Online Induction: Working to Meet the Needs of Beginning Science Teachers Joel D. Donna and Gillian H. Roehrig University of Minnesota Abstract: This paper explores how theory and data influenced the iterative design of an online induction program. In particular it looks at internal and external barriers that interfere with beginning teachers’ engagement within an online community of practice.

Background The retention of highly effective teachers is a topic of concern among policy makers, professional organizations, teacher educators, and schools districts. Studies suggest that teachers leave the profession at rates of up to 50% within the first five years of their practice (Smith & Ingersoll, 2004). These high rates of attrition, combined with retirement of baby boomers, have caused shortages of teachers in many regions of the country. As it is generally agreed that teachers become more proficient as they progress throughout their career, students in districts with high levels of attrition suffer the most as they have larger numbers of beginning teachers. Induction programs can work to alleviate some aspects of job dissatisfaction by offering support during the tenuous years of beginning teachers’ development. Induction programs also provide a critical bridge from teacher preparation to practice. While these programs usually include mentorship from a seasoned veteran, they vary significantly in length of support, professional development activities, and collaboration with peers. Although many schools have mentoring programs, not all beginning science teachers are matched with a content specific mentor who teaches the same science discipline. Finding these subject-matched mentors is challenging, particularly in smaller charter and rural schools. Online induction programs may help to solve the problem of providing subject specific support and professional development. These programs can connect science teachers across a larger region with subject specific mentors where face-to-face communication may not be possible. While there are compelling arguments for the development of online induction programs for beginning science teachers, there is a lack of empirical knowledge surrounding such work (Davis, 2006). One study suggests that face to face discipline specific induction may impact both beginning teacher beliefs and practices (Luft, Roerhig & Patterson, 2003); however there are few other studies in this area and none deal specifically with online induction. Missing from the knowledge base is an understanding of the unique challenges of learning to teach science and how interaction within an online induction community can be used to help meet these challenges and sustain professional growth. There is a need to understand the theory and practice of online induction programs for the science education community and the broader induction community. This paper examines the online induction program STEMMP (Science Technology Engineering and Mathematics Mentorship Program) for beginning teachers in Minnesota. In the 2006-2007 school year, this program served 50 beginning teachers, 35 of which were in science. This study looks to inform both the science education and the larger induction community as to how the interplay of theory and practice can be used to inform theory based designs of online induction programs. In particular we seek to understand how theory, participant interaction, and formative and summative assessment influence the iterative design of an online induction programs for beginning science teachers. This mixed-methodology study uses, participation frequencies, designer reflective journals, and qualitative and quantitative pre, mid, and post survey data to create a narrative on the iterative design of this program currently in its third phase of re-design. This study uses an emerging paradigm of study called design based research. Design based research seeks to modify learning theories through designing, implementing, and refining theory-based interventions in naturalistic, educational environments. This theory-driven, applied research uses methods from various paradigms to study interventions and produce new theories of teaching and learning (The Design-Based Research Collective, 2003). This study will analyze this data through the lenses of first order barriers (external contextual issues) and second order barriers (internal beliefs) that influence participation within this program (Ertmer, 1999) to better understand how these challenges interplay with this online community of practice.

