UCGIS Virtual Seminar - Fall 1998 [Back][Refresh][Options][Search] Course Wrap-Up [Edit*][Delete*] [Image] Wrapping things up Dawn Wright 11/29/98 [Image] Term paper: The Ethics of GIS Ronald William 12/15/98 Ward [Image] Term Paper Byong-Woon Jun 12/16/98 [Image] Term Paper: Simulating the "Real World" Bill Moseley 12/17/98 [Image] Term Paper: Identifying Problem Domains Erik Shepard 12/18/98 in Object-Oriented... [Image] Post new message in this thread ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: November 29, 1998 11:10 PM Author: Dawn Wright (dawn@dusk.geo.orst.edu) Subject: Wrapping things up Hello everyone, It's time now to bring things to a closure. I very much appreciate your participation this term! It will be useful for us to have some kind of evaluation of what went on (or didn't go on) this term. Please do TWO things for me if you would: (1) Take a look at the preliminary assessment of the course based on our panel discussion at GIS/LIS in Fort Worth. Click HERE. (2) Fill out a quick evaluation form. Feel free to post any additional comments in this thread. Thanks and good luck to those of you taking final exams! Dawn (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=3049) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: December 15, 1998 06:31 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: Term paper: The Ethics of GIS Term Paper - Ronald W. Ward "In determining whether an action will be considered unethical, unfair or unjust by a group in society or by society as a whole, we normally try to anticipate whether a consensus or large majority opinion of the group would hold the action to be unethical." Harlan J. Onsrud Introduction: The ethics of GIS include many complicated issues. A useful way in which to organize these many issues is to look at the ontological process that goes into a GIS, and to try and interject steps within the process where ethical issues might bear some consideration. According to Curry (1995) the following are typical steps in the development of a GIS: 1) collecting data and entering data into a system; 2) analyzing data; 3) representation and visualization of data; 4) interpretation of the output by the user. The earliest GIS were designed as instruments of policy making (Curry, 1995) and this remains their primary function today. Given this basic utility of GIS, I add 5) formulation of policy to this list of four basic developmental steps. At each step along the developmental path, accuracy can be an issue. Data collection that is inaccurate will cause subsequent inaccuracies at each step along the developmental pathway. In cases where inaccuracy stems from unconscious mistakes it may be difficult to argue that unethical conduct has occurred - after all, we all make mistakes. However, where the limitations of a data collection method are known, and subsequent steps in the development of GIS from said data fail to consider sampling limitations, ethical considerations are valid. Therefore, ethical issues arising from the use of data with known limitations are a necessary aspect of exploring the ethics of GIS in general. Beyond issues of inaccuracy, what are some of the ethical considerations arising at each step within the development of a GIS? Ethical considerations involving data collection and system entry might include the methods of data collection, the purposes behind collecting data pertaining to a certain demographic group or environmental issue, and the reasons behind choosing one type of commercially available GIS for data entry over another (Obermeyer, 1995). Regarding data analysis, the purposes for undertaking an analysis is also a logical choice for ethical consideration, and into this consideration can be added the influence of bias in choice of analytical methods. Representation and visualization are aspects of GIS development which are perhaps more subject to issues of accuracy than other developmental steps, and into this ethical consideration should be added the use of GIS output as propaganda for achieving political or corporate ends. Finally, the ways in which GIS are interpreted logically lead to policy formulation, and ethical issues arising from this final stage in the process could include access to information produced by GIS and marginalization of groups without adequate access to GIS information (Epstein, 1995; Sheppard, 1995). This list of considerations is certainly not inclusive of all aspects of the ethics of GIS, but does serve to illustrate the way in which ethical considerations in GIS tend to overlap into each step along the ontological process. GIS are ethically complex because of the interplay between overlapping and competing goals among GIS practitioners (Crampton, 1995), and because of the way in which ethical considerations within the GIS community are conceptualized from the standpoint of internalist (as in accuracy standards among users) and externalist (as in public opinion of the uses of GIS) judgments (Crampton, 1995). The interrelationships between competing goals and purposes, and internalist versus externalist ethical concerns, make identifying unethical conduct in GIS a difficult task. It is precisely for this reason that much of the conduct going on in the application of GIS today falls within a 'gray area' of being legal in the practical sense, yet unethical in the broader sense of general principle (Onsrud, 1995). Why is there a necessity to develop an ethical context for GIS? Sheppard (1995), in comparing the rapid development of GIS technology to that of the development of atomic power via the Manhattan Project, identified the area of widest concern. He states, "The potential for individual technologies to transform society must be taken seriously...Once longer term societal consequences become evident, the more prominent participants felt that they had changed the world in irreversible and undesirable ways." As practitioners of GIS, the time is right for us to mitigate what we might regret in the future by beginning to examine the ethical inconsistencies emerging in GIS of the present. The research question: What should be considered unethical conduct in GIS? As suggested by Onsrud (1995) at the beginning of this paper, before we can formulate a professional code of conduct for the development and application of GIS we must first try to compile a consensus as to what can be considered unethical conduct within the GIS community. Indeed, gray areas between ethical and unethical conduct exist in GIS largely due to the relatively new development of GIS technology (Onsrud, 1995). To arrive at a consensus Onsrud (1995) and the University Consortium on Geographic Information Science (UCGIS) suggest that opinions should be compiled and analyzed from members of the discipline (internalist considerations) as well as members of the public subject to policies stemming from GIS (externalist considerations). The present study will attempt to take the logical first step in formulating a code of ethical conduct for GIS: soliciting opinions from members of the discipline. The UCGIS Virtual Seminar is an ideal forum for soliciting members of the discipline for their opinions regarding unethical conduct in GIS. Student and faculty participants in the seminar work at several higher education institutions offering instruction in GIS technology. Each