The Role of Informal Support Networks in Teaching the Nature of Science

This study reports the participation of 13 secondary science teachers in informal support networks and how that participation was associated with their nature of science (NOS) teaching practices 2 to 5 years after having graduated from the same science teacher education program. The nine teachers who participated in informal support networks taught the NOS at high/medium levels, while the four non-participating teachers taught the NOS at low levels. The nine high/medium NOS implementation teachers credited the informal support networks for maintaining/heightening their sense of responsibility for teaching NOS and for helping them navigate institutional constraints that impede effective NOS instruction. Several high/medium NOS instruction implementers initially struggled to autonomously frame and resolve the complexities experienced in schools and thus drew from the support networks to engage in more sophisticated forms of teacher decision-making. In contrast, the NOS pedagogical decisions of the four teachers not participating in support networks were governed primarily by the expectations and constraints experienced in their schools. Implications of this study include the need for reconsidering the structure of teacher mentorship programs to ensure they do not promote archaic science teaching practices that are at odds with reform efforts in science education.

Keywords: Nature of science teaching; Teacher socialization; Mentoring and support

Appeals to accurately and effectively teach the nature of science (NOS) have a long history (Lederman 1992; Matthews 2012) going back more than a century, and persisting as students in the USA are now expected to distinguish science from engineering and learn how both fields work. Improved understanding of particular science ideas, greater interest in science research and careers, and more informed reasoning regarding socioscientific decision-making are just some of the many reasons that promoting NOS understanding remains an important goal for science education (Allchin et al. 2014; Herman 2015; Herman et al. 2013a, b; Hodson 2009). That said, accurate and effective NOS instruction is uncommon in elementary through post-secondary science education (Lederman 2007).

Researchers have investigated many issues regarding NOS teaching and learning to understand why purposeful attention to the NOS is so rare (Abd-El-Khalick et al. 1998; Backhus and Thompson 2006; Herman et al. 2013a, b; Lederman 1999). Scant devotion to accurate and effective NOS instruction is due to several factors, including teachers being inadequately prepared to understand and teach the NOS (Backhus and Thompson 2006), lack of teaching experience, insufficient science and NOS pedagogical content knowledge, not prioritizing NOS as an instructional objective, inability to effectively put into practice general reforms-based science teaching practices, and institutional constraints that downplay and interfere with efforts to accurately teach the NOS (Abd-El-Khalick et al. 1998; Bell et al. 2000; Brickhouse and Bodner 1992; Duschl and Wright 1989; Herman 2015; Herman et al. 2013a, b; Hodson 1993; Lakin and Wellington 1994; Lederman 1999; Schwartz and Lederman 2002).

Becoming an effective science teacher is a long and often bumpy journey extending well beyond the completion of a preparation program, and collegial environments are crucial for teaching success (Anderson and Mitchener 1994; Beck and Kosnik 2001; Coburn and Russell 2008). While teaching has been called the loneliest profession due to teacher isolation from adult interactions for much of their workday, social networks exist in the teaching profession. These include assigned mentors for new teachers, professional learning communities, grade level or subject matter teams, and other formal and less formal structures. Social networks can have multiple purposes as well. Instrumental social networks are focused on achieving organizational ends, and include professional learning communities, teams, and district mentor programs. Another form of network is the expressive social network, which exists to support affective relationships, and is often formed by participants themselves in an informal manner (Moolenaar and Sleegers 2010).

Research on social networks has shown that the presence of strong professionally oriented social networks among teachers within a school are associated with strong perceptions of school climate and school innovation (Moolenaar and Sleegers 2010). However, such collegial environments can be the exception, rather than the norm, for science teachers who attempt to implement reforms-based science instruction. Ihrig et al. (2014) showed that beginning science teachers often face severe constraints and pressures from colleagues and administrators to maintain the status quo in schools, and the devastating consequences on the implementation of research-based science teaching practices, including NOS instruction. Veenman (1984) and McGinnis et al. (2004) used the term reality shock to describe how teachers facing severe constraints often change their personality, attitudes, and teaching practices in ways that are at odds with the research-based practices learned in their teacher education programs, a finding supported by Ihrig et al. (2014). Thus, an institutional requirement to participate in an instrumental social network may become a negative influence on teachers' practices if the group norms are antithetical to reforms-based instruction (Atteberry and Bryk 2010; Daly 2010; Ihrig et al. 2014). When faced with such constraints, teachers can respond in a number of ways, including adopting the group norms, seeking support elsewhere, isolating themselves and keeping their practices from being known by others, or even leaving the school or the profession (Ihrig et al. 2014).

A model developed by Kegan (1994, 2000) assists in framing how adults respond in various ways to the reality shock of problematic situations and social contexts experienced in life, including the workplace (Drago-Severson 2007). Adults primarily fall into one of three progressively sophisticated ways of knowing (cognitively, interpersonally, and intrapersonally) called orders of consciousness. Kegan (2000) and Drago-Severson (2007) describe how individuals at different orders of consciousness negotiate institutional norms and expectations from others (what Kegan refers to as a "subject-object balance"). Individuals at higher orders of consciousness can deliberately contemplate and regulate norms and expectations in their environment in ways that those at lower levels of consciousness cannot. Kegan's model does not predict an individual's behavior in response to conflict, but it does account for how an individual frames the problem. Those at a higher order of consciousness frame problems differently (because they can "own and take them as objects") and thus have different options available to them when planning a course of action.

Table 1, adapted from Drago-Severson's (2007) work on teacher support and development, presents how teachers at second through fourth orders of consciousness frame their profession. Teachers at higher orders exhibit more sophisticated ways of knowing which enable them to balance their concerns in juxtaposition with the concerns of others, and engage in more sophisticated forms of socialization such as critically and objectively analyzing and responding what is requested of them in conjunction with their own values. For instance, fourth-order teachers are not owned by and subject to their environment (e.g., schools) and are able to make self-directed decisions because they can balance the complex demands of teaching through assuming multiple perspectives, managing relationships and stress, enacting constraint-coping strategies, and serving as supportive collaborators. Alternatively, teachers at lower orders of consciousness typically struggle to engage in self-directedness, and their decision-making often relies heavily on the expectations of their environment (e.g., schools) and others (e.g., standards). Progression through the stages presented in Table 1 occurs gradually and an individual may operationalize two ways of knowing simultaneously. For instance, he/she may exhibit third- and fourth-order tendencies. Typically, most adults transition from second to third order by their early twenties, and "at any given time one-half to two-thirds of the adult population would not be classified as having fully reached the fourth order of consciousness" (Kegan, 1994, p. 188, 191).

Three orders of consciousness and teachers' self-engagement at each order (adapted from Drago-Severson 2007)

The aforementioned literature collectively provides a robust theoretical framework for how and why some teachers may navigate the bumpy journey of teaching better than others. For instance, Reiman and DeAngelis Peace (2002) describe how success can be assisted by having teachers analyze social experiences and negotiate the perspectives of self and others. Similarly, Drago-Severson (2007) and Helsing et al. (2004) make clear that the gradual transition from simpler to more sophisticated ways of knowing and responding to complexities depends on the types of support available and challenges experienced. Thus, research has focused on how formal induction programs can assist new teachers to develop cognitive and social frameworks that enable them to navigate constraints and implement research-based practices (Luft et al. 2003; McGinnis et al. 2004; Veenman 1984).

