Examining Administrators’ Instructional Feedback to High School Math and Science Teachers

Defining Instructional Leadership

In its broadest sense, instructional leadership is described as a set of leadership practices that school administrators take to improve teaching and learning (Grissom & Loeb, 2011; Grissom et al., 2013; Hallinger, 2005; Horng, Klasick, & Loeb, 2010; Murphy, 1988; Neumerski, 2013). Recent conceptions of instructional leadership have noted that this practice involves establishing a clear vision for the school, setting high standards for students and staff, supervising classroom instruction and a school’s instructional program, managing resources, and cultivating relationships with parents and the community (Barth, 1986; Edmonds, 1979; Hallinger & Heck, 1996; Hallinger & Murphy, 1985). More recently, Leithwood and Louis (2012) suggested that instructional leadership also involves facilitating collaborative opportunities between classroom teachers, and, as is implied by the description, has a specific “focus on classroom practice” (p. 6). Furthermore, other research has suggested that school administrators’ managerial actions may be positively associated with improved student achievement and effective leadership practice (Grissom & Loeb, 2011). Within this study, I focus on the principal’s work as a supervisor of classroom instruction, which includes how they observe classroom instruction and provide feedback to classroom teachers about their instructional practice. Indeed, this reflects the view that “the principal is expected to understand the tenets of quality instruction as well as have sufficient knowledge of the curriculum to know that appropriate content is being delivered to all students” (Wahlstrom & Louis, 2008, p. 458).

Supervising Classroom Instruction

Supervision of classroom instruction is often considered a core instructional leadership behavior (Hallinger & Heck, 1996; Leithwood & Louis, 2012; Wahlstrom & Louis, 2008). Numerous models of supervision have been advanced in the literature with each featuring similar characteristics, including an explicit focus on providing classroom teachers with feedback about their instructional practice (Blase & Blase, 2003; Drago-Severson, 2004; Glickman, Gordon, & Ross-Gordon, 2009; P. F. Oliva & Pawlas, 2001; Zepeda, 2012; Zepeda & Kruskamp, 2007). Supervision is assumed to involve frequent observation of classroom teaching, ongoing dialogue with teachers about their instructional practice, and ongoing support for classroom teachers in the form of professional development, coaching, and collaboration. Most of these models presume that the principal is “capable of providing constructive feedback to improve teaching or is able to design a system in which others provide this support” (Wahlstrom & Louis, 2008, p. 458).

Many of the conceptualizations of instructional supervision treat supervisory behaviors generically, focusing primarily on leadership actions that seemingly cut across content areas without describing the nuances that might arise as administrators supervise different subjects. This represents an important disconnect within the literature about the contextual influence that content areas and departments might play, particularly for secondary teachers (Grossman & Stodolsky, 1995; Stodolsky & Grossman, 1995). Moreover, it appears to minimize the unique challenges of supervising instruction in departmentalized school structures that affect secondary school administrators (Siskin, 1991, 1994). Indeed, previous research has highlighted that there are significant differences across content areas which bear on teachers’ work (Ball, 1991; Grossman & Stodolsky, 1995; Siskin, 1994). These differences manifest themselves as “subcultures” (Grossman & Stodolsky, 1995, p. 5). A central assumption within this study is that school administrators must work within and across these subcultures if they are to effectively improve classroom instruction.

Within the educational leadership literature, surprisingly few studies have looked specifically at how school administrators navigate these subcultures within the context of teacher evaluation, particularly at the secondary level. In addition, minimal literature has considered how these contexts might influence feedback that school administrators provide to classroom teachers. Stein and Nelson (2003) noted that the field of educational leadership has not given adequate attention to the development of leadership models that consider the role of content in leadership actions. Such an understanding enables leaders to “know strong instruction when they see it, to encourage it when they don’t, and to set conditions for continuous academic learning among their professional staffs” (Stein & Nelson, 2003, p. 424). Within the context of secondary schools, classroom teachers’ work within subject subcultures bears significantly on their practice (Grossman & Stodolsky, 1995). These cultures influence how and what teachers believe about their content area, their instructional practice, and shape their identity as instructors (Ball, 1991; Drake et al., 2001; Siskin, 1991, 1994; Stodolsky & Grossman, 1995; van Veen et al., 2001). Such subject matter conceptualizations might also influence how teachers receive feedback from school administrators about their instruction and potentially shape their responses to administrators’ feedback, particularly when the administrator does not share a similar conception of the subject they teach.

