[ IIMS 92 contents ]
Computer based learning research at Edith Cowan University
Geoff Ring
Computer Education
Edith Cowan University
The research project
The general aims of the University's research program in Computer Based Learning are twofold: firstly, to use the "windows" provided by CBL environments to better understand how students think and learn; and secondly, to determine how best to use computers in education so that their contribution to the effectiveness of teaching and learning is optimised. While there have been numerous studies which claim to show the superiority of CBL environments over traditional learning environments (Kulik et al, 1983), much of the research has been criticised for assuming that different instructional media can represent meaningful experimental variables independent of content or strategy (Hagler & Knowlton, 1987). Very few studies have been able to clearly demonstrate that learning gains associated with CBL environments were due to characteristics unique to CBL. Many of the studies were primarily comparing "messages" rather than media because they tended to ignore the fact that different media place different restrictions on the kinds of messages they present.
The research team took the view that researchers in the field of CBL should discontinue this "media comparison" model, that CBL was clearly a valuable additional learning medium and that further studies comparing CBL with other media could only be of limited value. What was needed were more manageable and logical kinds of research designs which concentrated on the learning effectiveness of specific attributes of CBL environments which were known to be unique or believed to be advantageous.
The nature of CBL environments
CBL environments are unique in many ways: they can represent abstract or complex concepts with dynamic visual models; they can use screen "objects" to offer a world of "learning by doing"; they can offer interactive, individualised, learning modes which vary from directed instruction to open ended inquiry; and they can keep records on individual student performance, providing informative feedback for students and teachers. The essential elements which define a CBL environment are computers and learners. The broad definition adopted here, however, allows for the inclusion of other elements which are generally important to learning environments such as instructors and other curriculum materials. Further, the "computer component" is considered to have three facets: the human-machine interface; the learner interactions with the software "content"; and the "behind the scenes" management and record keeping functions. This conceptual framework may be represented graphically as shown.
Project components
The research project has several components and involves a team of researchers, each investigating specific attributes of CBL environments which fit within an overall conceptual framework, the basis of which is the common model of a CBL environment. The principal objective of each of the six research components is given below.
- To examine the relative effectiveness on learning of two forms of animation (computer created animation and computer controlled video images) on the learning of theoretical concepts associated with motor skills in sport.
- To study teachers' management of their learning of a courseware authoring system: influences of their self characteristics, environment, motives, goals and study perceptions.
- To determine the characteristics of student-computer interactions associated with particular implementation models using simulation and modelling educational software in secondary classrooms.
- To investigate the relative effectiveness of instructional design characteristics in Computer Based Learning implementations.
- To identify optimal characteristics of graphical user interface design, with particular reference to navigational aids within a CBL environment.
- To investigate the potential offered by CBL tools to provide an environment within which children can represent and explore their own conceptual models.
A brief description of the methodology to be used in each research component follows, together with the name of the principal researcher.
Component A
| The relative effectiveness of two forms of computer animation (computer created animation and computer controlled video images) on the learning of theoretical concepts associated with motor skills in sport.
Mr Rodney Ellis |
Aim
The aim of the project is to examine the relative effectiveness of two forms of animation (computer created animation and computer controlled video images) on the learning of theoretical concepts associated with motor skills in sport.
Knowledge required to facilitate motor learning and preparation for teaching motor skills is often obtained from written resources which are neither interactive nor capable of animated display. Even videotape, while being animated in nature, is not particularly interactive in terms of learner access and is disadvantaged by its linear form of presentation. On the other hand, computer software and devices such as videodisc players (under computer control) offer knowledge based resources that are both interactive and animated. The development of both of these types of instructional materials is expensive and it is important for physical education teachers and coaches to be aware of the relative effectiveness of each in terms of learning gains.
The expected research outcomes would be in the form of findings and recommendations concerning the relative effectiveness on learning of the two forms of computer controlled animation. A significant additional outcome would be the development of good quality CBL courseware which could be used as an interactive knowledge base with tutorial potential for complementing physical practice in the learning of motor skills as well as a resource for instructors.
