Abstract
Informal mentors likely play a substantial role in novice teacher learning, yet we know little about them, especially in relation to formal mentoring, which is the cornerstone to most induction programs. This study analyzes survey and interview data from 57 first-year mathematics teachers from 11 districts to investigate differences in the characteristics of formal and informal mentoring that can inform improvements in mentoring policy. Our findings suggest that informal and formal mentors sometimes serve similar functions but often provide compensatory and complementary support. Based on these findings, we identify a set of policy recommendations to improve new teacher supports.
Identifying ways to improve novice teacher support and learning is essential. First-year teachers comprise nearly 20% of the teaching workforce (Ingersoll & Merrill, 2010), with no signs that this trend will change anytime soon. Induction is the primary mechanism for new teacher support and learning, with formal mentoring, where a mentor is assigned by the school, district, or state, being by far the most common form of induction (Smith & Ingersoll, 2004). Nearly every state has a teacher mentoring policy, and many districts and schools provide their own additional mentoring programs (Rockoff, 2008).
Formal mentoring has been the focus of a substantial amount of case study and correlational research as well as a recent randomized control trial (Glazerman et al., 2010). These studies have identified key features of mentoring thought to improve teacher confidence, knowledge, and instruction; raise student achievement; and increase retention (e.g., Ingersoll & Kralik, 2004; S. M. Johnson & Birkeland, 2003; Kapadia & Coca, 2007; Stanulis & Floden, 2009; Stanulis, Little, & Wibbens, 2012; Villar & Strong, 2007). A recent review by Ingersoll and Strong (2011) found an overall support for links between induction and improved practice, retention, and student achievement, though there were some exceptions to these positive findings (e.g., Glazerman et al., 2010).
A missing perspective in many of these studies is an explicit consideration that novice teachers obtain mentoring from others besides those formally assigned to them—These informal mentors are people whom new teachers themselves choose to go to for help. The literature on informal teacher interactions, writ large, suggests that collegial exchanges can improve teachers’ work lives and increase retention and are critical to teacher learning (S. M. Johnson & Birkeland, 2003; Yee, 1990); yet, policy tends to ignore or underemphasize them. As Coburn (2001) states, “informal networks among teachers are largely unacknowledged by the policy world. Yet they have enormous potential to play an influential role in teacher sensemaking” (p. 163).
Informal mentors likely play a substantial role in novice teacher learning, yet we know little about them, especially in relation to formal mentors. No research that we know of directly compares novice teachers’ formal and informal mentoring. Several studies gather data on formal and informal mentors but do not allow simultaneous comparisons of formal and informal mentoring for the same teacher (e.g., Juarez-Torres, Hurst, & Hurst, 2007; Klug & Salzman, 1991).
In this analysis, we investigate differences in formal and informal mentoring that could inform improvements in mentoring policy. How do teachers balance the roles of formal and informal mentors? Do teachers choose informal mentors to compensate for the knowledge, skills, and supports that their formal mentors do not offer? We hypothesize a compensatory relationship, where informal mentors fulfill roles and functions not fulfilled by the formal mentor.
Knowledge of informal mentoring is important for explaining the success or failure of induction policies because informal interactions occur within the context of a broad array of novice teacher induction supports. Insights into how formal and informal mentoring compare could also help shape school and district mentoring programs to capitalize on the unique advantages of each, and thus maximize the impact of both types of mentoring on novice teachers.
Research Questions
We ground our research questions in a conceptual framework that identifies key characteristics of mentors and mentoring and suggests how informal and formal mentoring may differ. Our analysis answers the following questions:
Conceptual Framework
The literature on professional development and formal mentoring provides our conceptual framework for identifying the important characteristics and functions of mentors. We base our research questions, measures, analyses, and interpretations on this framework. The framework has three central dimensions: (a) characteristics of mentors, (b) the nature of teacher learning interactions with mentors, and (c) the quality of the mentor. We draw on the limited literature on informal mentoring to help shape our compensatory hypotheses about differences between formal and informal mentoring.
Undergirding our framework is the importance of subject-specific mentoring—specifically, the idea that mathematics content knowledge plays a critical role in math teacher effectiveness. As such, in each of the three areas that comprise our framework, we examine the role of subject-area emphasis, studying how formal and informal mentors differ in their math teaching experience (characteristics of mentors), their focus on math-specific activities (nature of teacher learning interactions), and their knowledge of mathematics (quality of mentors).
Characteristics of Mentors: Subject-Area Expertise, Local Knowledge, Location, and Time
Previous research has shown that having a mentor with subject expertise in the novice teacher’s field tends to produce mentoring that has more positive effects on teacher satisfaction and increases retention (Grossman & Thompson, 2004; Smith & Ingersoll, 2004; Youngs, 2007a). Mentors with mathematics expertise are more likely to be able to work with the novice teacher on multiple aspects of her mathematics instruction, including the amount and type of topic coverage, emphasis on cognitive demands (e.g., memorization vs. problem solving), quality of discussion, rigor of tasks, identification and treatment of student mistakes, and quality of student assignments (Ball, 1991; Carpenter, Fennema, & Franke, 1996; Hiebert et al., 1996). One of the critical goals of mentoring mathematics teachers is building both knowledge of mathematics and how to teach the subject (Ball, Sleep, Boerst, & Bass, 2009), which a mentor skilled in mathematics teaching is usually best able to provide.
Similarly, having a mentor who has experience in the novice teacher’s school, with the particular population of students, or with similar students in the same grade, has been shown to foster more productive mentoring relationships (Kilburg & Hancock, 2006; Rockoff, 2008). This congruence between mentor and novice teacher has been referred to as mentor–mentee match (Hobson, Ashby, Malderez, & Tomlinson, 2009).
Organizational characteristics also matter (Achinstein & Barrett, 2004). Locating mentors and mentees in the same school has been shown to be important for the quality of mentoring (Smith, 2007). Shared location matters because it provides more opportunities for the mentor and mentee to interact, and it also improves the match qualities mentioned above, such as knowing the school and students. Similarly, if mentors have ample time to meet with novice teachers, usually because they are given release time for mentoring activities, they are likely to spend more time with them, which increases opportunities for productive interactions.
Although not a mentor trait, another organizational characteristic that has been shown to influence the extent and nature of mentoring, as well as novice teachers’ success, is the level of challenge of their teaching assignment. This aspect of organizational context is largely influenced by student characteristics such as the extent to which they are English Language Learners or traditionally low-achieving students (e.g., M. L. Stein et al., 2008).
Nature of Interactions: Duration, Content Focus, and Active Learning
Duration, content, and active learning emerge repeatedly as aspects of teacher learning that are precursors to improving teacher knowledge, instruction, and student achievement (Desimone, 2009; Hochberg & Desimone, 2010; Penuel, Fishman, Yamaguchi, & Gallagher, 2007). Teacher learning activities need to occur for a sufficient number of hours—hypothesized to be a minimum of between 20 and 40 hr—and sustained over time across the year, to allow for feedback, follow-up, experimentation, and trial and error (Yoon, Duncan, Lee, Scarloss, & Shapley, 2007).
Furthermore, the importance of activities focused on subject-matter content, including how to teach the content and how students learn the content, has been emphasized in research on teaching and learning in mathematics (Ball, 2000; Valencia, Martin, Place, & Grossman, 2009). Activities that focus on mathematics lessons, activities, and how to teach particular content have been shown to be more effective than non-subject-specific instructional activities (e.g., Desimone, Smith, & Phillips, 2013; Garet, Porter, Desimone, Birman, & Yoon, 2001; Penuel et al., 2007).
However, for novice teachers, there is a debate about whether learning activities should initially focus on subject-matter instruction (e.g., Luft, Roehrig, & Patterson, 2003); some argue that novice teachers are overwhelmed by logistical issues in their first year and need supports to help them master emotional, organizational, and management issues before they can focus on instruction (e.g., Grossman & Thompson, 2004). We collect data on time spent with mentors focused on content-based, instruction-related activities as well as emotional/logistical activities, so that we can examine how this contrast plays out between formal and informal mentors.