institution offers both general introductory courses as well as courses specializing in particular aspects of GIS. The breadth of expertise among participants in the Virtual Seminar surely encompasses an equally broad range of opinions on what constitutes unethical conduct in the subdiscipline. Methods: Opinions were compiled by use of a questionnaire (Appendix A) posted in a discussion thread of the UCGIS Virtual Seminar. The questionnaire includes two sections of questions. The first section is composed of three ethical conflict scenarios (Onsrud, 1995) regarding privacy and geodemographic marketing, income and governmental policy, and data accuracy with regard to population analysis applications in less developed countries. For each scenario participants were asked to state whether they believe the conduct of the GIS user is unethical (yes) or not unethical(no). Participants were also asked to write brief comments on the factors they considered before answering yes or no. The second section of the questionnaire is a list of ten types GIS applications. In this case participants were asked to reply yes if they believe an application is unethical and no if they believe an application is not unethical. All scenario and GIS application questions included in the questionnaire are based on actual, present day conduct taking place within the subdiscipline of GIS. Responses were analyzed by means of multidimensional scaling (MDS). MDS is a descriptive statistical technique, relatively free of statistical assumptions, in which the basic data are dissimilarities between variables in the analysis. The purpose of MDS is to identify and model the structure of a set of stimuli taken from dissimilarities observed among responses to the set of stimuli. In the case of the binary type data (yes or no) constituting responses to the questionnaire, the first step in MDS analysis is to construct a distance matrix based on responses present in some observations but absent in others. For the purposes of this analysis variables were tabulated using a Euclidian distance function. The next step is to use the Euclidean distances derived from each observation to assign relative locations for each observation within a conceptual space such that the distances between observations reflect the dissimilarity among observations. This is done inside a two dimensional plot where observations located near one another reflect little dissimilarity and observations located farther apart reflect greater dissimilarity. Statistical analysis was undertaken using the Statistical Product and Service Solutions (SPSS) 7.0 software package. I constructed a MDS plot for the three scenarios in section one of the questionnaire such that each respondent's answers constitute an observation case (a data marker) and the distances between cases inside the plot constitutes the level of agreement on whether or not the conduct of the GIS users in the scenarios is unethical. Frequency distributions were used to determine whether or not there is a consensus on the conduct in each scenario. For the ten GIS applications listed in section two of the questionnaire I constructed a plot in which each application constitutes a data point. For this MDS model, applications located near one another reflect a higher level of agreement among respondents than applications located at greater distances from one another. Inferences on whether respondents believe each application is unethical or not unethical were made by compiling frequency distributions for each GIS application question. Results: By a period of four weeks after posting the ethics of GIS questionnaire (Appendix A) in the GIS and society discussion thread of the UCGIS Virtual Seminar, I had received seven responses from student participants in the seminar. To obtain enough responses, and thus be able to make meaningful statements about internalist opinions regarding ethical conduct of GIS practitioners, I posted the questionnaire in the GIS-L discussion list used by GIS practitioners around the world (Wright et al., 1997). I received an additional 13 responses from GIS-L list participants living as far off as Canada, Europe, and Australia. GIS-L list respondents included students, several university professors, and several persons working with GIS in the private sector. In total this analysis includes 20 respondents. Of the 20 respondents, all answered yes or no to scenarios one and two, and 19 answered yes or no to scenario 3 (Table 1). The single missing response to the conduct of the GIS firm in scenario three was stated by the missing respondent as being due to ambiguity in the scenario. Missing responses in section two of the questionnaire (GIS application questions) were also due to respondents feeling that the issue arising from the GIS application in question were either too ambiguous, or too complicated, to state whether the conduct of the GIS firm is unethical or not unethical. Missing responses for application questions one through nine ranged from zero to five (Table 1). Question ten of the questionnaire was omitted because most respondents (18) replied that the question of whether or not the conduct of the GIS firm was acting unethically or not unethically was unclear. Three respondents replied to the questionnaire by commenting that all of the GIS users in all of the scenario and application questions were acting ethically, and, therefore, these respondents were scored as ones (no) for all scenario and application questions. Each of these ********** Table 1. Summary results of the ethics of GIS questionnaire (n=20). Scenario/Application question yes/unethical no/not unethical missing responses Scenario 1/ 5 15 0 Scenario 2/ 8 12 0 Scenario 3/ 6 13 1 Application 1/ 2 13 5 Application 2/ 5 14 1 Application 3/ 12 8 0 Application 4/ 5 15 0 Application 5/ 2 17 1 Application 6/ 1 19 0 Application 7/ 2 15 3 Application 8/ 2 16 2 Application 9/ 12 7 1 ****** respondents stated that a GIS is a tool, and "in and of itself," there can be nothing unethical in using a tool. These respondents further stated that they felt there is nothing unethical about gathering information or performing analyses, but that they have reservations regarding the ethics of how a client might use the results of the GIS analyses described in some of the questions. One of these respondent's reason for answering no in all cases was that, "GIS service companies cannot be held to standards of ethical behavior based on how their client "might" (their emphasis) use the information provided." Another of these three respondents wrote, "there is nothing unethical about using a GIS for anything, " and went on to an analogy of the Holocaust stating, "what was unethical was the decision in the first place to move [jews], that's were the problem was, not in the tools used to move them." For scenario one, five respondents thought the conduct of the GIS user is unethical and 15 respondents thought the conduct is ethical. The scenario involved a geodemographic marketing strategy using license plate numbers, collected from customer's cars parked near a store, to obtain addresses for cross referencing with census data. Among the comments made by those who thought the GIS user was acting unethically, one respondent wrote, "The customers have not given permission for [license plate