Research design
The learner of a motor skill must have a reasonable grasp of the scope and temporal context of that motor skill before significant improvement in performance (motor learning) can be expected to occur. This largely cognitive process typically dominates the early stages of learning a motor skill. To facilitate this early cognitive stage of learning, learners and coaches usually seek information from a variety of sources. Training programs designed by physical education teachers and coaches typically require the trainee to demonstrate the ability to effectively describe the mechanical principles of sound technique and the appropriate strategies for dealing with particular game circumstances. The traditional means of supplying the material for this cognitive requirement is through 'wise council' from a recognised expert, complemented by written material in the form of textbooks and coaching manuals, or perhaps from viewing videotape reproductions. Expensive computer controlled resources have not been readily available to instructors, but this situation is changing. The use of computer driven interactive and animated graphics seems ideally suited to assist learners in their understanding of motor skills because of the dynamic nature of those skills. It seems likely that good quality CBL software employing animated graphics, together with display mediums such as videodisc technology, could provide valuable additional means to assist with the cognitive demands of learning motor skills and how to teach them to others.
The implementation of the project requires the creation of two software based sets of instructional materials on the techniques and strategies of tennis. The first set will include animated sequences created by using the graphics capabilities of an authoring system ("Authorware Professional"), while the second set will be identical except that the computer graphics will be replaced by video sequences stored in digital form.
Each of two evaluation phases will require that the a cohort of first year trainee secondary physical education teachers be used as subjects; the "computer graphics group" will be selected from the 1991 cohort and the "digital video group" will be selected from the 1992 cohort. Each group will participate in a 13 week (one hour per week) tennis unit in semester 2 of their first year of the secondary physical education program at Edith Cowan University. The actual group sizes to be considered will be determined after pre-testing both groups and matching subjects on their knowledge of tennis techniques and strategies. It is anticipated that group sizes will be in excess of 30. After supervised exposure to one of the sets of instructional materials, students in each group will be tested on their knowledge of techniques and strategies at the conclusion of the tennis unit and will be surveyed to find out their attitudes towards their computer based experiences.
Progress to date
Mr Ellis has completed a literature search and is currently developing prototype CBL software for use with secondary Physical Education student teachers at Edith Cowan University. The CBL model he is using provides for both a direct route through the subject content and the facility to browse at will. Therefore, in addition to the data related to the effectiveness of different forms of graphics, it will also provide information on the routes taken by students in their efforts to achieve the learning objectives and thus provide insights into student attitudes towards, and the effectiveness of, the different learning modes available.
Component B
| Teachers' management of their learning of a courseware authoring system: influences of their self characteristics, environment, motives, goals and study perceptions.
Mrs Lorraine Kershaw |
Aims
The purpose of this study was to investigate how teachers managed their learning during a one semester course of instruction in the use of a courseware authoring system which was employed by them to design learning activities for their students. It was concerned with the teachers' metacognitive strategies used in this learning and design process, their reflections about their learning and the quality of their learning outcomes. These aspects of their learning were explored in relation to their self characteristics, environmental influences, their motives, initial and revised goals for the course and their perceptions of their study over time. It was expected that the quality of the teachers' learning outcomes would be affected by their background experiences, knowledge, skills and beliefs. In addition, it was reasonable to expect that these learning outcomes, both cognitive and affective, would be mediated by how the teachers managed their learning.
The teaching environment is one where change is constantly occurring and it is vital for teachers to feel confident in their abilities to adapt to these changes. There is a need for teachers to develop, utilise and be aware of their own efficient learning behaviours in order to solve new problems which they will encounter not only during these change periods but also in their own professional development. Knowing about their own learning and the processes used to monitor this learning is an aspect of adults' learning which has received little attention in research studies.
This study will add to the limited knowledge we have about the influence and effectiveness of teachers' metacognitive strategies and beliefs on the quality of their learning outcomes. Results will provide additional guidance in the development of programs which train adults to use metacognitive skills and assist teachers to transfer their efficient self regulation skills to other domains where the level of problem solving difficulty is high.
Research design
This research into human behaviours was viewed from a social scientist's stand point and the ideas guiding the investigation of this topic emanated largely from the interpretivist's paradigm, but also to some extent from that of the critical social scientist. This approach therefore was concerned with describing and understanding influences on adults' learning with some engagement by them in a form of critical discourse (that is, self reflection).