The Importance of Principal Feedback in Instructional Supervision

An administrator’s ability to provide feedback to classroom teachers about their instruction is a central component of their supervisory practice (Blase & Blase, 2003; Danielson, 2007; Danielson & McGreal, 2000; Denner, Salzman, & Bangert, 2001; Kimball, 2002; Sergiovanni & Starratt, 2007; Stein & Nelson, 2003). Indeed, as Marks and Printy (2003) suggested, much of the instructional leadership literature presumes that school administrators are a “source of educational expertise” (p. 372), and that through their exchanges with classroom teachers they impart greater understanding of the various instructional approaches. This assumption reflects the view that school administrators should have some degree of understanding about the instructional areas that they supervise, how students learn the content, how teachers teach the content, and how teachers learn the content they present (Stein & Nelson, 2003). A complimentary view suggests that part of an administrator’s supervision of instruction must also be to empower teachers to make decisions that are in the interests of students (Hallinger, 1992; Leithwood & Louis, 2012; Sykes, 1990).

The literature about the feedback that administrators offer to teachers suggests that feedback must be provided at regular intervals, focused on specific teacher practices, and oriented toward the teacher’s long-term growth (Blase & Blase, 1999; Danielson, 2007; Glickman, 2002; Mathers, Oliva, & Laine, 2008; P. F. Oliva & Pawlas, 2001; Peterson, 2000). More specifically, feedback must be both formal and informal and may include a combination of modeling, inquiry, and praise. As Glickman (2002) suggests, teacher supervision should provide teachers and their evaluators a forum within which to reflect on their practice and identify aspects of their practice that need to be changed. Likewise, Sergiovanni and Starratt (2007) suggested that supervisory feedback should assist teachers in making decisions about the content of their instruction and their pedagogical behaviors. Feedback that falls short of this may be seen by teachers as “shallow and meaningless comments” that are “devoid of any connection to student learning” (Feeney, 2007, p. 193). Such feedback may, over time, result in diminished teacher capacity (Feeney, 2007).

Numerous studies of principal feedback to classroom teachers have been conducted (Arlestig, 2008; Blase & Blase, 1999; Kelley & Maslow, 2005; Kimball, 2002) Most of these studies indicate that principals spend too little time observing classroom instruction and providing teachers with detailed feedback about their instruction. Blase and Blase (1999) reported that teachers perceived that feedback should focus on observable classroom practices. These include how teachers pose questions to students to prompt thinking, discuss concepts related to the curriculum, and structure activities to promote student learning. Moreover, they believed that the feedback provided should be less about directing or restricting the teacher and more about empowering the teacher to make decisions that serves the needs of the students. Feedback was thus perceived as being important for both supervision and teacher development. Kimball (2002) noted that the credibility of the principal and the alignment that the feedback has with the school’s instructional program both shape the feedback provided to teachers. Moreover, more recent analyses suggested that the amount of time school administrators spend providing feedback to classroom teachers bears significantly on perceptions of their effectiveness as leaders and potentially the student achievement outcomes that a school obtains (Grissom & Loeb, 2011).

Focusing Feedback Through the Use of Relevant Data

The literature concerning the ways in which principals use data in their practice has grown considerably in recent years, particularly given heightened accountability expectations that require principals to inform their practice with data (Datnow, Park, & Wohlstetter, 2007; Marsh, Pane, & Hamilton, 2006; Snipes, Doolittle, & Herlihy, 2002). Lachat and Smith (2005) noted that there is “increasing attention to the need for more systematic uses of data to inform policy, management, and instructional changes that result in higher student achievement” (p. 333). Researchers have suggested that accountability, and its related emphasis on data-driven decision-making, focuses administrators’ attention on improvements in classroom instruction and that certain conditions in schools prompt administrators to use data in their leadership practice (Anderson, Leithwood, & Strauss, 2010; Goldring & Berends, 2009). The assumption in data-driven leadership is that leadership will become more specific—focusing primarily on student learning needs, classrooms, and teacher practice.