Active learning, as opposed to more passive formats such as lecture or sharing materials, is another dimension that is linked to better effects on teachers (Desimone, Porter, Garet, Yoon, & Birman, 2002; Garet et al., 2001). Active learning activities give teachers opportunities to explore ideas, ask questions, and try out strategies and receive feedback on them, none of which are usually possible with passive forms of learning. Many types of activities fall under the rubric of “active learning,” including a mentor observing a novice teacher and providing feedback, a novice teacher observing a mentor, a novice teacher participating with a mentor in a math activity, a mentor and mentee co-planning a lesson, and a mentor and mentee analyzing student work together (e.g., Jeanpierre, Oberhauser, & Freeman, 2005; C. Johnson, Kahle, & Fargo, 2007; Penuel et al., 2007).
Quality of the Mentor: Knowledge and Mentoring Ability
A third dimension of our conceptual framework is teachers’ perception of the quality of their mentors. Although teacher perceptions may differ from those of third-party observers (e.g., Tillema, 2009), how teachers feel about their mentor has been shown to relate to how often they interact with them, the types of issues they discuss with them, and how they judge the mentors’ effectiveness (Awaya et al., 2003; Lofstrom & Eisenschmidt, 2009). Especially important is the novice teacher’s perception of the mentor’s mentoring ability, knowledge of math, and knowledge of teaching (Achinstein & Barrett, 2004; Feiman-Nemser, 2001; Wang, Odell, & Schwille, 2008). In mathematics, as in other fields, subject-matter knowledge is distinct from knowledge for teaching (e.g., Ball, 1990, 1991).
The Role of Subject-Specific Knowledge in Mathematics Teacher Learning
There is a considerable literature documenting the strong and consistent relationships between math teacher effectiveness and teachers’ deep understanding of mathematics and how to teach it (e.g., Carpenter, Fennema, Peterson, Chiang, & Loef, 1989; Hiebert et al., 1997). Teachers with deeper content knowledge of mathematics and how students learn math have been shown to be more effective at teaching for understanding, including presenting problems with multiple and unconventional solutions, identifying and analyzing student mistakes, providing comprehensible explanations, and selecting appropriate representations of mathematical symbols and content (Ball, Lubienski, & Mewborn, 2001; Hill, Ball, Blunk, Goffney, & Rowan, 2007; Hill, Schilling, & Ball, 2004). Furthermore, teachers with more math knowledge have been shown to focus more on higher order cognitive demands (Charalambous, 2010), go deeper into fewer topic areas (Leinhardt & Smith, 1985), better understand students’ methods, and better grasp the structures underlying mathematics and how they connect (e.g., Ball, 1993; Carpenter et al., 1989; Ma, 1999; Thompson & Thompson, 1994). In contrast, teachers without deep math content knowledge tend to offer more fact-based lessons and routinized procedural work (Fennema & Franke, 1992), miss opportunities to make important connections (M. K. Stein, Baxter, & Leinhardt, 1990), and provide few meaningful opportunities for questioning, arguing, and explaining (Borko et al., 1992; Cohen, 1990).
Given that many teachers come to the classroom unprepared for high-quality math teaching, we believe that mentoring can play a significant role in building math teachers’ knowledge and practice. We hypothesize that orienting mentoring activities to focus on math content and how students learn that content has the potential to serve as a pivotal policy mechanism for teacher learning. Furthermore, considering that a primary reason new teachers leave the profession is because they do not believe their instruction is helping students learn (S. M. Johnson & Birkeland, 2003), we believe that making content the focus of mentoring and other induction activities could help retain successful teachers. Thus, while our framework provides an integrated reflection of critical aspects of mentoring, we are especially interested in understanding how informal and formal mentors differ in how they help novice teachers build their mathematics knowledge.
How Do Formal and Informal Mentoring Differ? Exploring a Compensatory Hypothesis Framework
Our analysis is grounded in the supposition that novice teachers interact differently with formal mentors than with informal mentors. The dimensions of informal teacher relationships that have been shown to be important include the structure of the relationship, the level of expertise, and the content of interactions (Coburn & Russell, 2008). Here, we discuss how we expect these dimensions to differ between formal and informal mentors.
As noted previously, formal mentors are assigned by the principal, district, or state. This process sometimes prevents matching mentors and mentees on important dimensions such as subject and grade-level experience or experience with the same population of students. Thus, it is reasonable to believe that teachers might seek out informal mentors who have experience with the students they are teaching, the particular curriculum, the grade level, and so forth. The idea of teachers trying to identify colleagues with similar experiences is consistent with Coburn’s (2001) conclusion, based on her case study of an elementary school, that “most teachers sought out like-minded colleagues to talk about their classrooms” (p. 158).
Furthermore, Coburn (2001) characterized conversations among teachers as being either out-facing, meaning focused on the administration and the district, or in-facing, emphasizing matters closer to practice (p. 158). We hypothesize that conversations with formal mentors outside of the school are more out-facing and those with informal mentors more in-facing. One reason to expect this result is because formal mentors outside of the school may provide a safe place for novice teachers to express any fears or frustrations they may have about the administration, which they do not feel comfortable expressing to their informal (or formal) mentors inside their school (Tillman, 2000).
We also hypothesize that informal mentors focus more on social/emotional issues than formal mentors. Emotional support can boost confidence of beginning teachers, increase their morale and job satisfaction, and reduce their feelings of isolation (DeWert, Babinski, & Jones, 2003). Formal mentor assignments are rarely able to consider personality and temperament, which can play a critical role in the effectiveness of a mentoring relationship (Engstrom, 2004). In contrast, novice teachers may choose informal mentors based on their perceptions of similar or complementary social or emotional characteristics, which may allow different types of connections, trust, and interactions (Bryk & Schneider, 2003).
In essence, we hypothesize a compensatory model, where novice teachers seek the help of informal mentors to compensate for what they are not receiving from their formal mentors, whether that is emphasis on emotional support or some other aspect of support they are not receiving from their formal mentors.
Conceptualizing and Defining Informal Mentors: Who Counts as an Informal Mentor?
We distinguish formal from informal in a way that is consistent with vocational, business, and industrial psychology literature, wherein “informal mentors are distinct from formal mentors in that the duties and personnel are not assigned by the organization” (Pollock, 1995, p. 144). Still, there are additional complications of defining “informal mentor” in the context of novice classroom teachers. Novice teachers interact with a host of educators, including their principal, other teachers in or outside of their school, math coaches, and so on.
Are there meaningful distinctions among these helpful colleagues, co-teachers, math coaches, principals who provide mentoring, and informal mentors? Obviously, within any of these dyads, mentoring relationships can form. For the purposes of our study, we rely on the novice teacher to arbitrate whether a helpful colleague or other educator is an informal mentor. That is, if the novice teacher designated that person as an informal mentor in response to our description of someone who provides them with instructional, emotional, or other types of assistance and support related to teaching, then we consider that person an informal mentor—no matter what his or her position is (e.g., principal or math coach). The spirit of the distinction between formal and informal lies in the inception of the relationship—whether the mentor was assigned through a school, district, or state policy or program, or whether the relationship developed organically.
Method
Data
Data used in this investigation come from a 5-year longitudinal study of natural variation in middle school mathematics teacher induction, supported by the National Science Foundation. The larger study involved a total of 66 beginning middle school mathematics teachers, as well as their principals, mentors, and district leaders, working in 11 school districts of varying size, wealth, and urbanity, in four mid-Atlantic and southeastern states.
The study used a mixed methods approach. Survey data provided comparable indicators of formal and informal mentor characteristics as well as the content and quality of interaction associated with each mentor. Data from structured interviews were audio-recorded and transcribed verbatim. They provided insight into the specific content and quality of mentoring and other induction experiences, as well as the genesis of relationships with mentors. Both surveys and interview protocols were developed in an iterative process, incorporating existing survey and interview instrumentation. Both were informed by a set of cognitive interviews and focus groups with teachers, mentors, and principals.
