Topic 8 – What to Consider When Scoping GIS
Beyond Mapping book
Both Dreams and Nightmares Are Born of Frustration — discusses the limitations of traditional cost/benefit analysis in evaluating the adoption of a radically new technology like GIS
GIS Is Never Having to Say You’re Sorry — discusses the human and organizational considerations in adopting GIS technology
A Tailored Plan and Curriculum Cure GIS Training Woes — describes and discusses the importance of effective education and training needed for successful GIS adoption
Note: The processing and figures discussed in this topic were derived using MapCalcTM software. See www.innovativegis.com to download a free MapCalc Learner version with tutorial materials for classroom and self-learning map analysis concepts and procedures.
Both Dreams and Nightmares Are
Born of Frustration
(GIS World, May 1992)
The dream is that GIS can do anything... the reality is that it isn't easy. With increasing fervor, technologists and users alike define and redefine the "unlimited" potential of GIS technology. These dreams are, at least in part, an expression of our hopes, as well as our science. When considering if GIS is for you, often you're biggest challenge is to carefully separate what you hear into two distinct piles-- the quixotic dream and the pragmatic reality.
Your first step in this process is establishing "where you are coming from." GIS means different things to different people. At least four distinct perspectives flavor both our expectations and our realities-- economic, organizational, visionary and emotional. The economic perspective is usually based on labor and time-savings considerations. Standard cost/benefit analysis is particularly appropriate in distilling the dreams from reality. A careful audit of your organization's current mapping and spatial data handling procedures establishes a reference to estimate the savings in moving "from pen to plotter and from file drawer to keyboard." If the savings are greater than the expenditures, you are economically irrational (foolish) if you don't implement GIS immediately.
There, that's easy. There is nothing to it. Just call in the accountants and they will identify the numbers to plug into the Cost/Benefit equation. The reality is that even a strictly economic perspective is not that easy. The comfortable feeling of quantifying the evaluation process is quickly lost to the pliable nature of the "yardsticks" used to measure the costs and benefits.
The time-span used in the analysis is critical. If it is too short, the stream of benefits is artificially truncated. The high front-end costs, combined with the confusion and frustration of implementing a new system, will far outweigh the benefits. It's like a bare-knuckle battle between Sylvester Stallone and a tiger cub. If it is delayed a few years, the outcome will likely be different. If you had used a two-week cost recovery period for word processing, would you have ever dropped your pencil?
So what time period should be used? That's a judgment call-- your judgment call. Like lying with statistics, you can choose the time period that insures the answer you want. In general, a longterm position favors the adoption of GIS.
Just as important (and "mushy") is how you identify and quantify the variables of the cost/benefit equation. Four cost considerations quickly surface-- hardware/software, data base development/administration, training and application models. The hardware figures are the easiest to quantify through a litany of parameters including MegaHertz, GigaBytes, RAM, SIMMs, MIPS (DIPs, DRIPS and SLIPS). The software specifications are a bit more difficult, yet factors, such as, point-in-polygon, buffering, coordinate accuracy, and transfer formats can be used.
Although relatively easy to quantify, these figures are fleeting and set you up for a bad case of "buyer's remorse." About the time you finally push through your procurement and take first delivery, your system is out of date. It's like that pocket calculator. Within a couple of months, the same expenditure gets you five more keys at half the price. The difficulty in nailing down the hardware/software cost component isn't in the definitions; it is keeping your footing in the quicksand of technology. Like shooting ducks, you had better have a good lead on your target. For large, bureaucratic organizations, it may be prudent to just set a budgetary figure for the "best available technology" and postpone the specifications to the moment of purchase. That may seem preposterous, but it may be more realistic.
Data base development, maintenance and management are not only larger expenses than hardware and software, but it is even more tricky and slippery to estimate. Rarely does a simple inventory of your current map and file cabinets multiplied times an estimate of encoding costs produce an acceptable cost figure. The differences between the digital and paper map make it too tricky for such a mechanical approach. It's prudent to launch an Information Needs Assessment (INA) to determine data base contents, structure, policy and costs (a later issue will focus on this process).
Even if you do get a good handle on the data base, you must develop, you're not out of the woods yet. How you obtain these data is slippery turf. Manual encoding, scanning or purchasing are your basic options. Not so long ago, in-house, manual encoding was your only option. More recently the scales have been tipping toward scanning and purchasing, as a room full of digitizer folks is a major cost and distraction from normal business activities. Also, many of the maps you might encode have time-bombs ticking within them. For example, if you encode (in-house or contract) a soils map, it will become invalid once the Soil Conservation Service's "authoritative" version is released. Its back to shooting ducks, you had better get your data requirements in line and lead them, or you will just be pumping pellets into the air.