To this end, there have been increasing calls for expanding the use of data within the context of teacher evaluations (see, e.g., Bill and Melinda Gates Foundation, 2012; Education First, 2015). Surprisingly, how administrators identify and use data sources within specific content areas to inform the feedback they provide to classroom teachers has not received considerable attention. Much of the recent discussions of principal’s data use is focused broadly at the school level and thus describes the relationship between data use and school improvement (Goldring & Berends, 2009; Wayman, Spring, Lemke, & Lehr, 2012). Earl and Katz (2002) noted that the purpose of using data within this context is to assess existing capacity within schools, identify areas for growth, and develop plans for long-term improvement in student achievement. A few studies have examined school administrators’ uses of informal and formal data in public school settings (Herman, Golan, & Dryfus, 1990). Herman et al. (1990) suggested that administrators may use and misuse data to support instructional improvement. However, their findings did not specifically consider how principals used data in relation to specific content areas for such purposes.

My review of the literature found few studies that have specifically considered how principals use data in relation to feedback they provide to content area teachers. Many studies focus on the ways in which principals work with classroom teachers to select data points to focus on (Abbott, 2008; Breiter & Light, 2006), how they make sense of these data points (Spillane & Miele, 2007), or even learn to use data software (Wayman & Stringfield, 2006). These studies provide ample descriptions of the data use process generally but have not thoroughly anchored such descriptions to specific content areas with which leaders might interact. There are exceptions, of course. For example, one study conducted by Coldren and Spillane (2007) focused on the ways in which an elementary school administrator collected and used data to inform her leadership practice in relation to literacy instruction. Two recent studies focused on the ways in which principals engaged in instructional leadership in relation to math and science, but these studies did not specifically consider how data use occurred (Gutierrez, 2012; Halverson et al., 2011). My review indicates that most of the existing research has tended to emphasize how classroom teachers use data to inform their decisions about specific instructional strategies (Little & Curry, 2009; Young, 2006; Ysseldyke, & Tardew. 2007). What is clear is that when data are infused within the supervisory process and coupled with meaningful feedback provided by principals, it promotes a structure wherein “teachers are more likely to internalize the feedback and make meaningful adjustments to improve their teaching” (Feeney, 2007, p. 195).

There appears to be a particular gap in our current understanding of the way that principals use data acquired within specific content areas to inform the feedback they provide to classroom teachers as part of the teacher evaluation process. Recent analyses suggest that principals rely heavily on observational data to inform their assessment of teaching practice (M. Oliva, Mathers, & Laine, 2009; Goldring et al., 2015). Yet these data have limited value unless successfully paired with appropriate feedback for classroom teachers. Indeed, a recent research brief prepared by researchers at the University of Virginia’s Center for the Advanced Study of Teaching and Learning suggested

Observational data contributes to professional development efforts only if it is shared effectively with teachers. Providing feedback to teachers about the results of their observations and helping them reflect on this feedback in productive ways provide the bridge between knowledge about what matters for students and changes in actual teaching practice. (Stuhlman, Hamre, Downer, & Pinta, 2012, p. 3)

Thus, while numerous models of instructional leadership advocate for the use of data in all aspects of school administrators’ practice, there is relatively little research exploring how data are used within the context of feedback provided to content area classroom teachers about their instructional practice and whether content area differences bear significantly on the feedback provided. Broadly, the existing research suggests that school administrators do not use data extensively when making decisions about human resources (Goldring et al., 2015; Schuermann et al., 2014). Based on the models that currently exist, it appears that most of this feedback is grounded in what school administrators see students doing as opposed to what students are actually learning (i.e., using assessment results and/or other forms of data beyond observations). In short, there appears to be a substantial disconnect in the existing literature between our current understanding of principal’s data use practices and the feedback provided to classroom teachers who are situated in specific content areas. Research in this area seems particularly important in secondary schools, where leaders likely encounter different interpretations of data because of the specialized nature of subjects being supervised and the varying interpretations of data found within them (Coburn & Turner, 2011).

Site Selection

I selected five high schools for this study based on recommendations provided by a state organization that aimed to the improve math and science achievement. The schools listed had each been actively involved with the organization in school-level reforms focused on math and science. The organization worked with these schools to improve the quality of their math and science program, mostly by providing teachers with professional development. Thus, staff within the organization had a working knowledge of high school math and science programs. The selection of the five schools was also based on the efforts described within each school’s improvement plan that were related to the improvement of classroom instruction in math and science. I obtained the school improvement plans from the schools’ websites and corroborated the information through informal conversations with staff at the state-level organization. More specifically, I selected schools for this study that were engaged in a variety of efforts to improve instruction in math and science, ranging from the adoption of a new project-based biomedical/health science curriculum developed by Project Lead the Way to a partnership with Carnegie Learning that introduced before and after school math interventions to intensive professional development facilitated by external professional development providers. Furthermore, I selected schools that were located in different regions of the state, served somewhat different student populations, and used different teacher evaluation models.