Survey data were collected 3 times during the teachers’ first year in the classroom. An initial survey, administered to teachers individually via mail in October and collected personally by a member of the research team at the time of the first interview, was used to obtain information on teachers’ backgrounds and work contexts. Additional surveys, distributed to individual teachers via mail or in person and returned in postage-paid envelopes in December and again in May of the teachers’ first year, were used to obtain detailed data on the content and quality of teachers’ mentoring and professional development. Members of the research team followed up with teachers whose surveys were not received or whose surveys were returned with missing or incomprehensible responses to ensure accurate and timely completion. Teachers’ formal and informal mentors were identified based on responses from the October survey and from brief phone interviews and e-mail questionnaires with the teachers in the late winter.
We asked teachers to report individually on each formal mentor but to aggregate their experiences across informal mentors. We took this approach mainly because we thought it was more important to collect data on individual formal mentors, who were working under policy directives. We believe that capturing the overall experiences with informal mentors gives us an appropriate contrast. Also, because we knew that some teachers had more than five informal mentors, reporting on each one would be unduly burdensome. We could imagine a study of informal mentoring that captures activities and content by each informal mentor, but that was beyond the scope of this study. 1
During their first year of teaching, teachers were interviewed in October or November and again in May or June. The structured interview protocol asked teachers to describe their first-year challenges and their formal and informal mentoring experiences, including the substance, nature, and quality of assistance they received, and the organizational and logistic challenges of working with formal and informal mentors; we also asked questions about other types of supports they received, including interactions with school administrators. These one-on-one interviews were conducted at the school site and tended to last between 1 and 2 hr.
A strength of our data is the combination of survey and interviews available for both formal and informal mentors; this permits a more thorough investigation than either method alone would allow. This multimethod approach builds on previous investigations of induction with a moderate sample size (e.g., S. M. Johnson & Birkeland, 2003). Our data enable us to undertake a detailed examination of teacher perceptions and experiences, though not at the depth that very small samples allow (e.g., Feiman-Nemser, 2001); at the same time, our data are on a larger scale than most other qualitative research investigating mentoring. Although these data are not representative of any district or state and should not be interpreted as such, we did recruit nearly all of the new math teachers in each of the 11 districts, and those districts spanned the southeast and middle Atlantic regions, suggesting that our results may apply to multiple settings.
Sample
We used data from interviews and surveys of 57 middle school mathematics teachers from the larger study’s three cohorts. These teachers began teaching in the 2007-2008, 2008-2009, or 2009-2010 school years. A total of 66 teachers participated in some aspect of data collection during their first year of teaching. However, because this investigation focuses on the role that informal mentors play as compared with formal mentors, we excluded the 4 teachers who reported that they did not have a formal mentor assigned to them. 2 We excluded an additional 5 teachers who completed only the October survey because survey items pertaining to the content and quality of mentoring were asked only on subsequent surveys. 3 Thus, our final analytical sample consists of 57 first-year teachers. In reporting, we use pseudonyms to maintain confidentiality.
One of the four states, which we call State 1, had an extensive teacher induction policy. It involved a three-person committee comprised of the principal, a resource teacher, and a teacher educator from a local university of college assigned to each new teacher. Each committee member was required to observe the new teacher 3 times and evaluate her against the standards set by the state for new teachers. The resource teacher was a practicing classroom teacher who had completed at least 4 years of “successful” teaching. The resource teacher’s responsibilities included 70 hr of consultation with the new teacher.
The three other states had less extensive policies. In State 2, each district was required to have a state-approved mentoring policy for all new teachers, as well as district-defined criteria for selecting mentors and training them in approved programs. Mentors were not required to receive compensation or release time. State 3 required districts to establish state-approved mentoring policies for all new teachers, but at the time, there were no training or mentor-mentee match requirements. State 4 required mentoring only for teachers who did not complete a traditional internship.
Layered on top of state policies are district policies, which showed a wide range of features. Two districts in State 1 did not have mentoring policies; one district had full-time, certified mentors for teachers in low-achieving schools or schools with two or more new teachers. Other districts had centralized programs, where mentoring was provided on an as-needed basis by district-level mentors working with up to 60 new teachers. In some districts, mentors received no training and provided support for both new and experienced teachers, while in other districts, mentors received training and supported beginning teachers only. Several districts had only school-based programs where the principal selected mentors. In no district was there a requirement for mentor-mentee match in subject or a required evaluative role for mentors, and only one district required a minimum number of contact hours. One district took a hybrid approach that paired instructional coaches selected at the district level with school-level master mentors selected by principals.
Survey Measures
The variables used in our quantitative comparative analysis were taken from our surveys of novice teachers. The surveys included measures of mentor characteristics as well as measures of the quality and nature of mentoring.
Characteristics of mentors
Based on our conceptual framework, we analyzed three characteristics of mentors: whether they (a) had math teaching experience, (b) were located in the same school building as the teacher, and (c) had time to meet with the teacher during the school day.
Math teaching experience
We created a dichotomous variable for whether each mentor identified on teachers’ surveys had math experience, assigning the variable a value of 1 if any of the following conditions were true: (a) the mentor selected “middle school math teacher” as a response to the question, “What is the mentor’s main position?” (b) in an interview, the mentor described himself or herself as having prior middle or high school mathematics teaching experience; or (c) the school or district website listed the mentor as a current or prior mathematics teacher.
Location in the same school building
We created a dichotomous variable for whether each mentor was located in the same building as the novice teacher. The variable had a value of 1 if the teacher selected “in my school” as a response to the question, “Is the mentor located in any of the following places?”
Time to meet
Similarly, we created a dichotomous variable for whether each mentor had free time in common with the teacher during the school day. The variable had a value of 1 if the teacher answered “yes” to the question, “Is there time during the school day that both you and the mentor have available to meet together?”
Quality and nature of interactions
We describe the nature of mentoring interactions using (a) the overall duration of mentoring interactions, (b) the content focus of these interactions, and (c) the extent to which the interactions incorporated forms of active learning.
Duration
Our estimates of the overall duration of teachers’ interactions with mentors are calculated from items asked on both the winter and spring surveys. For each formal and informal mentor, we asked teachers about the frequency and length of their formal meetings.
The frequency question was, “Since the beginning of the school year, how often have you had a formal, in-person meeting with Mentor X (other than a casual “hallway” conversation)?” Here we use “X,” but in the actual survey, we used the number assigned to each particular mentor, so teachers were referring to a specific mentor when they answered the questions. On the first-year spring survey, we replaced “Since the beginning of the school year” with “Since the date you completed the winter survey” and provided that date so that the reference periods would not overlap. Response options for both questions were as follows: 1 = never; 2 = once or twice (1-2 times); 3 = once a month (3-6 times); 4 = a few times a month (7-12 times); 5 = once a week (13-20 times); 6 = a few times a week (21-40 times); 7 = every day or every other day (more than 40 times).
The question on length of meetings was, “Since the beginning of the school year, how long on average was each of your meetings with Mentor X?” We used the same wording variation on the first-year spring survey as we used for the frequency question. Response options were as follows: 1 = less than 15 min; 2 = 15 min to 30 min; 3 = 31 min to 1 hr; and 4 = more than 1 hr.
To estimate the overall duration of mentoring teachers received from each mentor, we first recoded the responses to the frequency and length of meeting items. We used the midpoint of the range provided for each response option on the scale. For instance, we recoded “a few times a month (7-12 times)” to 9.5 times because 9.5 is the midpoint of 7 and 12. We recoded the response option corresponding to the greatest frequency as 60 times. As the response option was 40+ (meetings occurring every day or every other day) and each survey asked about roughly half the school year, 80 times would be an appropriate upper limit for an event that occurs “every day or every other day” over the course of a semester. Sixty is midway between 40 and 80, which is consistent with our general approach for using the midpoint as an anchor.
We used a similar strategy to convert responses to the length of meetings item. For example, we coded “15 min to 30 min” as 22.5 min because 22.5 is the midpoint of 15 and 30. We recoded the “more than 1 hr” response as 90 min because interview data indicated that 2 hr was as long as these meetings would realistically be, so we took the midpoint between 60 and 120 min. To generate our estimate of overall duration of mentoring time, we multiplied the estimate of frequency of formal in-person meetings by the meeting length estimate. The overall duration of mentoring for the entire first year was calculated as the sum of mentoring time in the winter and spring for each mentor.
