Post pictures of the students at the course site because it "humanizes" the distance learning process.

Asking questions of the students during the lessons is a good way to motivate the students to get involved with the material and focuses the student on what is most important.

Establish a style guide to ensure consistency on font sizes, use of colors, granularity of pictures, and other factors.

Develop lessons for 28.8KB modem bandwidth customers and avoid "razzle-dazzle" gratuitous use of multimedia.

Have only the number of students you can reasonably handle.

Raise the plane to a higher level and ask questions on evaluative or synthesis level and avoid the mere cut/paste dump of text materials. Role playing is fine; pseudo debates can be made to work; and chat rooms and discussions by BBS are not inherently bad (if the technology works).

The key is to create questions that maintain intellectually active and valuable discussions over the designated discussion time frame. This can be accomplished either through insights up front or active real-time intervention and guidance by the professor.

It is important to have team projects for graduate-level online courses. The Wharton School, for example, requires this of students in their online executive MBA program. The asynchronous part of this is a plus, not a minus. We just need better tools to enable students to work easily together asynchronously.

Where teamwork is involved, the lessons should be due on Tuesday, because the team leader needs Monday to combine the inputs from team members, who typically work on their lessons on the weekends.

Respond to completed student assignments within two or three days if possible. Respond to student email questions within two days.

Demonstrate to students that you are very knowledgeable in the subject--you will be judged on this.

With technology, change is constant; your material must be constantly revised to stay ahead of the change.

Have students limit their responses to one or two screens for most assignments. Special assignments can be longer. This helps to control typing demands and reading demands, which can be very time consuming in online courses.

Be sure to provide individual responses to each student on some of the assignments--to demonstrate your interest and that you read all of their inputs

When commenting on team products, demonstrate that you followed their online discussions.

Avoid shovelware. Try to include highly interactive activities such as discussion group exercises, chat rooms, etc.

Information technology is changing the way we play, work, and learn. There are numerous reports that chronicle how information technology has changed work, in particular. These reports offer testimonies of changes in work relationships and the processing. Some of these changes transform into savings for an organization, while others increase convenience for workers and the customers they serve. As in many other organizations, information technology is transforming the delivery of education in academic institutions.

Information technology in some academic institutions is rapidly taking off, while in others acceptance is slow. Interestingly, there are numerous studies show that there are no statistically significant differences in student learning between computer-media education (hereinafter-called distance education) and traditional education in university settings. The author suspects that this slow acceptance stems from cultural barriers in the academic community. Two, in particular, come to mind. These involve the change in the classroom environment—face-to-face versus "virtual" and the changing role of the teacher.

This first barrier involves the notion that the traditional or face-to-face classroom environment results in students that are better educated than the "virtual" or distance education classroom. This notion (perceived or real) is often accompanied by an explanation that a distance education class does not accommodate dialogue and argument—the interactions that stimulate synthesis and problem solving. For that matter, Gabriel Salomon (1981) examines the richness of reciprocal interactions that are inherent in traditional classroom communication. Implied is that this same richness cannot be incorporated into a distance education class. However, Moore and Kearsky (1996) would argue: "…what makes any course good or poor is a consequence of how well it is designed, delivered, and conducted, not whether students are face-to-face or at a distance."

The second barrier deals with the changing role of the teacher. In a distance education offering mediated by information technology, the teacher is no longer the "instructor." The teacher becomes a facilitator who enables the student to engage in a form of self-paced learning. In addition, the teacher now must become more concerned with the visual delivery of the course information. Limited computer screen space necessitates more carefully planned graphics and text. Lastly, since the computer replaces the traditional classroom method of interaction, the teacher must be creative regarding the interactions designed into a distance education course. Current technology and bandwidth do NOT enable an exact simulation of traditional classroom education.

Most of the studies, articles, and opinions on the effectiveness of distance learning have been from the learner perspective, and rightly so, since the learner is the objective of the education. However, the author believes that faculty opinions are very important for gauging acceptance of distance education. First, the degree of acceptance can "make or break" an educational institution’s ability to institutionalize distance education.

Successful institutionalization of information technology that facilitates or performs a major function in an organization often necessitates organizational and procedural changes. Faculty can also provide extremely valuable suggestions for organizational and process changes that need to occur to enable effective implementation of distance education.