Table 1 shows the demographic characteristics of the high schools included in this study. The schools enrolled between 1,381 and 1,960 students in Grades 9 through 12. The majority of the students enrolled at three of the schools were Caucasian, while the majority of students at two of the schools were from racial/ethnic minorities. The percentage of students eligible for free-or-reduced price lunch ranged from 25.5% to 81.5%. The schools also served varying numbers of students who received Special Education support (ranging from 9.0% to 15.9%) and students who are limited English proficient (ranging from 1.2% to 15.1%). The schools also graduated different numbers of students on time. In three of the schools, more than 80% of the students graduated in 4 years. The two remaining schools had much lower graduation rates, ranging from 49.6% to 74.4%. In total, the schools employed between 77 and 119 classroom teachers who had experience ranging from 11.6 to 14.6 years.

OPEN IN VIEWER

Table 1.

Characteristics of the High Schools.

	A	B	C	D	E
Enrollment	1,593	1,601	1,960	1,797	1,381
Student demographics
Asian/Pacific Islander	2.7	0	1.1	13.8	19.4
Black	3.0	1.1	1.6	4.7	21.2
Caucasian	75.7	84.6	13.2	64.9	26.0
Hispanic	16.8	5.4	80.7	13.6	25.0
Multiracial	1.2	8.9	2.6	2.5	6.0
Native American/Alaska Native	0.6	0	0.8	0.5	2.4
Free or reduced-price lunch eligible	28.2	25.5	80.1	40.9	81.5
Students eligible for special education	9.8	9.2	9.7	9.0	15.9
Students with limited English proficiency	3.1	1.2	15.1	3.8	3.8
Students who graduate on time (in 4 years)	82.9	95.3	74.4	88.1	49.6
Total number of classroom teachers	77	84	119	86	82
Average years of experience for teachers	14.7	14.6	12.7	11.6	12.5

Participants

Table 2 shows the number of participants by their roles and demographic characteristics. The study included a total of 51 participants, including 12 school administrators and 39 classroom teachers, with 28 women and 23 men ranging in age from 24 to 60 years included. Participants had, on average, 14.1 years of experience; however, participants ranged from teachers in their first year to veteran educators with 34 years of experience. The majority of the participants had a master’s degree. Similar to the demographics of the state, the majority of the participants were Caucasian. Fourteen of the participants were Hispanic. Two of the administrators in this study had previous experience as math teachers. One administrator had previous experience as a science teacher. The study also included 20 math teachers and 19 science teachers. As part of the recruitment process, I asked school administrators in each school to identify three to four classroom teachers in their math and science department who had been at the school throughout their tenure as the building administrator. Administrators were encouraged to recommend classroom teachers who they had recently observed in their classrooms or with whom they had recently interacted about their performance evaluation. The composition of the sample varied slightly between schools, reflecting the nature of the administrators’ recommendations and the availability of classroom teachers during the researcher’s site visit.

OPEN IN VIEWER

Table 2.

Study Participants by Role and Demographic Characteristics.

	A	B	C	D	E	Total
All study participants	9	11	10	12	9	51
Administrators
All administrators	2	3	3	3	1	12
With math experience	1	0	0	1	0	2
With science experience	0	0	0	1	0	1
Classroom teachers
All classroom teachers	7	8	7	9	8	39
Math	4	3	4	4	5	20
Science	3	5	3	5	3	19
Gender
Female	6	5	4	7	6	28
Male	3	6	6	5	3	23
Average age (years)	42	41	45	44	42	42.8
Average years of experience	16.5	14.6	13.5	12.5	14.1	14.1
Highest degree earned
Bachelors	1	3	3	2	1	10
Graduate degree	8	8	6	9	8	39
Did not report	—	—	1	1	—	2
Race/ethnicity
Asian or Pacific Islander	1	—	—	—	—	1
Black	—	—	—	—	2	2
Caucasian	8	3	6	8	7	32
Hispanic	—	8	3	3	—	14
Native American/Alaska Native	—	—	1	1	—	2

Data Collection

Data collection consisted of semistructured interviews with administrators and classroom teachers in each of the schools. I developed a common interview protocol for school administrators and classroom teachers (see Appendix A for the interview protocols). Existing literature on instructional leadership practices (Hallinger, 2005; Leithwood & Louis, 2012; Neumerski, 2013) and the study’s theoretical perspectives (Grossman & Stodolsky, 1995; Stein & Nelson, 2003) informed the development of the interview protocols I used. The questions included in the protocol sought to illuminate the administrator’s understanding of the content area they supervised, their understanding of the subculture within which that understanding was situated, and the teachers’ perceptions of the feedback they received. In addition, the protocol for administrators sought to identify what feedback administrators provided to classroom teachers in math and science. The protocol for classroom teachers asked them to describe the various ways in which administrators provided them with feedback about their instruction or to recall specific instances when an administrator offered feedback about their practice. The teacher interview protocols were adjusted to reflect the teachers’ specific content area.

