The major task of science is to describe, to reveal the nature of the world. The more fundamental goals of any branch of science are to understand, predict and control. It is believed that TEACHER EDUCATION can be made more efficacious by understanding the nature of TEACHER EXPERTISE. Current scientific research seeks to identify the characteristics which expert teachers possess in order to better understand the construct of teacher expertise. In an applied field such as education, the end goal of such research is to collect trustworthy data from which sensible predictions can be made and provide more control over the teacher development process. Of particular interest in the present study, is an examination of the characteristics of an EXPERT TECHNOLOGY TEACHER Currently, there is no educational research that examines or attempts to model the characteristics of an expert teacher of technology.
A relevant question to ask when developing a technology teacher training program is, "What does a novice teacher need to know to become a successful technology teacher?" Modified slightly, this question might be, "What is it that makes an expert an expert?" (Cushing, Sabers & Berliner, 1992). There is both theoretical and pragmatic value in examining teacher expertise. Investigations add to current literature on expertise, and also provide data about setting-specific expert teaching behavior from which we can make predictions about elements that should be included in a teacher development program. Identifying the characteristics of an expert technology teacher could lead to the development of models to which practicing teachers can be compared, but also to which future teachers can aspire. Questions to answer are: (1) can a defensible model can be built of the prototypical expert technology teacher?, and (2) can such a model be used to predict and control training events, thus informing and extending the organization and delivery of teacher education?
Process of Becoming Expert
Cognitive psychologists have discussed the amount of time it takes to become an expert. Pressley & McCormick (1995) suggest expertise requires many years of experience. Rich (1993) describes expertise as an evolving entity with gradations, or definable levels of competence over time. Norman (1980; 1992) estimates it takes a minimum of 5000 hours, or 2.5 years of 40 hours per week, 50 weeks per year, of focussed effort, for an individual to become an expert. He cautions that this estimate says little about the actual process of becoming an expert. Issues related to this process are practice, interest, opportunity and mentorship. Practice is not just repetition of the same skill or task over and over, but is distributed, purposeful, and focussed in the domain area. An individual generally has some motivation or drive to develop expertise.
Expert-Novice Research
Building upon findings from expert-novice studies, several researchers have proposed definitional models of the expert teacher. Pressley & McCormick (1995) define an expert teacher model which promotes characteristics of self-regulated learning ability, such as strategies for knowledge acquisition, procedures for problem solving and transfer of prior experience to new tasks. The authors also list additional elements of the "ideal teacher", such as being well-organized, alert to classroom events, concern for individuals and groups, and command of subject matter delivery. From a caring perspective, Agne (1992) proposes a "superior" teacher model based on a teachers personal belief system, which includes a teachers efficacy, locus of control, pupil control ideology, and stress management strategies. At best, these efforts to define the characteristics of the expert teacher can be regarded as representing "different pieces of the elephant". Although they describe certain elements of teaching expertise, none of these definitional models allow for variability in individual experts, or present a complete enough view to provide direction for the delivery of teacher training.
A Prototype View of Expert Teaching
In their position paper, Sternberg & Horvath (1995) seek a middle ground between the current definitional or ad hoc models of teaching expertise prevalent in educational research. They argue that no well-defined standard exists which all experts meet and no nonexperts meet. They propose instead a prototype view that provides a way of thinking about teaching expertise that incorporates standards (such that not every experienced practitioner is an expert) but allows for variability in the profiles of individual experts. The authors rationale for clarifying an expert teaching prototype is to distinguish teachers who are expert from teachers who are merely experienced at teaching students, in order to inform teacher education practices.
