2. Literature Review

Alternative methods of instruction

The first goal was to identify the various methods used for employee training, and compare their advantages and disadvantages.

Low technology methods of training use printed handouts and overhead transparencies. Overhead transparencies are primarily a labor saving device for classroom instruction. The teacher need not write the same information each time it is used. They allow easy presentation of color and graphics. The common attributes of print and overhead transparencies are affordability, reliability, flexibility, standardization of equipment, and ease of creation and use. However, they become unwieldy with large quantities of curriculum. They are static, and their distribution requires time and effort (Kearsley, 1990).

Television, videotape, and film have the advantage of ease of duplication and distribution to diverse audiences, but share a lack of interaction between the learner and the instructor. (Whetzel, 1996). Disadvantages to these formats also include high production cost, lack of involvement of local instructors, and learner boredom due to the lack of interaction (Kearsley, 1990).

Satellite training is a method of broadcasting curriculum to people in different locations simultaneously. Some systems have no feedback, some have audio feedback through a telephone line; which works well for question and answer periods. The feedback aspect solves some of the problems of one-way television broadcast. This system is suited for delivering consistent curriculum to widely scattered personnel, and is used by the postal service, the military, and large financial corporations ( Whetzel, 1996; Collis, Vingerhoets, & Moonen, 1997).

Videoconferencing using a special telephone is a format which has never reached widespread use, due to the bandwidth limitations of analog telephone lines and the lack of higher-bandwidth digital telephone lines (Collis, Vingerhoets, & Moonen, 1997).

Data communication includes a collection of computer-based applications such as e-mail, mailing lists, synchronous conferencing (e.g., "chat" rooms), and asynchronous conferencing (e.g., bulletin boards, forums). These allow great flexibility of location and time of learning, but require a certain level of equipment and skill of the users (Collis, Vingerhoets, & Moonen, 1997).

Teleconferencing consists of two-way communication. Audioconferencing is a low cost, easily implemented system, and may be set up using existing telephone equipment. Videoconferencing is more technically challenging and requires more specialized equipment (Whetzel, 1996; Collis, Vingerhoets, & Moonen, 1997). Computer conferencing uses existing computer equipment with the addition of microphones, but has been severely limited by bandwidth considerations (Kearsley, 1990).

Just-in-time training is short-term instruction provided just prior to its need. The learner has no need to retain the instruction, as it is always available. It is used in factory settings to provide workers with short videos demonstrating various assembly practices just prior to the worker performing the task (Kearsley, 1990).

Another form of just-in-time training is embedded training, which provides curriculum only if requested. One example is the Performance Support System, used in software help files to guide the user as needed (Bergman & Cheney, 1996). Another example is instruction contained within products, such as photocopy machines which have embedded help displays, called up in the event of malfunction to offer the user assistance in clearing a paper jam, for example.

Effects of CAI on learner knowledge

An early US Department of Labor study showed the introduction of CAI in the workplace to have mixed results. Some companies reported increases in worker knowledge, others reported no change or decreased knowledge. The study acknowledged that data from small companies was, at that time, not abundant; and suggested further research (Flynn, 1989). A later study found knowledge was enhanced with CAI curriculum, due to the quick feedback offered to the learner, and by the active learner participation required in navigating the curriculum (Hasty, 1993).

CAI grew into greater favor in the mid-1990s, when the US Department of Labor-sponsored National Alliance of Business reported small and mid-sized companies should embrace new technologies such as CAI, so they may use technology to cause needed change; rather than reacting as technology changes affect them. Their report showed CAI can assist in increasing worker knowledge, as it can be designed to provide consistent training in new standards, such as the quality standard ISO 9000. They also reported benefits from the individualized pace of training, and a better ability to accommodate an increasingly diverse workforce (Bergman & Kaufmann, 1995). The interactive aspects of CAI are seen to engage the learners in active thought and thereby increase knowledge retention. Cuoco and Goldenberg (1996) found in a mathematics curriculum, CAI offered the the learner the ability to tinker with concepts in order to visualize results. Learners who could manipulate formulae, variables, and models independently using a CAI-based tool gained a better working knowledge of these concepts compared to learners listening to the same concepts presented by lecture. Bergman and Cheney (1996) found CAI increases learner knowledge when it involves the synergy of multiple senses. Learners were found to retain new knowledge better when the curriculum was presented with a combination of formats of text, sound, graphics, and video. However, they noted drawbacks when CAI is not customized for a particular organization; and a tendency for instructors to not provide the same level of personal attention to the learners.

Effects of CAI on learner satisfaction

Satisfaction with a learning method is directly related to the learner's level of control of the experience. Learners in a community college electronics curriculum reported higher satisfaction with CAI due to the control they have over their individual learning agenda, the independence it offers, and the branching aspect to allow learners to follow different paths dependent on their responses. The learners also reported the graphic representations possible in CAI helped them understand abstract concepts better than traditional blackboard drawings (Hasty, 1993).