The interviews lasted 30 to 45 minutes with each classroom teacher and 50 to 60 minutes with each school administrator. All interviews were recorded using a digital audio recorder and were completed during a 2-day site visit to each of the schools. Prior to analysis, the audio files were sent to a professional transcriptionist for transcription.

Data Analysis

I completed a thematic analysis of the data (Saldaña, 2009). Throughout the analysis process, I used ATLAS.ti 7.0, a qualitative data analysis software package, to systematize and organize the analytic process (Paulus, Lester, & Dempster, 2014). My analysis began by reading and memoing each of the interview transcripts, whereby I noted the salient comments offered by the administrators and teachers related to the supervision of math and science instruction. I focused on those comments that described the feedback provided by the administrators or that characterized the type of feedback that administrators or teachers described as being important. Throughout this phase of the analysis, I sought out instances that were illustrative. I used generic codes at this stage to highlight key segments of the data for further analysis. These codes were used to narrow and organize the data set. As such, the codes I applied generically referred to “administrator comment,” “teacher comment,” or “example of feedback.”

Next, I applied a second round of coding to all the passages that were identified previously. I assigned descriptive codes (Saldaña, 2009) to passages of the data that were directly informed by the theoretical perspectives used in this study. These codes were developed a priori and specifically highlighted the administrators’ understanding of the content area, how they approached or interacted with classroom teachers to provide feedback, how the teachers perceived the administrators provided them with feedback, whether the administrators used data when providing feedback to classroom teachers, and how data were used (see Appendix B for an abbreviated coding scheme). The coding scheme was informed by the literature related to the influence of departments and subject subcultures (e.g., Grossman & Stodolsky, 1995; Siskin, 1991, 1994), the importance of content areas differences with respect to leadership practice (e.g., Burch & Spillane, 2003), descriptions of instructional leadership in practice (e.g., Coldren & Spillane, 2007), broader descriptions of instructional leadership (e.g., Grissom & Loeb, 2011; Hallinger, 2005), and the various components of leadership content knowledge (Stein & Nelson, 2003). After applying these descriptive codes, I then grouped the codes into broad categories that ultimately served as a basis for the final themes. For example, I grouped all the codes related to teacher perceptions of the administrators’ feedback into a “teacher perception” category that ultimately became part of a theme describing how administrators’ feedback focused on basic pedagogical moves. The three themes that were produced from my analysis are discussed in greater detail below.

Theme 1: Administrative Feedback Emphasized Basic Pedagogical Strategies

Administrators provided feedback to classroom teachers that focused on basic pedagogical strategies and sought to deemphasize the unique aspects of math and science content that teachers perceived were important. For example, administrators often equated good pedagogy in math and science with good pedagogy in other subject areas. As one administrator noted when asked how math and science pedagogy might differ from other subject areas, “I think good teaching is universal and it does not matter what subject area I supervise.” Another administrator offered a similar response, “I’d say that there are some moves in the classroom that are pretty common and I don’t think they differ by subject area.” As another administrator noted, “good instruction has clear and meaningful learning targets, a high level of student engagement, on-going assessment. . . . I’m thinking those are non-negotiable for me regardless of the content area.” Across these descriptions, the administrators positioned math and science as being similar to other content areas and in doing so positioned their understanding of the content area as being less important than their understanding of good pedagogical practice. As one administrator who supervised math teachers explained, “I focus mostly on good teaching . . . my observations are more explicit then and I can provide feedback to teachers that is more meaningful to them.” Indeed, several administrators explained that that they focused on pedagogy, as it was one of the few ways they could provide substantive guidance to the teachers they supervised. Overwhelmingly, administrators reported that they were less likely to engage teachers on matters related to their content areas.