numbers] to be collected (this is a privacy issue)." Others of these respondents commented that, "if [the GIS user] could get the customers addresses from the license plates than that is illegal, and therefore, unethical," demonstrating that legality is the ethical issue in this opinion. Comments made by respondents answering that the conduct of the GIS firm is ethical included, "license plate numbers are public information, and there is nothing unethical about using public information to conduct a study," and, "gathering information in this way is ethical." This respondent went on to say that, "in this case, the info is to be used to better meet the needs of the public, which is certainly ethical." Several of the respondents answering on the ethical side of the scenario made the distinction that using license plate numbers to collect address information is unethical, but that there is nothing unethical about using the information in a GIS. Scenario two involved a GIS firm that was asked by a government agency to use, among other criteria, mean annual income to eliminate high income areas from consideration for a landfill site. In this case, eight out of 20 respondents thought the conduct of the GIS user is unethical. Reasons given by respondents answering that conduct of the GIS user is unethical included, "what makes the conduct unethical is that the GIS firm somehow legitimizes income as a selection criterion by incorporating it into an analysis that gives the impression of being objective." Another respondent replied, "if you are taking the question as a purely ethical one, then the process is unethical and so anyone is tainted, and the GIS company has acted unethically in executing an unethical analysis." Comments made by respondents answering ethical varied from, "it is not unethical to use income in a GIS as a method of minimizing objections from the ignorant public," to, "the GIS firm is merely responding to the client's direction, which in and of itself, is neither illegal nor unethical." This later comment was echoed by most respondents answering on the ethical side of the scenario. Comments to this effect variously stated that the GIS user is acting ethically, but the ethics of the government agency directing the site selection criteria should be called into question. Six of 20 respondents replied that the conduct of the GIS user in scenario three is unethical. This scenario involved data accuracy as an issue for a GIS firm receiving payment for having conducted a population estimate in a less developed country. Accuracy was the main issue for respondents who answered that the conduct of the GIS user in the scenario is unethical. Comments included, "the sampling procedure is flawed and a 20% margin of error is too large, and the GIS firm should know better," and, "the question implies that the GIS firm was paid before conducting the sampling." On the side of respondents answering that the conduct is ethical, most wrote something to the effect of, "as long as [the less developed country's government] knows what they are buying, and the service is delivered in full, I don't see a problem." MDS analysis of the three scenarios shows a considerable amount of disagreement among respondents (Figure 1A, open mds.cdr). Respondents (observation cases) appearing on the positive side of the dissimilarity plot along Dimension 1 are those who answered mainly that the conduct of the GIS user in the scenarios is ethical, whereas respondents appearing on the negative side of this same dimension are those answering mostly unethical. Note the considerable spread of respondents along Dimension 1, which in cases of strong agreement among respondents would act to group like opinions together. Dimension 2 shows the amount of variation within groupings identified by Dimension 1. In this case respondents are also spread along the positive (ethical) and negative (unethical) sides of Dimension 2, meaning there is considerable variation in responses within the loose groupings identified by Dimension 1. Overall, with three scenarios, and yes (0) and no (1) responses possible for each scenario, for each respondent there are seven possible combinations of answers (000, 111, 011, 101, 001, 010, and 100). The distance between observation cases along both dimensions of the MDS model effectively shows that respondents answered in all possible combinations. Therefore, any respondent was as likely to answer ethical or unethical to each scenario as any other respondent, and thus, the MDS model shows no consensus of opinion as to the ethical conduct of the GIS firm in the scenarios. Most respondents, with the exception of two questions, replied that the GIS users in each application question (Appendix A) are acting ethically. I did not solicit comments from respondents for this section of the questionnaire, therefore results are summarized in terms of the number of yes (unethical) and no (not unethical) responses (Table 1). Five of these questions (one, five, six, seven, and eight) had less than five respondents each answering that the GIS firm involved acted unethically. Application questions two and four had five respondents each answering that the conduct of the GIS user is unethical. The exceptions to the general trend of respondents finding the conduct of GIS users in each application ethical were questions three and nine. Question three asks whether or not, "Applications involving corporate exchange of residential addresses (without permission of the addressees) for advertising purposes," are unethical, to which 12 of 20 respondents answered 'yes.' Question nine asked whether or not respondents thought, "Ecological GIS applications biased toward achieving a political end," are unethical. Again, 12 of 20 responded that such an application is unethical. MDS analysis of the relationships between responses to the nine application question shows general agreement that questions three and nine constitute unethical conduct on behalf of the GIS user, and that the remaining seven applications are ethical (Figure 1B). Furthermore, the distance between the seven applications for which respondents mostly answered ethical (positive side of Dimension 1) are minimal, showing that the same respondents made consistent decisions regarding the conduct in these applications. The distance between applications three and nine is also minimal along Dimension 1, showing a strong consensus of opinion that the conduct of the GIS firm with regard to these applications should be considered unethical. Dimension 2 shows little variation in the grouping of ethical applications, whereas applications three and nine are widely separated along this dimension, showing that the respondents who answered unethical to application number three are different than those who answered unethical to application number nine. Overall there is good agreement among respondents on both the ethical and unethical sides of the model, showing that there is a reasonable consensus as to which of the sample GIS applications should be considered ethical and which should be considered unethical. Discussion: The purpose of this analysis was to collect opinions on unethical conduct in GIS from practitioners within the subdiscipline of GIS. For this reason it should come as little surprise that, of the 12 questions used for