J Educ Change (2010) 11:2344

DOI 10.1007/s10833-008-9096-4

New teachers experiences of mentoring: the good,

the bad, and the inequity
Susan M. Kardos Susan Moore Johnson

Published online: 22 November 2008

Springer Science+Business Media B.V. 2008

Abstract Using a sample of 374 randomly selected first- and second-year teachers in
three states, this study examines new teachers experiences of official mentoring during
their first year. Descriptive analyses reveal that experienced mentors are generally present
in the work lives of new teachers. However, new teachers often have inappropriate mentormatches, and low percentages of new teachers are observed by or have conversations with
their mentor about the core activities of teaching. Low proportions of new teachers in lowincome schools and those in math, science, and technology have ideal matches and supports. The findings have implications for policymakers who look to mentoring as a strategy
to improve public schools and retain new teachers.
Keywords Beginning teachers
Equity
Induction
Mentoring
Retention

School improvement
Turnover

Introduction
Decades of research document new teachers turbulent landings into their classrooms (Gold
1996; Johnson and The Project on the Next Generation of Teachers 2004; Kauffman et al.
2002; Lortie 1975; McDonald and Elias 1983; Ryan 1970; Ryan et al. 1980; Veenman 1984;
Wideen et al. 1998). They are bumped about by the minute-to-minute decisions that determine whether they can maintain order among their students. They struggle deciding what to
teach and which resources to use, how fast or slowly to pace their lessons, how to engage
students with varied abilities and interests, and how to respond to demands for standardized
test preparation. They are confronted either by confusing and conflicting methods about how
S. M. Kardos (&)
The AVI CHAI Foundation, New York, NY, USA
e-mail: [email protected]
S. M. Johnson
Harvard Graduate School of Education, 431 Gutman Library, Appian Way, Cambridge,
MA 02138, USA
e-mail: [email protected]

123

24

J Educ Change (2010) 11:2344

best to teach their students or they are faced with a void: How on earth, they wonder, will
I teach this tomorrow? Over time, research has shown that most new teachers confront these
challenges alone, with little organized aid or assistance from their colleagues (FeimanNemser 1983; Goodlad 1984; Johnson 1990; Johnson and The Project on the Next Generation of Teachers 2004; Kardos and Johnson 2007; Little 1990b; Lortie 1975; McLaughlin
1993; Powell et al. 1985; Sizer 1984; Troen and Boles 2003).
Formal mentoring, which pairs new teachers with their veteran colleagues, is currently
the main strategy introduced by state and local policymakers to address new teachers
isolation, frustration, and failure. Mentoring programs are increasingly common in state
policy (McCabe 2006), but vary greatly in their design (Feiman-Nemser 2001). Some
programs are short-sighted and ill-conceived, difficult to implement and monitor, and
nearly impossible to evaluate (Feiman-Nemser and Schwille 2004; Fideler and Haselkorn
1999). Still, research has shown that, when done right, mentoring is good for new teachers
(Berry et al. 2002; Darling-Hammond 1999; Evertson and Smithey 2000; Feiman-Nemser
1996; Gold 1996; Holloway 2001; Humphrey et al. 2000; Ingersoll and Kralik 2004; Little
1990a; Smith and Ingersoll 2004; Wilson et al. 2001). It helps them deal with the barrage
of challenges they face, including the challenge of sifting through, sorting out, and succeeding in an ever-changing school environment. As much as observers have analogized
school change to the slow turning-around of a battleship, school contexts are, in fact,
dynamic and uncertain. Mentoring, when done right, can stabilize the shifting ground on
which new teachers try to stand.
Policymakers, teacher associations, school leaders, and new teachers, themselves, tend
to promote mentoring programs. Yet there is little research to document what new teachers
experience in them. We, therefore, solicited new teachers accounts of mentoring
arrangements by surveying a representative random sample of new teachers from three
states. Did they have official mentors assigned to them by their schools or districts? What
were the characteristics of these mentor matches? That is, did their mentors teach in the
same school, grade level, and subject? What was the nature of the interactions between the
new teacher and the mentor? Did the mentor observe the novice in the classroom and did
they discuss the critical aspects of teachingcurriculum, instruction, and student needs?
In general, three embedded stories about mentoring emerge from these data. The first
story is simple and promising: during their first year, 78 percent of the new teachers in
Florida, Massachusetts, and Michigan had official mentors assigned to them by their
schools or districts. These mentors were, in general, experienced teachers, with an average
of 15 years in the classroom. In short, experienced mentors are generally present in the
work-lives of most new teachers in these three states.
However, a second, troubling story emerges from the finer-grained details of the data.
Although nearly four-fifths of new teachers in these three states have official mentors, the
matches and types of interactions are far from ideal. For example, fewer than half of the
new teachers have a mentor in the same subject. Furthermore, less than 60 percent reported
having had three or more conversations with their mentor about classroom instruction,
about curriculum and lesson planning, or about classroom management and discipline.
Only 41 percent of the new teachers were observed teaching even once by their mentors.
A third and more troubling story emerges from comparisons between subgroups. Lower
proportions of new teachers in low-income schoolsschools that are traditionally hard to
staff and have the highest rates of turnover among new teachers1have official mentors
1

In their analysis of the nationally representative SASS data, Smith and Ingersoll (2004) found that school
poverty level was among the strongest predictors of likelihood of leaving teaching as opposed to staying.

123

J Educ Change (2010) 11:2344

25

during their first year than their counterparts who have begun teaching in high-income
schools. Furthermore, they are less likely to have mentors in their same school, at their
same grade level, and in the same subject area. They are less likely to have even three
conversations with their mentors by the spring of their first year about any of the core tasks
of teaching: classroom management, lesson planning, and classroom instruction.
Similarly, new teachers in math, science, and technologytraditionally hard to staff
subject areas also plagued by high attrition rates2have official mentors during their first
year in lower proportions than their counterparts teaching other subjects. They are less
likely to have mentors in the same school and in the same subject area, though they are
more likely to have mentors in the same grade level. They are less likely to have been
observed teaching at least once by their mentor, and they are less likely to have conversations with their mentors about any of the core activities of teaching: classroom
management, lesson planning, and classroom instruction.3