The statements offered by administrators contrasted with those offered by math and science teachers, who tended to privilege their content area and seek advice that recognized the unique aspects of the content they taught. Their comments highlighted the specialization that was required to effectively support teachers in math and science. To be effective supervisors, teachers indicated that administrators needed substantial understanding of the content in order to offer meaningful feedback. Their comments often placed their content area at the center of their work and tended to lessen the emphasis on basic pedagogical strategies. For example, a classroom teacher noted,

I’m focused on the sequencing of ideas in ways that consistently build on each other in mathematics. I may use common questioning strategies or assessment techniques that all teachers are being encouraged to use, but it really is less about those techniques than it is about the content that I’m hoping students are getting. For me, teaching is about mathematical thinking. Helping them think through problems using concepts. Helping them know when to use a particular concept or to do a particular calculation. Most administrators I’ve worked with don’t see this as being as important as the other strategies that I’m using. It’s like they focus on what they can see me doing or what they’re familiar with and don’t tend to focus as much on what’s happening behind that. It’s that part behind it though that makes math different. They need to see that more than they do.

A science teacher offered another perspective noting that, “when you teach a class with a lab you have to be thinking not only whether students have the conceptual understanding but also set up the experiments so that it reinforces that understanding.” In both examples, the teachers perceived that their teaching was fundamentally about conveying content-based ideas to students. As such, their pedagogical choices had nuances that they felt administrators should be able to recognize, as without recognizing these nuances administrators would be unable to provide substantive feedback. As another teacher explained, “. . . principals should understand something about the content. The principal needs to understand how my teaching is allowing the students to access the mathematical concepts that they must know.” Collectively, the teachers’ comments highlight the extent to which their content area shaped their belief about effective pedagogy, as well as the role that administrators should take in providing feedback. Consistent with previous research (Ball, 1991; Ball et al., 2008; Drake et al., 2001; Siskin, 1994; Stodolsky & Grossman, 1995; van Veen et al., 2001), the comments offered by teachers tended to emphasize their identity as content experts and their expectation that administrators engage with them in a way that honored their expertise.

Yet when teachers described helpful feedback they received from administrators, their examples often represented the generic types of pedagogical recommendations that they appeared to critique administrators for providing. For example, a veteran mathematics teacher at one school described feedback she received from an administrator who had previous experience in math. As the below example illustrates, while the administrator had experience in the teacher’s content area, their comments were largely focused on pedagogical strategies and not specifically aimed at the teacher’s content understanding. As the teacher reported,

The closest I got was my first or second year here when I had Stanley. He [the school administrator] tended to be a little more practical. He taught IB math. So his was . . . and I can tell you specifically what it was, and it had to do with doing something at the end to pull everything together. Some type of closure. And while I don’t do it every day, he was exactly right. There are some things that, when presented, by the end of the period, even if it’s a worksheet you gave to the kids, you can always go around and take a quick look at the worksheet to see where they’re at and see if you actually got somewhere. Or, throw questions out right at the end of the period. That was the only really valuable feedback that I feel I’ve gotten since I’ve been here.

A teacher at a different school offered a similar assessment, recalling an exchange with an administrator early in her career. She noted that this particular administrator provided detailed feedback that focused on “all aspects of her teaching.” Yet this administrator’s examples offered no direct connection to the teacher’s content area and much as the previous example highlighted, suggested that the feedback was mostly focused on pedagogy. As the teacher explained,

Well it was one administrator and it was early on in my career, she actually gave me concrete feedback on all aspects of teaching. She was frequently in my room doing informal observations; she worried less about the formal process than just providing me with feedback that was guiding me and helpful and every time she would leave my classroom, I can still see these little sticky notes in her handwriting . . . just a couple little things, you know, some feel goods and a suggestion. That was really helpful because I felt like it was a second set of eyes. Some of her comments were, you know, gosh, the students were on task, I asked a couple students questions about the material, they demonstrated their understanding and you might try this as you transition into the next part of your activity or something like that, so very concrete, targeted specific . . . it was very helpful.

What is striking about each of these quotes is that the feedback provided to the teachers was not about their content area but was instead primarily about their pedagogical practices. In both cases, the examples illustrate that while teachers may want specific feedback that aligns with their content area, they may not be completely dissatisfied when the feedback provided to them is more specifically focused on their pedagogical practices than their content area. Furthermore, the teachers’ comments illustrate that even when an administrator has knowledge of the content area they may not provide feedback differently.