Version 1 (V.1) – (0ctober 2006 – Feb 2007) V.1 - The knowledge base informing the design The design of this program was greatly informed by the body of work regarding general and science specific induction. Reviews of induction research (Wojnowski, Bellamy, & Cooke, 2003) support the idea that induction programs can help improve novices teaching practices, can lower attrition rates, as well as have benefits to the mentor as well. For science teacher induction, Luft (2003) and others (e.g. Wojnowski, Bellamy, & Cooke, 2003) have indicated the importance of having a subject specific mentor and science specific induction programs to support these novice science teachers. This subject specific focus may be best at helping these novices develop greater pedagogical content knowledge (PCK). Online induction programs have other practical benefits as well. In a review of the literature regarding general ementoring programs for students, teachers and other professionals, Knapczyky, et al (2005) found that e-mentoring programs have added benefits such as providing a private place to discuss matters, support programs that can be used more frequently and at times convenient to the users. In addition, asynchronous discussion areas, may allow for more thoughtful dialog than face to face communication due to their delayed nature. One of the larger online induction programs for science teachers, eMSS, was developed by the New Teacher Center, Montana State, and the National Science Teacher’s Association in 2002. The project’s main focus is to "increase student achievement in science and math by providing early-career secondary school teachers with content understanding and disciplinespecific pedagogy" (Grimberg, 2006). The program offers support mainly through public and private asynchronous forums. There has been limited published research from this program, however unpublished dissertations (Bice, 2006) and conference presentations (McAleer, 2006 and Taylor & Mike, 2006) generally lend support to the program goals for the development of PCK as well as developing reflective, problem solvers. V.1 - Design Goals STEMMP, supported through a Minnesota Department of Education’s National Governors Association Grant, was designed to provide online mentoring support for untenured secondary math, science, career and technical education teachers throughout Minnesota. The goals for STEMMP were to improve teacher job satisfaction to increase retention of effective teachers and to increase teacher effectiveness. STEMMP used Moodle, an online, open-source learning management system, as a platform that provides both synchronous chats and asynchronous forums. Each novice teacher was matched with a mentor with teaching experience in the same content area and grade level. Each mentor/mentee pair was also grouped with 5-7 other pairs who taught similar content to form a small learning community (SLC). Conversations between mentors, mentees, and their peers took place in several areas of the STEMMP site and have different goals, as described below: Mentor-Mentee Chat: This was a private area for conversations between mentors and mentees where unstructured and some structured conversations took place. The mentees were required to meet at least weekly with their mentor in either a private synchronous chat room or an asynchronous private discussion board. In this support, the design goal was to provide content specific mentoring to help develop PCK through coaching. Small Learning Communities: The SLCs were organized by subject matter and grade level to support mentees by providing access to resources to use in the classroom as well as space to engage in conversations. The mentees asked questions and posted concerns in this area through either synchronous chat or an asynchronous discussion board. The mentees were required to meet at least monthly with their SLC synchronously within a chat room. Professional Development Inquiries: The first PDI was designed to improve lesson planning through an 8-week learning cycle. This was done by helping new teachers focus on student learning and student work, by critically examining their own teaching in relation to their beliefs and commitments, and develop the skills of data collection, analysis and reflection. The mentees were to complete three PDI’s thorugh the course of the program. Blogs: Mentees completed weekly blog entries designed to help the mentees reflect on their practice and as a way for mentors to better understand the needs of the mentee. They also provide a tool for mentees to reflect back on their growth over the year. Mentees were to post once per week. V.1 - Implementation Prior to the September 20, 2006 mentor training, 58 mentees had been recruited – 38 science, 13 mathematics, and 7 FACS/CTE teachers. Of these, 12 sought credit by participating online, 13 sought credit in conjunction with a