this analysis, responses to all but two questions (application questions three and nine) were overwhelmingly on the side of 'no unethical conduct on behalf of GIS users.' Over half of the twenty respondents stated that, in most cases, they regarded a GIS as a tool, and any use of a tool, regardless of methods of gathering information or performing analyses, cannot be unethical. In a similar study, Wright et al. (1997) analyzed a debate on the GIS-L list serve in which participants argued either that GIS is a tool or that GIS is a science. Participants arguing that GIS is merely a tool claimed that using a tool does not constitute science, but rather, science is conducted by the GIS user through many means, and using a GIS is just one of these means. In the case of the present study, respondents who stated that GIS is a tool argued that merely using GIS 'the tool' does not constitute unethical conduct. Respondents who commented that GIS is a tool, and using the tool is not unethical, fall into two groups. One set of these individuals answered based on whether they thought both the GIS firm and the GIS firm's clients in each case were acting unethically, and thus, answered each question based upon the conduct of GIS user and client together. These respondents were as likely to answer that the GIS firm is acting unethically as they were to answer that the GIS firm is acting ethically. This group of respondents also includes respondents who answered that the GIS firm is acting ethically in all cases because there is nothing unethical about the conduct of either the GIS firm or the client. In this way, the responses of this group are consistent with their individual philosophies which regard the conduct of GIS user and client together in each scenario and application question. An inconsistency arises with a second group of respondents who regard the conduct of the client in some of the scenarios and applications questions as unethical, yet, see nothing unethical in the conduct of the GIS firm performing the analysis. For this second group of respondents the fact that they regard a GIS as a tool separates the GIS user from the motivations of the client asking for the service. For example, one individual among this second group of respondents commented, "for the most part, the GIS firm is collecting, compiling and analyzing information and or observations, it is how these data are used by the hiring [client] that is sometimes unethical." With regard to the ontological process of a GIS, this group of respondents feel that all steps along the developmental path are ethical, and only the motivations of the client, before collecting information and beginning an analysis, and upon making policy decisions based upon an analysis, are open for debate over unethical conduct. This type of inconsistency has previously been critiqued by several social theorists including Shepard (1995), Miller (1995), and Curry (1997, 1995). Shepard (1995, p. 14) states that "GIS is not simply a tool for processing geographical information, but a social technology incorporating an entire institutional and intellectual infrastructure...as such, its development cannot be understood apart from the social context in which [its development] occurs." Miller's (1995, p. 100) observations are a bit more direct in this regard when he writes, "...GIS practitioners and theorists should not be merely technical functionaries, but cognizant, socially-aware actors...with the responsibility of considering the ultimate deposition of their efforts, rather than simply to follow orders." Curry (1995) observed that ethical inconsistency in the subdiscipline of GIS is inevitable due to the 'competing goals' of the GIS practitioner and society. Certainly this inconsistency is reflected in the opinions of respondents who answer that the GIS firm is acting ethically when performing an analysis on behalf of an unethical client, in which case the goal of the client is an unethical result, whereas the goal of the GIS user is simply 'to do work.' However, this inconsistency is rather unlike that outlined by Curry (1995), because in this case the different goals of the GIS practitioner and society are accentuated by respondents who feel that the GIS firm should not be held to the same ethical standard as their client. These respondents make this distinction apparently because the former is motivated by work while the later is motivated by some unethical end. Conclusion: Ethics, whether regarding the use of GIS, or pertaining to other segments of society, are to a large extent a matter of personal philosophy. MDS analysis of ethical scenarios one, two, and three (Figure 1A) revealed that there is little consensus among the 20 GIS practitioners surveyed as to the ethical nature of the conduct in each scenario. Indeed, all possible combinations of answers were obtained by surveying only 20 GIS practitioners. Therefore, trying to construct an ethical code of conduct for the subdiscipline of GIS, with its thousands of practitioners, all of which have differing personal philosophies about ethics in general, might be an impossible task. However, results of application questions three and nine show that there is strong agreement among the 20 respondents that exchanging personal addresses without permission of addressees, and including known bias in a GIS application, are both unethical. In addition, respondents show strong agreement that the conduct of GIS users in the seven other application questions is ethical (Figure 1B). Both the above observations mean that, when given fairly specific examples of ongoing GIS applications, GIS practitioners are able to arrive at a consensus. A broader based study, inclusive of a far greater number of GIS practitioners, should be conducted to assess whether the results of section two of my questionnaire are repeatable for a larger group of GIS practitioners. The difficulty entailed by such a study is that many GIS practitioners are resentful of the mere mention of ethical inconsistencies in the use and application of GIS (personal observation). That some GIS practitioners are inconsistent in their belief that a client is acting unethically, but a GIS user acting on the client's behalf is not acting unethically, is part of the inevitable inconsistency of GIS (Curry, 1995) where the goals of the GIS user and the goals of society are different. This study was an attempt at compiling the opinions of GIS practitioners regarding the ethics of GIS. As suggested by Onsrud (1995), the matter of ethics in GIS should include obtaining opinions from not only members of the subdiscipline, but also from persons outside the subdiscipline of GIS (the public) who are subject to policies developed via GIS analyses. Surveying should be conducted among the non-GIS public to test Curry's (1995) assumption that the goals of GIS practitioners are different from those of the broader group of individuals making up the society we live in. References (see attachments) Questionnaire (see attachments) Attachments: MDS.cdr, Gisref.wpd, Gisquest.wpd (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=3259) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: December 16, 1998 07:06 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Term Paper My term paper is in the title of "Web-based GIS Approach to