Content focus
On each survey, we asked about the extent to which mentoring focused on particular topics. Similar to the items used to estimate duration of mentoring time, we adjusted wording in the stem on the Year 1 spring survey to ensure that reference periods did not overlap. These items were asked about all individual formal mentors. However, we asked teachers to consider all informal mentors together in describing the content focus of their mentoring. The specific Year 1 winter survey item asked, “Since the beginning of the school year, thinking about all of your meetings, formal and informal, with Mentor X, how much of a focus was each of the following general topics/areas?” There were 14 topics listed below. Response options were as follows: 0 = not a focus; 1 = minor focus; 2 = major focus.
We used the estimate of overall duration of mentoring with each mentor to estimate the amount of time allocated to particular topics. Topics were grouped into seven categories, five of which (classroom management, expectations of teachers in the school/district, parent involvement, emotional support/stress management, and technology) consisted of one item only for each mentor. The remaining items were categorized as math instruction or general instruction (i.e., topics not specific to math instruction) based on results of prior principal components analyses. Four items (how students learn math, deepening content knowledge in math, individualizing instruction in math, and analyzing student work in math) were included in the math instruction composite, which had a Cronbach’s alpha internal consistency of .86. The five items included as the general instruction topic (assessment/testing, what math content to teach, pacing instruction, planning lessons, and working with low-achieving students) had a Cronbach’s alpha internal consistency of .79.
To generate the estimates of mentoring time allocated to each content area, we first computed the sum of the “amount of focus” responses for each of the 14 individual topics, counting a major focus as 2 and a minor focus as 1; this weighting allows us to look at relative focus in a meaningful way. Ideally, we would like to have had information on number of minutes spent on each topic, but that was too burdensome to collect in the context of our study.
Next, we computed the sum of the “amount of focus” response for the two composite topic categories—math instruction and general instruction—individually and used actual responses for the remaining five topic categories defined from single items. We then divided the sum of foci for each topic category by the sum of foci across all the categories to get the percentage of mentoring that focused on each category. We multiplied this percentage by the total duration of mentoring estimate for each mentor to create a more interpretable measure of the relative amount of mentoring provided in each area. We completed this procedure separately for the winter and spring surveys. When a teacher had the same mentor or mentors all year, we estimated the total amount of time spent on each topic by adding the number of minutes per topic estimated from the winter and spring surveys.
Active learning
The survey items used to measure the extent to which teachers and mentors engaged in active learning activities were similar to the items used to determine the relative emphasis on different content areas. On each survey, we asked, “Since the beginning of the school year/since the date you completed the winter survey, how often did you engage in the following activities with Mentor X?” The activities listed include being observed and receiving feedback; observing the mentor teach; co-teaching a mathematics lesson; participating in an activity focused on mathematics such as a workshop or inservice; collaboratively planning a math lesson; and analyzing student work together. Teachers were asked to record the number of times they engaged in each activity with each formal mentor, and with all informal mentors combined, and were limited to two digits (i.e., a maximum of 99) per activity. 4 As with the duration and content focus of mentoring, active learning responses were added across the winter and spring surveys to give an estimate of the total amount of active learning activities with each mentor across the entire year.
Mentor quality
At each survey point beginning with the December survey in the first year, teachers were asked to rate each formal and informal mentor on three categories: (a) mentoring ability, (b) mathematics knowledge, and (c) knowledge of teaching. Each of these ratings had four response options: 1 = not very good, 2 = adequate/average, 3 = good, and 4 = very good. Teachers could also indicate don’t know as a response, which we treated as missing. In our results, we report mentoring quality using the three categories, as well as a mentor quality composite, computed as the average of the three categories, which has an internal consistency of .79.
Challenge of the teaching assignment
On the spring survey, we asked teachers, “To what extent do the following limit your math teaching?” The aspects of teachers’ work contexts that they rated included eight factors pertaining to their classes: students with special needs (e.g., physical, cognitive, emotional disabilities); low-achieving students; English-language learners; class size; student mobility; student absenteeism; uninterested students; and student behavior. Each factor had four response options: 1 = not at all; 2 = to a small extent; 3 = to a moderate extent; 4 = to a great extent. We aggregated the eight ratings to form an index, with the mean 17.12 and standard deviation 7.26.
Analytic Approach
For our analysis, we pooled data across the study’s three cohorts of teachers. Our primary analytic approach was to examine descriptive statistics, using t tests to examine the statistical significance of comparisons between formal and informal mentoring. Throughout our analyses, we examined data distributions for potential outliers. We used a .05 significance level as the cutoff, though given our small sample size, we mention marginally significant findings between .05 and .10.
Our methods here reflect our research questions, which are primarily descriptive. The importance of information garnered through descriptive analyses has been well established by previous research (Figlio, Hart, & Metzger, 2010; Hill, 2007; Ingersoll & Perda, 2010), and is perhaps best elucidated by Reardon and Galindo (2009), who say that “just as in medicine, where epidemiological documentation may stimulate the discovery of a cure, so too in educational research, a detailed description . . . may lead to a better understanding of . . . causes and solutions” (p. 854).
For the qualitative data analysis, we followed the procedures outlined by Miles and Huberman (1994), Huberman and Miles (1994), Patton (1990), and Coffey and Atkinson (1996). Our conceptual framework and research questions served as the basis for our initial coding framework for interview transcripts (Alexander, 2001). We then added more themes and subthemes as called for by our ongoing analysis of the transcript data. We used the constant comparative method to develop the codes (Glaser & Strauss, 1967; Strauss & Corbin, 1998) so that ideas from the transcripts were used to expand and refine the coding system. Through this iterative process, we changed, adapted, and integrated categories or themes (Goetz & LeCompte, 1984). In this way, we were able to interactively identify themes using both our conceptual framework as well as the transcript data. This method enabled us to use the data to inductively test our hypotheses, as well as to deductively allow other themes and explanations to emerge that we had not anticipated in our conceptual framework (Emerson, Fretz, & Shaw, 1995; Green, Dixon, & Zaharlock, 2002).
We illustrate each overarching theme with key quotes as exemplars (see Atkinson, Coffey, & Delamont, 2003), a technique that, according to Ryan and Bernard (2003), is “a widely used method for describing themes . . . that lead the reader to understand quickly what it may have taken the researcher months or years to figure out” (p. 282). If more than two teachers mentioned a topic, we added that topic to our coding system so that we could explore the extent to which other teachers commented on it. We indicate if an issue came up for only a small handful of teachers. In essence, while we avoid reporting on anomalies that occurred for only one teacher, we took a conservative approach in including themes and ideas that two or more teachers mentioned. Given the modest size of our sample, we thought it prudent to try to reflect the continuum of teacher experiences rather than focus only on those issues that “most” teachers discussed. Overall, however, most key themes were reflected in the majority of teacher interviews.
Results from the interview analysis generally supported survey findings. For economy of reporting and to emphasize new information, we focus our reporting of interview findings on areas that (a) provide insights into understanding survey findings or (b) reveal differences not detected with our survey data. This results in uneven application of interview data to variables examined in the survey analysis. However, it represents where our interview data added insights and nuances, which we would not expect to be spread evenly across every variable measured in our survey.
Aggregate Experience Across All Mentors or Average Experience With One Mentor?
Our tables show results of comparing formal and informal mentoring in two ways. One way is by aggregating a novice teacher’s experiences with her formal mentors and aggregating her experiences with her informal mentors, and comparing those two sets of aggregated experiences. The second is to compare a novice teacher’s average experience with one formal mentor to her average experience with one informal mentor. Both approaches are important from a policy perspective.
In instituting and shaping mentor policies, districts and principals want to know, in general, what a formal mentor is likely to provide to a novice teacher. We take the perspective that such decisions can and should be shaped by the support and experiences an informal mentor is likely to provide so that the formal mentor role can be designed to complement informal mentoring. Thus, comparing the average experiences provided by formal versus informal mentors offers a valuable perspective in helping to shape mentor policy. These comparisons are complicated by the reality that novice teachers can and often do have multiple formal and informal mentors. Thus, policymakers would also want to know the full experience a novice teacher has across all of her formal and informal mentors.