While computing environments in general may have been familiar to some participants in the study, the environment provided by the authoring system "Authorware Professional" was unknown. In this sense it was useful to observe and monitor the strategies which were accessed and used in this problem solving setting. It is acknowledged that this study had some limitations in that these strategies were only observed within the context described. No comparisons can therefore be made about the possible transfer of skills between domains.
The sample required for the study consisted of two groups of teachers. Each group was undertaking a BEd unit at Edith Cowan entitled Courseware Authoring which was based on the use of the "Authorware Professional" authoring system. One group was studied in each semester of 1991. At the beginning of the unit the purpose and design of the study was explained to the teachers and they were asked if they would agree to participate in the research project. The researcher acted as a participant observer for the duration of the unit. Data was gathered by observations, through the administration of questionnaires and associated interviews. The researcher maintained a journal and each teacher was also asked to keep a journal to monitor their own perceptions of their learning.
Progress to date
Mrs Kershaw completed a literature search in the area and established the research methodology for her study during the first half of 1991. Suitable instruments for data collection were either identified from other studies or designed by the researcher. Three questionnaires, containing both closed and open ended questions, were administered to each group during the course. Group and individual observations were made. Participants were provided with a focus set of questions for their journal entries which were made on four occasions. The researcher maintained a weekly journal which included details of any critical episodes occurring during each lecture session. At least two interviews and a final telephone interview were conducted with each participant. Initial data analysis has commenced and a conceptual heuristic model has been developed for reporting on the data. This model will enable findings to be discussed in relation to a computing culture which is seen as being shaped by, but also determining, the cognitive and metacognitive approaches and attitudinal responses of the teacher acting in the learning situation.
Component C
| The characteristics of student-computer interactions associated with particular implementation models using simulation and modelling educational software in secondary classrooms.
Mr Paul Newhouse |
Aims
The study was designed to determine what features of particular implementation models, especially those involved in group work strategies, enhance student-computer and student-student interactions. The number of educational software packages available is increasing at an exponential rate, the design quality of the software is improving rapidly and the number of computers available in schools is increasing. Despite these trends, CBL products are generally not having the level of impact on classroom learning predicted by many, and the intention was to explore possible links between specific CBL implementation models and the quality of the student-computer and student-student interactions which result. The results of the research are expected to give new insights into the nature of classroom based CBL environments.
Research design
A small number of existing software packages, along with a number of special purpose packages developed using the authoring system "Authorware Professional", are being trialled in secondary classrooms. Each trial involves a significant portion of classroom time (about half a school term) in which the researcher acts as adviser to the cooperating teacher with regard to implementation strategies. An implementation model is constructed for each trial, based on the wishes of the cooperating teacher and the advice of the researcher.
Implementation models are being designed to maximise the control the student has in the task environment and to maximise the quality of the student-computer and student-student interactions. The teacher and researcher reach consensus on the intended interaction patterns and on the nature of the classroom environment. In each trial the purpose is to determine to what extent the actual interaction patterns and classroom environments reflected the original intentions. Any discrepancies will be analysed in terms of the original implementation model. Data is being gathered by observation, questionnaires and interviews. The data analysis is based on a triangulation method to determine characteristics of the interactions for each implementation model. Implementation models will be grouped according to variables concerning intended teacher, computer and student roles, software category and hardware access configuration. Differences in interaction characteristics between different model types are being noted and will be evaluated.
Progress to date
Mr Paul Newhouse has completed an extensive literature review and has gathered data from six CBL implementation sequences, five associated with a "Maths in Practice" Year 11 class at a secondary school and one associated with a class of academically talented students at a primary school. Two questionnaires and a structured interview process were trialled during 1991. Initial analyses on three of the CBL implementation sequences have been completed. Early feedback as to the success of the classroom strategies has been encouraging. All the student projects were successfully completed and many student projects were of a standard well in excess of expectations. Most students indicated, both verbally and through questionnaires, that they had gained a high degree of satisfaction and enjoyment from their use of the computer in classroom based activities.
Component D
| The relative effectiveness of instructional design characteristics in Computer Based Learning implementations.