Distributed Geospatial Data and Geoprocessing". Here is a summary of my term paper. The rise of the Web technology has been reshaping the ways of data access, sharing and dissemination. It has further facilitated four major changes in traditional geographic information systems (GIS): public access to data, distribution of data, access to GIS functionality, and visualization of multimedia data. Web-based GIS is rapidly evolving along with the advancement of spatial information technology and the popularity of the Web. Web-based GIS is a networked-centric GIS tool that uses the Web as a major means to access and transmit distributed data and analysis tool modules, and to perform analysis and visualization. It represents an evolution from traditional GIS solutions, in which proprietary data models and monolithic software functions are made interoperable and distributed. Applications which adhere to the objectives of Web-based GIS are more able to access and use various types of distributed data, and to utilize multiple geoprocessing tools and services. Using an integrated client/server architecture in communication, Web-based GIS concerns the balance between the weight of server application and client application in design and implementation. Currently, two widely used design methods include heavy server and thick client. Common Gateway Interface (CGI), GIS plug-In, ActiveX control, and Java applet are major implementation techniques. This research addresses the potential and use of Web-based GIS in accessing and visualizing distributed geospatial data, and performing geoprocessing. In this context, two prototype Web-based GIS applications developed in this research are implemented. One illustrates a Web-based GIS to support public awareness for a selected Georgia watershed database using a server-side strategy which uses a GIS map server for accessing and visualizing it. Another focuses on a Web-based location-allocation modeling for spatial decision supporting using a hybrid strategy that uses the Arc/Info software as a GIS server and CGM viewer as a client-side plug-in. This study represents the potential of Web-based GIS in supporting public awareness for environmental data and developing open spatial decision supporting system. In addition, this paper shows that Web-based GIS is a useful vehicle in accessing and visualizing the distributed geospatial data among general public, and conducting geoprocessing among the particular user community. For more detail, click here. http://www.arches.uga.edu/~bwjun/paper/wgis.html I wish you all Merry Chrismas and Happy New Year! Jun (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=3283) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: December 17, 1998 07:53 AM Author: Bill Moseley (wmoseley@uga.edu) Subject: Term Paper: Simulating the "Real World" Title: Simulating the "Real World": A Case Study Examining the Impact of a GIS Exercise on Students' Understanding of Food Security Issues in Africa Author: William Moseley, Department of Geography, University of Georgia, Athens, GA 30606 Introduction "I felt like a real 'scientist' by using real data." Comment of a Study Participant In many university geography departments, students learn about Geographic Information Systems (GIS) by taking a specific course on the subject. The objectives of these courses often include: 1) student understanding of how to perform geographic analysis using a number of different GIS software packages; and 2) heightened student awareness of the key issues in GIS. While these courses often require students to undertake practical exercises, the goal of such activities is typically to become proficient in certain aspects of the GIS software rather than to better understand the particular problem being analyzed. For a variety of reasons, GIS is less frequently used in non-GIS courses as a hands-on learning tool. This represents a missed opportunity as a much broader segment of the student population will take a geography course (often because it is required) than will ever take a course specifically related to GIS. Simple, customized GIS programs could be developed for hands-on learning exercises. If these programs were developed, would they enhance students' understanding of a subject? This paper presents the findings from a classroom exercise designed to answer three, slightly more specific, questions related to this general query. 1) Does one's understanding of a subject differ if it has been learned through a hands-on GIS exercise (versus learning through reading, lecture or discussion)? 2) Would the ability to run "what if" scenarios improve students' understanding of dynamic systems? 3) What are student attitudes towards learning via interactive, GIS exercises? This line of inquiry is related to a larger body of research dealing with 'GIS and society' issues as well as cognition of geographic information. Background Introductory environmental geography courses often address food security issues within the context of discussions regarding world hunger and population. Undergraduate student attitudes on this subject tend to mirror those of the U.S. public. A caricature of the public's general understanding of food security in Africa might include the following assumptions: the people of the continent are monolithically poor, racked by drought, starving, dependent on government handouts, and incapable of helping themselves. In teaching environmental geography I have often sought to convey a more sophisticated understanding of these issues. Examples of key concepts that I have striven to convey in class include: (1) environmental and economic phenomena have varying food security impacts on different types of rural households; (2) contrary to conventional wisdom, rural households in Africa engage in many more activities than farming in order to feed their families (e.g. petty commerce, urban employment, etc.); (3) in the face of food shortages, households are not passive but actively attempt to cope in a number ingenious ways; and (4) famine is a complex phenomena that is related to factors other than drought. The key to understanding many of these issues is gaining an understanding of how rural food systems function in developing countries. Since most students have never been to a developing country, or even spent much time in a rural area, it is often difficult grasp the dynamics of these systems. The use of interactive, computer simulated rural economy dynamics may be one way to improve understanding of these systems. RiskMap: A Food Security Software Package The RiskMap software package was developed by Save the Children (UK), a British non-profit agency, and was programmed using Visual Basic. This software was developed for use by policy makers in order to improve their decision making capabilities regarding food aid and rural development assistance (SCF 1997). The software includes a recently compiled database for 13 African countries. The software models the structure of household economies in Africa, and allows one to predict the effects of a variety of situations (crop loss, price changes, employment changes, etc.) on the food security of poor, middle, and wealthy households. The aforementioned household structures are geo-referenced and the modeling results are displayed cartographically. Exercise in Relation