On most dimensions, the aggregate and average analyses were very similar. We provide both sets of results in our tables. We also calculate and include the ratio of formal to informal mentoring at both the aggregate and average level, to facilitate the comparison of the aggregate to average results. The aggregate and average ratios were not statistically different from each other (not shown) except in one case, which we further report. In the text, we present the average results, as arguably they are most consistent with previous work on mentoring, which focuses on what one mentor may provide to a novice teacher. We discuss the few cases where our aggregate findings seem to differ meaningfully from the average findings.
Findings
RQ1: How Do the Characteristics of Informal and Formal Mentors Differ?
Here we investigate how many of each type of mentor each novice teacher had, personal characteristics of mentors in terms of experience teaching mathematics, and organizational characteristics such as location in or outside the novice teacher’s school building and amount of time during the school day to meet.
Number of formal and informal mentors
As Table 1 shows, teachers reported having from one to five formal mentors. Of the 57 teachers in our analytic sample, 25 reported having one formal mentor, 8 reported having two, 19 reported having three, 3 reported having four, and 2 said they had five formal mentors. Sixteen of the 19 teachers who reported having three mentors were in a state that had an induction policy requiring teachers to have three formally assigned mentors, including their principal, someone from a teacher preparation institution, and a more traditional school-based mentor. Four of the 5 teachers who reported having four or five formal mentors were in that same state, so one can interpret this to mean they had the three assigned to them through the state induction policy, and then an additional one or two mentors assigned by their district or school. Table 1 also shows that novice teachers reported having varied numbers of informal mentors, with most having from one to three informal mentors.
Number of Formal and Informal Mentors Teachers Reported in Their First Year (N = 57).
Our analysis excludes the four novice teachers in our sample who did not have any formal mentors.
A relationship might be compensatory, in that teachers with fewer formal mentors would seek out more informal mentors, and vice versa. Figure 1, a scatterplot of formal by informal mentors, shows that this was not the case. The numbers of formal and informal mentors novice teachers reported were not correlated (r = .03, p = .84).
Scatterplot showing number of formal mentors by number of informal mentors (N = 57).
Location, time to meet, and math teaching experience
To examine how formal and informal mentors differed, we compared the number of each who (a) were located in the same school as the novice teacher, (b) had time to meet with the novice teacher during the school day, and (c) had experience teaching mathematics.
Table 2 compares formal and informal mentor characteristics for teachers who had both types of mentors. 5 For these teachers, there was little difference between their formal and informal mentors as to whether they were located in the same school (78% vs. 84%, respectively, p = .41) or had time to meet during the day (86% vs. 94%, respectively, p = .16). Eighty-six (86%) of novice teachers had a mentor with math teaching experience, compared with 80% who had an informal mentor with math teaching experience (p = .47). Fifty-one (51%) of teachers had at least one formal and informal mentor with all three of these characteristics (in the same school, time to meet, and experience teaching math).
Characteristics of Formal and Informal Mentors for First-Year Teachers Who Had at Least One Informal Mentor.
Note. The denominator in percentage calculations in the table is 51. Six teachers did not have at least one informal mentor and were, therefore, excluded.
Summary of mentor differences
We found little support for our compensatory hypothesis in terms of mentor characteristics. The number of formal mentors was not systematically related to the number of informal mentors. Neither did formal and informal mentors differ on whether they were in the novice teacher’s school (about 70% of each type were). The main difference we found was that formal mentors were significantly more likely to have math teaching experience.
Our interview data did not provide new or added insights into informal/formal mentor differences in terms of characteristics; thus we do not report them here.
RQ2: How Does the Nature and Quality of Informal and Formal Mentoring Interactions Differ?
To answer this question, we used the 51 teachers in our sample who had at least one formal and one informal mentor. We compared informal and formal mentoring on (a) the duration of interaction time, (b) the content of those interactions,(c) the extent to which those interactions included active learning opportunities, and (d) novice teachers’ ratings of the quality of the formal and informal mentor.
Interaction time and content of interactions
Table 3 shows that novice teachers spent more time interacting with their informal mentor (1,068 min/17.8 hr) compared with their formal mentor (679 min/11.3 hr); this difference was significant (p = .044). The teachers did not spend significantly more time with either of their mentors on general instruction, math instruction, classroom management, or technology, but they did spend significantly more informal mentoring time on expectations for teachers (p = .044), parent involvement (67 min compared with 33 min, p = .043), and emotional support (115 min compared with 53 min, p = .009). 6
Comparisons of Formal and Informal Mentoring Time and Content (n = 51).
Analyses of percentage of total mentoring time allocated to different topics used a sample of 49 teachers; two teachers who had spent no time in formal meetings with any informal mentors had missing values on the percentage variables due to division by zero.
p < .05. **p < .01. ***p < .001.
Another way to compare these interactions is to examine whether the proportion of time spent on different content areas is different. This tests our compensatory hypothesis—for example, a compensatory hypothesis would suggest that if novice teachers spend 80% of their time with formal mentors on classroom management, we would expect them to spend a larger percentage of time with informal mentors on other topic areas, such as math instruction. Results, shown in the third main column of Table 3, did not support this hypothesis; they indicated that novice teachers spend about the same proportion of time across different mentoring activities with their formal and informal mentors.
Active learning
Table 4 shows that, on average, active learning was used in twice as many activities with a novice teacher’s formal than informal mentor (8.6 activities compared with 4.8, p = .000). Teachers reported that formal mentors more often observed them and gave them feedback (3.8 times compared with .43 times, p = .000). 7 In contrast, novice teachers participated in more math activities with their informal mentors (1.2 times compared with .65, p = .008). There were no significant differences between formal and informal mentors in terms of the number of times novice teachers co-taught, 8 co-planned, and analyzed student work with them. We conducted several other correlational analyses to test for compensatory relationships but did not find any. 9
Comparisons of Formal and Informal Mentoring Active Learning (n = 51).
Note. Active learning composite includes six activities: observing mentor teaching, receiving feedback from mentor’s observation of mentee, co-teaching with mentor, participating in a math professional development activity with mentor, co-planning a lesson, and analyzing student work together. There is not a constant relationship between the aggregate number of activities and the number of activities per mentor because (a) some teachers had different numbers of mentors in the fall and spring and (b) the activities didn’t happen evenly during both halves of the year.
p < .05. **p < .01. ***p < .001.
The interview data provided insights into the one compensatory relationship the survey data revealed—that novice teachers are more often observed by their formal mentors. Novice teachers explained why this was important to them, and why it happened more often with their formal rather than informal mentors. Beginning teachers generally indicated that having their formal mentors observe them allowed the mentors to be “more detailed about what I do well, what I can work on” (Alice), while they say that their informal mentors “have never been in here to observe me and see how I teach.” As acknowledged by the novice teachers, formal mentors are more detailed because they have to “write down all those comments about me and ask all these questions” (Alice).
Teachers overwhelmingly appreciated being observed and getting feedback and wanted more of this; in interviews, 39 teachers gave a clear indication that they value some aspects of being observed. They talked about the advantages of having someone come in to watch their class to help diagnose their challenges, rather than self-assessing their challenges, which they do with their informal mentors. For example, one teacher said that he values advice he received after his formal mentor observed his class because it’s one thing me telling you what I need to do, and what I see in this class, and what’s going on for me, as it is to actually sit back and take in what’s going on, and therefore be able to give me even more constructive criticism. (Rob)
Another said, “she saw my teaching style, so she was able to be more personal in her suggestions because she knew what would work for me, and what wouldn’t” (Rachel).
Another teacher noted the difference between general advice from informal mentors compared with the advice formal mentors were able to give based on their observation of the new teacher, saying that “[my colleagues] were all like have you tried this, whereas [my formal mentor] was like this is what I think would work [for you]” (Jen).