monthly face to face induction program. These teachers were evenly distributed among rural, suburban, and urban districts. Although a majority of the beginning teachers were in traditional middle and high school settings, 12 beginning teachers were in alternative or charter schools. Many of the participants were first (n = 22) and second year teachers (n = 19), yet there were a number that had thee or more years experience (n = 11). In October, 2006, the mentors and mentees were brought together for an initial face to face meeting. Each attendee was provided financial support for a substitute teacher and mileage. The goals of the meeting were to provide an overview and requirements of the program, to building community and trust amongst the mentor and mentee and small learning communities, and to provide background information on the challenges of learning to teach and the technologies employed in STEMMP. Unfortunately, not all of the mentees were able to attend this meeting. In the initial months of the program, participation within the program began to wane. Although monthly mentor logs showed that there was contact between mentors and mentees, the engagement with the rest of the supports seemed to falter. In addition, some of the mentees who never attended the face to face meetings began to drop out of the program. In addition, it was a near impossible task for the program designers to both provide technical support for over 100 users as well as facilitate conversations within 12 small learning communities and PDI investigations. In January, the midpoint of the year-long program, it became obvious that a program redesign was necessary. These changes could be introduced at the second face to face meeting in February. In preparation for the re-design, a follow up needs self assessment, a number of both quantitative and qualitative survey questions sent to the active mentor mentee pairs. 45 of the 50 active mentees (90% response rate) responded. V1. External Barriers towards engagement Although mentees were generally satisfied with the program, several barriers towards engagement emerged. The primary external barrier towards engagement was time. The participants, who were mainly first year teachers, were overwhelmed with the time requirements of planning instruction, grading, setting up laboratories and felt that there was not enough time to fully engage within the program. When queried about barriers towards engagement with the PDI, 13 out of the 34 who did not fully engage with the PDI cited some form of time constraint towards engaging with the activity. For the SLC chats, it was a challenge for the groups to arrange a time to consistently meet each month when they were all available. This form of synchronous chat, while allowing for immediate response and feedback does present a challenge for beginning teachers who are already overwhelmed with the difficult challenges of learning to teach. Furthermore, if the beginning teachers do not see this use of time as valuable, they will not make the extra effort to engage unless there is some form of extrinsic reward/punishment. The technology used to support this community also acted as a barrier towards engagement for some of the participants. For those who did not attend the face to face training, they found interaction with the site challenging; for those who did attend, there was opportunity to ask questions and work through technical issues together. In addition, for those who chose to access this site from home, some had issues with the chat room application. Many commented that the chat room was slow and often crashed. While later site updates helped to resolve these issues, these technical issues, coupled with the time barriers, may have helped to explain why many mentees preferred to email their mentor instead of using the chat room or site forum (see Table 1). Some switched to phone calls or inperson visits to communicate with their mentor. While email can be used, it does draw participants away from the community space and supports. In addition, email conversations are challenging for program designers and evaluators to track. Table 1 - Mid Survey - As a mentee, which of the following have you sought support through…? Yes Count No Count Mentor/Mentee Chat Room Meetings 81.4% (35) 18.6% (8) Mentor/Mentee Email 79.5% (35) 20.5% (9) Mentor/Mentee Forum Messages 51.2% (22) 48.8% (21) SLC Chat Room Meetings 45.5% (20) 54.5% (24) SLC Forum Messages 39.5% (17) 60.5% (26) Other Mentor Communication (F2F, phone, etc) 22.0% (9) 78.0% (32) V1. Internal Barriers towards engagement While time and technology may be significant barriers towards engagement within an online community of practice, it is important to investigate internal barriers to engagement, such as participant beliefs about community support. While strong mentor/mentee collaboration is positive as it gives credence to the importance of content specific support, it is also important to note that current guiding principles for the design of induction programs stress the

importance of community based supports such as peer groups and professional development activities (Ingersoll, 2003). In particular, it was important for the designers to understand the second order barriers towards the other supports such as the SLC Chats, Blogs, and PDI activities. An analysis of the participant surveys revealed that beyond the technical and time constraints, many participants did not see the value in such community supports. The blogs and SLC chats were viewed as a ‘venting session’ that duplicated the work done with their mentor. Also, because the blogs did not have the capability for participant commenting, some felt that the blogs “feel isolated and artificial, as though I am writing to no one.” Blogging was ranked as the least satisfying component. In addition, for many participants, they did not see the value or understand the value of the PDI investigation. For some, it duplicated prior or existing professional development activities. While some participants discussed the PDI work with their mentor within their weekly chat sessions, some who did participate by posting the forum began to question if they were posting in the right area. Without interaction with other members, the value of working within a community to learn from and with quickly dissipates. It becomes a direct method of teaching (read and react) instead of a group dialog. It became clear to the designers that online communities of practice do not self sustain; they need to be facilitated. In addition, participants must buy into a community of learners approach to increase engagement.