Dr Ron Oliver |
Aims
The overall aim of this research project is to investigate the relative effectiveness of instructional design characteristics, particularly those which relate to the level of learner control on navigation through an instructional sequence. CBL materials can be designed with varying levels of flexibility of navigation. Traditionally these have been characterised by the tutorial mode which provides little choice or option to the user in completing an instructional sequence, through to modes where the user is provided with a choice of varying hypermedia presentations. Through this research project it is intended that the effectiveness of different instructional design formats will be determined to provide empirical evidence to assist software designers to choose the most appropriate form of instructional design for particular functions or aspects of a CBL environment.
Research design
The authoring system "Authorware Professional" is being used to generate parallel CBL materials which differ with respect to instructional design and format and can be used by students in an independent mode to achieve specific learning goals. The experimental study is comparing the effectiveness of the different forms of instructional design used. Third year students studying computer education at Edith Cowan University will be asked to participate in the investigation and will be exposed to one of the different pieces of software. A pre-test will be used to assess initial knowledge and understanding on the chosen topic, the students will be randomly assigned a package for use and a post-test will be used to determine subsequent levels of knowledge. Standard statistical tests will be used to compare the learning differences which may occur and these differences will be analysed in the light of the different levels of flexibility of instructional sequences which were used in the programs and the nature of the learning differences observed to occur.
Progress to date
Dr Oliver is developing an Authorware based shell for instructional programs which traces and records the user's selections and reactions in completing an instructional sequence. The software is being designed to record such things as navigational choices and time spent in sections of the program. He intends to apply the structure of this shell to the creation of two forms of computer based instruction and to compare the influence of differences in flexibility of instructional design sequence on learning outcomes.
Component E
| Characteristics of CBL user interface design, with particular reference
to navigational aids within a CBL environment.
Ms Jan Ring |
Aims
The overall aim of this research component is to identify optimal characteristics of graphical user interfaces, with particular reference to navigational aids within a CBL environment. It is intended that this study will contribute new knowledge to the area of user interface design as applied to CBL materials by identifying the ways in which learners navigate CBL environments and the effectiveness of those navigational styles according to the age, gender and prior experience of the learner. It is expected that the new knowledge will assist developers of CBL materials in maximising the effectiveness of the navigational models they employ.
The specific aims are to determine:
- how learners visually interrogate computer screens exhibiting a graphical user interface;
- the characteristics of effective navigational aids within a CBL environment; and
- age and sex related differences in the use, perception and preference of navigational aids.
This research is drawing on the disciplines of educational psychology, the psychology of reading, ergonomics, graphic design and computer science. The focus of the research is on learners and how they interact physically, emotionally and cognitively with the computer interface. The study is expected to identify optimal characteristics for user interface design across a broad age range. The results of the study will have application for software developers in general and CBL authors in particular.
The two primary sources of information regarding the subjects' behaviours while using CBL software will be gathered using the CEDRIC eye tracking device and a CBL software package developed by the researcher using the authoring system "Authorware Professional". The other sources of information are interviews and questionnaires. The CEDRIC eye tracking device is mounted above the computer output screen and tracks eye movements on the screen using an infrared beam in a technique known as 'pupil to corneal reflection vector measurement'.
Data to be gathered by the CEDRIC eye tracking device include:
- subjects' eye movements: the location of gaze as an X:Y coordinate of the screen will be captured every tenth of a second. The resultant data stream can later be 'replayed' or drawn over the software image giving a clear view of how the user visually interrogated the screen. As the data is timed, it is also possible to determine where and for how long the user paused or fixated on particular areas of the screen.
- keyboard responses: which keys were pressed and when
- mouse actions: position of the mouse cursor on the screen and when mouse clicked
Data to be gathered by the software package include:
- the position in the hierarchy of the software domain
- keyboard responses: which keys were pressed and when
- mouse actions: position of the mouse cursor on the screen and when mouse clicked
- specific interactions: which option chosen, response times
- results of the task (where a set task has been set)
- time taken on task
The two capture systems will be synchronised and their data combined to give a full picture of user behaviour over time. The area of duplication between the two systems, keyboard responses and mouse actions, will provide a check to the synchronisation of the data flows. As the data flow is time tagged, it will be possible to see how long the subject spent on various aspects of the CBL software.