The costs of training your people to use GIS can easily outstrip the combined costs of hardware/software and data base development. Early successes in using GIS were often more a function of the zealots using it than the technology itself. Like The Little Train That Could, GIS could do a lot. The pitfalls that accompany any new technology are overcome by innovative "work-a-rounds" of committed users. In a wholesale adoption, however, the user community is expanded to "I don't think I can" and "I am damned if I will" outlooks.
One reaction to this reality is to form a GIS division. On the surface it is a plausible alternative. All you have to do is train a small cadre of experts. There, that's both efficient and effective. But it rarely works for two reasons.
First, the GIS product produced is just that-- a GIS product, not the direct expression of the final user. In the late 1970's I had an opportunity to observe a large timber company's centralized implementation of GIS. Most of the field personnel merely dismissed the "computer jerk's" forest management maps handed to them through the glass windows of the computer center. "What do they know about the @#*^! forest anyway?", was the rallying cry. If the maps were used at all, they became the center of attention for the short period it took to locate that "one" forest stand in the middle of a lake. This meant wasted effort on both the GIS and user sides, a situation that could be helped with sufficient investment in training.
If costs of training are identified at all, they are usually associated with vocational instruction on system operations. But GIS is a challenging new way of thinking, as well as a new sequence of buttons to push. The mechanics of translating what you currently do with maps into a GIS is straight forward. In fact, colorful icons and mouse clicking can make it almost fun. However, most of the potential GIS applications within and organization are yet to be discovered.
The development of application models is the other reason for failure of a centralized approach. How the new technology leads to new ways of doing things is the least understood cost (and benefit) of GIS technology. It's like your son or daughter dumping the tin of tinker toys on the floor. The mechanics of how the pieces fit together is fairly simple. What ought to be built with the individual pieces is the difficult part. The tinker toy makers (vis. GIS experts) can supply some ideas, but they certainly do not cover all of the possibilities. Vocational training develops an awareness of the GIS "owner's manual" description of the pieces and parts, but beware "some assembly is required" before you are up and running.
The creative assembly is entirely up to your people. If you ignore or skimp on training and application model development, you will incur opportunity costs at the minimum. More likely, you will generate a backlash of confusion and apprehension that quickly outweighs the set benefits you identify. A couple of strategically placed anti-GIS terrorists will wreak havoc with the even your best laid plans.
A strict economic perspective is the first step in scoping GIS technology. Identification (and ultimately quantification) of the costs and benefits sets the stage. However, organizational, visionary and emotional perspectives are needed to complete the picture-- whether a dream or a nightmare. That gives us something to discuss in the next issue.
As with all Beyond Mapping articles, allow me to apologize in advance for the "poetic license" invoked in this terse treatment of a complex subject. Readers interested in more information on "scoping" GIS should begin with a paper by Robert J. Lima of the Boshe Institute in the Conference Proceedings of URISA '91, Vol 2:8-13, Urban and Regional Information Systems Association (URISA), 900 Second Street, NE, Suite 304, Washington, DC, USA 20002 (phone, 202-289-1685).
GIS Is Never Having to Say You’re
(GIS World, June 1992)
The previous section began the discussion of the broad considerations in implementing a GIS with economic cost/benefit analysis. This is where most organizations begin their first step of what seems to be a thousand mile journey to GIS implementation. At first glance the seductive appearance of a rigorous, quantitative analysis is quickly lost to the pliable nature of the "yardsticks" used to measure the costs and benefits. At best, a cost/benefit analysis sets the stage for further investigation into the full impact of implementing a GIS. Even the most favorable C/B ratio should be further scrutinized in terms of the organizational and human impacts of GIS. Whether real or imagined, the perceived threats of GIS technology form the actual mine field that you must traverse.
The organizational structure (both formal and informal) is an important concern, as it is the direct expression of the "corporate character"-- the most basic element of any organization. If extensive individual latitude and autonomy best describes the current character, GIS will likely have a rocky-road to implementation. Within this environment, data often are viewed as the medium of exchange for power brokers at all levels. Simply stated, "if you must pass through me to get to important data in my map cabinets and file drawers, then I am as important as the data I keep." However, if GIS places my data in some central repository accessible to all by a single mouse click, my corporate worth has been severely devalued. The result, as viewed by some, is an electronic end-run around the current data gatekeepers and a direct assault on the existing organizational structure. It may be a benefit to the organization to have a corporate data base, but to many it represents a personal loss of influence. If your implementation plan ignores this reality, you'll be sorry.