Quality ratings
We compared teacher judgments of their mentors’ quality—overall, and for mentoring ability, knowledge of math and knowledge of teaching. Novice teachers rated informal mentors slightly higher than formal mentors in all the three categories, as Table 5 shows. Those differences were just barely significant for mentoring ability (3.31 compared with 3.53, p = .052) and knowledge of teaching (3.62 compared with 3.77, p = .049) but not for knowledge of mathematics (3.21 compared with 3.26, p = .770). On a composite of all the three ratings, novice teachers rated their informal mentors significantly higher (3.54) than their formal mentors (3.38, p = .026). Notable was that on average most ratings were quite high; they ranged an average of 3.21 to 3.77 on a scale where 1 = not very good, 2 = adequate/average, 3 = good, and 4 = very good. 10
Comparisons of Teacher Ratings of Formal and Informal Mentoring Quality (n = 51).
Note. One teacher was excluded from this analysis because the teacher responded “Don’t know” on this item for each of the teacher’s formal mentors, resulting in missing data on this particular variable. Quality composite includes 3 teacher ratings of mentoring quality (α = .79)—mentor’s mentoring ability, math knowledge, and knowledge of teaching—all on a scale of 1 = not very good, 2 = adequate/average, 3 = good, 4 = very good.
For ratings of mentor’s mathematics knowledge, n = 50.
However, analysis of the interview data identified two areas, not measured on the survey, where novice teachers indicated that formal mentors outperformed informal mentors: focusing on performance standards and willingness to initiate interactions.
Many new teachers commented that they found their formal mentors’ focus on performance standards helpful, as one teacher commented, the interaction is usually directly tied into the [state] teaching standards and where I stood on each one, where I’d improved from the last visit and what we need to do to go in the direction to keep moving upward. (Kailey)
In contrast, no beginning teachers indicated that their informal mentors calibrated their interactions to performance standards, or charted their progression so systematically.
In addition, a substantial number of novice teachers indicated that while they greatly benefited from having a nearby informal mentor whom they could ask questions, they appreciated having a formal mentor who initiated interactions with them. It was quite common for formal mentors to initiate interactions, whereas with informal mentors, it was usually the beginning teacher who initiated the interaction. As novice teacher Joe explained, when he interacts with his informal mentor “it’s all me initiated, he personally has never come and said, hey, I’ve got something for you . . . with [my formal mentors], they initiate a lot of stuff which is wonderful.” He goes on to discuss the advantages of having a formal mentor, explaining that to me I really think it’s a positive because you have someone you can go to and not only that but [the formal mentor will] come to you as well. You’re not always running to [him]. [He’ll] come to you and see what’s going on. What do you need?
Summary of mentoring differences
We find that novice teachers interacted more with informal mentors than with formal mentors, but the interactions themselves were very similar. Teachers did not spend more time on math instruction with their informal mentors, but they did spend more time with their informal mentors on non-math-related instruction, classroom management, expectations, parent involvement, and emotional support. The proportion of time they spent on different topics was generally the same, and novice teacher ratings of their informal mentors were only slightly higher than of their formal mentors. Both were generally high. Formal mentors outperformed informal mentors in that formal mentors were much more likely to observe novice teachers and provide them with feedback. Novice teachers greatly valued this activity. Furthermore, formal mentors were more likely to help novice teachers improve their responsiveness to performance standards and to initiate interactions.
RQ3: To What Extent Do Certain Organizational, Structural, or Personal Characteristics Explain Differences Between the Nature and Quality of Formal and Informal Mentoring?
From the survey and interview data, we analyzed whether differences between formal and informal mentors can be explained by being located in the same building, having time to meet, the mentor’s math teaching experience, and the level of challenge of the new teacher’s teaching assignment. We also examined the interview data for insights about additional factors beyond those measured in the surveys, which might explain differences in formal and informal mentoring.
Location
We hypothesized that novice teachers may spend more time with informal mentors because informal mentors are more often located in the novice teachers’ school, whereas many formal mentors work in different schools or at the district office. To examine this hypothesis, we held location constant and compared time spent with formal and informal mentors.
Table 6 shows that novice teachers spent more time with mentors who are in the same building as them, and this applies to both formal and informal mentors. Teachers spent an average of 459 min/7.65 hr with their formal in-school mentor, but only 221 min/3.68 hr with their formal out-of-school mentor (p = .040). Similarly, novice teachers spent 864 min/14.4 hr with their informal mentors located in the same school as they are but only 205 min/3.4 hr with their out-of-school informal mentors (p = .002). When both informal and formal mentors were located in the school, novice teachers spent more time with their informal mentors (864 min/14.4 hr) than their formal mentors (459 min/7.65 hr). This difference was significant (p = .027). In contrast, when both informal and formal mentors were located outside of the school, there were no significant differences between time spent with formal and informal mentors.
Mean (SDs) Comparisons of Formal and Informal Mentoring by Location, Time to Meet and Math Experience (n = 51).
p < .05. **p < .01. ***p < .001.
Not only does working in the same school correlate with more time with a mentor but interviews revealed that working in the same school also influences the nature and quality of mentoring in ways not measured on the survey; and these differences seem driven by the location of the mentor, not whether they are formal or informal. Thirty-six teachers talked positively about seeking help from mentors who are familiar with their school, curriculum, and/or students.
Interviews indicated that in-school mentors, whether formal or informal, were (a) more likely to offer pacing and lesson planning advice because they were teaching or had taught the same material as the new teacher; (b) more apt to share materials and less apt to engage in extended discussion, (c) able to offer child-specific behavioral management advice, and (d) ideally situated to offer “in the moment” advice and feedback.
Sharing ideas about curriculum and pacing
Many teachers indicated that in-school mentors were able to offer specific advice about the curriculum and pacing because they were teaching or had taught the same material. For example, one teacher described the help she got from her informal mentor as being about “what are we doing with this subject this week, you know, or what’s the flow? What are we teaching, what, with the scope and sequence; how are we staying with it?” (Janet). Another teacher added that we discuss, we’re about on the same level as far as pacing in our books, so we discuss a lot, how are we gonna do this lesson, how can we tweak this unit to, you know, we only have a month in this unit . . . what can we do to tweak and just, to speed it up like that, and what normally gets cut down. (Beth)
Offering materials versus discussing ideas
The interviews also revealed that in-school mentors, whether formal or informal, were much more likely to have a “use this” method of mentoring, compared with out-of-school mentors, who were more likely to focus on ideas and discussion. As one teacher described of her in-school formal mentor she gives me basically her lesson plans every week and gives me what resources she’s going to use and I get to look through there and see what I like and kind of tweak it to my personality and use it that way. (Kristin)
Another said that her formal in-school mentor for the most part she’s done good about taking care of the practical stuff like creating the tests and giving me a copy and you know, having the plans and letting me just use what she has planned and you know, so that’s been great. (Susan)
This is in contrast to out-of-school formal mentors, who often seemed to focus more on deeper understandings, as reflected by one teacher who said that her formal out-of-school mentor “[makes] sure I understand exactly what was supposed to be taught in each topic” (Debbie). Another said her formal out-of-school mentor “[goes] over lesson plans about what, you know. what you need to have prepared, what might work, what might not work, you know questions you might want to have for homework” (Michelle). Another said that “we’d still bounce ideas off of each other . . . and are your kids getting it?”
Providing classroom management advice focused on particular students
Teachers in both the surveys and interviews indicated they receive classroom management advice from both their formal and informal mentors. The one substantive difference identified in the interviews was that informal mentors, or formal mentors in the beginning teacher’s school, were able to provide student-specific advice because they were often familiar with the students in the beginning teachers’ class. This was represented by one new teacher talking about several other teachers she considered informal mentors: We talk a little bit about classroom management maybe but we talk more about the personalities of the kids and what works with this kid for this teacher and what doesn’t for another teacher. We try to find that one common thread that maybe all the teachers can use to get through to that one kid. (Anna)
One teacher explained that a main difference between the help she gets from her informal and formal mentors was that “the difference would be of course [my informal mentors] understand the kids that we’re working with because they’re here in the building” (Christine). Another said that her informal mentor provided “more detailed, more specific help because she knows her kids” (Tonya). Another teacher, referring to out-of-school mentors who didn’t know her students, said, even though they try to throw ideas out there, every class, every student, every year is completely different. So although you’ve worked with at risk students, it’s have you worked with these students. Even though you say in your classroom you could do this, how can I do it with these students, if that makes sense. (Vivian)
In-the-moment feedback
We found that novice teachers do seek out informal mentors to satisfy a need that most formal mentors cannot provide: “in-the-moment” advice and feedback. Twenty-two teachers commented that they had a preference for help that was nearby and consistently available. Teachers generally agreed with the teacher who said that “I get help from those close by” and “I go to those who have a schedule that leaves them available when I need them.”