Socialization, constraints, cognitive development, and other factors may account for why few studies report instances of accurate and effective NOS instruction, and the characteristics of teachers who do implement those practices. For instance, Herman et al. (2013a) recently reported that 12 of 13 teachers they studied were overtly teaching the NOS (nine at medium to high levels) 2 to 5 years after completing an intensive research-based science teacher education program. Those teaching the NOS at higher levels were more adept at implementing general reforms-based science teaching practices and capitalized on both planned and spontaneous instances of NOS instruction (Herman et al. 2013b). Furthermore, higher NOS implementers' reflections about their NOS and general science teaching practices were thorough, situated in classroom activities, and accurately addressed what the researchers observed in those teachers' classrooms (Herman et al. 2015). Higher NOS implementers' reflections also indicated the following:

• They grasped key aspects of effective NOS pedagogy including the need to overtly draw students' attention to the NOS, have students think about the NOS (Lederman 2007), and scaffolded across multiple contexts (Allchin et al. 2014; Clough 2006).
• They based instructional decisions on how people learn (e.g., provided concrete representations before abstract ideas) (Bransford et al. 2000).
• They highly valued and had compelling reasons for NOS teaching.

Conversely, lower NOS implementers provided uninspiring rationales for NOS instruction, and their teaching reflections were shallow, often misrepresented their actual practices, and made evident their struggles to effectively teach science and NOS based on how people learn.

The study reported here builds upon this previously published work and explores a phenomenon—the presence of support networks—that we noticed during data collection in our prior studies (Herman et al. 2013a, b, 2015) noted above. We explored this variable to seek greater clarity on why nine of the 13 teachers profiled in the prior research were highly successful at implementing NOS instruction and maintaining the teaching frameworks and values necessary for successful NOS implementation 2 to 5 years after leaving the supportive environment of their science teacher education program. The participants' science education program consists of several required and tightly connected science education courses (including a Nature of Science and Science Education course) employing a highly socialized cohort model. Knowing that strong cohorts rarely occur spontaneously and that teachers often view teaching as an "individualistic affair" (Beck and Kosnik 2001; Lortie 1975), the program faculty overtly model and promote a collaborative approach and emphasize teaching as decision-making (e.g., Clough et al. 2009; Dewey 1929) based on how people learn and science education goals appearing in reform documents (AAAS 1990, 1993; Bransford et al. 2000; NRC 1996).

The Nature of Science and Science Education course promotes teaching and learning the NOS through questions (Clough 2007) rather than tenets so that a more contextual and nuanced understanding of NOS ideas is encouraged. The course also repeatedly draws students' attention to the value of NOS understanding for both teaching and learning science and promoting more informed socioscientific decision-making. The instructor, drawing from NOS pedagogical research and his extensive past experience teaching NOS to secondary school students, emphasizes and overtly models how to effectively teach the NOS primarily in the context of teaching science content. See Clough (1997, 2004, 2006, 2011, and 2015) or contact the third author for more detailed information regarding the NOS course.

None of the participants in the study reported here experienced a formal induction program that promoted NOS instruction or other research-based teaching practices consistent with goals appearing in science education reform documents. Thus, we sought to determine how teachers from a science teacher education program that promotes professional collegiality and support navigated teaching environments often fraught with significant constraints. We learned that study participants stayed in contact with one another and that these connections resulted in the formation of informal support networks. Therefore, our initial two research questions were as follows:

• What is the structure and function of informal support networks that were created by study participants?
• What differences, if any, exist between high/medium and low NOS implementers in their utilization of support networks?

In analyzing data relevant to these two questions, we revisited information that had been collected regarding participants' reflections about teaching and the social strategies they employed to cope with institutional constraints. These variables were examined because during interviews we noticed differences in how participants negotiated the dynamics of their social networks and how those networks were related to their teacher decision-making and managing workplace constraints through various ways of knowing (Drago-Severson 2007; Kegan 1994). Thus, the following research question emerged after having analyzed the findings regarding the first two research questions.

• What relationship, if any, exists between participation in informal support networks and study participants' (a) order of consciousness, (b) sense of responsibility for teaching NOS, and (c) coping strategy used to respond to institutional constraints?

Aligning with prior investigations, this third research question postulates that the sophistication of an individual's way of knowing may operate in tandem with several other factors such as teacher socialization patterns, attitudes, and behaviors to impact NOS teaching practices (Drago-Severson 2007; McGinnis et al. 2004; Reiman and DeAngelis Peace 2002; Veenman 1984).

A majority of studies on social networks in education have focused on within-school or within-district networks and tend to lack information about how networks form and change over time (Coburn et al. 2010). This study examines the influences of networks developed by the teachers themselves that may cross school and district boundaries, and provides insight into the development and changes in these networks over time. Answers to these three questions address whether NOS implementation practices are associated with the use of informal support networks, and what role, if any, these networks and teachers' order of consciousness play in shaping practices—perhaps supporting or eroding what was taught in their teacher education program.

Thirteen experienced secondary science teachers, four women and nine men, agreed to participate in the study (see Table 2). Their teaching experience ranged from 2 to 14 years and none were teaching in schools that expected or promoted attention to the NOS. Of the 13 participants, ten earned their teaching license in a 15-month post-baccalaureate MAT program, two earned their teaching license as undergraduates, and one completed an MS in science education. Participants completed their science education coursework, including a required Nature of Science and Science Education course, in cohort groups, a feature of the science teacher education program that was purposely designed to encourage supportive collaboration among its students within cohorts and between cohorts.

A detailed description of the study participants' NOS understandings reported in Table 2 appears in Herman and Clough (2016). Briefly summarizing, the participants' NOS views were assessed based on their responses to ten modified SUSSI NOS constructs (Liang et al. 2008). Informed ratings indicate that at least 70% of a participant's SUSSI Likert and qualitative responses aligned with the consensus views about the NOS (e.g., McComas et al. 1998; Eflin et al. 1999; Abd-El-Khalick 2012). For instance, informed NOS understandings include rejecting ideas such as that scientific research utilizes a set procedural method, entails objective observations and inferences, and eschews imagination and creativity. Participants' NOS understanding was categorized as naive if at least 70 % of their SUSSI Likert and qualitative responses were fully incongruent with consensus NOS views. Naive NOS understandings include accepting that science proceeds via an objective and set methodological approach devoid of imagination and creativity. The transitional (i.e., ideas have merit) category was used for participants whose NOS understanding fell in the middle ground (e.g., a combination of accurate and inaccurate NOS views) between those rated informed and naive. For instance, transitional NOS understandings may recognize that scientific research not only lacks a rigid scientific method but also presents caveats such as research shares many common steps and methods and science proceeds from observation. One participant's NOS understanding was rated unclassifiable because he completed an insufficient number of SUSSI responses.