Another concern which may run amuck with the corporate character is the imposition of data standards. In many organizations, mapping standards are either non-existent, or merely address geographic registration and data exchange formats. But this is just the tip of the chilling iceberg of standards. The ability to export a map from one GIS package and swallow it in another is basic and rapidly becoming a non-issue. Likewise, the ability to convert projections, rectify and register maps is commonplace (although not necessarily easy). The confusion and frustration isn't in the locational (where) set of standards, but in the informational (what) set.
A corporate data base consists of three levels of maps based on their degree of abstraction-- base, derived and interpreted. Base maps are usually physical data we collect, such as roads, water and ownership boundaries. They have minimal abstraction, and as much as possible, represent a scale model with all of the detail of a flatten model train set. Definitions and procedures for mapping most these data are in place— but not all.
Consider a map of cover type. Is Forest/Non-Forest a sufficient standard? Or should the Forest class be further divided into Conifer and Deciduous? And the Conifer, in turn, subdivided into Pine, Fir and Hemlock? What about age and stocking classes? Should you identify a lone pine tree in the middle of a meadow as a Conifer Stand? Two, three, four, five trees-- what does it take to form a forest stand? Ask a forester, ecologist and recreation scientist and you'll get at least three different responses. Or maybe four or five different responses depending on how they decipher different applications. You'll be sorry if you don't tackle these questions before you implement GIS.
For example, a wildfire had the audacity to burn across the boundary of two National Forests. Maps of cover type were encoded for both Forests, but they couldn't be edge-matched. One Forest had six classes of age and stocking for Douglas Fir, the other had eight. The GIS was able to account for locational adjustments during encoding, but not the differences in informational content. A common classification standard for cover type had to be established and encoded. The struggle for whose classification scheme was the best eclipsed the mundane tasks of reconstruction and encoding a compatible cover type map. The challenges to human and organizational interests run much deeper than those encountered at the digitizing tablet.
Vested interests in the definitions of map categories go beyond base data. Derived maps, such as slope, visual exposure and proximity to roads, are physical things. However, the data are too difficult to collect, so we use the computer to calculate them. Even something as simple as slope calculation has several algorithms, each with its pros and cons. For something as complex as visual exposure, there is a quagmire of assumptions, approaches and procedures. Which will you entrench in your system? Rest assured that the choice won't be by consensus, nor the dissenting voices reserved.
Even more volatile are the assumptions embedded in interpreted maps. These data are the most abstract, as they are conceptual renderings of expert opinion. Taunts of "my elk habitat model is better than yours" reverberate through the halls whenever two wildlife ecologists are cornered in the same room. It is naive to assume that elk models will edge-match across two forests, much less an entire region. And certainly not match across the paradigm chasm of two experts.
So whose derived and interpreted maps capture the standards in the corporate data base? The question of standards runs a lot deeper than just geographic registration and encoding effort. It involves organizational and individual perceptions, reputations and vested interests. You'll be sorry if your implementation plan ignores these elements. Sure, they will get sorted out later— after you and the GIS system fail.
A GIS implementation strategy has to go beyond simply scoping system design to nurturing a receptive environment. This passes the baton from the system engineers and GIS specialists to the sociologists and human relation professionals. As continually reminded in this column (possibly to the point of being shrill) GIS is not just automating what you do, but changing how you do things. Sensitivity to the full impact of these changes, human as well as procedural, is paramount.
Table 1. Institutional and Individual Threats and responses.