When asked about the advantages of close-by mentors, one teacher said “well, they’re here. I can go to them quickly. I can talk to them out in the hallway or whatever and we can, you know, I can have an answer in seconds during a class break” (Chris). Similarly, a teacher compared her formal with informal mentors, saying, “I mean with her [formal mentor], you know, having to schedule time I kinda had to wait to address it, but with [my informal mentors], of course, I can address it immediately” (Tomika). Even when a formal mentor is in the same school, often teachers will choose to interact more with their informal mentor, as explained by one teacher: it’s harder for me to go try and find [my formal mentor] cause we don’t have the same lunch time, we don’t have the same planning. It’s harder for me to seek him out when I can just ask someone right next to me, so I mean I hardly ever see him. (Megan)
This sentiment was echoed by many teachers, who indicated that their formal mentors, though in the same school, were not as convenient as the “teacher next door.”
Time to meet during the day
Results for “time to meet during the day” have a similar pattern as the results for “in-school.” As seen in Table 6, novice teachers spent more time with their mentors, formal and informal, if they had time to meet during the day, as we would expect. They spent 451 min/7.51 hr with their formal mentors with time to meet, compared to 228 min/3.8 hr with mentors who reported not having time to meet during the school day (p = .029); similarly, novice teachers spent 877 min/14.6 hr with their informal mentors who had time to meet during the day and only 191 min/3.18 with informal mentors who did not have such free time (p = .000). Considering only mentors who had time to meet during the day, teachers had more informal mentoring time (877 min) than formal mentoring time (451 min, p = .013). There was not a statistically significant difference in formal and informal mentoring coming from mentors who did not have time to meet during the day.
Time to meet also influenced the nature of informal and formal mentoring: It served as a structural constraint to informal mentors for observing their mentee. We reported earlier that formal mentors were much more likely to observe novice teachers’ classrooms; a natural hypothesis is that this occurs more often because observation is usually a requirement of the mentor-mentee relationship (where informal relationships have no requirements), and also there is more likely time in formal mentors’ schedules to allow them to observe. Our interview data support this hypothesis, as one teacher explains, [my formal mentor] could look directly at my classroom and say this is what I think you should try. None of my other colleagues have been able to step into my classroom and watch me teach because they’re teaching at the same time. (Brian)
Math background
Teachers spent more formal mentoring time with formal mentors who had math teaching experience (503 min/8.3 hr) than with formal mentors who did not have math teaching experience (175 min/2.9 hr, p = .002). This was not the case for informal mentors, however; the amount of time spent with informal mentors was not related to the informal mentors’ math background.
Challenge of teaching assignment
To see if novice teachers with more challenging teaching assignments received more mentoring from their formal or informal mentors, we correlated the variable measuring teacher reports of how challenging their teaching environment was, with amount of time and type of mentoring (not shown). More challenging assignments were significantly correlated with spending more time with the formal mentor (.29, p = .05) and spending more time with the formal mentor on math instruction (.27, p = .04), classroom management (.29, p = .03), and being observed and getting feedback (.30, p = .02). In contrast, we found no significant relationships between time spent with informal mentors and the level of challenge of the teaching environment.
Summary: Explaining the nature and quality of mentoring differences
In terms of explaining differences among formal and informal mentors, we find that being in the same school, with time during the day to meet, is associated with spending more time with a mentor, both formal and informal. Furthermore, novice teachers spent more time with their formal mentors if the novice had more challenging classrooms, or if their formal mentors had mathematics teaching experience. These differences did not explain time spent with informal mentors.
Discussion
This study investigates an issue largely ignored by mentoring, induction, and professional development policy: understanding differences between formal and informal mentoring, which can inform how we shape teacher learning opportunities. Ours is the only examination that we know of that compares how novice teachers interact with formal and informal mentors.
Interpreting Results
Several issues should be considered in interpreting results. Our data are limited in that we do not have data on each informal mentor separately, so in cases where novice teachers have multiple informal mentors, data are aggregated. We are not able to explore how role types (e.g., coach, principal) may influence the content of interactions with informal mentors, though we did analyze role types with formal mentors (not shown) and found no systematic relationships. We focus on dyad relationships between novice teachers and their mentors, rather than communities of practice or cohorts of novice teachers, both of which can also be important forms of support (e.g., Cuddapah & Clayton, 2011; Pogodzinski, 2012).
We look at mentoring only in the teacher’s first year. We do not address the stages, transitions, or evolution of mentoring relationships (e.g., Kram, 1983; Missirian, 1982; Pollock, 1995). Very few teachers in our study, however, had formal mentors beyond their first year.
Other studies have emphasized the critical role that principals can play in the support of new teachers (e.g., Pogodzinski, Youngs, Frank, & Belman, 2012; Youngs, 2007b); we do not separately analyze principal support in the constellation of novice teacher induction.
We focus on mathematics, which is a strength in that it allows us to interpret findings in light of what we know about mathematics teacher learning, but mentoring in other subject areas might work differently. Similarly, our analysis includes data from novice teachers only; future studies might compare perspectives of mentors and mentees.
Another consideration is the extent to which state and district official induction and mentoring policy impact the results. In analyses reported elsewhere we found significant variation between the state’s intended and enacted policies (Polikoff, Desimone, Hochberg, & Porter, 2013). We did find that teachers in State 1, with a required observation and contact hours component, were more likely to experience those components than other teachers. However, there was still considerable teacher-level variation. Thus, we find that most of the results reported here cannot be explained by the “intended” policies but rather depend on local context, as previous research would predict (e.g., Desimone, 2002; McLaughlin, 1987).
Discussion of Feelings
We discuss what our findings imply for guiding and improving informal mentoring, teacher education, and formal mentoring.
Implications for informal mentoring
Perhaps the most obvious implication is that informal mentoring should be considered explicitly as part of the “constellation” or “mosaic” of induction supports for new teachers (e.g., Kram, 1985) and integrated to complement the role of formal mentors. A “mentor coordinator” could assess the specific types of assistance the novice teacher might be missing from her formal mentoring relationship and other induction activities and help her identify and build informal relationships to compensate for what she is missing. Such a coordinator could be the principal or a lead or veteran teacher in the school. Integrating the role of informal mentors into the array of induction supports is consistent with the idea of developing a more coherent system of supports for teachers across teacher education, induction, and professional development (Wilson, Rozelle, & Mikeska, 2011).
A second implication is that the complementary roles of informal and formal mentors point to the desirability of having both. In considering the bundle of activities that may comprise good induction programs (e.g., D. L. Bickmore & Bickmore, 2010; D. L. Bickmore, Bickmore, & Hart, 2005; S. T. Bickmore & Bickmore, 2010), informal sources of support can help meet the personal needs of teachers, while formal mentoring is more appropriately directed to the professional needs of teachers (Gold, 1996) and, as such, serves as a critical mechanism for teacher learning and professionalism (Darling-Hammond, 2003; Darling-Hammond & Berry, 2006).
Also, informal mentors can help address the tension between mentoring assistance and assessment (Feiman-Nemser, Carver, Schwille, & Yusko, 2000). Our results are consistent with the idea that novice teachers seek support for management and emotional issues from people who are not officially evaluating them. These findings are also consistent with private-sector studies showing that informal mentoring was stronger than formal mentoring on psychosocial dimensions (Blake-Beard, 2001; Ragins & Cotton, 1999). Thus, in states and districts where formal mentors play an evaluative role, informal mentors could play a productive role in providing mentorship free of any accountability component.