The authors describe a defensible, rather than a definitive, prototype based on psychological research of expertise across a variety of domains. They identify three ways in which experts differ from novices: experts (1) bring knowledge to bear more effectively on problems than do novices, (2) solve problems more efficiently and do more in less time, than do novices, and (3) are more likely to arrive at novel and appropriate solutions to problems than are novices. Knowledge, efficiency, and insight, are the primary features upon which the expert teacher prototype is founded. Concept Map: Prototype of Expert Teacher
The authors suggest that when such features are organized around a prototype, two equally valid members of the expert teacher category may resemble each other much less than they individually resemble the prototype. The prototype is to be viewed as a central exemplar, to which expert teachers can be compared, and is consistent with the existence of features that are necessary (but insufficient) for membership in the category. Sternberg & Horvath (1995) acknowledge the pretheoretical nature of their argument by emphasizing a prototypical view, rather than a prototypical model of teacher expertise. They hope to have a generative effect on the conceptualization of teaching expertise in both research and practice. In practice, the authors hope a prototypical view will suggest new approaches to the recruitment, training, and selection of teachers, as well as the evaluation of systems directed towards these activities. In research, the authors call for both validation and modification of the prototype, specifically, research examining those features that are important in peopleÕs judgement of expert status and how these features are weighted and combined in the structure of the expert teacher category. Review of Sternberg & Horvath's (1995) Paper
Phase 1: Defining Characteristics of Domain: Teaching Expertise
A review of the research literature on Teaching Expertise was conducted and three articles describing expert teaching were selected. Sternberg & Horvath (1995) call for research that will build, extend and possibly validate their current prototypical view. This author proposes that the prototypical view needs to be extended to include self-regulation and a personal belief system. Three models, based on articles identifying different characteristics of the expert teacher, were constructed using concept maps. In addition, a fourth model, which combines the three individual models into one Expert Teacher Prototype, was constructed.
The belief guiding the present study is that no one of the three models discussed offer a complete picture of the expert teacher. By combining the three models into a five-part prototype, one seems to get a more comprehensive picture of teacher expertise. Another limitation of the three models is that they describe expert teaching in a content-independent context, which does not offer insight into possible characteristics that might be unique to the domain of teaching with technology.
Phase 2: The Present Investigation
The present study attempts to build and extend upon the models proposed by Sternberg & Horvath (1995), Pressley & McCormick (1995), and Agne (1992) in a content-dependent, exploratory manner. The Expert Teacher Prototype is considered within the domain of technology. The goal of this study is to describe and compare three expert's conceptualizations of the expert technology teacher. The hypothesis guiding this research is that by describing the characteristics of an expert technology teacher, one can provide more trustworthy data for predicting and controlling which elements to include in a technology teacher training program.
It was predicted in the present study that expert technology teachers, even those trained in the same University setting, would vary in their conceptualization of domain-specific characteristics of expert technology teaching. This expectation fits well with Sternberg & Horvaths (1995) prototype view that provides a way of thinking about teaching expertise that allows for variability in the profiles of individual experts. The prototypical view provides a central exemplar to which expert teachers can be compared. A clear description of points at which these individual expert teacher profiles overlap may suggest new approaches to the education and training of future technology teachers.
Using computer-based, repertory grid elicitation techniques developed by Shaw & Gaines (1989), the underlying conceptual structures of three expert technology teachers were elicited and compared.
Element Consensus & Conflict
Margo - Gail: Margo seems to use element "Self-Regulating" in the same manner to describe the same concepts as Gail (4.2% greater than 82.1). There is conflict on 14 of the 24 elements, which indicates that Margo may use the same terminology as Gail, but in reality may be discussing different concepts.
Gail - Margo: Gail reached consensus on two of the elements when using Margos constructs. Both "Organization of classroom events" and "Responsive to potential student problems" were rated above 80.0. There is conflict on 16 elements, which indicates that although Gail uses a similar terminology to Margo, she is referring to different concepts.
Gail - Michele: Gail reached consensus on five elements when she completed the grid using Micheles constructs. The five elements are: "Self-Regulating", "Responsive to potential computer problems", "Personal belief structure", "Responsive to potential student problems", and "Student control". Conflict arose on 9 elements, 8 of which were the same as those in conflict when Gail completed Margos grid (Strategies related to structure of technology, Assimilates new/novel ideas, Accomodates new/novel ideas, Effective, Efficient, Insight into prior student knowledge, Projects and integrates change, and Beliefs about the value of computers).
Michele - Gail: Michele achieved consensus on three elements using Gails constructs; "Theoretical knowledge", "Assumes responsibility for student learning", and "Personal belief structure". Conflict arose on 11 elements when Michele used Gails constructs to elicit the grid.