The Information Resources Management College Experience

The Information Resources Management College of the National Defense University prepares military and civilian leaders to direct the information component of national by providing graduate-level education in critical information resources management subjects. The predominant method of education is via the traditional classroom. However, the College also is leveraging information technology to deliver education at a distance.

The College’s approach has been a caution one. However, pressures from the customers (students and their bosses) and the testimonies of success by other institutions continues to push the College toward increasing numbers of courses offered from a distance. It is important that the organization, people, processes, and technology adapt to the changing environment.

The College is going into its third year of implementing distance education. During this time, no substantial organizational or process changes have occurred. In addition, there is no organized framework that guides the College’s distance education efforts. To that end, the author developed a questionnaire to solicit faculty and staff opinion on such subjects as recommended organizational and process changes, pluses and challenges of distance education, preferences in teaching milieu, and opinions on the adequacy of distance education for graduate-level courses.

The questionnaire is a combination multiple choice and fill-in in the blank that required no more than fifteen minutes to complete. The questionnaire is found at enclosure 1. It was given to thirty-six (36) faculty and four (4) staff (academic dean, dean of students, etc.) members. Thirty (30) faculty, about an 83% return rate, and three (3) staff, a 75% return rate, responded. The graduate level teaching experience of the respondents is 5-to-9 years (the median) and the mean is 8.3 years. Of the faculty members responding, fifteen (15) are experienced as course managers for a distance education course, fifteen (15) as lesson developers, and fifteen (15) as lesson facilitators (e.g., facilitating a lesson during the conduct of a course). Four (4) faculty members have experience in administration (other than course manager) and seven (7) faculty members have no experience at all with the distance education format.

The answers to the questions were consolidated by hand. This raw data may be found at enclosure 2. The prevailing opinions are summarized. These prevailing opinions are the basis for conclusions on the degree of acceptance of distance education in the College.

The first question asked faculty and staff their opinion on whether distance education, as we do it, is suited for undergraduate education, graduate education, or neither. They had the option to check all that applied. Approximately 73% indicated that distance education (as we do it) is suited for undergraduate education, while approximately 67% indicated graduate education. Approximately 6% (2 respondents) clearly indicated that this form of education was suited for neither level, and one respondent penciled a response that this form of education was suited for "training" only. Twelve respondents made comments. The prevailing opinions focused on the limitations of technology for delivery (transmission) and the limitations of technology in providing a means for interaction between students and between students and the teacher.

The next question asked if distance education is "as good as," "an adequate alternative to," or "inadequate alternative to" (traditional) classroom education. Four (4) respondents (12%) indicated that distance education is "as good as" classroom education and twenty (20) respondents (60%) indicated that distance education is "an adequate alternative to" classroom education. Eleven (11) respondents (37%) indicated that distance education for graduate-level courses is "an inadequate alternative to" classroom education. This high percentage suggests a significant degree hesitation by faculty in embracing distance education for graduate level education.

The results of the next question naturally follow the results of the previous. Faculty members were asked their preference for teaching: distance education, in the classroom, or other. An overwhelming 77% (23) prefer teaching in the (traditional) classroom. Only one respondent preferred distance education and 20% (6) have no opinion or prefer either mode. This further supports a high degree of hesitation by faculty in embracing distance education.

Two questions dealt with average hours to convert a (traditional) course to a distance education format. The median is 4-to-8 hours for every classroom hour and mean (after discarding one very high response) is approximately 7.3 hours. The companion question dealt with the average hours required to revise the course for the second and third offerings. The median is 2-to-3 hours for every classroom hour and mean (after discarding one very high response) is approximately 2.6 hours. Having first hand experience with converting graduate level courses, the author would take the higher mean a better estimate of time. Adding the two together, results in approximately 9.9 hours to convert to distance education for every classroom hour. Therefore, a graduate course of 42-to-45 hours of classroom would take more than 400 hours to convert.

The next set of questions deals with the advantages and challenges of implementing distance education. The advantages are categorized under student convenience, student participation and self-paced learning, and the richness of on-line information resources. The challenges are categorized under interaction, technology, and teacher resources.

Student convenience. An overwhelming 59% of the responses (37 out of 63) related to convenience for this student. Many comments referred to the asynchronous capabilities that provide schedule (any time) flexibility for students and location flexibility.