Analytic framework
New teachers
Novice teachers struggle mightily with the challenges of their new jobs. Despite their
idealism, they describe daily difficulties with the students in their classrooms (FeimanNemser 1983; Johnson and The Project on the Next Generation of Teachers 2004;
McDonald 1980; McDonald and Elias 1983; Rust 1994). They describe being lost at sea,
not knowing what to teach or how to teach it (Kauffman et al. 2002). At the same time,
they are trying to acclimate to the norms and modes of professional interaction in their new
workplace (Kardos et al. 2001) and internally and externally imposed school reform initiatives and mandates. It is important to pay attention to new teachers and their struggles,
since teachers early experiences determine their sense of success with their students
(Johnson and Birkeland 2003), their long-term performance in the classroom (FeimanNemser 1983; Gold 1996; Rust 1994), and their decisions about whether or not to stay in
teaching (Adelman 1991; Feiman-Nemser 1983; Gold 1996). Research has found that
teacher turnover rates are high (Grissmer and Kirby 1997; Lortie 1975) even compared to
those in many other occupations including teachings closest cousin, nursing (Ingersoll
2001). In 19992000, 29 percent of all first-year teachers left their schools: 14 percent left
teaching altogether, and 15 percent moved to other schools (Smith and Ingersoll 2004).4
Other studies have estimated that 50 percent of new teachers leave within the first five
years (Huling-Austin 1990; Ingersoll and Smith 2003; Murnane et al. 1991).
2

Although their coefficients were not statistically significant, Smith and Ingersoll (2004) also found, for
example, that math or science teachers were about 10% more likely than teachers of other subjects to leave
teaching.

Throughout this analysis, findings about mentoring refer to new teachers self-reports of their experiences
of the mentor assigned to them by their school or their district during their first year of teaching.

The national random sample of new teachers used in Smith and Ingersolls (2004) analysis of the 1999
2000 SASS data included teachers in regular public, charter, and private schools. The percentages presented
above are for all beginning teachers. Twenty-seven percent of regular public school teachers left their
schools (11 percent left teaching, and 16 percent moved to other schools). Thirty-seven percent of charter
school teachers left their schools (24 percent left teaching, and 13 percent moved to other schools). Thirtysix percent of private school teachers left their schools (26 percent left teaching, and 10 percent moved to
other schools).

123

26

J Educ Change (2010) 11:2344

Mentoring
For the most part, new teachers experience their early career challenges in professional
isolation (Johnson and The Project on the Next Generation of Teachers 2004; Kardos and
Johnson 2007; Lortie 1975). Thus, at the time when they are most in need and most
vulnerable, they are left to sink or swim. Mentoring programs for new teachers are often
introduced to moderate teacher isolation and to support new teachers in their early years
(Berry et al. 2002; Little 1990a; National Commission on Teaching and Americas Future
2003). Analysts and observers conclude that mentoring not only is good for new teachers
(Berry et al. 2002; Darling-Hammond 1999; Evertson and Smithey 2000; Feiman-Nemser
1996; Gold 1996; Holloway 2001; Humphrey et al. 2000, Ingersoll and Kralik 2004; Little
1990a; Smith and Ingersoll 2004; Wilson et al. 2001), but also has benefits for mentors
themselves (Drago-Severson 2004; Feiman-Nemser 1996, 2001; Holloway 2001). In this
context of school-based accountability and high-stakes testing, policymakers, central office
administrators, and school leaders have sponsored and supported mentoring programs to
improve new teachers effectiveness and reduce their attrition (Geschwend and Moir 2007;
Useem 2001). Recent national data show that induction and mentoring practices for new
teachers have greatly expanded over the past decade (Smith and Ingersoll 2004).
Support for mentoring is grounded in evidence that it works. Studies show that new
teachers who are mentored early in their career are more effective teachers (DarlingHammond 1999; Evertson and Smithey 2000; Feiman-Nemser 1983; Humphrey et al.
2000) and are likely to remain in their school or in teaching longer than those who are not
mentored (Humphrey et al. 2000; Smith and Ingersoll 2004). Without the proper support,
new teachers will resort to survival instructional strategies (Berry et al. 2002, p. 4;
Feiman-Nemser and Floden 1986; Huling-Austin 1990), which, in the long term, do not
serve them or their students well (Feiman-Nemser 1983; Gold 1996; McDonald 1980;
Rosenholtz 1989).
However, mentoring is likely not the silver bullet that schools, districts, or states often
hope it will be. Researchers who study mentoring caution against assuming that the new
teachers induction needs at the school site will be fully met if only they are assigned
experienced teacher mentors (Berry et al. 2002; Feiman-Nemser 1996; Little 1990a;
Stansbury and Zimmerman 2000). Our qualitative interview study of 50 new teachers in
Massachusetts also concluded that one-to-one mentor assignments, in themselves, cannot
meet the myriad needs of new teachers (Johnson and The Project on the Next Generation of
Teachers 2004; Kardos et al. 2001). Although certain programs that offer training for
mentors report measurable success in their mentoring program (Gless and Moir 2001),5 it
would be a mistake to assume that all mentoring programs are thoughtfully organized and
that all mentors know what they are supposed to do and how they are supposed to do it.
When mentoring assignments are haphazard, they may well be ignored and avoided by
both the new teacher and the mentor. Unwilling mentors and negative professional cultures
within schools may make mentoring assignments burdensome and even counter-productive
(Johnson and The Project on the Next Generation of Teachers 2004; Kardos et al. 2001).
Furthermore, new teachers experiences with their mentors vary wildly (Feiman-Nemser
2001; Feiman-Nemser and Parker 1993; Little 1990a), even within schools (Johnson and
The Project on the Next Generation of Teachers 2004; Kardos et al. 2001).
What might mentoring of new teachers look like in the ideal? Models of mentoring and
induction programs exist (Breaux and Wong 2003; Fideler and Haselkorn 1999; Huling5

For example, for more information on the New Teacher Center, visit www.newteachercenter.org.