Margo - Michele: The highest incidence of consensus on elements, and lowest incidence of conflict, occurred when Margo completed the grid using Micheles constructs. There was clear consensus on 9 elements, and clear conflict on 4 elements. The nine elements in consensus were: "Confidence with partial knowledge", "Personal belief structure", "Self-regulating", "Beliefs about value of computers", "Persistence", "Assumes responsibility for student learning", "Responsive to potential student problems", "Problem solving" and "Student control". Not surprisingly, three of the four elements that were in conflict were the same elements that Margo considered to be identical on her original grid; "Content knowledge", "Practical knowledge", "Pedagogical knowledge", and also "Efficient".
Michele - Margo: There was clear consensus on five elements when Michele completed Margos grid; "Strategies related to structure of technology", "Effective", "Theoretical knowledge", "Responsive to potential student problems", and "Organization of classroom events". There was conflict on 14 elements in this exchange.
Summary of Element Consensus
Three additional comparisons were made between exchanged grids. For example, the grids where Gail and Margo used Micheles constructs (as well as Gail-Michele on Margos, and Margo-Michele on Gails constructs) were analyzed using SOCIO in KSS0. The results generated a combined list of 13 separate elements which showed clear consensus. Results from the nine SOCIO comparisons can be used to separate specific elements from the original list of 24 about which the experts share a common terminology and common understanding. Elements that 2 or more experts reached consensus upon are:
Elements (characteristics) from this list that seem to be unique to an expert technology teacher are "Strategies related to structure of technology", "Beliefs about the value of computers", and "Confidence with partial knowledge (about computers)".
Construct Consensus
Gail & Margo did not reach consensus on terminology or concepts when they exchanged grids. However, a very interesting result occurred when Margo & Gail construed Micheles constructs. Margo & Gail reached consensus on four of Micheles constructs; "HCI-Learning theory", "rigid-flexible", "classroom events-personality", and "computer related-self related".
Both Michele and Gail created a constructs which differentiated between technology and humans. Gail created "technological-human", and Michele created "HCI-Learning theory", and "about technology-about humans". Michele uses the terminology "human-technological" to describe the same concepts as Gail. Gail uses the terminology "HCI-Learning theory" to describe the same concepts as Michele. Margo reached consensus on two of Micheles constructs, "about technology-about humans" and "HCI-Learning theory", which seems to indicate that all three experts use this terminology to describe the same concepts. This could result from a shared experience with similar courses and or instructors at the university. Additionally, this result could indicate that this terminology clearly differentiates between concepts in the domain.
Construct Correspondence
Margo & Gail, and Margo & Michele, did not reach correspondence on any terminology when they exchanged grids. The only correspondence was between Gail & Michele.
Gail and Michele reached correspondence for two sets of terminology. Gail and Michele use different terminology, "others-self" and "external-internal", to describe the same concepts. Gail created the construct "external-internal", and Michele created the construct "others-self".
Gail and Michele also reached correspondence for another pair of constructs when they exchanged grids, "HCI-Learning theory" and "technological-human".
Construct Conflict
There was a great deal of conflict between the three experts, in which they used the same terminology to describe different concepts. This result was predicted at the beginning of the study.
Gail and Margo generated seven constructs to differentiate between elements. Both experts generated "external-internal" as a construct, however they use this terminology to describe different concepts. This result is quite interesting considering that the two experts teach together, and often work together on software development projects. When Gail & Margo exchanged grids, they conflicted on terminology four and five times respectively.
Michele appears to use similar terminology with Gail, but to describe different concepts, and vice versa. Michele reached conflict on five of Gails constructs, and Gail reached conflict on three of Micheles. An interesting result is that Gail and Michele created what appeared to be similar constructs, "Lesson Planning-Lesson Execution" and "planning-execution", however on both exchanges, the two experts use this terminology to describe different concepts.
Construct Contrast
There was also a great deal of contrast among the three experts, in which they used different terminology to describe different concepts.
Much of the terminology that Gail and Margo use is different, and describes different concepts. This would be of great benefit to students that they team-teach, because those students will get a broad perspective of the domain. However, this also may result in confusion for some students, in that they will not be able to bring together a picture of the domain into some kind of cohesive framework.
When Michele completed Gails grid, it revealed that she uses different terminology than Gail to describe different concepts. Michele contrasted with Gail on three pairs of constructs; "diagnosis-solution" is different from "Learning Theory-HCI"; "global-detailed" is different from "abstract-concrete"; and finally, "critical-optional" is contrasted with "computer related-self related".