A detailed description of the study participants' NOS instructional practices reported in Table 2 appears in Herman et al. (2013a). To summarize here, over a 4-month period, participants' classroom practices were observed and teaching artifacts collected. These data were analyzed using the "Nature of Science Classroom Observation Protocol" (NOS-COP) instrument, and participants' NOS instruction was categorized high, medium, or low. Participants receiving a high NOS implementation rating consistently taught the NOS accurately and in a manner reflecting established science education literature (Abd-El-Khalick and Lederman 2000; Clough 2006; Khishfe and Abd-El-Khalick 2002; Khishfe and Lederman 2006). They addressed NOS ideas through decontextualized activities, moderately contextualized laboratory experiences, and contextualized experiences such as readings and videos that depicted the actual work of scientists. Discussions scaffolded students' thinking across these contexts, and connections to NOS ideas were made through both planned and unplanned opportunities in the classroom. Medium NOS implementers' NOS portrayals were mostly accurate, sporadically planned, rarely spontaneous, and they sometimes struggled to have students deeply reflect about NOS ideas through questions. Low NOS implementers, at best, taught NOS in a haphazard fashion that was almost exclusively devoid of science content. Furthermore, these participants' instruction was rife with NOS inaccuracies and devoid of nuanced reflection about NOS ideas.

Using qualitative methods grounded in a naturalistic inquiry approach with emergent design flexibility (Lincoln and Guba 1985; Maxwell 2005; Miles and Huberman 1994), the study reported here addresses how study participants co-generated and participated in informal support networks, and how those support networks may have impacted their NOS instruction. This examination followed from our prior research (Herman et al. 2013a, b, 2015) discussed above where we initially studied participants' NOS instruction over the course of a semester and factors that might account for their attention to the NOS (e.g., utility value for NOS instruction, NOS PCK, self-reflection abilities, general reforms-based science teaching practices).

Participants were observed teaching science on several occasions over a 4-month period and unstructured interviews conducted prior to and after observing each lesson. A 2-hour semi-structured interview occurred with each individual participant after all classroom observations were completed (Posner and Gertzog 1982). To learn about study participants' support networks and how those networks impacted their teaching practices, we examined all responses to interview questions related to resources participants were using and support they sought for their NOS and general science teaching practices, and we also focused on instances where participants discussed these issues spontaneously or in response to other interview questions. Interview responses were coded and analyzed following a semi-inductive approach (Miles and Huberman 1994). As data collection and analysis proceeded, emergent themes were pursued, including the role of informal support networks that we report here.

We began the in-depth analysis of interview responses using provisional coding schemes that were developed based on data collection and interpretation that occurred during the beginning stages of the study. Through multiple analyses and data triangulation, we modified and refined our coding categories until they were saturated and represented by transparent exemplars. Upon refining the coding categories, each researcher used these coding categories to independently review and analyze the study participants' interview responses, and then cross-compared their coding. This process was repeated until full consensus was achieved regarding the coding of each teacher's interview responses and clear patterns emerged elucidating how informal support networks organically developed (research question 1) and how they were associated with study participants' NOS implementation (research question 2). This same process was then used to analyze study participants' responses to determine what relationship, if any, exists between their informal support network participation and their order of consciousness as described by Kegan (1994) and Drago-Severson (2007), sense of responsibility for teaching NOS, and response to institutional constraints (research question 3). Exemplar quotations from the participants illustrating our assertions are presented in the findings section of this work.

Two to 5 years after graduation, nine of the 13 study participants were in contact with at least one individual from the same program. Figure 1 depicts the structure of the informal support networks used by these nine individuals. The bold arrows represent substantial collaborations that regularly occurred among study participants. For instance, these participants periodically talked in person or on the phone and meet for dinner several times a semester to discuss NOS activities and workplace constraints. Dotted arrows represent collaborations that began during or shortly after the teacher education program, but were only somewhat active at the time of this study. These interactions included sporadic contact to exchange ideas, activities, and borrow materials. Matching numbers after each name indicate the study participants that were in the same school at the time of this investigation. One teacher, Matthew, was in contact with one other person from the program; John, Sharon, and Mark had two; Luke and Carey had three; Peter had four; Isaac had five; and Andrew had six. These collaborations included individuals from different cohort groups and outside the same school.Study participants informal support networks and their NOS implementation levels

PHOTO (COLOR)

The nine study participants involved in informal support networks stated during interviews that they sought to improve their practice and noted that improvement would be promoted by seeking out others who understand reforms-based science instruction. The networks serve a similar function as an instrumental social network, yet the purpose is to improve the teaching of the individuals, rather than the improvement of their institutions more broadly. These relationships appeared to have begun as professional acquaintanceships, but then grew organically to become supportive co-mentorships. Therefore, it appears these teachers initially sought out and were inducted into the informal support networks principally based on the extent they shared common visions for teaching and learning (e.g., a deep and compelling utility value for NOS instruction), not because of pre-existing personal relationships with those already in the networks. For instance, Luke described how his supportive relationship with Andrew evolved as follows:

With Andrew it was much more of a mentor relationship. Not so much now, but that's how it started. Mostly because he was a year behind me [in the science teacher education program]. So when he was in the MAT, he came and saw Mark and I giving talks at conferences and things like that. (Luke, 17:36)

Similarly, Andrew expressed his sentiments about why he continues to collaborate with Luke, Peter, Carey, and Matthew, who all work at different schools, by stating:

I want to help them become the best teacher they can be, and I want to be the best teacher I can be and those two things go hand in hand. If I help them I will become better. So it wasn't so selfish a reason, it is just so we can all become better. (Andrew, 1:53)

Sharon, Isaac, and John—who completed the program at different times, but work in the same large school with other more traditional teachers—also conveyed they sought each other out because of their mutual goals for students and understanding of teaching and learning:

Even if John wasn't my mentor in that official capacity he would be a role model for me. I ask John's advice multiple times a day. (Isaac, 19:35)

I didn't feel that there were any like-minded professionals in the building that I was working at, and so I knew that if I wanted to continue and maintain and build my own strategies I needed to get myself connected with two like-minded people who value certain strategies and teaching like I did. So I found them (John and Isaac) and they couldn't get rid of me. (Sharon, 2:37)

Notably, Sharon had moved from an adjoining middle school where her colleagues were experienced, yet quite traditional, in order to work more closely with Isaac and John. Furthermore, John's official role as an assigned mentor was a secondary consideration for why Isaac and Sharon closely collaborate with him. Rather, their collaborative relationship is valued and successful because they share, as Sharon points out, a like-mindedness for teaching and learning, which appears to require two elements: (1) a shared strong understanding of research-based instruction aligned with science education reform documents and (2) a strong desire to improve.

Four of the 13 study participants chose not to be part of any informal support network, and are considered isolated actors (Daly 2010). Unlike those in the network, these four individuals did not pursue and appeared not to value such collaborations for improving their teaching practices.

I kind of lost touch with most of the graduates. I was pretty much on an island. I felt like I had to just use whatever knowledge and I had and try to remember the ideas that I had from my methods and nature of science courses. (Philip, 24:30)

I don't have the views of Mark. I just don't have the strong views of [others] from the MAT program. (Thomas, 26:34)

The latter quotation is particularly telling because Mark and Thomas taught in the same school just down the hall from one another. But Thomas did not share Mark's respect for and desire to put into place research-based science teaching practices aligned with reform documents.