Version 2 (V.2) – (Feb 2007 – May 2007) V.2 - The knowledge base re-informing the design With the approach of the second face to face meeting in February, we felt that this was an excellent opportunity to modify the program design. Using the knowledge gained from the surveys, participant interaction with the site, and our design notes, we returned to the knowledge base surrounding this work. Returning to the literature we once again focused on improving our theoretical understanding of communities of practice literature. The development of a fully online community of practice is a difficult design challenge (Barab et al., 2003) as face to face meetings are necessary to build trust required for sharing and engagement. In addition, facilitators within online communities of practice can play a critical role in helping push the thinking of all participants further by making connections to theory and encouraging further discussion (Collision, 2000). In addition, literature surrounding professional development and adult learning theory helped to inform the redesign of this project. Adult learners are motivated to engage if they feel that the activities they engage in are relevant to their own context and that they have a sense of choice in their activities in order to ensure that their individual needs are met (Wlodkowski, 1999). To ensure the redesign of the activities could work to meet the needs of the larger STEMMP community while working to improve their practice, we closely examined the changing needs as assessed through the mid-year survey (Table 2) by using a beginning teacher needs assessment (Gordon, 1991). It became clear that additional supports were necessary for classroom management and discipline concerns. Table 2 – Top 8 Needs of Beginning Teachers – Choose the response that indicates your level of assistance needed High/Very High Need (Oct) High/Very High Need (Jan) Maintaining student discipline 58% 64% Obtaining instructional resources and materials 62% 62% Motivating students 47% 60% Diagnosing student needs 55% 57% Organizing and managing my classroom 43% 57% Managing my time and work 40% 48% Assisting students with special needs 56% 46% Planning for instruction 39% 45% V.2 - The Re-Design of STEMMP Using this knowledge informed by theory and practice, we set out to re-design the program to better meet the needs of the participants and to increase the interaction of the participants within this online community. During the face to face meeting in which all mentor/mentee pairs would attend, it was important to address issues of buy in, provide a space for community development, provide additional structure and clarity, and provide choices of supports that could be more responsive to individual needs. To address the issue of time and structure, we modified the program requirements. (to see a copy of the program requirements visit http://stemmp.googlepages.com). We decreased the mentor/mentee chats to twice a month,

decreased the blogs to once a month, and provided the SLC’s with the option to hold their discussion within an asynchronous forum instead of the synchronous chat room.; all but two of the SLC groups opted for the forum-based discussion. In addition, a month by month checklist was provided to each participant to help them understand what they had to do and when. As well as decreasing the number of requirements and improving structure, it was important for issues of buy in to re-address the purposes and relevance of the activities such as blogging. We need to stress that blogs are a reflective tool that could not only be used as an individual to look back across growth, but as a place to read others posts and take comfort that they were not alone in their struggles. To increase SLC work, we re-designed the group meetings around a common topic of concern. Using the data from the mentee needs assessment (Table 2) we developed case discussions. Allison Mike, who was a facilitator in the eMSS program, shared the content and structure from the eMSS program called Dilemmas. Dilemmas are illstructured problems, modified from actual teacher cases. The structure provided for these activities was that the mentees were to explore the problem first by posting a response to the case and then responding to another mentee. After two weeks the mentors would enter the discussion. The SLCs were also given the option to create their own dilemmas and have the group work through them. Dilemmas were centered on topics such as classroom management, disruptive individual students, and students with learning disabilities. These topics were chosen to match the needs from the participant survey data. The PDI was also revised in terms of structure, content, and time. The PDI continued to follow a learning cycle approach, however we added a choice of topics that we felt were relevant to the mentees needs as beginning STEM teachers. Modified lessons from Marzano’s (2001) Classroom Instruction that Works provided the mentees with a choice of topics such as cooperative learning, non-linguistic representations, inquiry based instruction, vocabulary, and skills and processes. Mentees were to choose their topic at the face to face meeting and began work with their mentor in order to scaffold them into this new system. In addition to the program activities and structures, each PDI and SLC group was assigned a facilitator. These facilitators served as members of the group who would help spur on discussion amongst group members by posing questions, linking responses, and summarizing discussion. In addition to increasing group participation, the facilitator’s role was to push the thinking of the participants further. The facilitators would also introduce the PDI and Dilemma topics to their groups. V.2 - The Implementation after re-design After the early February face to face meeting, the work of the participants on the STEMMP site was carefully monitored by the design team to see if there was increased interaction within the site. Results showed that there was a substantial increase in the participation directly after the February meeting. The overall participation within the blog and SLC activities, as judged against the new requirements, increased; the SLC posts increased by 18% and the blog posts by 20%. Within the PDI, there was an increase from 46% to 73% of mentees participating within some aspects of the PDI. While these numbers are encouraging, there were still some who struggled to engage. Also, as the end of the school year approached, the participation began to wane. There certainly still existed other barriers towards engaging fully with the community. In addition to the participation data, a follow up mentee survey was given at the end of the program. There was a 95% (N = 45) response rate. This survey asked the participants to explain what they did and did not value about every aspect of the STEMMP program as well as to explain what they think may have hindered their engagement with each of the supports. Overall, the mentees were much more satisfied with V.2. (65% - very satisfied or satisfied) as compared to V.1 (39% - very satisfied or satisfied). Again, while these results help us to justify the changes to the program, the open ended responses suggest there were other barriers to overcome. V.2 - External Barriers Revisited While the participation and satisfaction did increase from V1 to V2, time continued to be a barrier for all participants. Although the changes in the program seemed to impact participation, the fluctuation with the program may be the result of the changing demands on time and energy as the year progresses. For example, in the beginning of the year, much of the time may have been devoted to curriculum preparation these first year teachers were teaching courses for the first time. One teacher comments that “Unfortunately, as school got more stressful, I had less time to try STEMMP. If I had a light teaching load, I would have been more engaged. But, ironically, it might have been less helpful.” For many, the lack of time was intertwined with barriers related to the technology used to support this online community of practice. Due to the limited time and technology constraints at their schools 69%