123

J Educ Change (2010) 11:2344

27

Austin 1990; Stansbury and Zimmerman 2000; Villani 2002; Zeichner 1979). A composite
of their successful features suggests that, in the ideal, new teachers have mentors who help
them meet the challenges of being a beginning teacher (Feiman-Nemser 1983; FeimanNemser 2001; Gold 1996; National Commission on Teaching and Americas Future 2003;
Veenman 1984) in the context of a strong, trusting relationship (Daloz 1986; Gold 1996;
Villani 2002). Mentors help novice teachers decide what to teach and how to teach,
advising them about how to choose, adapt, or create appropriate materials and instructional
practices. Mentors help new teachers learn to manage their classrooms and develop
strategies for succeeding with particular students. Mentors observe them in their classroom,
model good teaching, and share materials and ideas. In short, the mentors work with the
new teacher is focused on the central components of teaching: classroom instruction,
curriculum and lesson planning, and classroom management (Darling-Hammond 1999;
Evertson and Smithey 2000; Feiman-Nemser 1983; Holloway 2001; Humphrey et al. 2000;
Wilson et al. 2001). Mentors help new teachers learn the modes of professional practice in
the school and adjust to their school and the families it serves (Kardos et al. 2001; Villani
2002). Mentors help new teachers understand and adjust to new reform and school change
efforts, both from inside and outside the schools (Geschwend and Moir 2007; Useem
2001). In addition, effective mentors know their proteges well enough to realize when to
grant them their privacy and independence (Little and McLaughlin 1993).
New teachers need mentors who know and understand the school in which they work
and the students and subjects they teach (Berry et al. 2002; National Commission on
Teaching and Americas Future 2003; Youngs 2002). They need mentors who themselves
are or have been expert and reflective teachers. New teachers need protected time and
space to meet with their mentors (Feiman-Nemser 1996), whose purpose is to support them
in their development as teachers (Moir et al. 1999). They need to work within a professional culture where colleagues anticipate their need for help, and not only encourage them
to seek it out, but also, routinely provide easy access to assistance (Feiman-Nemser 1996;
Johnson and The Project on the Next Generation of Teachers 2004; Kardos et al. 2001;
Little 1990a; Rosenholtz 1989). As Little explains (1990a), Given the structural and
cultural constraints on mentoring, its salience is likely to depend on the degree to which it
is congruent with other forms of support in the lives of beginning teachers (p. 331). Thus,
the context in which the mentoring is embedded will influence its usefulness to new
teachers. Most mentoring programs, however, rely primarily on the good will and generosity of experienced teachers motivated only by a small stipend and their own generous
readiness to share their accumulated wisdom with a novice colleague (Berry et al. 2002).
With a mentoring ideal in mind, we sought to describe in a generalizable way what new
teachers experiences are with their official mentors and compare these experiences across
various subgroups of teachers and schools. We were most concerned ultimately with
support and retention of new teachers. Therefore, in addition to providing descriptive
statistics of the entire sample, we focus on three important subgroups. We compare the
experiences of new teachers in low-income and high-income schools because low-income
schools often have difficulty attracting and retaining strong candidates. We compare the
experiences of math, science, and technology teachers with novices who teach in other
subject areas because, again, math, science, and technology have proven to be disciplines
in which it is hard to attract and retain strong teaching candidates (Ingersoll 2000). Since
high schools are organized very differently than elementary and middle schools, we
compare the experiences of new high school teachers with elementary/middle
school teachers to examine whether new teachers at these school levels have different
mentoring experiences.

123

28

J Educ Change (2010) 11:2344

In the following discussion, we present the questions that guided this inquiry, the
research design, and a summary of the findings. We explain how a cursory look at mentoring presents a more positive picture than really exists. Embedded in the detailsthe
characteristics of the match and the nature of the interactions between the mentor and the
new teacher (i.e., classroom observations, conversations about teaching)is a more
worrisome reality of the support new teachers receive from their mentors. We then
compare and contrast three subgroups of new teachers and examine the inequities present
between new teachers experiences of mentoring in high- and low-income schools; the
differences in the experiences of new math, science, or technology teachers and those who
teach other subjects; and the general similarities between new teachers in high schools and
new teachers in elementary and middle schools.

Focus of the inquiry

The inquiry focuses on the presence of mentoring, the characteristics of the mentor match,
and the nature of the interactions between the mentor and the new teacher.
Presence of Mentoring: What proportions of new teachers have official mentors during
their first year?
Characteristics of the Mentor Match: How much experience do the mentors have? Are
the mentors in the same schools, grade levels, and subject areas as the new teachers they
are assigned to mentor?
Nature of the Interactions between the Mentor and New Teacher: What proportions of
new teachers are observed by their mentors while teaching? Do they talk about
classroom instruction, curriculum and lesson planning, and classroom management and
discipline?
Research design6
Sites
This study draws upon data from three states: Florida, Massachusetts, and Michigan.7 All
are experiencing some degree of teacher shortage; all have alternative routes to certification; all have charter school legislation; all have adopted standards in core subjects; all
use criterion-referenced assessments aligned to standards; and all are collective bargaining

We worked with Edward Liu to collaboratively design this 4-state study and the 1-state New Jersey study
on which it is built (Kardos 2001; Liu and Kardos 2002). We extend special thanks to John B. Willett and
Robert S. Peterkin for their assistance with this study.
7

This study is part of a 4-state survey study of a representative random sample of new teachers that also
included California. We omit California in this analysis for three reasons: (1) demographic and other
relevant characteristics about CA make it unusual (for example, California has greater than four times more
public schools and greater than six times more public school students than Massachusetts); (2) concerns
about potential bias in the CA sample resulting from low response rates in certain subgroups in CA (for
example, in California, the group of responding schools included a much lower proportion of middle schools
than the group of non-responding schools); (3) a swamping effect resulting from the sample weights used to
correct for the study design. Because the responses of California teachers were given much greater weight
than the teachers in the other three states, population estimates for the four states together look almost
exactly like the estimates for CA alone.

123

J Educ Change (2010) 11:2344

29

states. Table 1 summarizes data from the state reports of Education Weeks Quality
Counts 2003 (2003).
The states vary in geographic size, average size of schools, composition of student
populations, student achievement, per pupil spending, teacher salaries, teacher preparation,
and number of charter schools. They are located in different regions of the country.
Massachusetts has the smallest number of schools, teachers, and students, and Florida has
the smallest percentage of small schools. Nearly half of Floridas students are members of
minority groups, while nearly one-quarter of the students in both Massachusetts and

Table 1 Summary of selective state level characteristics for Florida, Massachusetts, and Michigan, 2002
FL

MA

MI

Size
Number of public schools

3,231

1,898

3,743

Number of public school teachers

136,000

69,000

97,000

Number of public school students (pK-12)

2,500,000

980,000

1,734,000

Percent of students in elementary schools

with 350 or fewer students (2001)

3%

27%

28%

Percent of students in high schools with

900 or fewer students (2001)

6%

33%

37%

Percent minority students

47%

24%

25%

Percent children in poverty

22%

14%

17%

Percent students with disabilities

15%

16%

13%

Percent english-language learners

10%

5%

3%

Selected student demographics

Selected student achievement and spending

Percent of 8th graders scoring at or above proficient
on NAEP math (2000)

N/A

32%Total

28%Total

37%White

35%White

8%Black

2%Black

14%Hispanic

9%Hispanic

Statewide graduation rates

55%

73%

N/A

State average education spending per student

(adjusted for regional cost differences)

$8,429

$6,161

$6,512

Average teacher starting salaries, adjusted

for the cost of living (2001)

$27,387

$27,198

$30,188

Average teacher salaries, adjusted for cost

of living (2001)

$40,604

$41,773

$51,868

Number of teacher participants in state

alternative route programs

180

200

N/A

Percent of teacher graduates from

NCATE-accredited teacher education
programs (2001)

79%

76%

69%

232

47

186

Selected teacher pay and preparation

Charter schools
Number of charter schools

These data come from the state data and state reports of Education Weeks Quality Counts 2003: The
Teacher Gap (2003)
N/A appears where data was not available