A 1996 comparison study on frog dissection in a traditional lab and by a CAI simulation found users reported higher satisfaction levels using the simulation. While some of the satisfaction was due to the ability to perform a dissection without requiring an actual animal, users also reported satisfaction with the branching ability of the instruction, the ability to make their own choice on navigating the dissection, and the ability to back up and correct mistakes (Kinzie, Larsen, Burch, & Boker, 1996). A study to determine factors affecting successful training programs found learner's enjoyment of the training experience is one of the indicators of a successful training program (Klimczak & Wedman, 1997).

The TeleScopia Project, a joint study of European universities and industries on learning methods, found CAI problem-induced learning resulted in higher satisfaction due to the minimizing of constraints set by the instructor and curriculum. The learners attained greater satisfaction when choosing their own route through the curriculum. The study found learners reported "generally very positive" levels of satisfaction with CAI and other technology-based delivery methods (Collis, Vingerhoets, & Moonen, 1997). The course developers had planned the curriculum to fit the medium, rather than just converting existing materials to CAI. A high priority was placed on enriching the communication aspects of the courses, to provide the learners various methods of interaction with the instructors.

Effects of CAI on cost

It is generally accepted that CAI has a higher development cost than classroom training, but it can be recovered by use with a large number of users over time (Bergman & Cheney, 1996). The TeleScopia Project report stated the increased flexibility of CAI and other technology-based systems may result in less direct costs for employers, as measured by training cost and the ability to reuse curriculum easily; as well as less opportunity costs, as measured by reduction in idle time of workplace equipment, and less time away from the workplace for learners. Initial high costs of development and equipment were found to be eventually recovered. Break-even points were determined for each course as expressed by the number of students required for each course. (Collis, Vingerhoets, & Moonen, 1997).

A study contracted by the US Office of Technology Assessment measured the cost-effectiveness of various technology-based instruction formats, and found a given curriculum delivered by CAI consumed 30-40% less time than in a traditional classroom. This reduced time was a result of most persons learning faster by themselves, rather than at the slower pace of a group in a classroom; which tends to progress at the rate of the slowest learners (Kearsley, 1990).

A report on CAI use in the United Kingdom states a common process is to monitor the usage time of CAI to calculate a cost per hour of training delivery. Any CAI which allows the user to branch or bypass already-mastered material reduces usage time, thus reducing the cost per hour (Houldsworth & Hawkridge, 1996).

If this cost recovery can be documented, industry is willing to accept its implementation. A study on factors affecting successful training programs found the ability of a training program to have a documented good effect on a company's bottom line is one of the indicators of a successful training program (Klimczak & Wedman, 1997).

The cost recovery period may favor smaller companies, if they can produce their own CAI curriculum, minimize development time, and roll it out to employees quickly. A recent survey found smaller companies have higher CAI budgets relative to their training budgets than larger companies. The survey found an annual per-employee courseware expenditure to be $5.72 for training of manufacturing personnel. When the respondents were asked if CAI had increased or reduced their overall training cost, 34% reported a cost reduction, 20% reported a cost increase, and 46% did not know (1999 CBT Report).

Bergman (1997) showed that medium size companies can better contain costs of CAI by using a number of specific tactics. Examples provided include sharing curriculum development with other companies, and renting out time by bringing in learners from other local companies to use the CAI system during times of low demand. The cost of not implementing CAI may also be shown. A market research study reported that while the cost of hardware and software for CAI continues to drop, the cost of not employing technology-rich learning environments continues to rise; as measured by the cost of maintaining poor skill levels of employees. While the majority of business training budgets have been expended on management, CAI is seen as a tool to open up workplace education to the majority of blue collar workers. CAI is seen as a method to provide just-in-time training, seamlessly joining learning and work requirements (Benson, 1994).

Technical resource materials

The literature review was used to locate additional materials to assist in the development of the CAI curriculum used in the study.

Assistance in organization of material and concepts for presentation was found in a guide written for design of adult instruction of similar topics for the military. William Montague, of the Navy Personnel Research and Development Center; and Frederick Knirk, of the University of Southern California; collaborated on the guide for curriculum developers to build instruction for the theory, assembly, and maintenance of military machinery (Knirk & Montague, 1999); similar in scope and audience to the project at NESLAB.

Assistance in technical aspects of course development was found in a report on CAI implementation for the US Air Force. The report stressed investment of time in designing the appearance and structure of the CAI, as it is difficult to change these aspects once released. Graphics must be carefully adapted to minimize file size and resulting network transfer time. The topics must be considered for appropriateness of CAI delivery. The content should be universally readable across various web browsers and operating systems (Riley, 1997). Assistance in usability testing and evaluation was found in a National Alliance of Business guide written for businesses to evaluate CAI programs for purchase (Bergman & Bixler, 2000).