All study participants taught in schools that did not expect or support accurate and effective NOS instruction. The nine study participants who engaged in the co-generated informal support networks were unanimous in claiming that such participation was crucial for coping with the stresses of teaching and navigating the many institutional constraints that interfered with or were hostile toward accurate and effective NOS instruction, and more general research-based science teaching practices aligned with science education reform documents. For instance, when discussing the impact of constraints without the support of others in their support networks, those participants responded:

Well, I'll be a little blunt. I don't think I would be here. (Andrew, 1:42)

I don't know if I would still be here. I would probably look more like the other teachers here. (Matthew, 22:18)

I think I would be more traditional. (Isaac, 19:33)

Study participants who engaged in an informal support network also credited that network for helping them effectively teach the NOS and science more generally. Informal support network participants provided many examples illustrating how their numerous collaborations influenced their decisions about science and NOS teaching.

Andrew, Carey, Luke, and all those people, and probably even Isaac. It's not a question of whether or not we're going to talk about the nature of science. It's about: How long are we going to talk about it? And what are you going to do this year kind of stuff... I have specifically taken things from other teachers from the program such as Andrew, Carey, and Luke. I take those activities and try as hard as I can to make them fit or change things up. (Peter, 25:15)

Peter, Andrew, and I will sit down and have dinner and just throw nature of science ideas around such as: What are you doing for nature of science? And then we bring those ideas right back to our classroom. So that definitely impacts the way we bring in nature of science into the classroom. And it also impacts how we are bringing inquiry into the classroom... Peter made up a lesson plan for the Gestalt switches and what he does, and I literally used it two days later. (Carey, 21:41)

Well with some of the planning Andrew and I would talk about teaching inquiry and how do you take cookbook labs and make it into an inquiry lab. We were thinking about what our goals are for students, and what actions do we look for...I borrowed some Rutherford trays that I teach NOS with. (Matthew, 22:16)

I think that it helps to have colleagues (John and Sharon) that also appreciate it, that also believe it is important to explicitly and accurately teach the nature of science...[referring to a science department meeting where NOS curriculum and instruction were deeply discussed]. (Isaac, 19:12, 16)

Study participants engaged in the informal support network wanted to improve their science teaching practices and were adamant that obtaining help from others who understood effective science teaching was a critical way to achieve this. This was the case among study participants whose teaching practices were already well-aligned with science education reform documents. For these participants, the social capital resided in the network that extended beyond their school, rather than their colleagues in the school. In other words, the resources that exist in the social relations was seen as stronger in the network, and this became the resource for teaching improvements.

In contrast, the four individuals who were not involved in an informal support network did not appear to understand how such a support network could positively impact their teaching practice. When talking about the informal support network that Mark was involved in, Thomas stated, "If they want to commiserate, that would be great. [But] I am not sure I would've needed them..." (Thomas, 26:40). Mary worked in the same school as Andrew, but said, "...a lot of times I just don't see him that much.... As a whole, [he's] not much of an impact [on my practice]" (Mary, 18:52). Despite her instruction departing significantly from that promoted in science education reform documents, Mary was very confident in her teaching ability and saw little benefit to be gained by working with Andrew or others.

As noted earlier in Table 2, of the 13 study participants, nine were categorized as medium or high NOS instruction implementers and the four remaining participants were categorized as low NOS instruction implementers.

All nine study participants who engaged in informal support networks with others seeking to employ science teaching practices aligned with science education reform documents demonstrated either high or moderate levels of NOS instruction. Figure 1 presents study participants' level of NOS instruction and a map of their engagement in the informal support network. As noted earlier, collaborations occurred between study participants within and across different cohorts and schools (matching numbers after each name indicate the participants were in the same school when this study took place). Solid lines represent informal collaborations that began when or shortly after study participants were in their science teacher education program and were still regularly (e.g., weekly to monthly) utilized at the time of this study. All four high NOS instruction implementers and four of the five medium NOS implementers were still engaged in these types of collaborations with at least one other individual in the network. Those participating in a support network indicated that if these collaborations were stopped, they would have experienced substantial negative impacts on their general teaching practices and their NOS teaching practices. For instance, high and medium NOS implementers had the following to say regarding how their teaching would be impacted in the absence of the support networks:

I think I would've been a hybrid [between reform and traditional practice]. It is really attractive to see these multiple-choice tests, and see teachers done with their grading in 20 minutes. And it is really attractive to see teachers watching students work on worksheets for the last 20 minutes of class. (Luke, 17:30)

But if I didn't have any of that support I would say that probably would still try to teach some nature of science but it probably would've fallen away. And then I would say I would probably become, purely out of time and effort, a more traditional teacher. You need that support system because without it, it all falls away......Because without having that support system I would fall into the trap of okay I'll do just what you guys [my traditional colleagues] do. (Carey, 19:33)

Dotted lines in the figure represent collaborations that began when study participants were in their science teacher education program, continued at least 1 year beyond the program, but at the time of this study (2-5 years after completing their teacher education program) were no longer vibrant and regular. Despite some collaborations having dwindled, study participants claimed those previous interactions early in their career were crucial to their developing NOS and general science teaching practices. For instance, Mark spoke favorably about his early career collaborations with Andrew and Luke:

Just reflecting on how students learn and how you take that into account and philosophically refining your research-based framework in your head. I think the same goes for developing a deep understanding of what the nature of science is and how we implement it in the classroom. (Mark, 27:24)

The four remaining study participants who chose not to participate in the informal support network (shown at the bottom of Fig. 1) implemented NOS instruction at low levels. All four individuals made few attempts to teach the NOS, and when such attempts were made, the portrayals often contained inaccuracies. Thomas and Mary, despite having colleagues from the same science teacher education program in their respective schools who implemented NOS instruction at higher levels, chose not to engage with them.

Based on the iterative design of this study, the findings related to participation in support networks warranted further investigation. The researchers recognized that an informal support network is a reflection of something deeper—in this case, a desire to improve and a perception that finding colleagues who understand effective science teaching is one way to accomplish that goal. For instance, the low implementers in this study could have sought the assistance of others who they perceived shared their understanding of effective science teaching, but chose not to for reasons noted earlier. An ex post facto data analysis in light of our findings regarding research questions 1 and 2 showed that relationships existed between the study participants' participation in informal support networks and their (a) order of consciousness, (b) sense of responsibility for teaching the NOS, and (c) means of coping with institutional constraints.

All participants categorized as operating at the fourth order of consciousness engaged in informal support networks, acted in a manner indicating they were not "owned by" and "subject to" the demands and constraints of their workplace, and maintained a high personal responsibility to implement NOS. Conversely, teachers categorized at lower orders of consciousness appeared to be "subject to" workplace demands and constraints, and struggled to make decisions when value systems conflicted (e.g., when a more senior colleague exerted authority on that teacher to abandon research-based teaching practices in favor of traditional practices). Of crucial importance, participants coded as transitional third/fourth-order consciousness implemented NOS instruction at medium to high levels, but only if they participated in an informal support network of high NOS implementation teachers. The network appeared to provide a source of values and meaning-making that fostered a sense of responsibility to implement NOS and the ability to navigate teaching constraints. Detailed accounts follow regarding the relationship between the study participants' participation in informal support networks and their: (a) order of consciousness, (b) sense of responsibility for teaching the NOS, and (c) means of coping with institutional constraints.