THE THREAT OF GIS
- Organizational -- "There's only one problem having all this sophisticated equipment... we don't have anyone sophisticated enough to use it" (Beetle Bailey)
- Status Quo -- "If it ain't broke, don't fix it"
- Overload – “Torture numbers and they will tell you anything"
- Stifling -- "Imagination is more important than information" (Einstein)
- Awareness -- "Technobabble... that seemingly endless drone masking what would otherwise be a clear understanding of a new technology's concepts and use"
- Intimidation -- "It's like new math, I'm just too old"
- Power -- "Experience used to be worth something, now you just dazzle them with color"
- Dependence -- "Middle management is an endangered species... they are information brokers hooked to the computer jerk down the hall"
COPING WITH GIS THREATS
CHOICES -- "Dammed if you do, dammed if you don't"
Fight It — Ignore It — Face It
- Grassroots Support -- "They don't know what they are doing"
- Understanding -- "You know, this GIS stuff isn't so bad after all"
- Proof-of-Concept -- "Oh, now I see, I could use something like that"
- Commitment -- "What do you mean, learn it in my spare time"
- Tough Love -- "Like it or not, unless you have retired, your job has evolved for the better"
- Lingo -- "Sticks and stones my break my bones, but arcane terminology will never hurt me"
- Continuing Education -- "The era of the four-year smart pill is over"
- Leadership -- "If the boss can handle this stuff, then I guess anybody can"
- Long Haul -- "Like other new technologies, GIS is something that is best understood backwards... but must be learned forwards"
Table 1 outlines some of the threats and responses which need to be addressed. The outline is designed to stimulate discussion in a workshop setting, but hopefully they will trip some thoughts in your mind. As you look over the outline, try some "free associations" with the points. Conjure up some of your own threats and possible coping responses. It is a lot of fun at the workshops and sparks a broader perspective on GIS implementation. At minimum, the exercise should encourage you to go beyond a focus on the mechanics of GIS technology to its institutional and human implications... if you don't, you'll be sorry.
As with all Beyond Mapping articles, allow me to apologize in advance for the "poetic license" invoked in this terse treatment of a complex subject. Readers interested in more information on the human impacts of GIS should consult "The Threat of GIS," by Dennison Parker and Evan Valchos in the 1989 GIS/LIS conference proceedings. A reprint is available from GIS World.
A Tailored Plan and Curriculum
Cure GIS Training Woes
(GIS World, August 1992)
…why waste time learning when ignorance is instantaneous (Calvin & Hobbs)
The last couple of sections have encouraged you to go beyond a focus on the mechanics of GIS to its institutional and human implications. Like most new technologies, the technical aspects of GIS are the easiest part. It's the non-technical implications and impacts that ultimately determine success or failure. These have little to do with bits, bytes, buffers and even bucks (US$'s). It's a corporate "warm-fuzzy" feeling about GIS.
So how does one insure such acceptance? Let's start with the easier question... how do you insure failure? That's simple. Just deliver crates of computers and masses of shrink-wrapped manuals. Within hours, anti-GIS terrorists will have torched all the managerial offices and be moving toward the Board Room. Defensive positions, such as, "we're doing it for you" and "try it, you'll like it" will crumble like papier-mâché bunkers. General rebellion and anarchy will sap any remaining vestiges of the "corporate good idea." An emotional and intellectual wasteland will lie at the feet of the sterile grey boxes and brightly colored CRTs.
Like the "Christmas Future" vision in Dickens’s novel, A Christmas Carol, there is an alternative. It's a commitment to education. Without it, ignorance prevails, confusion is rampant and negative rumors abound. However, at least as much effort is involved in planning and implementing an effective educational program, as in the scoping of hardware and software requirements. Matching instructional approach with the various skills required is similar to matching the appropriate platform and functionality to the information needs of an organization.
But why not leave training to the universities? It's their job isn't it? Two major points come to mind. First, most universities have tightly defined programs for traditional degree-seeking students. It is hard enough to get a GIS course into a recognized major, let alone canonized as a new program of study. However, even if the academic tanker was turned overnight, you can't wait until the GIS matriculates rise to top decision-making positions. Technology moves faster than scholarly debate or openings in your organization. You're left with the non-traditional student-- your current employees with all of the warts and scars left from their last brush with the "four-year" smart pill. A snarly bunch, but they're the key to the success or failure of GIS in your organization.
The first consideration in GIS training is recognition that there is both a formal and informal process. Attention to the informal process must be made throughout the implementation of GIS by nurturing "in-house zealots." In the early stages, these individuals not only provide over-the-shoulder instruction, but legitimize the technology. They enthusiastically demonstrate that "one-of-us" can use the damn thing. Your challenge is to create situations that quickly identify these individuals. Undertaking a couple pilot projects, early in the GIS scoping process, is a good strategy. However, be certain that the volunteers can focus their full attention on the project, receive ample support and are given complete freedom in their approach. A good rule of thumb is "If you can't afford this involvement, the chances are you can't afford the technology." Another good rule is "One in-house GIS zealot is worth a dozen out-house specialists."
The nucleus of zealots provides leadership and credibility, but is inefficient in conveying basic procedures and concepts to the masses. This is where formal training comes in. Three instructional approaches are involved-- awareness, vocational and educational. General awareness instruction provides a non-technical overview of GIS's capabilities and limitations. It counters GIS ignorance, but stops far short of a working knowledge of the field.