Questionnaires will be used Prior to each stage to determine prior experience levels and personal preferences related to the use of navigational elements. Interviews with subjects following the computer sessions will be used to determine personal preferences regarding navigational styles and to determine their perceptions of the CBL domain.
Research design
The subjects form a 2 x 3 matrix, with gender (M/F) by age group (10-14, 17-25, 35-45). Each group will consist of 30 subjects. The age groups have been chosen to represent the groups commonly being confronted with CBL in the course of their work and education. The younger age group represents the upper primary-lower secondary child, the second group the average undergraduate and the third group, the professional in the work force.
The study will involve three stages. Stage One will investigate subjects preferred and actual navigational methods in a goal free situation. Stage Two will introduce a set task and Stage Three will present subjects with a set task to solve in an optimal environment as defined by results of Stage Two. Stages One and Two of the research will use the same subjects. Stage Three will have the same groups but a new set of subjects.
The data will be analysed to determine the effectiveness of the various navigational elements. Effectiveness will be determined in terms of directness of the navigational pathways used, degree of task completion and time taken to accomplish the task. The relationship of prior experience to performance will also be determined. Differences between groups will be analysed. The dependent variables are age, gender and prior experience and the independent variables are task score, time and navigational effectiveness.
Progress to date
Ms Ring has completed a literature review of the area and spent the latter half of 1991 working with a pre-release version of Authorware Professional for Windows to prepare a research environment involving CEDRIC. CEDRIC has been calibrated and set up for a dual environment situation in which the researcher and the subject each have their own screen, and the researcher sees what the subject sees.
Component F
| The effectiveness of computer based modelling environments in
promoting conceptual understanding in young children.
Mr Martyn Wild |
Aims
The overall aim is to investigate the potential offered by software tools to provide an environment within which children can represent and explore their own conceptual models. The study will use software tools that allow children of a primary school age to construct, modify and explore their own mental models with respect to one or more concepts. In the processes involved in these activities learning can be expected to occur and it is the quality of this learning which will be evaluated.
The main questions of interest in the study are:
- is understanding facilitated by children representing and exploring their own models of a concept or concepts?
- what teaching methodology makes best use of modelling activities, as described above?
Associated questions will include:
- what are the characteristics of children's mental models in a specific domain or area?
- how do software tools that permit modelling activities help children's learning; in other words, how do children use appropriate software to represent their own model, evaluate that model and then modify it?
The research is expected to give new insights into the nature of children's conceptual learning in CBL environments.
Research design
The rationale for this study is based on the importance of modelling activities in children's conceptual learning. The work of Piaget provides an important theoretical basis for the investigation. Piaget was interested in how knowledge develops and the laws that govern this development. He was interested not in the "rightness" or "wrongness" of children's ideas but rather in how children come to form those ideas. Piaget showed that children construct their own models of reality in order to understand and that these models are very different to those created by adults.
In many teaching-learning situations the teacher will present children with a model of a concept or concepts which the children are expected to assimilate and understand. However, many children fail to move from their own mental model of the same concept to that offered by the teacher and in this failure lies misunderstanding. It is suggested that certain software tools may provide a successful medium through which children can represent and explore their own conceptual models by testing their operation. In this process children might be expected to come to understand the potential and the limitations of their own models. In other words, children's learning may be enhanced through involvement in the following three stage process: construction and exploration of a model; editing and modifying that model; and comparing that model to other models, created by either their peers or their teacher.
The role of the educator is often interpreted in terms of providing learners with appropriate representations or models in a range of domains of knowledge. To facilitate this role it is important to provide learners with accurate and consistent models that are appropriate to their needs. Learners' misunderstandings can often be interpreted to be the result of a failure to move from their own models of a concept or situation to the conceptual model held to be correct or desirable by the educator. Much of the current rhetoric of educational research is founded on the premise that it is essential to start the learning process from the learner's experience and understanding. However, there appears to be little thought given to the need and the methods to move the learner from his/her own experience and understanding to that deemed necessary by the educator. These educational issues are all relevant to the study proposed. The contention is that children can successfully learn by representing and exploring their own models and those of others.