123

30

J Educ Change (2010) 11:2344

Michigan are. Florida has the highest percentage of children in poverty and the highest
percentage of English-language learners of the three states. Florida spends the most per
student, and Michigan has the highest average teacher salary. All three states have similar
starting salaries (adjusted for the cost of living), with Michigans being slightly higher.
In 2002, at the time of data collection, both Massachusetts and Michigan had statewide
induction programs for new teachers; Massachusetts both required and funded these programs, Michigan did not. Both states also had required mentoring programs; Massachusetts
required districts to provide one year of new teacher mentoring and Michigan required
three (Quality Counts 2003). Quality Counts data were not available for Florida; however,
Florida previously had a mandated statewide induction program which, by the time these
data were collected, was no longer funded and no longer required. Districts were instructed
to provide mentoring for new teachers.
The sample
The sample consists of 374 first-year and second-year full-time, K-12 public school
teachers (including foreign language and special education, excluding arts and physical
education) from three states, Florida, Massachusetts, and Michigan. To draw the sample we
used two-stage stratified cluster sampling (Levy and Lemeshow 1999; Light et al. 1990;
Rea and Parker 1997), which we had pilot tested in New Jersey (Kardos 2001; Liu and
Kardos 2002). In Stage 1 of our sampling process, we stratified the sample by state, school
level (elementary, middle, high),8 and school type (charter, non-charter), in order to ensure
adequate representation along each stratum. Working from lists of schools from the U.S.
Department of Educations Common Core of Data, we drew a total of 199 schools: 58 in
Florida, 62 in Massachusetts, and 79 in Michigan. We over-sampled in the smaller states
and under-sampled in the larger ones to enable us to conduct supplementary analyses
within each state. In addition, we drew more schools in Michigan, because Michigan was
experiencing less of a teacher shortage than the other three states, and had fewer new
teachers per school. We also over-sampled charter schools to facilitate future subgroup
analysis. In our analyses, we incorporated sampling weights to correct for the over- and
under-sampling.
We drew the sample of schools in proportion to the number of students in each school,
which served as a proxy for the number of new teachers, an unknown quantity (Levy and
Lemeshow 1999). We contacted principals in each of the schools, by phone and through
the mail, and asked for the names and teaching assignments of all first-year and secondyear teachers in their schools. Seventy-two percent of the selected schools agreed to
provide lists of teachers.9
All new teachers in each randomly selected school were included in the sample (Stage 2
of the sampling process). We were given the names of 564 first-year and second-year
teachers in Florida, Massachusetts and Michigan. We mailed each new teacher an introductory letter, quickly followed by the questionnaire with an accompanying cover letter.
As an incentive to participate, all respondents who returned completed surveys were sent a
8

Elementary, middle, and high schools differ in ways likely to impact professional culture such as organizational structure (Rowan 1990) and the ways in which new teachers interact with their experienced
colleagues, both formally and informally. It is, therefore, important to ensure that the sample does not
contain a disproportionate number of high schools, which might result from just sampling proportional to
size.

The school response rates for each state were: 71 percent in Florida; 82 percent in Massachusetts; and 71
percent in Michigan.

123

J Educ Change (2010) 11:2344

31

$15 gift certificate to a major online bookseller. We sent a series of reminders to nonrespondents over the course of the late winter and spring of 2002.
We achieved a teacher response rate for these three states of 66% (374 teachers) using
strategies that proved effective in our pilot study (Dillman 1991; Kardos 2001; Keiley
1996; Liu and Kardos 2002).10 Analysis of patterns of response and non-response suggests
that we have a reasonably representative sample. To explore possible sources of selection
bias, we used data from our survey and public sources to compare the group of responding
schools to the group of non-responding schools,11 and the group of responding teachers to
the group of non-responding teachers.
Chi-squared analysis revealed that there are no statistically significant differences
between responding and non-responding schools in the following measures: average faculty size, average size of student population, percentage of students eligible for free or
reduced-price lunch, eligibility for Title I funds, and percentage of Black and Hispanic
students. This is true for both the 3-state sample and the individual state samples. One
possible source of bias, however, is school level. In Florida, the responding schools
included a higher proportion of elementary schools and a lower proportion of middle
schools than the non-responding schools. However, in Massachusetts and Michigan, there
appear to be no significant group differences in school level.
In Stage 2 of our sampling process, at the level of the individual teacher, there are no
(or very minor) differences between responding teachers and non-responding teachers in
terms of the following: gender, teaching experience (first-year or second-year), school type
(charter school or conventional), grade level, primary teaching assignment, and school
locale (urbanicity). One notable exception is that, in Michigan, non-respondents were more
likely than respondents to teach in urban schools and schools with higher proportions of
Black and Hispanic students.12
Measures and data analysis
The instrument used in this analysis is a mail survey containing 18 general information
questions, 18 items on teacher satisfaction, 8 items on teacher career decision-making, and
92 items about professional culture. The instrument also contains a section on teacher
hiring, not used in this analysis. The questions are based on a review of the literature and a
prior qualitative study (Kardos et al. 2001). The instrument was pilot tested in a study of
110 New Jersey teachers (Kardos 2001; Liu and Kardos 2002) and further refined after a
series of focus groups.13
The section on professional culture includes questions on formal and informal mentoring, classroom observations, official and informal meetings, teacher interaction, novice
status, collective responsibility, and the principal. Only data from questions about mentoring are presented and interpreted here. Most items about mentoring were categorical
(the majority were dichotomous) and a few were continuous measures.

10

Individual response rates in each of the three states are as follows: 63 percent in Florida; 67 percent in
Massachusetts; and 69 percent in Michigan.
11
School-level data were obtained from the U.S. Department of Educations Common Core of Data and
from state departments of education. We consulted state databases to fill in gaps in the Common Core of
Data.
12

Additional information on sampling and response is available from the authors.

13

Additional information on instrument development is available from the authors.

123

32

J Educ Change (2010) 11:2344

We sought to describe, in a generalizable way, new teachers experiences of official

mentoring during their first year: Did they have a mentor? What were the characteristics of
the match? And to what extent did new teachers and mentors have interactions that focused
on the core activities of the new teachers teaching (either through classroom observations
or through conversations about classroom management, lesson planning, or classroom
instruction)? We conducted descriptive analyses of the questionnaire data and summarized
new teachers experiences in a series of comparative tables which describe the individual
items and differences by certain individual and school level subgroups.