Student participation and self- paced learning. About 16% of the responses (10 out of 63) stated that this format enables everyone to participate equally. Students have the time to consider their responses more carefully and students cannot coast (they must participate). Several comments stated that this format enabled students to absorb and reflect at their own pace.

Richness of on-line information resources. There were three comments regarding the richness of on-line resources. One comment is quoted: "Rich on-line resources are available-beyond what could be presented in a class to respond to individual need and interests.

Interaction. An overwhelming 43% of the responses (30 out of 69) indicated that a major challenge was overcoming the poor interaction between students and between students and professors. Some of the comments suggest that social interaction is important in developing intellectual interactions. Several comments are worth quoting: "Faculty cannot see faces of students to determine lack of understanding…or when the light of understanding comes on." "Difficulty in covering complex material with limited feedback [to students]." "Shallow learning on the part of the student and the illusion of education on the part of the school administration and policy makers." The comments go on. These comments imply that nonverbal communication is vital to the education process. For that matter, there are some who say that nonverbal skills account for 2/3 of communication If this is true this might suggest that nonverbal skills are vital to teaching and learning. Without the ability to use nonverbal skills, learning, in fact, may be shallow.

Technology. This next category covers the issues of bandwidth, technology constraints and support. Eighteen (18) responses or 26% suggest that the bandwidth is inadequate to provide the video and audio needed to provide a richer environment. In addition, technical support and technical frustrations in using the information technology were also cited as challenges.

Faculty resources. This last category of challenges includes faculty time to learn new tools, to adapt to new types of student/faculty interactions, and to develop and present the courses. Ten responses or about 15% were made in this category area. Several comments stated that more time is spent teaching on-line. In a tradition classroom setting (with say 20 students), there is a 1-to-20 ration of interaction with students. In a "virtual" classroom this become a 1-to-1 twenty times. In essence, each student is afforded the luxury of a private tutor.

The final set of comments relates to organizational and process changes needed for successful institutionalization of distance education. Since it appears by the answers, some of the respondents were confused about the difference between organizational change and process changes, both these are treated as a single summary. The comments are categorized under policies, procedures, and methods; technology and training; telecommuting; organization, and personnel rewards.

Policies, procedures, and methods. Almost 30% of the responses (23 out of 78) suggest changes in policies, procedures and methods. These include the following:

• Standardize on a consistent process, look, and feel—Agree on a functional architecture—practice what we preach.
• Create templates.
• Provide "how to" guides.
• Provide for on-line registration and improve registration data (three comments).
• Agree on a core application.
• Provide for electronic notifications to students (registration, read-aheads, etc.).
• Develop a special end-of-course survey that is tailored to the delivery format. This suggests that students should be asked specific questions regarding the delivery medium.
• Eliminate paper-based course reviews.
• Change the faculty productivity metrics to something more suited to distanced education—other student contact hours (two comments)

Technology and training. Approximately twenty (17) responses (22%) related to technology and training. These included:

• Hire more technical personnel skilled in distance education technologies who are available to assist faculty more readily.
• Work toward goals of increased reliability in the technology.
• Provide training to faculty on how to design and implement their courses in a standard distance education format.
• Have the administrative office be the central location for collecting papers and keeping track of late submissions

Telecommuting. Approximately nine (9) responses (12%) related to more flexible work hours for faculty. This also included the option to telecommute.

Organization. More than eight (8) responses (10%) suggest organizational changes. These included:

• Create a separate distance education office (that may include the technical support, as well as other administrative support).
• Create users’ group (for information sharing, lessons learned, and configuration management).

Personnel Rewards. Two of the responses suggested that a personnel rewards structure be established to reward outstanding efforts in the area of distance education.

Faculty opinion is important to an educational institution’s implementation of distance education. Institutions not only gain suggestions on changes that need to be made, but also opinions on such subjects as job satisfaction as it relates to delivery education via information technology.

Faculty accustomed to traditional classrooms may not readily adjust to "virtual" classrooms. Skills they view as vital to education, student and teacher dialogue and debate, may no longer continue to be the predominant skills needed for a successful "virtual" classes. This interaction with students is a key component of faculty job satisfaction.

