences, pp. 282–289, (Mahwah, NJ, Lawrence Erlbaum Associates), also online at http://www.umich.edu/ ⬃ icls/proceedings/pdf/Wilensky.pdf.