of the participants engaged with STEMMP at home. However due to dial up access or lack of computers at home, some had to engage during or after school. Many felt that it was more efficient to engage with their mentor on the phone or in email as indicated by Table 3. One participant echoed this as they stated “I just didn't have the extra time to participate in STEMMP. I just wanted to pick up the phone and talk to my mentor, but we were supposed to do everything online which seemed to slow. Maybe I'm old fashioned, but I would rather talk to someone on the phone - I can cover more that way.” Again, these preferred methods of support take away from engagement with the larger community. Table 3: What is your preference for support? Response Mentor/Mentee Email Other Mentor Communication (face to face, phone, etc) Mentor/Mentee Chat room Meetings Mentor/Mentee Forum Messages Small Learning Community (SLC) Forum Messages Face to Face Meetings at TIES Small Learning Community (SLC) Chat Room Meetings

Count 12 9 9 5 5 4 0

Percent 26.1% 19.6% 19.6% 10.9% 10.9% 8.7% 0.0%

The local context in which beginning teachers taught also functioned as an external barrier towards engagement with this community of practice. For mentees with district induction programs, the program duplicated some of the work that they did locally. While preliminary analysis revealed no significant correlation between the number of local supports and the frequency of participation of the beginning teachers, it is certainly a consideration for future program design. When designing comprehensive online induction systems we must account for the individual teachers’ local support systems and contexts towards developing an “a la carte” system that accounts for local contexts so efforts are not duplicated. One of the primary external barriers towards engagement within this online community of practice was an extrinsic motivator for participation; namely credit. Table 4 shows the average level of participation for the different levels of credit. The participants enrolled in the “Credit – Monthly Face to Face Meetings” participated in the same activities as the other two groups except for the PDI activities. Unsurprisingly, having an external reinforcement such as credit tied to the quantity and substance of postings does impact the level at which they engage in the community of practice, especially in the face of limited beginning teacher time and energy. However, this does raise the issue of having some form of external reward for participation, perhaps in the form of a stipend or free credit, as this becomes a costly programmatic venture. Or, in this case, in which participants had to purchase these credits, we must pause to think about their own motivations for engaging in this program. Do they see this as mainly a convenient tool to get credit to advance their salary? Do they see the value in such a community based program in supporting their continued development and the credit is a bonus? Is there value in participation in either case? While these questions are beyond the scope of this paper, it does highlight the importance of making sure that we can align the participant’s expectations with our own expectations of engagement within a community of practice. Table 4: Blog and SLC Posts by Credit Level (Expected 26 posts across program) No Credit Mean SD