Summary of findings: the good, the bad, and the inequity

The good news is that a large proportion of new teachers had official, experienced mentors.
This appears, on its face, very promising. If supporting new teachers meant simply supplying them with a mentor, policymakers and educational leaders have assured that most
new teachers were well supported. However, a more detailed examination of new teachers
experiences with their mentors reveals a more complicated, and less reassuring story.
The bad news lies in the finer-grained details of the new teacher-mentor experience: (1)
the characteristics of the mentor match and (2) the nature of the interactions between the
mentor and the new teacher, particularly about interactions specifically dealing with
teaching practice. When we examined the details of the mentor matchwhether the
mentor was in the same school, same grade level, and same subject areawe found large
proportions of new teachers in what seemed to be less than ideal pairings. When we
examined data about the nature of the interactions between the mentor and the new
teacherwhether the mentor observed the new teacher and whether the mentor and the
new teacher had conversations about the core activities of teachingwe found large
proportions of new teachers who had never been observed by their mentors or did not have
at least three conversations with their mentors about classroom management, lesson
planning, and classroom instruction.
Finally, there are also important differences between the experiences of new teachers in
high-income schools and the experiences of new teachers in low-income schools. New
teachers in low-income schools (where more than 50 percent of the students qualify for
free or reduced-price lunch) had what would be considered desirable matches in lower
proportions than their counterparts in high-income schools (where less than 15 percent of
the students qualify for free or reduced price lunch). In addition, new teachers in lowincome schools had substantive discussions with their mentors about teaching in lower
proportions than their counterparts in high-income schools. Likewise, we found important
differences between the experiences of new teachers who teach math, science, or technology and those who do not. New math, science, or technology teachers were less likely to
have desirable matches or substantive interactions than their counterparts who teach other
subjects.14 These differences have a range of implications for low-income schools and the
disciplines of math, science, or technology. Of great concern here is whether these schools
and these disciplines will be able to attract, support, and retain new teachers in sufficient
numbers and whether the new teachers they do attract are able to understand and enact
school change efforts.

14
Notably, we found few significant differences between high school teachers and elementary/middle
school teachers.

123

J Educ Change (2010) 11:2344

33

Presence of a mentor
Seventy-eight percent of new teachers in Florida, Massachusetts, and Michigan had official
mentors during their first year. Seventy-eight percent may seem high overall, but new
teachers in the traditionally hard-to-staff schools and disciplines had official mentors in
lower proportions. New teachers in low-income schools had mentors in lower proportions
(65 percent) than their colleagues in high-income schools (91 percent). Math, science, and
technology teachers had mentors in lower proportions (61 percent) than their colleagues in
other teaching assignments (79 percent). Elementary/middle school teachers had official
mentors during their first year in slightly lower proportions (77 percent) than high school
teachers (79 percent).

Characteristics of the mentor match

The characteristics of the mentor match suggest, at first, that new teachers mentors had
what they needed to support new teachers well. The mentors were experienced teachers
who, for the most part, taught in fairly close proximity to the new teachers they were
supposed to serve. On average, new teachers in Florida, Massachusetts, and Michigan had
mentors with approximately 16 years of experience (se 0.8). Although there was some
variation among the subgroups of new teachers, in all categories examined, new teachers
had mentors with an average of at least 10 years of experience. Furthermore, a large
proportion of them had mentors who teach in their school (68 percent). At the very least,
having a mentor in the same school would seem to increase their access to assistance from
an experienced colleague.
However, new teachers were far less likely to have a mentor matched by grade-level
(44 percent) and subject (48 percent). This is notable, given reported findings that
beginning teachers with a same-subject mentor are less likely to leave their school and to
leave teaching after one year (Smith and Ingersoll 2004). It is further notable that new
teachers who work in low-income schools (where more than 50 percent of the students
qualify for free or reduced-price lunch) had an official mentor teaching in the same
school, grade level, and subject in lower proportions than new teachers in the wealthiest
schools (where fewer than 15 percent of the students qualify for free or reduced-price
lunch).15 Math, science, and technology teachers had same-school and same-subject
matches in lower proportions than teachers of other subjects; however, they had gradelevel matches in higher proportions.16 High school teachers had same-school, samegrade-level and same-subject matches in higher proportions than their counterparts in
elementary/middle schools, with a particularly large difference for same-grade-level
matches.17

15
School economic status breakdowns follow those published in the Education Watch State Summaries
by the Education Trust (2003). High-income schools are defined as schools where less than 15 percent of the
students qualify for free or reduced-price lunch. Low-income schools are defined as schools where greater
than 50 percent of the students qualify for free or reduced-price lunch.
16

These differences were not statistically significant.

17

The differences for same-school and same-subject mentor matches were not statistically significant. The
very large difference for same-grade-level matches was statistically significant.

123

34

J Educ Change (2010) 11:2344

The nature of the interactions between the new teacher and the mentor
In general, new teachers did not have the sort of face-to-face interaction about the core
work of teaching that one might expect from a mentoring relationship. Only 41 percent
were observed by their mentor in their classroom. By the spring of their first year, only 58
percent had even 3 conversations with their mentor about classroom management and
discipline, only 58 percent had conversations about curriculum and lesson planning, and
only 56 percent had conversations about classroom instruction. Although new teachers in
low-income schools were more likely to be observed at least once by their mentors (42
percent) than teachers in high-income schools (31 percent), they were less likely to have
even three conversations about classroom management (43 percent compared to 69 percent), lesson planning (47 percent compared to 69 percent), and classroom instruction (47
percent compared to 61 percent).18
New teachers of math, science, or technology were less likely to be observed by their
mentors than new teachers of other subjects (24 percent compared to 43 percent). They
were also less likely to have had three conversations about the core areas of teaching. They
had conversations in lower proportions about classroom management and discipline (54
percent compared to 58 percent), curriculum and lesson planning (43 percent compared to
60 percent), and classroom instruction (42 percent compared to 57 percent).19 Comparing
new teachers in high schools with those in elementary/middle schools, however, we found
no significant difference in the nature of the interactions with their mentors (classroom
observations and conversations about classroom management, lesson planning, and
classroom instruction).

Inequities among schools

These data, summarized and contrasted in Table 2 below, indicate consistent, statistically
significant differences in new teachers official mentoring experiences depending on
whether they are in high-income or low-income schools. In this analysis, high-income
schools are defined as schools where fewer than 15 percent of the students qualify for free
or reduced-price lunch. Low-income schools are defined as schools where more than 50
percent of the students qualify for free or reduced price lunch.
When pooled, 78 percent of all new teachers in Florida, Massachusetts, and Michigan
were assigned official mentors by their school or district during their first year. Ninety-one
percent of new teachers in high-income schools had official mentors, while only 65 percent
of new teachers in low-income schools had official mentors. The difference of 26 percentage points is statistically significant at the 0.01 level (see Table 2, row 1). Although
this is certainly a stark and potentially important difference, it is possible that the mere
presence of an official mentor may matter less to a new teacher than the characteristics of
that mentor or the nature of the interactions between the mentor and the new teacher.
When we examine the characteristics of the mentor matchwhether the mentor teacher
was located at the same school (Table 2, row 2), in the same grade level (Table 2, row 3),
18
All differences were statistically significant except the difference in classroom observation and conversations about classroom instruction.
19
Differences in classroom observation and conversations about curriculum and lesson planning were
statistically significant. Differences in conversations about classroom management and discipline and
classroom instruction were not.