to the Broader Research Context The results of this exercise may be related to a larger body of research concerning GIS and society as well as cognition of geographic information. A major question raised in the UCGIS White Paper on GIS and Society was: "how does use of GIS affect users' social practices and their views of society and nature?" (UCGIS 1998a) Similar questions are raised in the literature regarding cognition of geographic information. For example, "[h]ow does exposure to new geographic information technologies alter human ways of perceiving and thinking about the world?" (UCGIS 1998b) Some research indicates that individuals build personal mental models of the world (Medyckyj-Scott and Hearnshaw1993: 55). These models are largely conditioned by experience. Interactive GIS and computer simulation may have a positive influence on students' mental models of the world. Research indicates that computer simulations may influence "learner's attitudes and action modalities." (Linard 1994:57) Faryniarz and Lockwood found that"[s]imulations can enrich a science curriculum by allowing students to repeatedly and interactively explore and solve a problem." (1992: 468) An imported related question concerns "[h]ow geographic information technology can be used to improve education in geography...?" (UCGIS 1998b) Many authors have asserted that GIS improves geographic education (Palladino 1992, Bishop et al. 1995, Nellis 1994) More specifically, it has been said to help students address a broader range of spatial questions (Nellis 1994) and enhance the way students visualize and interpret information about the world (Palladino 1992) In spite of all these claims. Audit and Abegg (1996:21) noted an absence of direct links between the educational and GIS communities in the literature. Furthermore, proving the effectiveness of the classroom use of GIS and computers exercises has been elusive (Proctor and Richardson 1997). It has also been noted that "[t]he expansion of?technologies tends to outpace the associated knowledge base concerning learning and teaching." (Audit and Abegg 1996: 21) Methodology The research was conducted during a 50 minute class period with an honors section of an introductory level environmental geography class at the University of Georgia. Thirty students undertook the exercise, although only 27 students fully completed it. The mean age of students was 19. The class included 19 female students and 11 male students. Every student in the class had previous experience using computers. In all cases this experience included the use of word processing packages. Some of the students had also used spread sheets, statistical packages, and desktop publishing programs in the past. None of the students had prior experience using GIS software. Each student in the class worked independently at their own computer. Students were given five minutes at the beginning of class to answer background questions as well as four open-ended questions in a pre-test. The four questions in the pre- and post test are listed in table 1. The instructor then walked students through the first part of the exercise (see exercise in appendix I) in order to familiarize them with different aspects of the program. This guided "walk-through" involved an introduction to food security themes and an inspection of the structure of the food system in one area of Malawi, a southern African country. Students used the second half of class to undertake a self-guided food security scenario simulation. In this simulation, the students created a crop shortage and then examined: 1) how this initially impacted different households in the rural community, and 2) how different households attempted to deal with this shortfall. The instructor was available to answer questions during the simulation exercise. During the last five minutes of class the students answered a four question post-test (the same four questions that were asked in the pre-test) and completed a three question evaluation of the exercise. Table 1: Content of Pre- and Post-Test 1) Food security is generally defined as access to enough food at all times for an active, healthy life (World Bank 1996). The opposite of food security is food insecurity. Extreme food insecurity may result in famine. What do you think some of the major causes of food insecurity in Africa might be (please list)? 2) What types of activities do you think African households pursue in order to acquire food (please list)? 3) When a drought or some type of other disaster strikes a community in Africa, do you think the food security of all households is impacted equally or differentially? Explain your answer. 4) When faced with a food shortage, how do you think African households react? Please explain your answer. Of the four questions in the pre- and post-test, the first question served as a control. While the content of this question was discussed generally in class, and the instructor answered student queries related to this question, the computer exercise did not explicitly address it. The second question was addressed in the computer exercise, but was not an integral component of the simulation. Finally, the third and fourth questions were most closely associated with the output of the simulation exercise. Distinct from the pre- and post-test were 14 questions embedded in the exercise. The purpose of these questions was: 1) to encourage the students to reflect on specific issues, and 2) to ensure that the students participated fully in the exercise. Three of these questions were related to the control question, while the remaining 11 were related to the non-control questions in the pre- and post-test. In designing the exercise, the investigator sought to: 1) provide the students with a learning experience, and 2) be able to distinguish between three types of learning. These types included: 1) learning that occurred irrespective of the computer exercise, 2) learning that occurred as a result of material covered in the computer exercise, and 3) learning that occurred as a result of the simulation aspects of the exercise. Results and Discussion The results in this section are presented in relation to each of the three research questions. These questions were as follows. 1) Does one's understanding of a subject differ if it has been learned through a hands-on GIS exercise (versus learning through reading, lecture or discussion)? 2) Would the ability to run "what if" scenarios improve students' understanding of food system dynamics? 