6.0 4.4

Credit – Face to Face Hybrid 11.7 4.3

Credit – Fully Online 20.1 8.7

V.2 – Internal Barriers Revisited Participant beliefs about connectedness within in an online environment impacted their engagement. Beyond the efficiency of face to face or phone conversations with mentors, some mentees did not engage because they did not believe that online environments afford community development. One participant noted that “nothing matches or comes close to personal interaction, I’m quite strongly against using technology as a substitute.” This comment reveals that they did not feel a sense of connection towards their mentor or the larger STEMMP community in spite of attempts to build community at the face to face meetings and through the support activities. In order for participants to engage with this community they must feel a sense of trust and that it is a community that is supported through technology. Ice-breakers that occur face to face and online may serve as an important role in building trust and making connections between the members in order for their work to make stronger connections to their practice.

When asked about the most important aspects of the program one mentee stated, “The [Face to Face Meetings were] very encouraging to me. Knowing I was not “alone” in the fight was very useful” While others had similar comments referring to online learning environment, this mentee specifically cited the face to face meeting. Indeed, it was a time of tears frustration and relief for some as we interacted with these beginning teachers and really checked to see if things were going ‘fine’. For some teachers, this deep emotional reaction may not have been possible in an online environment. Online chat rooms and forums miss the subtle voice and body cues that can tell when things are not ‘ok’. This alone may give credence to the important, yet expensive challenge, of connecting teachers occasionally in a face to face environment beyond the potential of the online systems. It is important to note that even though online learning communities, and future versions that connect others through VOIP video conferencing, have power in connecting communities in an efficient way, there is sometime no substitute for handing someone a tissue when they are at the their lowest point. As designers, we cannot forget that this is a ‘community of practice supported though technology’ not a ‘technology based community of practice’. Finally, a less obvious barrier towards engagement was an incompatibility of goals and purposes between the designers and the mentees. Across the participant surveys, the mentors/mentee relationship was seen as the most valuable aspect of the program. Referring back to Table 3 we note that the mentor based supports (76%, n = 35) were the preferred method of support. Open ended responses indicate that the mentees see their mentor relationship as providing timely resources and advice to help them survive by providing curricular activities and “tricks of the trade.” The danger in this thinking is that this puts the mentor/mentee relationship as one of the mentor as the giver and the mentee as the receiver instead of a more cognitive approach that pushes them past the deficit notion of survival towards thriving in their first years of teaching. While these mentor relationships may build short term PCK and may help provide encouragement during the difficult times, care must be taken that their work helps induct these teachers into the communities’ way of thinking and knowing. In addition, there is danger if the advice or curriculum is either non-reform based or is taken out of context. With the heavy reliance on mentors as a primary means of support, it becomes critical to address the selection and professional development of mentors. In particular it is important to find mentors that are subject specific mentors that have the ability to implement reform based science instruction. Extensive training is required in both cognitive/standards based mentoring as well as reform based science instruction. As designers, we must explore and challenge the individual beliefs that our beginning teachers and even mentors bring to the community that may be counter to reform based instruction. If we are to be a true community of practice that connects pre-service, inservice, local and national teachers organizations we must have a common, yet evolving set of goals and aims that is open to thoughtful discourse based in praxis. Great care must be given to the design of online communities in order to help our newest members become full participating members of the community in which they can work to help bring others into the community to reform teaching practices of all. Implications for future designs Our current V.3 design, Project TIN (Science Teacher Induction Network) uses the knowledge of these barriers towards developing a more effective technology enhanced community of practice. It uses more a formative system of mentee self evaluations to guide the supports that the mentees choose. There are tighter context-specific learning communities and additional supports such as a learning object repository with reform based lessons and activities. In addition, we added collaborative technologies to promote the co-development of lessons between mentors and mentees and annotated video technologies to help facilitate discussion of practice. Also, we have changed the mentor/mentee chat rooms to synchronous VOIP chat rooms that function much like a phone call. In addition, because we are no longer STEM-focused but science focused we can tailor the dilemmas and PDI investigations the specific concerns of science teachers. It is critical to keep in mind the external demands placed on beginning teachers as they transition from a relatively safe environment of student teaching towards an environment in which they must perform the same tasks as a 30 year veteran. These demands placed on these teachers weigh heavily on their available time and energy. However, induction programs must be seen by beginning teachers as a help towards their development during these challenging times and not merely a tool for survival. They must see how online communities of practice, in which they learn from and with each other, can meld “theory and practice, doing and reflection, the individual and community, in a manner that transforms all components” (Barab et al., 2002, p. 530) to help them both survive and thrive.