It is imperative that all personnel participate in this training phase-- from clerical to technical to professional to managerial. Ideally, the presentation is made in mixed audiences and discussion is encouraged. It is not so much a tutorage, as it is a forum. Sure some basic concepts and terminology slip in, but mostly the gathering introduces GIS and sets the stage for its implementation. The mixed audience provides recognition of in-house zealots and reinforces management's commitment. The absence of GIS experts and/or top management is dysfunctional as it makes the meeting merely perfunctory.
Vocational instruction develops operating skills in the procedures and practices of a specific system. It is designed to show you how to use the system. Educational instruction, on the other hand, develops spatial reasoning skills through understanding of basic concepts and theory. It is designed to show you why you might want to use the system. Think of it this way-- awareness instruction is similar to a newspaper article; vocational instruction is similar to a manual; and educational instruction is similar to a textbook. Each approach or item is directed to a different audience, presents different material and produces different "products." An inappropriate match either bores or overwhelms the audience, with either case rendering your training a waste of time and money.
Enough of the academic hyperbole …what's the reality? The reality is that most implementation plans focus on vocational training alone (if at all). Remember when your organization implemented word processing? Did you receive instruction, or just left to your own devices? How did you do? How would you do it differently? GIS is like word processing-- only different. Compared to manual techniques both are faster, easier and can quickly generate more piles of paper than wall space and surface area of furniture combined. But is it better? Probably not, until you use a new technology in new ways. That's the big difference-- GIS presents an entirely new way of doing things. In addition to the mechanics of how to work the thing, new analytic concepts and spatial reasoning skills must be developed.
The balance between vocational and educational approaches depends on which skills are addressed. Four distinct GIS skill levels can be identified—
- Database Development Technician-- encodes and maintains the spatial and attribute databases.
- Data Center Manager-- coordinates data integration, information flow and maintains the system.
- Application Specialist-- facilitates the development of application-specific models.
- General User-- uses GIS in both routine activities and decision-making.
Note that the first three roles are in support of the User and his/her ability to do their job. That's not a moot point. Without frequent reality checks, any new technology can take on a life of its own. Also, the listing is ordered in terms of the GIS technical knowledge required (from most to least). In general, more technical/vocational training is needed at the top of the list. The balance shifts to more conceptual/educational training at the bottom. So what? At a minimum, you're put on alert that one "comprehensive" short course may not be sufficient for all GIS skill levels.
So what's the appropriate balance? As usual, "it depends"... mostly on your information needs (i.e., the User). If routine mapping and data base management demands dominate system use, then training is best focused on the top of the list. Training in the proper care and feeding of the database is paramount. The Specialist and User interests will revolve around GUI interfaces that make access and retrieval a "piece-of-cake." The bulk of Automated Mapping and Facilities Management (AM-FM) applications fall into this group.
However, if your GIS needs lean more toward Decision Support Systems (DSS), then your training requirements are significantly altered and move toward a more conceptual/educational focus. In this environment, GIS is less a tool to extend the hand, than and a medium to extend the mind. Creative uses, such as spatial modeling, require both a proficiency in system operation and a thorough understanding of system functionality. Effective dialogue between the application Specialist and the general User is rooted in a common understanding of GIS capabilities and limitations in expressing spatial relationships. Even memorization of the "User's Manual" won't cover these bases. An organizational commitment to education will.
Ignorance is instantaneous and a lot cheaper... but is it really? This discussion should have dispelled the notion of GIS training through "immaculate conception." If your personnel struggled with word processing, choked on spreadsheets and gagged on data base management, expect to be in intensive care with a massive head wound with GIS. Preventive medicine, in the form a tailored training plan and curriculum is advised. At least as much thought (and ultimately, direct investment) should go into training as in the scoping of the hardware/software and database requirements.
Also, this and past discussions should have brought you (and me) to the realization that the mechanics of GIS are a lot easier to wrestle than the amorphous hulk of its institutional and human implications. Next issue we will get back to the basics-- analytics.
As with all Beyond Mapping articles, allow me to apologize in advance for the "poetic license" invoked in this terse treatment of a complex subject. Readers interested in more information on technology's broader impacts should consult "Communication of Innovation," by Rogers and Schoemaker, 1971, Free Press, New York. Information on a National Curriculum for GIS may be obtained from the National Center for Geographic Information and Analysis (NCGIA), University of California, Santa Barbara, California, phone (805) 893-8224.