Theme 2: Administrators’ Feedback Anchored in Their Past Experiences as Teachers

Probing further into the administrators’ comments about their feedback, I sought to understand what influenced the feedback the administrators provided. Much as teachers’ views about feedback were anchored in their content area experience, I found that administrators’ views about math and science pedagogy were often anchored in their past experience as classroom teachers. This experience informed what the administrators looked for in the classrooms they supervised and how they presented their comments to classroom teachers. For example, one administrator noted,

As a former English teacher, I often use my own experience in the classroom to determine what to look for . . . good instruction is essentially the same as what I would do in my classroom and so I am looking for things that I would be doing.

Another administrator noted that his experience in social studies served as an important framework from which to provide meaningful feedback to teachers he supervised. As he noted, “You go into those conversations knowing your own content area. . . . Feedback you provide [to teachers] is shaped by that knowledge and so it reflects the kinds of things you recommend to other teachers.” As this comment highlights, he clearly viewed his experience and expertise as both informing the feedback he provided to classroom teachers he supervised, as well as serving as a source of “knowledge” that could be used to understand and identify good pedagogical practices. As this administrator further explained, “You enter classrooms as an administrator having had specific experiences before that influence what you see and how you see it. When I observe a math class, I still have my social studies teacher hat on.” These comments were not dissimilar from other administrators I interviewed and suggest that the administrators often bound their feedback within the context of their own content area. The administrators’ comments suggest that they approached their supervision of instruction by drawing upon their content area expertise—even when this content area was not math or science. In effect, their own subject subculture became a basis for their instructional leadership.

I found that administrators who had previous experience in math and science tended to offer slightly different views from those who were not previously teachers in these subjects. Yet their views about instruction were still bound within their content area. What distinguished their views from other administrators was the perception that their comments had “credibility” based on their previous experiences. The administrators with math and science experience perceived that this credibility gave them the ability to engage with teachers about the content they were presenting at a level that was more detailed than administrators without math or science experience. As one administrator with math experience explained:

For a principal, you know, the one thing with math and science, if you want credibility with math and science teachers, I think you get more credibility in your feedback if you have taught math and science classes, I really do. I think this is much truer for math or science than like for history or English. . . . I think for math and science, the teachers really value somebody who’s actually taught their classes before. I think they take feedback more explicitly and believe what you’re saying if you actually have taught it. I had this experience, I would go into a classroom and I started talking about a bunch of advanced biology or chemistry stuff, right, then all of a sudden it’s like the teacher believes you because you know their content, they trust that what you’re saying, that it isn’t just, you know, coming in and observing. They actually believe what you’re saying because you can actually speak to it in a level of detail.

The administrator’s comments illustrate both the importance of engaging with classroom teachers on a detailed level, as well as interacting with them as knowledgeable colleagues. The administrator felt he had “credibility” with math and science teachers because he had formerly taught the classes and also could speak with them at a level of detail since he knew the content. As an example, the administrator noted his recent experience with a classroom teacher who he observed as part of the district’s formal evaluation process, sharing:

I went into the room and did a formal observation. . . . It was part of the evaluation system and I was there to gather data for the evaluation to complete the evaluation form. But as I watched, I kept noticing that the teacher was teaching, she was a Calculus teacher, and she was trying to teach the students about trigonometric functions and derivatives. It was clear that no matter how she explained it to the students she was not connecting with them. . . . So I went back in after the class and then explained to her how I would re-teach the concept the next day. I explained that it was not uncommon for students to get hung upon on the differences between sines and cosines and that when I taught the class I’d often need to do this.

As the administrator’s comments highlight, his knowledge of the subject matter enabled him to offer guidance about how he would teach a specific concept. The other administrator with a math background offered comments that were similar, noting that her experience as a math teacher afforded her the opportunity to engage with teachers in these departments on substantive issues that other administrators could not.

While administrators with math and science expertise perceived that they were better able to interact with classroom teachers about their content area, the teachers did not offer comments that confirmed this perspective. Rather, teachers perceived that administrators with math and science experience often enforced their own views about instruction. When asked whether it was helpful having an administrator with math experience as her supervisor, one teacher described it was both helpful and unhelpful, noting:

My assistant principal has math experience. Yes, I think that’s right, she has a math endorsement or math minor, so that means that some of her suggestions are based on what thinks should happen. You know, I was teaching like this and so I expect you to do this, too. That sort of thing. Now, she hasn’t taught in years and doesn’t participate in our PLCs or department meetings or really see herself as a math teacher anymore. She is very much an administrator and so her feedback . . . focuses on what the district wants us to do. She doesn’t, like, provide feedback that would be something like a colleague in the department would offer me. So, I guess, your question about whether it’s helpful. Well, it’s not unhelpful. It’s not helpful, though, either. It’s not like she providing me the same kind of feedback that my department chair does. I don’t go to her when there is a problem that I need help with.