We used the orders of consciousness framework of adult development (Kegan 1994; Lahey et al. 1988) to classify the extent each study participant was self-directed when framing and solving problems regarding teaching and learning. Table 3 presents data for research question three, including participants' order of consciousness. Five of the nine study participants who were involved in an informal support relationship exhibited fourth-order consciousness and the remaining four exhibited third/fourth-order consciousness. Three of the four study participants who did not participate in an informal support relationship exhibited third-order consciousness and the fourth exhibited third/fourth-order consciousness. Among our thirteen study participants, those exhibiting fourth and third/fourth-orders of consciousness appeared more likely to engage in informal support networks. Moreover, a higher order consciousness was associated with NOS instruction implementation.

Three of the four high NOS implementers and two of five medium NOS implementers in our study were coded as operating from a fourth-order consciousness. Because of the way they framed problems and solutions, these individuals worked in a self-directed manner and were not "owned" by and "subject to" the constraints found in their school environments. For instance, Andrew discussed the dissatisfaction he has with his colleagues who teach in archaic ways and expect him to do the same. He then proceeded to reflect on how he views the problem.

So what I have come to realize is this. They [my colleagues] have misconceptions just as students, and I want to help them overcome their misconceptions just like students. I subconsciously assume because they are educated, that they should know better but they don't. So that is a fault of mine. They just haven't had the education that I have and I am lucky in that way. I'm glad I'm not starting where they're starting. But I think I need to help them get to a place where they are moving away from their misconceptions, but it's very emotional for them because they stick a lot of their pride on their professional life and how they teach. So for me to say, "Here, why don't you try it this way" is a subtle message they perceive is "You aren't good enough." I'm not saying anything about who they are as a human being. I'm just saying what they're doing isn't matching with what research is saying is most effective. (Andrew, 1:30)

Indicative of a fourth-order consciousness, Andrew not only recognized the position he took as something he could objectively manipulate, but also provided ways he could do so. He also provided reasons why the colleagues he works with react the way they do, without his being manipulated by those reactions. This statement is representative of the other fourth-order individuals in this study. These individuals use their own system of meaning-making to navigate and solve the many conflicts they face within the teaching profession. These fourth-order individuals may be the only study participants who, without significant support, could overcome the pressures they faced to extensively implement NOS and other teaching practices aligned with science education reform documents.

One high NOS implementer, three medium NOS implementers and one low NOS implementer in our study exhibited traits of both third and fourth-order consciousness and thus were categorized as third/fourth-order consciousness. These teachers alternated between drawing from their own meaning-making and the meaning-making of their workplace and outside sources to make decisions and frame problems. Thus, teachers exhibiting third/fourth order consciousness who partook in informal support networks with individuals valuing and teaching NOS did teach the NOS at medium to high levels. Conversely, the teacher classified as third/fourth order consciousness who did not partake in those informal support networks implemented the NOS at low levels. For instance, Carey exemplifies how teachers between third and fourth order coped with implementing reforms-based practices in a school with fierce institutional constraints:

I really don't talk about things that I've tried that I thought was good but didn't go well with that group [other teachers in her school] because they will say, "See, it doesn't work." So it's not, "Maybe you should try this next time." It's, "See, it didn't work." So you stay as positive as you possibly can which also brings you down because it is hard to keep improving when you can't ask somebody, "How did you do it in this case?" You're not getting that reflection such as, "Oh why don't you try this one? Why don't you try this one?" (Carey, 21:36).

On one hand, Carey frames the relationship she has with colleagues as something she can objectively conceptualize and manage (fourth order) when discussing how she regulates her feelings and limits the amount of information she divulges about teaching. On the other hand, Carey conveys third-order thinking when she reveals her distress over the situation and desire for external sources to frame and solve the problems she is facing.

Thomas, a low NOS implementer, also feels the conflict between what he wants to do and being subject to the rules, opinions, and expectations of the school environment he worked.

It's terrible. Well, this gets to national standards and all that. And I'm not for national standards. I am for a national plan. I love the NSF program that they have, and also the big book—the Atlas [Project 2061]. I love having that, because that is great to use because that is based on research and I think that is good. And I like the freedom to be able to take that and use it the way I want. I do not want people telling me what to do. Teaching is become more and more of that. You just look at the new state standards. I'm not a big fan of it and I can already see it is getting to be a bureaucratic nightmare. (Thomas, 26:42)

He conveys characteristics of fourth-order thinking when he states:

But I am a very independent person and I will just deal with what's right of course and there has to be some sort of moral compass that you have, not even just moral but what you know is the right way to do things. I think a lot of that has to be your guiding principle. Where do you get that I think it comes from you as a person. (Thomas, 26:42)

However, the way he responds to conflict conveys third-order thinking. For instance, when asked how the curriculum standardization has influenced the way he teaches, Thomas replied, "Yeah but if that influence is good, then it's good and that means it's influenced me. If it's a bad influence, it is too bad because it is killing me. It is killing creativity in me". In the latter portion of this statement, Thomas makes clear that he yields with little, if any, resistance to the external pressures he faces, and this was apparent in his teaching practices.

The other low NOS implementers in our study were coded as third order. These individuals' decision-making and problem solving was largely subject to the schools they worked in. For instance, Maddy explained how she would respond to the implementation of a standardized department-wide test at her school—something she claims she is strongly against:

I don't know, just roll with the flow, you can't fight it. It is not worth it to get upset about. That's the job that you are in. I mean we are in an institutionalized teaching job where you are going to be told to do stuff all the time and it is going to change. And if you get into a big huff about it, your job will be miserable. So, roll with the flow. (Maddy, 23:36)

The nine high/moderate NOS instruction study participants embodied Fullan's (1994, 2001) assertion that teachers need to realize the moral and ethical purpose behind their work. These study participants demonstrated this responsibility by expressing frustration at institutional constraints and maintaining in the face of those constraints that NOS and reforms-based teaching practices were just the right thing to do.

That's me, I'm just kind of that way. It is I guess at some point you get so fed up with seeing that teachers continuing to maintain status quo and I care about kids. I am sorry, but what teachers are doing is not good for kids. So I get really upset by that and therefore I emotionally react when people question the things that I'm doing. (Luke, 17:19)

You don't even have to convince yourself [to teach the NOS]. You just have to. The NOS is not being taught out there and it is my responsibility to just kind of do it....for my kids...I'm not teaching for a paycheck, and I'm not teaching for my administration or other teachers and not to impress the people that know me. (Peter, 25:15, 25)

This strong sense of responsibility was promoted in the science teacher education program that all study participants completed, and the co-generated informal support networks appear to have sustained, and for some renewed, that responsibility. The nine study participants who taught the NOS at high/moderate levels claimed that this responsibility was crucial to their NOS and general science teaching effectiveness and manifested in the form of guilt or embarrassment if they were "caught" by others in their support network poorly teaching science and neglecting to teach the NOS. The crucial nature of this positive pressure to implement effective NOS and general science teaching practices is exemplified by Carey:

The professors say they don't want you to punt on the students. It gives you a sense of pressure not to fail the students. You are almost embarrassed if you talk to your classmates [from the science teacher education program] and you are like, "Oh yeah, I had them read a book today because I was exhausted." Whatever it might be, you feel like you're not only letting your own students down but how much you are letting the teacher education program down....I don't know. Maybe it is a support system because I know that maybe it's the embarrassment that if I told Peter or Andrew what I was doing they would be like: Really? You are really doing that? (Carey, 21:31).