3) What are student attitudes towards learning via interactive, GIS exercises? The first question I sought to answer was whether a student's understanding of a subject differs if it has been learned through a hands-on GIS exercise versus learning through reading, lecture or discussion. I attempted to answer this question by comparing change in student performance on question 1 (the control) versus change in performance on questions 2 and 4. As explained earlier, the content of question 1 was discussed during the lab exercise but it was not dealt with explicitly in the computer exercise. In contrast, the content of the questions 2 and 4 were addressed in the computer exercise. In all three questions, improvement may be measured by an increase in the number correct answers provided by the student. The results of question 3 will not be examined in this section as the structure of the question made it difficult to compare with questions 1,2 and 4. If a hands-on GIS exercise is a relatively better way of learning, then one would expect to see a larger improvement (between pre- and post-test) for questions 2 and 4 than question 1. Table 2 presents mean response findings for questions 1, 2, and 4. The number of correct answers did not increase between the pre- and post-test for question 1. In other words, on average, students listed 2.6 correct answers for this question in the pre-test and 2.6 correct answers on the post-test. In contrast, the mean number of correct answers did increase for questions 2 and 4, with question 2 increasing from 2.9 to 4.2 correct answers and question 4 increasing from 1.5 to 3.1 correct answers. Table 2: Mean Response Findings for Control vs. Comparable Simulation Questions (n=27) Question # No. of Correct Answers on Pre-test No. of Correct Answers on Post-Test 1 (control) List major causes of food insecurity 2.6 2.6 2 List activities African households pursue in order to acquire food 2.9 4.2 4 Alternate means of acquiring food in times of shortage 1.5 3.1 An examination of the mean number of correct responses for a particular question is limited in its helpfulness as it does not indicate if and how much individual students improved between the pre- and post-test. Table 3 presents an analysis of the same three questions. For each question, the table indicates what proportion of students showed no change in the number of correct answers provided, as well as the percentage of students who increased the number of correct responses by 1, 2, or 3 or more. It is interesting to note that 77.8% of students did not improve their response for question 1 (the control). This compares with an average of 22.7% of students who did not improve their responses for questions 2 and 4. Similarly, only 14.8% of students increased the correct number of responses by 1 for the control, versus an average of 35.8% of students for questions 2 and 4. Based on the evidence presented in tables 2 and 3, I would conclude that GIS assisted learning led to significantly better results than that produced by limited discussion. Table 3: Answers to Control Question vs. Comparable Simulation Questions (n = 27) Question # Change in number of correct answers between pre- and post-test No Change + 1 +2 +3 1 (control) List major causes of food insecurity 77.8% 14.8% 3.7% 3.7% 2 List activities African households pursue in order to acquire food 18.5% 40.7% 29.6% 11.1% 4 Alternate means of acquiring food in times of shortage 26.9% 30.8% 15.4% 26.9% Mean of 2 and 4. 22.7% 35.8% 22.5% 19.0% Difference between control and mean of 2 and 4. - 55.1% + 21.0% +18.8% +15.3% I finally note that there was very little correlation between how a student performed on the control question versus how they performed on questions 2 and 4. When question 1 (control) and question 2 responses are compared, Pearson's r = .3225. When question 1 and question 4 are compared, Pearson's r = .0115. The next question I sought to answer was the extent to which "what if" scenarios or simulations might improve student understanding of food system dynamics? While somewhat difficult to answer, one may examine levels of improvement made in questions 3 and 4 (the questions that correspond most closely to the dynamics of the computer simulation) and compare this to progress made between pre- and post-tests for questions 1 and 2. As discussed earlier, question 1 is a control and question 2 dealt with more static elements of the computer exercise. While improvements for question 4 were significantly higher than that of control (see table 3), responses did not improve at a greater rate than question 2. It is more difficult to compare the progress made on question 3 as compared to questions 1 or 2. However, one may deduce from table 3 , that 22.2% improved their answer for question 1 (the percent who did not stay the same), whereas 81.5% improved their answer for question 2. This compares 34.6% improvement on question 3 (see table 4) when it is assessed quantitatively. When answers were assessed qualitatively for question 3, it was found that 84.6% of students improved their answer; an improvement over both questions 1 and 2. Table 4: Questions 3 (n=30) Question: When a drought or some type of other natural disaster strikes a community in Africa, do you think the food security of all households is impacted equally or differentially? % who said impacted differentially in Pre-Test: 65.4% % who said impacted differentially in Post-Test: 100% Quantitative Improvement: 34.6% Qualitative Improvement in Answer: 84.6% Answers to question 4 also improved on a qualitative basis in many instances (see examples in table 5). It was not unusual for a student to switch from the view that Africans are completely helpless in the face of adversity to a recognition that Africans are quite resourceful in these circumstances. In relation to this point, one of the students commented in the evaluation: "We were able to learn ways Africans acquire their food. The stereotype is that they accept handouts and that poor are only sector affected." In sum, while learning through simulation definitely produced better results than the control question, it is difficult to conclusively state if this form of learning was more effective than learning that was derived from more static presentations on the computer. This inconclusiveness is based on the fact that only one of the two simulation based questions out-performed question #2 which was based on static presentation in the computer exercise. Table 5: Examples of Pre- and Post-Test Qualitative responses to question 4 Question: When faced with a food shortage, how do you think African households react? Please explain your answer. Student 1 Pre-Test: "With no set means to deal with the problems they take whatever they can get." Post-Test: "African households seek both government support and alternate means of acquiring food as well as employment." Student 2 Pre-test: "I think they rely on government and international support because there are no other sources of food." Post-test: "They can sell their livestock, find wild foods, and purchase crops at the market if money is available." Student 3 Pre-test: "I imagine that there is not much they can do. Many, I'm sure, die." Post-test: "They pursue other mean like additional employment, trade, and use of cash and capital assets." The final question I will address is student attitudes towards learning via interactive, GIS exercises. This question is largely assessed based on the results of a mini-evaluation at the end of the exercise. As indicated in table 6, 3.3% if students found the exercise to not be helpful, 60% found it helpful, 36.7% found it very helpful. In response to a separate question, a full 86.7% of participants recommended that the RiskMap exercise be used again as a teaching tool. As such, these numbers indicate that students were very favorably disposed towards this GIS exercise. Table 6: Evaluation Questions (n=30) Question: How helpful or not helpful do you think the RiskMap program was in learning about food security issues in Africa? Response Not Helpful Helpful Very Helpful 3.3% 60% 36.7% The qualitative comments shed further light