As the teacher’s comments highlight, even while the administrator had experience in math, the teacher did not view her feedback as being more helpful than other feedbacks she received from colleagues. In fact, the teacher’s comment that the administrator provided feedback based on what she thought should happen suggests that, in some ways, the feedback was actually more of a reflection of the administrator’s past experience. What these comments illustrate, then, is that even when an administrator comes from the same subject subculture as the teachers he or she supervises, teachers may still resist the feedback they provide.

Theme 3: Seeking Ways to Make Administrators’ Feedback Relevant

While each of the administrators I interviewed perceived that providing feedback to classroom teachers was an important part of their instructional leadership and teachers generally agreed with this view, most of the administrators struggled to find ways to make the feedback they provided relevant to the teachers they supervised. Administrators saw various types of data—including student achievement data, classroom-based assessments, and examples of student work—as one possible way to create opportunities to dialogue with classroom teachers about their instruction. Most of the administrators, however, indicated that they relied primarily on data they collected during classroom observations. The data they collected referred to classroom conditions, teacher or student behaviors, or instructional moves that were observable. When asked to provide examples, principals often described teachers’ questioning strategies, student behavior, on-task behaviors, and in one case, the orderliness and cleanliness of the classroom itself. As one administrator explained, “I typically focus on the big picture when I do an unannounced observation. . . . I only get more specific when it’s a formal observation and I have asked the teacher what to look at.” As this statement highlights, what administrators looked for during the course of the observations differed depending on whether the observation was announced or unannounced.

Despite relying mostly on classroom observations, each of the administrators signaled that they were interested in providing more substantive feedback to the teachers they supervised. I found this true of all the administrators regardless of their background in math or science. Specifically, the administrators saw recently enacted statewide criteria for teacher evaluation, which increased the evaluation procedure’s emphasis on teacher growth, as being a positive driver for changes in their evaluation practice. For example, one administrator, who did not have previous experience in math or science, was actively rethinking his approach to providing feedback to teachers. He noted that

the shift toward new evaluation criteria for classroom teachers is really causing me to rethink how I provide feedback to them. The criteria are still sort of general but I feel like I need to have a better understanding of the instruction and better grasp on what the teachers are doing.

Asked to describe how he was rethinking the feedback he provided, the administrator recognized both the importance of using data and differentiating the feedback provided to meet the needs of each teacher, sharing:

One thing that I have done is to expand the observation process to include discussions with teachers about their assessment data. Our district has invested a lot in end of unit assessments and end of course assessments and so I am trying to think about what it means to include that in my conversations with teachers. You know, basically asking them to explain their work using the data. So, like, when I meet with the Algebra teachers I ask them to bring their end of unit assessment data and we look at the results and we dialogue about what works and what doesn’t work and what they will do. I care about what they will do because that’s what will help us get the results. But I also know that I can’t keep saying the same things to every teacher because each teacher needs specific feedback from me about what’s working and what’s not working. . . . It’s really like trying to target what they are doing.

As the administrator’s comment suggests, he was intentionally using student achievement data to help teachers explain what they were doing in their classrooms to help students achieve and to specifically target aspects of their instruction that he wanted further information about. This approach lessened his need to understand the specific content area and increased his ability to question and support teachers on specific instructional strategies. As the administrator suggested, “I use the data to help the teacher translate from their specific content area to a more general conversation around pedagogy.” A teacher at the school offered comments that suggested that this practice was helping her have more meaningful conversations with the administrator about pedagogy. She stated,

One thing I’ve noticed is that he is using data in our conversations a lot more. He is sort of using the data to help us have a shared understanding or shared language or whatever. So, for example, he met with us about two weeks ago at the end of the unit and we looked over the assessment results and we all shared what we were doing to help kids. And he sort of pushed us to think, you know, about whether that would really help kids understand. He really doesn’t get in specifics like questioning whether the concepts relate or whatever but he does use data to ask about how we are supporting the students. . . . This is the first high school I’ve worked in where the principal has done this and it has been helpful.

As the teacher’s comments suggest, the administrator’s use of data was unique and provided him with a means to question teachers about the support they were providing to students who were experiencing difficulty in the units.