An interesting result when Margo and Michele exchanged grids was that they used similar terminology in different ways, to describe different concepts. For example, Margo uses "skills-personality" and "classroom events-personality" in a contrasting manner. Michele uses "flexible-structured" and "flexible-rigid" in a contrasting manner.
The present study attempted to build and extend upon the expert teaching models proposed by Sternberg & Horvath (1995), Pressley & McCormick (1995), and Agne (1992) in the domain of technology teaching. Three expert's conceptualizations of the expert technology teacher were described using 24 elements, and compared using expert-generated constructs. Because of the changing nature of the field, as compared to a more stable field such as mathematics or history, the expert technology teacher may have different characteristics than other teachers. Three specific and unique elements were generated by the experts, and reached consensus in the SOCIO analysis.
The present study may have been limited as a result of the number of elements in the elicitation. Twenty four elements are almost unmanageable for the elicitation process. For two of the grids, there were 3 times the number of elements as constructs. Constructs from the original three grids indicated that several of the elements were overlapping, and were being considered as same, or closely related, characteristics. Because of the exploratory nature of the study, it was not a goal to limit the number of elements generated, or establish a minimum number of constructs on which to rate them. However, as a result of the findings in this study, a future elicitation could be limited to the list of the16 elements on which the experts reached clear consensus, as well as an extended list of constructs gathered by this study.
Although Gail, Margo and Michele share many characteristics in common, they use domain-specific terminology and language to describe expert teaching characteristics in different ways. From the outset of the present study, this result was expected and was valued as a true representation of the variability of teaching expertise.
The literature review, concept mapping and repertory grid elicitation were all useful for an exploratory examination of the characteristics of expert teaching of technology. The literature review revealled a host of definitional and ad hoc models of the "expert teacher", many of which are incomplete and too rigid to allow for variability among experts. However, the protoypical model proposed by Sternberg & Horvath (1995) is a more complete view, with a "central exemplar" to which various and variable expert teachers can be compared. The prototypical view can be profitably extended by adding two components, the self-regulatory characteristic of expert teachers (Pressley & McCormick, 1995), and the personal belief structure (Agne, 1992).
Concept mapping was useful to visualize the different models, and layout a plan for this research project. The maps were also a valuable way to represent the topic to the two experts who graciously participated in this study. The initial brainstorming session to generate characteristics (elements) was a valuable exercise, however, the three of us were defining characteristics that were both too broad and too detailed. During the second session, the three experts examined the concept maps of three different expert teacher models, and were able to clarify, modify, and extend the original list of characteristics we felt best represented our shared understanding of expert teaching.
The repertory grid elicitation process itself was useful both for the grids created for examination, but also for the conversations shared around the computer. In a future elicitation, it would be valuable to tape record the comments and conversations shared as experts generate elements, and then create and examine their own grids. Often, Margo, Gail or myself would comment, "I do not really think that way", or even "I did not know that I thought that way about certain concepts!". Capturing the "live process" of element generation, discussion, negotiation and grid examination would be an interesting study by itself. It would be exciting and informative to meet again with the two experts who participated in this study, to discuss the results and further refine the list of elements chosen as characteristics of expert technology teachers.
WebGrid versus KSS0
I was very fortunate to have the participation of Margo and Gail in this project, both because of their experience in the domain, and their previous experience with the repertory grid elicitation process. Feedback on WebGrid, as compared to KSS0, was positive. Both Gail and Margo indicated that the WebGrid interface was easy to use. Gail mentioned that she liked the ability to drag and arrange objects in KSS0 better than the pop-up menus in WebGrid. All three of us were frustrated by the limitations of running NetScape and WebGrid on the Mac IIs in the KSI lab. We often had difficulty viewing the FOCUS results on the Mac IIs. Both Margo and Gail were very interested in viewing their results immediately after completing their grid, and our attempts to do this online were thwarted by slow, dinosaur technology. However, printouts of the grids were given to Margo and Gail soon after they created their grids, so all was not lost. It should be made clear to WebGrid users what version of NetScape to use, and also the minimum hardware requirements for a smooth display of FOCUS results.