Luke more light-heartedly reflected how he would be embarrassed if others from his science teacher education program saw videos of his initial teaching attempts.

It made me reflect and want to improve out of my own competitive spirit. If Andrew was to see one of those videos I would be like, I would be embarrassed for him to see one of those videos! (Luke, 17:39, 40).

Peter reflected about how he persevered, despite experiencing fierce constraints, in his effort to teach science in a manner consistent with science education reform documents due to his connections with Carey and Andrew:

I could not have done it without them (Carey and Andrew) because it is too easy to be an island out here. It is too easy to cave into institutional constraints and have the person signing a paycheck dictating what you do, or having parents who have the ear of the person signing the paycheck determine what you do. If I would not have had them I don't think I would ever be a traditional teacher, but I would not be the teacher I am today. I would not feel guilty about my questioning [pattern]. I wouldn't feel guilty about not teaching the nature of science. (Peter, 25:33)

The previous quotations were from study participants who taught in different schools, but held each other accountable through interactions that included meetings, phone calls and email messages. This was also evident among John, Sharon, and Isaac who taught at the same school:

I would say mostly our discussions maintain a positive pressure on continually justifying what we're doing in relation to how students learn. It's helped maintain pressure on some goals (e.g., NOS) that I think sometimes get lost in the shuffle. (John, 20:20)

I guess I'm questioning myself a lot more now than I did initially because I am examining what I do much more closely. As a group of professionals we push each other and put pressure on each other, when if we didn't have each other that may or may not be happening. (Sharon, 2:36)

They (Sharon and John) expect me to teach in a way that is best for students. ...if my colleague John walked in here and every day I was showing a film or students were reading out of books he would let me have it, and justifiably so. (Isaac, 19:23)

The four study participants who chose not to participate in any informal support relationship did not convey any responsibility to teach science in a manner that promoted accurate NOS understanding and other goals aligned with science education reforms documents. Instead, they referred to these goals and congruent teaching practices that were promoted in their science teacher education program as "too ideal," "challenging," and "unrealistic." Rather than taking responsibility for teaching in such a manner, they instead maintained that factors beyond their control (e.g., time, students' abilities, district/school standards and benchmarks, and colleagues) made implementing such practices impossible. These teachers appeared to easily abdicate their responsibility and quickly succumb without a struggle to common constraints that interfere with efforts to promote needed reform in science teaching and learning.

This sense of time and how much time you can spend on things seem very idealistic in our preservice program. So sometimes you maybe have to find a filler or something for that day that can, it gives you a chance to breathe. There are some days you just put in a movie and you need to just breathe. (Mary, 18:30)

It is kind of content driven and obviously there's not a lot time for the NOS, which I know is a horrible excuse.... Another reason is I kind of teach what everybody else in my department is teaching and I am really trying to do closely with what one of my other colleagues does.... I am not making excuses, but it's just some of things that I just do is what they do.... If I did everything I was taught [in my science education program] to do, I would be dead. I have so much other stuff going on right now. Sometimes you just need the kids to get out their vocabulary charts and make their note cards and their flashcards. Which is maybe not the most effective teaching strategy but you know how that is. You have to concede sometimes or you're going to kill yourself. (Maddy, 23:14)

That is difficult at times, to teach the NOS and inquiry, where you get so caught up in content and you need to cover this and this and this with the state curriculum coming down....I think the content that I teach; it is definitely a memorization oriented course. There is almost no way around it. (Philip, 24:10)

I leave them [the students] behind because I have to move on. I mean there's material we have to cover.... I have been dragged in this but, we have sat down as biology teachers and we have basically very methodically set: "We will cover this, this, and this first semester and we are going to cover this, this, and this and the second semester." (Thomas, 26:29).

All of the participants in this study experienced significant institutional constraints including pressure from administrators and colleagues to superficially cover large amounts of science content, pressure from students and sometimes parents to erode the intellectual level of coursework, and lack of time. However, several study participants worked at the same school, presumably faced similar constraints, yet implemented NOS instruction at different levels. For instance, Andrew (high NOS instruction) and Mary (low NOS instruction) worked at the same school. Mark (medium level NOS instruction) and Thomas (low level NOS instruction) had classrooms in close proximity at another school. John (high level NOS instruction) and Sharon and Isaac (both medium level NOS instruction) also taught in the same school. Therefore, the presence of institutional constraints was not associated with NOS instruction implementation levels.

However, study participants did employ different strategies in response to the institutional constraints they faced. Based on interview responses, we determined that study participants primarily employed one of three coping strategies in response to constraints. We classified these coping strategies as assert, navigate, and concede. The first two strategies were explicitly taught to study participants in their teacher education program to help prepare them to cope with common institutional constraints. Science education faculty overtly addressed how new teachers should navigate potential constraints until they gain experience and credibility. For instance, new teachers might say things to their colleagues and administrators that imply they are conforming to expectations, but cleverly implement NOS instruction and other reforms-based practices in a manner that does not draw attention. After establishing themselves as competent teachers and perhaps also getting beyond their probationary period, new teachers may then be in a position to more overtly address constraints and exhibit (assert) their teaching practices and rationales to others. Acquiescing (concede) was overtly stated to be unprofessional, unethical, and unacceptable because doing so harms children and places teachers on a path of mediocrity that will likely persist. The need for teachers to maintain contact with others who share their understanding of and passion for research-based practices aligned with science education reform documents was emphasized throughout the teacher education program that participants completed, particularly when addressing institutional constraints.

As noted previously in the findings related to research questions one and two, all high and medium NOS implementers indicated that their participation in an informal support network was crucial for dealing with the institutional constraints they faced, implementing NOS and general reforms-based science teaching practices, and remaining in the teaching profession. They were better able to cope with the institutional constraints they faced by implementing coping strategies learned in their science teacher education program. At the time of this study, 2 to 5 years after having completed their science teacher education program, all high NOS instruction implementers and one medium NOS instruction implementer took firm stances against school-based efforts to erode reforms-based and NOS implementation practices and were therefore coded as using the strategy "assert." As illustrated by the following statements, these study participants implement proactive measures that thwart teaching constraints and openly explain their research-based rationale to students, colleagues or parents.