References Barab, S. A., Barnett, M., & Squire, K. (2002). Developing an empirical account of a community of practice: Characterizing the essential tensions. Journal of the Learning Sciences, 11(4), 489-542. Bice, L. (2005) The construction of knowledge about teaching practice and educating students from diverse cultures in an online induction program. Unpublished doctoral dissertation, Montana State University, Bozeman. Collison, G., Elbaum, B., Haavind, S., & Tinker, R. (2000). Facilitating online learning: effective strategies for moderators. Madison, WI: Atwood. Davis, E. A., Petish, D., & Smithey, J. (2006). Challenges New Science Teachers Face. Review of Educational Research, 76(4), 607-651.

Design-Based Research Collective. (2003). Design-Based Research: An Emerging Paradigm for Educational Inquiry. Educational Researcher, 32(1), 5-8. Ertmer, P.A. (1999). Addressing First and Second Order Barriers to Change: Strategies for Technology Integration. Educational Technology Research and Development, 47(4), 47-61. Gordon, S. P. (1991). How to help beginning teachers succeed. Alexandria, VA: Association for Supervision and Curriculum Development. Grimbrug, B. (2006, April). The Structure of Teacher’s Online Discourses. Paper presented at the meeting of the National Association for Research in Science Teaching, San Francisco, CA. Ingersoll, R.M. (2003). Turnover and shortages among science and mathematics teachers in the United States. In J. Rhoton,, & P. Bowers (Eds.), Science teacher retention (pp. 1-12). Arlington, VA: NSTA Press. Knapczyk, D. R., Hew, K. F., Frey, T. J., & Wall-Marencik, W. (2005). Evaluation of online mentoring of practicum for limited licensed teachers. Teacher Education and Special Education, 28(3-4), 207-220. Luft, J. A. (2003). Induction programs for science teachers: What the research says. In J. Rhoton, & P. Bowers (Eds.), Science teacher retention (pp. 35-44). Arlington, VA: NSTA Press. Luft, J. A., Roehrig, G. H. H., & Patterson, N. C. (2003). Contrasting landscapes: A comparison of the impact of different induction programs on beginning secondary science teachers' practices, beliefs, and experiences. Journal of Research in Science Teaching, 40(1), 77-97. Marzano, R.J., Pickering, D.J., & Pollock, J.E. (2001). Classroom instruction that works: Research-based strategies for increasing student achievement. Alexandria, VA: Association for Supervision and Curriculum Development. McAleer, D (2006, April). Coding Dialogue at the Individual Post Level in the Content Oriented Discussion Areas in an Online Mentoring Program. Paper presented at the meeting of the National Association for Research in Science Teaching, San Francisco, CA. Smith, T. M., & Ingersoll, R. M. (2004). What are the effects of induction and mentoring on beginning teacher turnover? American Educational Research Journal, 41(3), 681-714. Taylor, P. & Mike, A. (2006, April). An examination of the effect of facilitation training on the improvement of dialogue quality in the electronic mentoring for student success program: addendum- another year of analysis. Paper presented at the meeting of the National Association for Research in Science Teaching, San Francisco, CA. Wlodkowski, R. J. (1999). Enhancing adult motivation to learn: A comprehensive guide for teaching all adults (rev. ed.). San Francisco: Jossey Bass. Wojnowski, B. S., Bellamy, M. L., & Cooke, S. (2003). A review of literature on the mentoring and induction of beginning teachers with an emphasis on the retention and renewal of science teachers. In J. Rhoton, & P. Bowers (Eds.), Science teacher retention (pp. 22-33). Arlington, VA: NSTA Press.