Other complementary features are that informal mentors are usually the ones able to provide in-the-moment feedback, while formal mentors are more likely to follow-up at specific intervals, and initiate interactions, which is critical given that novice teachers are often hesitant to call attention to their struggles (Gratch, 1998). Both types of interactions are important. Similarly, advantages of an in-school mentor, articulated by teachers in our study as well as by other researchers (e.g., Coburn & Russell, 2008), include insights into school norms and culture and experience with the specific curriculum and students. But out-of-school mentors also have advantages, including the potential to offer a different perspective and to provide a safe place for teachers to discuss their struggles and concerns that they may not feel comfortable sharing with people inside their schools (e.g., Murrell, Blake-Beard, Porter, & Perkins-Williamson, 2008).
While we emphasize below that formal mentors should be matched on subject area when possible, when this is not possible, a mentor coordinator might link a novice teacher with an informal mentor with expertise in mathematics. Our conceptual framework is grounded on the centrality of mathematics knowledge at the core of new teacher learning, indicating that it is essential that mathematics teachers work with teachers experienced in teaching mathematics as part of their induction activities.
Implications for teacher education
The Association of Teacher Educators has guidelines for mentoring programs (Bey & Holmes, 1992); this and other forms of guidance could be strengthened by including provisions for formal mentoring to be responsive to novice teachers’ other learning opportunities, including informal mentoring. Also, it could be helpful to raise awareness among preservice teachers of the different pieces of school and district organizational structures concerned with teacher induction, as well as common models for mentoring, so they can be more deliberate about which kinds of support they seek from which people, where they might turn if support is lacking in a specific area, and so on.
In addition, teacher education programs that do not already focus on the importance and value of subject-focused observation and feedback might do so; this could help create a culture where novice teachers seek out such activities and continue to value them as they advance in their careers. Similarly, teacher education programs whose constituents are primarily inservice teachers working toward their masters’ degrees might include explicit instruction on how to mentor, including the idea that novice teachers’ informal mentoring experiences should be considered in assessing the types of support beginning teachers need.
Our findings offer further support for the frequent call to break down the silos between teacher preparation and K-12 (U.S. Department of Education, 2011). For example, the continuum and interchange between formal and informal mentors is fluid—teacher educators who serve as formal mentors during preservice training may serve as informal mentors during the first few years of beginning teaching. Creating opportunities for formal and informal mentors to work together to better complement each other could be a powerful way to efficiently improve induction support to beginning teachers. It might also contribute to fostering a culture in which mentoring is valued and prioritized (Solomon, 2009).
Much mentoring time was spent on social/emotional and classroom management issues, which suggests that often teachers are unprepared for the challenges of running their own class (see also Desimone, Bartlett, et al., 2013). This suggests that teacher preparation programs should provide a more extensive apprenticeship experience where preservice teachers can build classroom management skills under the guidance of a veteran teacher.
Implications for improving formal mentoring
Teacher learning policy is greatly influenced by state policy (e.g., Phillips, Desimone, & Smith, 2011) as well as district policy (e.g., Desimone, 2013; Desimone, Garet, Porter, Birman, & Yoon, 2001), and our findings point to several recommendations for shaping policies at these different levels. Our conceptual framework has subject-specific (mathematics) mentoring at its core, and our findings support the importance of having a mentor with mathematics experience and expertise in mathematics. Novice teachers spent more time with their formal mentor if that mentor had math teaching experience. This is consistent with research that shows such a match on content-area expertise is related to the quality and outcomes of mentoring (Lopez, Lash, Schaffner, Shields, & Wagner, 2004; Villar & Strong, 2007) and is necessary for fostering productive interactions and teachers’ instructional growth (e.g., Ball, 2000).
Our findings support the idea of making formal mentors more accessible, for example, by providing release time for mentors to meet with and observe novice teachers (see also Feiman-Nemser, 2001). Teachers extolled the virtues of being observed and receiving feedback. We know from previous research the value of observation and feedback for providing meaningful improvement in teachers (Carver & Feiman-Nemser, 2009) and, specifically, that it is essential for building teachers’ skills in teaching mathematics for understanding (Ball, 1990), as well as providing a mechanism for building a learning community (e.g., Harrison, Dymoke, & Pell, 2006). Thus, we recommend structuring novice teachers’ first year so that they can be observed multiple times by their mentors. Furthermore, novice teachers met with their formal mentors twice as much if those mentors had time to meet during the day and were in the same building. Although our study does not link teacher or student outcomes with the amount of time novice teachers spent with mentors, we consider more time with mentors to be a desirable feature, based on previous research (Glazerman et al., 2010; Rockoff, 2008).
We also recommend increasing formal mentor training. Novice teachers rated their informal mentors as slightly better in two out of three categories. One might expect novice teachers to view formal mentors, who often have more experience and mentor training, as more expert or effective, but this was not the case in our data. We suspect this may be because novice teachers see formal and informal mentors as each having advantages over the other, as suggested by our interview data, and that perspective averages out their ratings; for example, novices and their informal mentors may be better matched on personality and temperament given that those relationships emerge organically, while formal mentors may provide needed structure and observational expertise. Still, we might expect (hope?) that formal mentors on average would have more mentoring expertise. Our finding is consistent with other work that indicates coaches and others in mentoring roles need more training and expertise to maximize the potential of their role (Coburn & Russell, 2008; Stanulis & Floden, 2009).
Conclusion
Our findings that informal and formal mentors sometimes serve similar functions but often provide compensatory and complementary support suggest that novice teachers would benefit from having both. Our study identifies a set of policy variables—mixing in-school and out-of-school mentors, both formal and informal, who provide feedback based on classroom observations, as well as classroom management and emotional support—as potential levers to improve learning opportunities for novice teachers. Furthermore, paying more attention to teachers’ informal mentoring could have direct and relevant implications for formal mentoring policy; for example, schools might develop a monitoring and feedback mechanism that would allow them to shape ongoing formal mentoring to explicitly complement the activities and interactions occurring during informal mentoring. Further understanding the dynamics between formal and informal mentors could help districts and schools improve their array of induction supports for novice teachers.
Footnotes
Appendix
Correlations Between Formal and Informal Mentoring in Teachers’ First Year.
| Using time/activities aggregated across mentors | Using time/activities divided by no. of mentors | Using percentage of time on each topic (averaged across mentors) | |
|---|---|---|---|
| Total mentoring | .33* (p = .012) | .17 (p = .20) | — |
| Math | .30* (p = .024) | .17 (p = .22) | .30* (p = .034) |
| General instruction | .38** (p = .0040) | .17 (p = .21) | −.14 (p = .33) |
| Class management | .43*** (p = .0009) | .34** (p = .0098) | −.0010 (p = .99) |
| Expectations | .30* (p = .024) | .15 (p = .27) | .19 (p = .18) |
| Parent involvement | .08 (p = .57) | .05 (p = .72) | .25 (p = .086) |
| Emotional support | .28* (p = .034) | .18 (p = .18) | −.07 (p = .62) |
| Technology | −.03 (p = .84) | −.13 (p = .35) | −.10 (p = .51) |
| Active learning (composite) | .24 (p = .074) | .29* (p = .027) | — |
| Active: observation and feedback | .26 (p = .055) | .13 (p = .35) | — |
| Active: teacher observe mentor | −.30* (p = .025) | −.18 (p = .18) | — |
| Active: co-teach | −.17 (p = .21) | −.10 (p = .48) | — |
| Active: math activity | .39** (p = .0029) | .48*** (p = .0002) | — |
| Active: co-plan | .08 (p = .57) | .13 (p = .33) | — |
| Active: Student work | .30* (p = .022) but becomes .13 (p = .33) after removal of one high leverage point | .55*** (p = .0001) but becomes .08 (p = .57) after removal of one high leverage point | — |
Authors’ Note
The opinions expressed are those of the authors.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research reported here was supported by the Institute of Education Sciences, U.S. Department of Education, through Grant R305B090015, and by the National Science Foundation Grant DRL-0554434.