123

J Educ Change (2010) 11:2344

35

Table 2 Comparison of new teachers official mentoring experiences during the first year in high-income
and low-income schools in FL, MA, and MI (n = 374)
Difference

All new
teachers

New
teachers:
high-income
schools

New
teachers:
low-income
schools

78% (5.0)

91% (3.1)

65% (9.6)

26%** (10.0)

Has a mentor who is in the same

school

68% (5.6)

82% (6.0)

53% (10.4)

29%* (11.9)

Has a mentor who is in the same grade

level

44% (3.9)

61% (5.4)

28% (5.6)

Has a mentor who is in the same

subject

48% (4.7)

60% (8.1)

40% (7.6)

20%* (11.0)

Was observed at least once by

a mentor

41% (5.1)

31% (6.5)

42% (8.8)

11% (11.1)

Has at least three conversations

with a mentor about classroom
management and discipline

58% (5.2)

69% (6.9)

43% (9.0)

26%* (11.3)

Has at least three conversations with

a mentor about curriculum and
lesson planning

58% (4.6)

69% (5.0)

47% (8.6)

22%* (10.0)

Has at least three conversations

with a mentor about classroom
instruction

56% (4.8)

61% (6.7)

47% (8.8)

14% (11.0)

Presence of a mentor
Percentage of new teachers who have
a mentor
Characteristics of the mentor match

33%*** (7.7)

Nature of interactions

All statistics take into account the complex nature of the survey sample; standard errors are in parentheses
* p \ .10, * p \ .05, ** p \ .01, *** p \ .001
High incomeSchools with less than 15 percent of students eligible for free or reduced price lunch
Low incomeSchools with more than 50 percent of students eligible for free or reduced price lunch

or in the same subject as the new teacher (Table 2, row 4)we see that new teachers in
high-income schools experienced higher proportions of match on each. While 82 percent of
new teachers in high-income schools had same-school mentor matches, only 53 percent of
new teachers in low-income schools did. The 29 percentage point difference is large and
statistically significant at the 0.05 level. While 61 percent of new teachers in high-income
schools had same-grade level mentors, only 28 percent of new teachers in low-income
schools did. And while 60 percent of new teachers in high-income schools had samesubject mentors, only 40 percent of new teachers in low-income schools did. The 33
percentage point difference for same-grade level mentors is statistically significant at the
0.001 level, and the 20 percentage point difference for same-subject mentors is statistically
significant at the 0.10 level.
Although having a mentor in the same grade level or same subject certainly would not
guarantee an ideal match, it might increase the chances that the mentor and the new teacher
would share students or have common curricular concerns. It is important to note, however,
that while same-subject or same-level mentoring might increase the chance for interaction
between a new teacher and a mentor, there is less chance for meaningful exchange when
teachers responsibilities are not also entwined.

123

36

J Educ Change (2010) 11:2344

It is also important to examine the nature of the interaction between the new teacher and
the mentor, and the extent to which they talk about the substantive challenges of teaching:
classroom instruction, curriculum and lesson planning, and classroom management and
student discipline. New teachers in high-income schools had at least three conversations with
their mentors about all three core activities of teaching (classroom management, lesson
planning, and classroom instruction) in higher proportions than new teachers in low-income
schools (Table 2, rows 68). It is important to note that the bar in these measures is set
extremely low. It is not much for a new teacher to expect to be observed by a mentor at least
once or to have at least three conversations about classroom management, lesson planning,
and classroom instruction by the spring of the new teachers first year. What this means, then,
is that many new teachers are getting virtually nothing from their mentors (i.e., have not even
cleared this very low bar). They have not been observed even once in their classroom by their
mentors (59 percent of all new teachers), and they have not had even three conversations
about classroom management (42 percent of all new teachers), lesson planning (42 percent of
all new teachers), or classroom instruction (44 percent of all new teachers) with their mentors
by spring of their first year. For many new teachers, formal mentoring is the only official
structure that the school or district has in place to assist them.
Although 69 percent of new teachers in high-income schools had conversations with
their mentors about classroom management and discipline, only 43 percent of new teachers
in low-income schools did. The 26 percentage point difference is statistically significant at
the 0.05 level (Table 2, row 6). Sixty-nine percent of new teachers in high-income schools
had conversations with their mentors about curriculum and lesson planning, while only 47
percent of their counterparts in low-income schools did. The 22 percentage point difference
is statistically significant at the 0.05 level (Table 2, row 7). Sixty-one percent of new
teachers in high-income schools had conversations with their mentors about classroom
instruction, while only 47 percent of their counterparts in low-income schools did. The 14
percentage point difference is large, though not statistically significant (Table 2, row 8).
These low percentages for both sub-groups, but particularly for new teachers in lowincome schools, are cause for concern.20 These teachers are brand new, and decades of
research document the many challenges they face as novice teachers. Indeed, they need
ongoing conversations with experienced teachers about the core work of teaching. Experienced teachers have experimented with and refined classroom practices and strategies. They
have a sense of what works, and many of them already hold the bag of tricks for which new
teachers yearn. In addition to their instructional expertise, experienced teachers hold a sense
of institutional history, organizational knowledge, and community understanding. Knowledge and wisdom accumulate in schools over time, and must be interpreted and handed
carefully to new teachers. New teachers bemoan their professional isolation, particularly
when there are others nearby who already know some of what they need to know.21
20

Notably, 100 percent of both groupsnew teachers in high-income and low-income schoolsshould be
reasonably expected to have at least three conversations with their mentors about these three substantive
topics related to their teaching by the spring of their first year. Therefore, while the percentages in highincome schools are higher than those in low-income schools, arguably, they are also not high enough.
21
It is true that some schools experience large influxes of new teachers and, therefore, face shortages of
experienced teachers able and willing to support the incoming beginners. Still, our qualitative study of new
teachers (Johnson and The Project on the Next Generation of Teachers 2004) suggests that this new
generation of teachers differs from the generation of teachers described by Lortie (1975) in his landmark
sociological study. Lortie described teachers as preferring privacy and autonomy. While this new generation
of teachers values their independence, they also want feedback on their teaching, opportunities for collaborative work, and classroom observations. The new teachers in our study said they wanted experienced
teachers to help show them the way.