As job satisfaction decreases, these experts (faculty) in communication and education begin to leave for better consulting jobs. What evolves is an institution made up of administrative personnel charged with the task of writing/developing and administering the presentation of "virtual" courses. Information technology becomes the predominant "teacher and communicator." This changes the face of the education entirely. Is this so bad? This is an area that is ripe for future research.

The following table shows how students in AMP 19 use the Internet for electronic commerce (John Saunders’ survey):

Organizations recognize the potential for finding information that can be used in decision-making buried in huge data mines. The challenge, however, is finding the precious data nuggets. Data mining attempts to produce new knowledge that someone can use by building a model that describes patterns and relationships of data. This article provides an overview of the technology of "data mining."

"Data mining, as a methodology, is a set of techniques used to uncover obscure or unknown patterns and relationships in very large databases." It is a key part in the process of discovering knowledge in databases (KDD). Data mining applies algorithms to databases in an attempt to extract patterns automatically from large databases. "Historically the notion of finding useful patterns (or nuggets of knowledge) in raw data has been given various names, including knowledge discovery, information harvesting, data archeology, and data pattern processing. Today, a commonly accepted view is that KDD refers to the overall process of discovering useful knowledge from data, while data mining refers to the application of algorithms for extracting patterns from data."

Extracting knowledge is described as a five-step process: 1. Select application domain; 2. Select target data; 3. Preprocess data; 4. Extract information/knowledge; 5. Interpret and evaluate. The first three steps are important preparation for data mining. Step four actually performs the data mining, which includes deciding on the type of operation, selecting the technique, choosing the algorithm, and mining the data. The fifth step is the interpretation and evaluation of the discovered patterns. The focus of this paper will be on the fourth step in the process, that of actually mining the data.

Once the pattern is identified it can be used in a multitude of ways, such as becoming the content of a report, serving as the training input to a neural network, or being encoded as a rule into an expert system. However the actual data mining process is considered complete at the point in which the pattern is discovered. "Analytical approaches that search data sets on the basis of known patterns are not doing data mining, although they may use inputs from data mining exercises to form the basis of target matches."

Data mining uses sophisticated modeling techniques to uncover patterns and relationships that might otherwise remain hidden in organizational databases. It helps to find the needle in the haystack or see the forest through the trees. There are two main kinds of models used, predictive and descriptive. Predictive models use data with known results to predict values for different data. An example of usage might be assessing loan applications based on prior payment history. Descriptive models describe patterns in existing data, which may guide decisions. They may be used to construct predictive models.

Data mining does not eliminate the need to understand the data or the analytical methods. It is simply a tool that helps to find patterns and relationships in data. The patterns must still be verified and interpreted by an analyst who knows the business. It does not replace skilled managers but rather equips them with a powerful tool. The techniques vary considerably in complexity and ease of use. Neural networks, for instance, are not easily interpreted because there is no explicit rationale given for the decisions or predictions that the network makes. The data mining success is also largely dependent on the quality of the data being mined. Large databases often contain missing and miscoded data, as well as many different variable types. These difficulties present challenges in identifying relationships. For example financial institutions often lack the data management required to maintain and access data across all asset classes. Recent crises were traced to old market data obtained from vendors and exchanges that kept potential disasters hidden from view. The institutions lacked not only the data mining technology but also the infrastructure that could effectively enable its use.

The most common categories of operations are classification, regression, link analysis, segmentation, and deviation detection. Typically two or more of these are combined for a given data mining application. Classification maps data items into predefined classes. Regression maps data items into real-valued prediction variables based on statistical methods. Both are used for prediction. Link analysis establishes connections between data records, such as products to sales. Segmentation identifies clusters of records that exhibit similar characteristics, and finally deviation detection focuses on identifying patterns that do not fit with previously measured norms.

There are a variety of techniques used to mine for data nuggets. Depending on the type of operation being performed some techniques are better suited than others to achieve optimal results are. The induction technique develops a classification model from a "training" set of data examples. Once the model is trained to identify patterns, it can classify data records automatically. Neural net methods develop classification models from algorithms. Nearest-neighbor and case-based reasoning are other types of example-based methods. These approximate a classification model. Associated and sequence discovery techniques discover rules to identify similarities among a collection of items. Clustering technique segments data into clusters, which share common properties of interest. Visualization algorithms enable users to analyze structures of multidimensional data by displaying patterns and trends in multivariate data.