I would have to say, "Okay, I'm going to show you or try to help you understand why I'm doing what I'm doing." And then tell them, "Read this and then get back to me and tell me how that fits in with what you would like me to do." That is what I would like to do after teaching three years. I'm in a position where I can do that. (Matthew, 22:41)

I have had a lot of things like parents, institutional constraints, and colleagues attempt to affect my practices. For instance, I had parents call me and say, "My daughter is getting a B in your class and it is because she doesn't participate in class." And I say, "Okay, and?" They said, "I don't think she should be getting a B because she doesn't participate in class." And I say, "I understand that, but my expectations are very clear that I expect people will participate in class and that you cannot earn an A in my class if you do not participate.... It is much better to learn those lessons now than when they are in high school or college or in their job." I say, "So this is a goal that I have for my students and I think you'd agree that it is an important characteristic for your daughter to have and I am trying to push her and give her those expectations so she develops that." (Luke, 17:60).

We have maintained pressure on this administration in this district for many years and they see the value in what we're trying to do with the kids when they come in and observe us. They see our questioning strategies and see our approach to trying to get the kids to learn things over time and they value it. I am not saying we didn't have to fight for it. I've been called into the principal's office multiple times asking and being asked to back up or provide rationale to what I am doing in my class. Fortunately, I've been aware enough to do that. (John, 20:26)

The context of John's comment is important because at the time of this study, he was Isaac's mentor. Because of this, Isaac could primarily use the "assert" strategy because John minimized many of the constraints and reactions Isaac would otherwise have likely incurred. Furthermore, John appears to support Isaac when he does have to cope with constraints. Therefore, despite being a relatively new teacher, Isaac is in a position few new teachers have and he can more overtly face the institutional constraints he encounters.

The four remaining medium NOS implementers (Sharon, Mark, Peter, and Carey) indicated they primarily cope with institutional constraints by "navigating" around them. For instance, if pressured to teach like their colleagues, they would often "appear" to do so in order to deflect attention, criticism, and more intense pressure to conform. For example, Peter explained how he "appears" to teach highly abstract science content, but is really using instructional time more profitably to teach the NOS.

How do we know at the end of the day that all classrooms are addressing the same thing? Like in our chemistry department people think it is important to draw orbital location and a lecture on filling diagrams and I hate those. I understand the value, but that is because if you do chemistry for a living, to understand how things interact with each other. Otherwise, they're not important. When I had to teach it I spent a week talking about the old guys who figured it out rather than the math behind it, because there's little value behind students learning orbital notation. I feel there is behind knowing how they figured it out. So there are ways to get around it. (Peter: 25:41)

As another example, Carey was teaching a class of students in which student attendance and misbehavior were problems. She received little support from her administration and colleagues, but she navigated this by working to deftly alter lessons that she felt pressure to use so that they were more engaging:

Some of the other issues go away like behavioral issues. Some of the issues like attendance issues. If the kids really know that something really neat or we are in the middle of a project and they will tell me that they don't want to miss this class....This is because they are engaged. (Carey, 21:65)

Three of the teachers rated as "navigators" (Carey, Peter, and Mark) went out of their way to pursue school leadership positions. This was done for political (making clear they wanted to work with others and be part of the school culture) and practical reasons (i.e., change the culture so that institutional constraints would be minimized).

I did apply for a director of the department job, which is a curriculum job. A big reason is to change up the curriculum and make sure that it matches the new state curriculum and get people on board with that. (Carey, 21:60)

In contrast, low NOS instruction implementers rarely worked to overcome institutional constraints. Their "concede" strategy was apparent in multiple statements, some which include the following:

You can't talk at the kids for an hour and a half. You need to throw activities and discussions in, and peer sharing or something like that and that takes time up and it cuts away from the amount of content you can do. I'm not saying that those are not valuable experiences because they should be and they are, but you got to get the content out. So we decided that we need to take it off the [90 minute] blocks so we have time to get the content in.... I kind of teach what everybody else in my department is teaching and I am really trying to do closely with what one of my other colleagues does. Her lessons don't reflect the NOS either. (Maddy, 23:43,62)

With the new state curriculum coming down. There so many concepts there. I think the nature of science and inquiry take a little bit longer to teach sometimes. To get through some materials and with all that coming down I think that is going to be a difficult transition. (Philip, 23:48)

I think the nature of science, it's trying to show the idea that it's every day and you can incorporate some of those ideas into everyday life.....But you just can't get to all that. That is just too much. (Mary, 18:76)

I know there are some things that I did in my NOS course that I haven't actually implemented. My first year I used a ton of stuff straight from my NOS course and that was some, those pictures when you're looking at those drawings, you can see it in different ways and all that stuff. So I used a lot of that and to be honest with you I don't do that as much because that's the part that they [the students] don't take seriously. They think it's just "I'm off on a tangent" or something like that. (Thomas, 26:56)

Unlike many studies of social networks in education that seek to determine impacts of within-organization networks on organizational goals (Atteberry and Bryk 2010; Coburn et al. 2010), this study revealed the presence of instrumental social networks that extend beyond the single school and/or district that serve to overcome organizational goals that conflicted with reforms-based science instruction. The informal support network connections among nine of the 13 study participants grew organically out of their desire to improve their teaching performance and a conviction that surrounding themselves with individuals who understood reforms-based science instruction would promote that end. This "perception that others are similar to you on some consequential dimension" (called homophily) (Coburn et al. 2010, p. 33) was present for the participants in the support network and was cited as a reason for creation of the network. The program completed by study participants was a cohort model and fostered interdependency and collaboration to the extent that graduates had knowledge of how to support one another and who could do so. Thus, not surprising is that the support network included fellow program graduates; these were available individuals who could support their development of expertise. However, unlike the Coburn et al. (2010) study, proximity did not seem to play a role in the formation or sustainability of the network; the networks seen in this study crossed school and even substantial geographic boundaries. Coburn et al. (2010) also found that over time, within-school networks became larger and more diverse when the school emphasized a desired reform, and smaller and less diverse, yet more expert when the school initiative faded, but teachers continued to seek out ways to improve. Consistent with that study, the number of ties in the network decreased over time and the participants increased in expertise.

These informal network connections appear to have played a crucial role in enabling these nine individuals to teach the NOS in spite of institutional constraints that did not support and often worked against accurate and effective NOS instruction. Study participants received ideas, support, protection from constraints, advice, access to more sophisticated ways of knowing, and other benefits from the network and stated that the network connections were crucial to their success at teaching the NOS. Thus, the network supported individual change rather than organizational change. Others have argued that instrumental social networks are marked by lower trust and are less durable than expressive social networks (e.g., Uzzi 1997), but in this study, the networks existed for multiple years (although network change did occur) and participants' comments exhibited a high level of trust in one another.

Our findings illustrate that support networks can function across and within schools to support NOS instruction. However, working within the same school (proximity) does not guarantee that a teacher will be involved in a support network, even if it is available and composed of individuals who understand and implement reforms-based science instruction. Low NOS implementers in our study were not amenable to working with more effective colleagues, did not convey any meaningful sense of responsibility to accurately and effectively teach the NOS, over-assessed their teaching practices, and saw little need to improve their existing teaching practices. Perhaps the gap between the practices of medium/high implementers and low implementers is large enough that low implementers are embarrassed by the thought of being "mentored" by peers from the same program. As Moolenaar and Sleegers (2010) note, trust is an important feature in an instrumental social network. Also likely is that they do not share a vision for research-based teaching and a strong desire to improve their practice, as evidenced by many comments from low NOS implementers—a situation compounded by their over-assessment of their teaching practices.