123

J Educ Change (2010) 11:2344

37

One exception to the pattern in these data is in the proportion of new teachers who were
observed by their mentors. Although the difference is not statistically significant, a larger
proportion (42 percent) of new teachers was observed by their mentors in low-income
schools than in high-income schools (31 percent). These findings are surprising, since one
might expect that schools with higher proportions of economically disadvantaged children
might lack sufficient resources and, thus, that the teachers (including mentors and mentees)
might be overscheduled and have less time to observe and meet with colleagues. On the
other hand, since only 53 percent of new teachers in low-income schools had same-school
mentors, nearly half of the new teachers (47 percent) had mentors from outside their
school. We have no data on how the assignments are made, who the mentors are, how they
are trained, and how their work is structured. Therefore, it is impossible to determine what
this might mean and whether these observations are supportive, evaluative, useful, or
counter-productive for new teachers. Since school level poverty concentration has been
found to be associated with an increased risk of new teachers leaving teaching after their
first year (Smith and Ingersoll 2004) it is important to better understand all aspects of new
teachers experiences in low-income schools.

Differences between new math, science, or technology teachers and new teachers
of other subjects
New teachers of math, science, or technology, as Table 3 row 1 shows, were less likely to
have official mentors assigned to them by the school or district (61 percent) than new
teachers of other subjects (79 percent). The 18 percentage point difference, though large, is
not statistically significant.
Beyond the mere presence of an official mentor, new teachers of math, science, or
technology were less likely than teachers of other subjects to have a same-school mentor
(60 percent compared to 69 percent) or same-subject mentor (42 percent compared to 49
percent). By contrast, new teachers of math, science, or technology had same-grade-level
mentors in higher proportions (52 percent compared to 43 percent) than their counterparts
in other subjects. Notably, these differences were all relatively small (less than 10 percentage points) and not statistically significant. Thus, while these proportions for teachers
of all subjects are low overall, they are not significantly different depending on whether the
new teacher teaches math, science, or technology or some other subject. Still, they are
noteworthy, given recent findings (Smith and Ingersoll 2004) that new math and science
teachers are about 10 percent more likely than other teachers to leave teaching.
As the last three rows of Table 3 show, new teachers of math, science, or technology were
somewhat less likely than their counterparts, whose primary teaching assignment was in
other subject areas, to have at least three conversations with their mentors about classroom
management and student discipline (54 percent compared to 58 percent), curriculum and
lesson planning (43 percent compared to 60 percent), and classroom instruction (42 percent
compared to 57 percent) by the spring of their first year. The 4 percentage point difference in
the category conversations about classroom management and student discipline is negligible (Table 3, row 6); however, the differences in the other two categories are large.
Teachers who teach subjects other than math, science, or technology had conversations with
their mentors about classroom instruction at a proportion that is 15 percentage points higher
than the group of math, science or technology teachers (Table 3, row 8).22 And teachers of
22

This difference, though large, is not statistically significant.

123

38

J Educ Change (2010) 11:2344

Table 3 Comparison of new teachers official mentoring experiences during the first year for math, science, or technology teachers and teachers of other subjects in FL, MA, and MI (n = 374)
Difference

All new
teachers

New math,
science, or
technology
teachers

New teachers who

teach subjects other
than math, science,
or technology

78% (5.0)

61% (16.8)

79% (4.0)

18% (15.0)

Has a mentor who is a mentor

in the same school

68% (5.6)

60% (16.6)

69% (5.2)

9% (15.4)

Has a mentor who is a mentor

in the same grade level

44% (3.9)

52% (15.4)

43% (3.5)

9% (14.6)

Has a mentor who is a mentor

in the same subject

48% (4.7)

42% (13.9)

49% (4.4)

7% (12.6)

Was observed at least once

by a mentor

41% (5.1)

24% (9.4)

43% (5.0)

Has at least three

conversations with
a mentor about classroom
management and discipline

58% (5.2)

54% (15.6)

58% (4.9)

4% (14.9)

Has at least three

conversations with a mentor
about curriculum and lesson
planning

58% (4.6)

43% (12.2)

60% (4.2)

17%* (11.5)

Has at least three

conversations with a mentor
about classroom instruction

56% (4.8)

42% (12.7)

57% (4.5)

15% (11.9)

Presence of a mentor
Percentage of new teachers
who have a mentor
Characteristics of the mentor match

Nature of interactions
19%* (9.6)

All statistics take into account the complex nature of the survey sample; standard errors are in parentheses
* p \ .10, * p \ .05, ** p \ .01, *** p \ .001

subjects other than math, science, or technology had conversations about curriculum and
lesson planning at a proportion that is 17 percentage points higher than their counterparts
(Table 3, row 7). The difference is statistically significant at the 0.10 level. Finally, only 24
percent of new math, science, or technology teachers were observed in their classes at least
once by their mentors, compared to 43 percent of new teachers in other subjects. The 19
percentage point difference is statistically significant at the 0.05 level (Table 3, row 5).
Low proportions of both groups of new teachersthose who primarily teach math,
science, or technology and those who teach other subjectsreported having interactions
with the mentors that focused on their teaching. As much as new teachers, in general, were
left alone to face the difficult challenges of teaching, the math, science, and technology
teachers were even worse off.

Similarities between school levels

When we compared new teachers experiences of mentoring across school levels, we found
little difference between those in high schools and elementary/middle schools. As Table 4

123

J Educ Change (2010) 11:2344

39

Table 4 Comparison of new teachers official mentoring experiences during the first year in high schools
and elementary/middle schools in FL, MA, and MI (n = 374)
Difference

All new
teachers

New high
school
teachers

New elementary
or middle school
teachers

78% (5.0)

79% (7.6)

77% (5.4)

2% (9.4)

Has a mentor who is a mentor

in the same school

68% (5.6)

76% (8.3)

67% (6.1)

9% (10.3)

Has a mentor who is a mentor

in the same grade level

44% (3.9)

70% (10.1)

42% (3.9)

28%** (10.9)

Has a mentor who is a mentor

in the same subject

48% (4.7)

59% (12.4)

47% (5.0)

12% (13.4)

Was observed at least once

by the mentor

41% (5.1)

37% (7.5)

42% (5.5)

5% (9.3)

Has at least three conversations with

mentor about classroom
management and discipline

58% (5.2)

61% (6.5)

58% (5.7)

3% (8.6)

Has at least three conversations with

mentor about curriculum
and lesson planning

58% (4.6)

63% (5.1)

58% (5.0)

5% (7.2)

Has at least three conversations with

mentor about classroom instruction

56% (4.8)

55% (8.4)

56% (5.2)

1% (9.9)

Presence of a mentor
Percentage of new teachers who have
a mentor
Characteristics of the mentor match