on the students attitudes towards learning with a GIS exercise in class (see table 7). In general, the students felt like the exercise made the issue more "real." This notion is supported by such comments as "I'd say it was helpful in illustrating more precisely the ways in which adversity affects a given population." Other examples of this type of comment were: "It helps to visualize what is actually happening when we say food shortage or famine" or "Good hands on experience, makes issue more real." Other authors have noted that GIS technologies "lend and air of authenticity to what students can study." (Audit and Abegg 1996: 22) Another theme that emerged from the comments was an increased understanding of the larger socioeconomic context in which food security problems occur. Examples of these type of comments were: "It was helpful because it puts things in perspective and showed what Africans stand to lose should tragedy/disaster occur" or "It helped expose more of the social and economic conditions in Africa." Finally, students generally seem to like the change of pace and the opportunity to work with computers. For example, "Something different than lecture can be helpful to students" or "Brings food problems to life. Our generation needs to use computers as often as possible." In sum, a quantitative and qualitative analysis of responses indicated that students were favorably disposed towards learning in an interactive GIS environment. Table 7: Sample of Positive and Negative Comments Question: How helpful or not helpful do you think the RiskMap program was in learning about food security issues in Africa? Positive Comments ·"Helpful! I felt like a real 'scientist' by using real data." ·"This opened my eyes on the distribution of employment and who grows the crops in Africa, the different class structures." ·"Reaity check to what one problem can do to an entire country" ·"I'd say it was helpful in illustrating more precisely the ways in which adversity affects a given population." ·"It was helpful because it puts things in perspective and showed what Africans stand to lose should tragedy/disaster occur." ·"Very helpful in giving visual grasp of food security issues." ·"It was helpful in allowing me realize where the people get their food and income from and how they are affected by disaster." ·"The analysis abilities are impressive." Negative Comments · "The only drawback in the program could be the data. Is it frequently updated?" ·"It was kind of helpful, a video would have been better." Question: Would you recommend that the RiskMap computer program be used again as a teaching tool in this class? Positive Comments · "Brings food problems to life. Our generation needs to use computers as often as possible." · "Something different than lecture can be helpful to students." · "It is more descriptive in showing where the problems come from and what alternatives are used. It is also useful in showing who is more affected." · "Gives actual numbers" · "Able to learn ways Africans acquire their food. Stereotype is that they accept handouts and that poor are only sector affected." · "It helps to visualize what is actually happening when we say food shortage or famine." · "It helped expose more of the social and economic conditions in Africa." · "Good hands on experience, makes issue more real." · Negative "The only thing enjoyable was looking at the two pictures." · "It was kind of boring I like learning about real numbers I can see." · "No, because it is all based on statistics which do not give us a sense of the problem. It is based completely on emprical evidence. Makes problem seem far removed." Conclusion This paper presented the results of a classroom experience with students using an interactive GIS software package known as RiskMap. RiskMap allows the user to run food security scenarios for different countries in Africa. The author sought to answer three questions. 1) Does one's understanding of a subject differ if it has been learned through a hands-on GIS exercise (versus learning through reading, lecture or discussion)? 2) Would the ability to run "what if" scenarios improve students' understanding of dynamic, food systems? 3) What are student attitudes towards learning via interactive, GIS exercises? The exercise revealed a quantifiable improvement in understanding that could linked to the GIS exercise. The evidence was inconclusive as to whether learning by engaging in a simulation was superior to general learning on the computer package in general. Finally, evaluation results revealed that students were favorably disposed to learning via an interactive GIS exercise in class. The "realness" of the experience seemed to account for some of this appeal. References Audit, Richard and Abegg, Gerald. 1996. "Geographic Information Systems: Implications for Problem Solving." Journal of Research in Science Teaching. Vol. 33. No. 1: 21-45. Bishop, MP, Shroder, JF and Moore, TK. 1995. "Integration of Computer Technology and Interactive Learning in Geographic Education." Journal of Geography in Higher Education. Vol. 19. No. 1: 97-110. Faryniarz, Joseph and Lockwood, Linda. 1992. "Effectiveness of Microcomputer simulations in Stimulating Environmental Problem Solving by Community College Students." Journal of Research in Science Teaching. Vol 29. No. 5: 453-470. Linard, M. 1994. "From Learner's Styles to Learner's Activity - Lessons from Various Learner Centered Research." Transactions in Computer Science and Technology. Vol. 46: 57-79. Medyckyj-Scott, D. and Hearnshaw, D. 1993. Human Factors in GIS. London: Belhaven Press. Nellis, Duane. 1994. "Technology in Geographic Education: Reflections and Future Directions." Journal of Geography. Vol. 93. No. 1: 36-39. Palladino, S. 1992. A report on GIS in the schools workshop. NCGIA Secondary Education Project. Santa Barbara: National Center for Geographic Information and Analysis, University of California. Proctor, JD and Richardson, AE. 1997. "Evaluating the Effectiveness of Multimedia Computer Models as Enrichment Exercises for Introductory Human Geography." Journal of Geography in Higher Education. Vol. 21. No. 1: 41-55. Save the Children Fund (SCF). 1997. RiskMap 1.2 (computer database and analytical model for 15 African countries including Mali). London: Save the Children (UK). Seaman, John. 1996. "How RiskMap Works." Mimeo. London: Save the Children Fund (UK). UCGIS. 1998a. "White Paper on GIS and Society." Unpublished Mimeo. University Consortium for Geographic Information Science. URL: www.geog.umn.edu/umucgis/research_priorities/GISSOCIET.html. UCGIS. 1998b. "White Paper on Cognition of Geographic Information." Unpublished Mimeo. University Consortium for Geographic Information Science. URL: www.geog.umn.edu/umucgis/research_priorities/cogwhite.html. Appendix I: Copy of Exercise (Not included in this version) (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=3284) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: December 18, 1998 07:41 AM Author: Erik Shepard (shepard@uga.edu) Subject: Term Paper: Identifying Problem Domains in Object-Oriented... Please refer to my website for my term paper, entitled Identifying Problem Domains in Object-Oriented GIS Applications Development http://www.arches.uga.edu/~shepard/oogis (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=3306)