It is intended to study children between the ages of eight and eleven years. These children will work within both quantitative and qualitative modelling environments, with small groups of them being followed for periods of up to a year in a series of activities that occur naturally within their teaching program. This relatively long time span will help to diminish the 'honeymoon effects' of introducing new technology to teachers and children. A variety of software environments that facilitate both exploratory and expressive activities are to be used: exploratory activities will involve a learner's exploration of a given model; expressive activities will involve learners in representing their own models. Each of these activity types will be classed as separate for analytical purposes but in actuality both can be expected to occur within the same learning context. Data will be collected that will facilitate analysis of children's conceptual understandings and associated teaching methodologies. Analysis of data will be conducted according to historical and ethnographic traditions.
Progress to date
Mr Wild commenced his research towards the end of 1991 and is currently reviewing the literature and evaluating software that has the potential to provide the necessary learning environments for the subjects in his study, given their ages, computer experience and curriculum circumstances. He has completed a "pre-pilot" study in a primary school to ascertain the relevance of his work to curriculum opportunities in the upper primary section of Western Australian schools and is currently considering the relative merits of spreadsheets and authoring systems in developing the CBL environments needed for his study.
Overview of progress during 1991
The overall research and development project was based on a 3 year plan (February, 1991 to December, 1993). The original estimates for completion dates and timelines for each stage of each research component are given in the table below and they have been largely adhered to by each of the researchers.
| Component |
Lit.
Review | Preparation* | Implement-
ation | Data
analysis | Pub. of results |
| A |
Jun '91 | Jun'91 & '92 | Dec'91 & '92 | Jun '93 | Dec '93 |
| B |
Mar '91 | Mar '91 | Dec '91 | Jun '92 | Dec '92 |
| C |
Oct '91 | Jun '92 | Dec '92 | Jun '93 | Dec '93 |
| D |
Jun '91 | Dec '91 | Mar '92 | Jun '92 | Dec '92 |
| E |
Jun '91 | Dec '91 | Jun '92 | Dec '92 | Jun '93 |
| F |
Dec '91 | Jun '92 | Dec '92 | Jun '93 | Dec '93 |
| * software development, instruments, procedures, etc |
The major outcomes of the project are expected to be in the form of research findings and recommendations concerning the effects upon learning of specific attributes of CBL environments. A more immediate outcome has been the establishment of an R & D environment which is creating new research opportunities for University staff and students. The project has also provided a sound basis for the establishment of a multimedia courseware development centre at the University to complement the research dimension.
Conclusion
Keeping a balanced perspective on the value of technology in education is difficult. It is probably true to say that some of the early high expectations have not been realised and that two of the major reasons for this have been the world wide shortage of high quality educational software and the lack of knowledge about the attributes of effective Computer Based Learning (CBL) environments. If educators are to maximise the advantages of computer technology, a substantial research and development effort is needed to address these deficiencies. It is hoped that the CBL research and development program described in this paper will be able to make a valuable contribution to this effort.
References
Hagler, P. and Knowlton, J. (1987). Invalid implicit assumption in CBI comparison research. Journal of Computer Based Instruction, 14(3), 84-88.
Kulik, Bangert and Williams (1983). Effects of computer based college teaching: A meta-analysis of findings. Review of Educational Research, 50, 525-544.
Schoenfeld, A. H. (1983). Beyond the purely cognitive: Belief systems, social cognitions, and meta-cognition as driving forces in intellectual performance. Cognitive Science, 7, 329-363.
| Author: Dr Geoff Ring, Head, Computer Education, Edith Cowan University, Perth.
Please cite as: Ring, G. (1992). Computer based learning research at Edith Cowan University. In Promaco Conventions (Ed.), Proceedings of the International Interactive Multimedia Symposium, 267-278. Perth, Western Australia, 27-31 January. Promaco Conventions. http://www.aset.org.au/confs/iims/1992/ring.html
|
[ IIMS 92 contents ]
[ IIMS Main ]
[ ASET home ]
This URL: http://www.aset.org.au/confs/iims/1992/ring.html
© 1992 Promaco Conventions. Reproduced by permission. Last revision: 4 Apr 2004. Editor: Roger Atkinson
Previous URL 7 Apr 2000 to 30 Sep 2002: http://cleo.murdoch.edu.au/gen/aset/confs/iims/92/ring.html