Almost all domains of human activity could benefit from data mining. This includes not only those activities in which a lot of data is already available but also the ones where data must be simulated in order to extract profitable knowledge concerning the field. Some examples of the many current uses are market analysis, customer segmentation, fraud detection, pattern detection, medical diagnosis, and power system management. Data mining may identify buying patterns based on characteristics of consumer choices. There are many documented business examples that credit the data mining capability with providing a significant competitive advantage, thereby easily justifying the return on investment.

Applications in the public sector have been primarily focused in the areas of law enforcement, space and scientific applications, and health care. The Defense Intelligence Agency has used data mining to aid in drug interdiction. The Financial Crimes Enforcement Network division of the Treasury Department has been able to detect money laundering. NASA’s Jet Propulsion Lab performs analysis of Venus images and classifies sky objects. The U.S. Defense Department’s Health Affairs organization performs benchmarking, utilization forecasts, and identification of high-risk patients using data mining techniques.

Most marketing managers although eager to learn more about applications for data mining are also skeptical, noting that the availability of the tools has preceded the development of a guide to their proper application. The nature of data mining poses concerns some concerns. Many databases are collected with no particular analysis in mind. Data miners are willing to analyze such serendipitous data, which has led some to regard data mining as "snooping or dredging". This undisciplined research is like a large-scale "fishing expedition". Another potential concern is identifying and acting on relationships among variables that are thought to be important when they really are not. Similarly important relationships may be missed. If data mining is not performed wisely, it may have some very negative impacts and these could create fear of the technology which would impede market growth.

"A recent Computerworld survey found that companies now using or planning to use data mining expect the number of mining users to triple in the next year. A similar, Forrester Research Inc. poll found that four times the number of projects currently actively are planned to go live this year." A recent survey of corporation with data warehouses conducted by consulting firm Meta Group indicated that 54% of the respondents plan to purchase data mining tools this year. That is a 20% increase since 1996. Currently only 8% use data mining software. Palo Alto Management Group, a California based research firm, estimates that by 2002, companies will spend $113 billion to analyze customer data, including mining it.

These growth estimates are based largely on the recognition that technological enhancements have improved the ability to capture and store immense amounts of data. Micro-processor devices are extending the data-capture capabilities. At the same time data warehouses are providing the capability to mine huge centralized databases. As IT departments figure out ways to funnel data from mobile and embedded computing devices into data repositories, such as data marts and warehouses, data mining will provide the potential for rich new sources of information. IBM, one of the vendors marketing products for mobile computing and data mining, believe that this offers the capability of assessing data for a million organizations and a billion people connected to a trillion devices. The real potential in pervasive computing comes from a company’s ability to manage and mine the data generated by mobile and embedded devices. Data mining can identify new relationships that will provide more information about customer behavior than has ever before been available. IT organizations are feeling the pressure to recognize and respond to this need to make use of the data that is being made available through these combined techniques of mobile computing and centralized repositories.

It is estimated that Fortune 500 companies manage over a terabyte of electronic information each day, with annual growth projected at 57%. Most companies use mining technologies to track business information that is stored in these large databases, called data warehouses. Data mining use is increasing due to a variety of factors including the trend toward data warehouses, the explosion in the amount of information captured electronically, the dramatic price decreases in data storage hardware, and the focus on knowledge management in organizations to gain competitive advantage.

In conclusion, as data continues to become increasingly available through electronic media such as the Internet, so too will the need increase to improve techniques and tools for managing the data and enable enterprises to recognize its true potential.

Data mining is a very promising technology that offers tremendous return on investment when performed successfully. However, care must be taken to not oversell the capability. It cannot succeed without the proper supporting infrastructure. The technology itself is very complex and not something that can be easily adapted to an organization that is not technically and culturally prepared to implement it. It is currently in the very early stages of its evolution with an extremely promising future. All forecasts support the need for such capabilities; however in order to fully recognize the benefits of data mining, a considerable amount of additional research and development of the technology is needed. Currently, the industry is recognizing fantastic growth making it an attractive investment opportunity, which may in turn fuel the increase of research.