Of particular interest is the role that support networks potentially play for teachers who exhibit elements of third-order consciousness. As Kegan (1994) and Drago-Severson (2007) note, individuals at a third order of consciousness are highly influenced by the surrounding environment, because they need that environment to support their decision-making. When conflicts arise, third-order individuals will likely conform to the environment in which they find themselves, but this may be mitigated if they have access to a different environment that provides the needed support. Thus, our study participants who operate at a third to third/fourth-order consciousness are particularly vulnerable to institutional constraints, but for those who sought out a more supportive environment (i.e., like-minded individuals in a support relationship), it appears to have enabled them to make decisions that promoted accurate and effective NOS instruction as well as other goals aligned with reforms-based science teaching. This "pulling" function of the network is supported by research on teacher attitudes conducted by Cole and Weinbaum (2010). Study participants who exhibited third or third/fourth order consciousness and were not in a support network did not make decisions that promoted NOS instruction or other goals aligned with science education reform documents. Thus, third and third/fourth order individuals appear particularly in need of an environment that values and promotes NOS teaching and learning.

We are aware that multiple variables are at play in determining whether a teacher will promote the NOS and other goals aligned with science education reform documents. For instance, what would have caused some third to third/fourth-order study participants to seek support for reforms-based science teaching practices, while others did not? This may be a function of a personal commitment to improve (seen in all medium/high NOS instruction implementers), and a deep sense of responsibility to teach NOS (also seen in all medium/high NOS implementers). Perhaps because the four low NOS instruction implementers lacked these two convictions, they saw no need to collaborate with others. Lacking collaborations with others who seek and promote reforms-based science teaching practices, the prevailing school environment became their undoing. They learned to fit in and explained away discrepancies from what they learned about effective teaching by claiming the institutional constraints they faced cannot be overcome and that reforms-based science teaching is too idealistic.

Thus, efforts to change science teaching practices demands more than assisting teachers to understand research-based practices, including NOS content and pedagogy, aligned with science education reform documents. Teachers are in school environments that have issues of authority, procedures, expectations, and power. Those environments necessitate preparing teachers for institutional constraints and how to navigate complex professional relationships. But more is required as evidenced by the four individuals in our study who, despite receiving this preparation, were low NOS instruction implementers. Also crucial is promoting among teachers a deep sense of responsibility to teach the NOS and other goals aligned with science education reform documents. Study participants who taught the NOS at medium or high levels all felt personally responsible to do so. Kegan (1994), Drago-Severson (2007) and others argue that higher orders of consciousness can be explicitly taught and developed over time. This is particularly encouraging given that Olson and Finson (2009) found that most undergraduate teacher education students in their study were not yet at a cognitive level to handle the demands of complex reflection tasks. Perhaps teacher education programs should be giving more attention to the ways in which teacher candidates frame experience and make decisions, and provide supports to enable them to do so in ways that prevent simply conforming to the norms of the school. This is particularly important given the current state of science education (Banilower et al., 2013) and the need for change to occur.

An implication that should not be drawn from this study is that teachers should be assigned to support networks to promote improvement. Forced mentorships and collaborations face myriad problems including often promoting traditional practices over those that are reforms-based and causing mentee teachers a great deal of frustration (Feiman-Nemser 2001; Feiman-Nemser and Parker 1993; Ihrig et al. 2014). Prevailing school environments from which mentors are frequently haphazardly assigned often inculcate thinking and practice aligned with the status quo. Thus, special attention must be afforded to ensuring mentors and collaborators (a) are advocates of reforms-based teaching and learning, (b) understand and support research-based teaching practices aligned with desired ends, and (c) possess sophisticated cognitive and social frameworks for assisting others. The latter is important because teacher-mentors functioning at lower levels of cognitive complexity struggle to facilitate constructive dialog during mentoring discussions and typically promote a singular prescribed approach to teaching (Reiman and DeAngelis Peace 2002; Reiman and Gardner 1995).

In summary, the implications from this study are as follows:

• Teacher education efforts directed at promoting accurate NOS understanding and effective NOS pedagogy are necessary, but insufficient, for promoting such instruction. Also required are promoting a deep sense of responsibility to teach the NOS, preparing teachers for potentially fierce institutional constraints, and addressing the importance of building a support system to assist in NOS teaching efforts.
• Support networks are important for promoting accurate and effective NOS instruction, particularly when teachers find themselves in organizations that do not support such practices and/or for teachers who exhibit elements of third-order thinking.
• Support networks that promote accurate and effective NOS instruction consist of individuals who value such instruction, are committed to teach NOS, understand what such instruction entails, and share a desire to improve their teaching.
• Support networks that effectively promote accurate and effective NOS instruction need not exist within a school, but can exist among teachers from different schools and districts.
• Support networks may promote accurate and effective NOS instruction among individuals, even without concomitant organizational change.
• Because accurate and effective NOS instruction is so rare and not often valued in schools, forced mentorships and collaborations are unlikely to promote, and will most likely hinder, accurate and effective NOS instruction efforts.
• Schools that require mentorship and collaboration among science teachers must ensure that mentors advocate for, or at the very least do not hinder, accurate and effective NOS instruction efforts.

The support networks observed in our study were created by the nine participants who sought others having an understanding of research-based science teaching and learning, including NOS. Our participants spoke frequently in their interviews about their frustrations with colleagues who did not possess fundamental research-based knowledge about teaching such as how people learn, the crucial role of teacher behaviors, and decisions regarding the selection of content and activities that promote deep conceptual understanding. Often, these frustrations were targeted toward colleagues who were assigned as mentors merely because of their longevity and fit with common, but archaic, notions of teaching and learning. Affording authority in this way likely stifles fledgling teachers' efforts to implement research-based practices (Ihrig et al. 2014).

Critics of preservice teacher education programs claim they make little, if any, difference in teacher effectiveness. Reflecting this perspective, alternative licensure pathways that reduce or bypass preservice programs have proliferated (Olson et al. 2015). However, the study reported here as well as other studies (Ihrig et al. 2014) make clear that beginning teachers are often expected to maintain the status quo in schools where they are hired, and are severely admonished to fall in line if they earnestly work to implement research-based teaching practices. The evidence reported here and elsewhere (Herman et al. 2013a, b; Ihrig et al. 2014) makes clear that a well-conceived and implemented preservice teacher education program advances both understanding and implementation of research-based decision-making that promotes deep and robust understanding of science and NOS content as well as other noble goals of education. But more than discipline-specific content knowledge and pedagogical and pedagogical content knowledge are required to effectively teach science. A fervent rationale for desired ends along with an equally intense desire to accurately self-assess and improve practice in the face of intense constraints is required. However, the pervasive schooling system is too often unwelcoming of, and at times hostile toward, change. Thus, the socialization experience of beginning teachers must be framed in ways that enable them to either manipulate the many forces that work against research-based teaching, or have systems in place that support such practices.