Nature of interactions

All statistics take into account the complex nature of the survey sample; standard errors are in parentheses
* p \ .10, * p \ .05, ** p \ .01, *** p \ .001

shows, consistent with the overall percentage of all new teachers who had official mentors
assigned to them during their first year (78 percent), new teachers in high schools and new
teachers in elementary/middle schools had official mentors assigned to them in virtually
the same proportions (79 percent and 77 percent respectively).
There were notable differences in the characteristics of the mentor match between new
high school teachers and their counterparts in elementary and middle schools. New
teachers in high schools had same-school mentors in higher proportions (76 percent) than
their counterparts in elementary or middle schools (67 percent), a 9 percentage point
difference (not statistically significant). This might be explained by the fact that high
schools are generally larger than elementary and middle schools and thus, may have a
larger pool of in-school mentors to draw from. New high school teachers had same-gradelevel mentor matches in much higher proportions (70 percent) than their counterparts in
elementary or middle schools (42 percent). The 28 percentage point difference is statistically significant at the 0.01 level. New high school teachers had same-subject mentor
matches in higher proportions (59 percent) than their counterparts in elementary/middle
schools (47 percent), a 12 percentage point difference which is large, but not statistically
significant. The larger proportions of math/science/technology teachers with same-grade
level and same-subject mentors makes sense, since math and science assignments often
differ by grade and subject (e.g., 9th grade Algebra, 10th grade biology, 11th grade

123

40

J Educ Change (2010) 11:2344

chemistry), and high schools might have the human resources and the incentive to pair new
teachers with mentor teachers who teach the same subject or grade.
Overall, the percentages of new teachers who were observed in their classrooms and
who had substantive conversations with their mentor teacherwhose job it is, presumably,
to support and advise as the novice learns to teachare low by any reasonable standard in
both high schools and elementary/middle schools. New teachers were observed by their
mentor in low proportions (approximately 40 percent), and low proportions of them had at
least three conversations about classroom management, lesson planning, and classroom
instruction. Nearly 60 percent of the new teachers were not observed even once by their
mentors, and approximately 40 percent did not have even three substantive conversations
in the three core topic areas of teaching with their mentors by the spring of their first year.
However, there is little difference between the experiences of high school and elementary/
middle school teachers.

Conclusions and implications

Summary
The good news is simply that most first-year teachers (78 percent) are assigned experienced teachers as mentors by their school or district. In fact, a fairly high percentage (68
percent) has mentors in the same school. While this would seem to provide good support
for new teachers, it obscures the low percentages of new teachers who have same-gradelevel and same-subject matches with their mentors. Also, very low percentages of new
teachers are observed by their mentor and low percentages of new teachers have conversations with their mentor about the core activities of teaching. Furthermore, a
comparative examination reveals serious inequities between schools serving different
economic communities. The proportions of what might be considered ideal matches and
substantive teaching-focused interactions are lower for new teachers in low-income
schools than they are for new teachers in high-income schools. They are generally lower
for math, science, and technology teachers than for teachers of other subjects. Thus,
schools and subject areas that have been hard to staff and have had difficulty retaining
teachers (low-income schools and the subjects of math, science, and technology), are those
where new teachers are least likely to report substantive interactions with their mentors
about classroom teaching.
Implications for further research
This research assumes that novice teachers are better off with an experienced colleague by
their side than they would be if they had to go it alone. With these data, we can describe
new teachers experiences with their mentors in simple dichotomies: do they have mentors
or not? Do they teach in the same subjects or not? We can also describe, in some detail, the
nature of their interactions, but again only by dichotomies: are they observed by their
mentors or not? Do they meet with their mentors or not? Do they talk with their mentors
about their classes and their teaching or not? These descriptions are useful, since they are
both pointed and generalizable. Based on the large body of research about mentoring, we
can identify likely short-circuits in efforts to provide mentoring support for new teachers.
Furthermore, the data point to inequities in hard-to-staff schools and subjects, which are
likewise plagued by attrition problems.

123

J Educ Change (2010) 11:2344

41

While we present generalizable data on new teachers experiences with their mentors,
we do not know what effect this might have on new teachers sense of job satisfaction,
sense of efficacy, their effectiveness in the classroom, or their ultimate retention. Thus,
further research is required to test these relationships in more complex analyses.
The quality of mentoring may, in fact, affect many other aspects of new teachers work;
most important is its possible influence on new teachers effectiveness. It may also have an
impact on new teachers learning, professional growth, commitment to students, sense of
self-efficacy, and motivation. Given the recent upsurge in mandated mentoring, it is
important to understand whether and how differences in what is provided for new teachers
affects their work and satisfaction. It is particularly important to understand the consequences of deficient mentoring arrangements for new teachers in low-income schools and
in hard-to-staff subject areas.
Implications for policy and practice
Analysis of national data has shown that participation in induction programs (predominantly mentoring programs) increased dramatically between 1990 and 2000 (Smith and
Ingersoll 2004). Whereas, approximately half of public school teachers participated in
official support programs for new teachers in 1990, over 80 percent participated in 2000.
This has led to increased public spending and reallocation of human and administrative
resources to provide new teachers with mentors.
It is promising that educational leaders and policymakers are taking seriously the need
to support new teachers and that most new teachers have been assigned experienced
mentors. Funding and endorsement of policies and programs aimed at assisting teachers is
a hopeful sign, particularly in this current climate in which local, district, state, and
national school reform efforts are rapidly introducedmany of them with high-stakes for
students and for teachers. However, policymakers and educational leaders need to be
careful not to conclude mistakenly that mentored new teachers are uniformly wellsupported.
Those who endorse mentoring programs assume that the interactions are helpful to new
teachers and, in some way, support them in their first year of teaching. It makes sense that a
logic of confidence23 (Meyer and Rowan 1978) about the benefits of mentoring would
develop, given policymakers and educational leaders (and even new teachers) enthusiasm for the idea of mentoring and their real desire to support and retain high-quality
teachers in public schools. However, the bureaucratic structure of formal mentoring may
be disconnected from the relationships that form between new teachers and their mentors
and the interactions that occur in the context of those relationships. These findings suggest
that policymakers hopes and expectations for mentoring may not yet be realized, especially for new teachers in low-income schools. It is important to study these programs
closely in order to understand whether they deliver what they promise and, if not, how they
can be strengthened so new teachers can be better served.
Acknowledgments Research for this paper was conducted under the auspices of the Project on the Next
Generation of Teachers at the Harvard Graduate School of Education. Funding was provided by the Spencer
Foundation; however, the analysis and conclusions presented here are solely those of the authors.

23
This term refers to the existing belief that all participants are acting in good faith (Meyer and Rowan
1978).

123

42

J Educ Change (2010) 11:2344

References
Adelman, N. E. (1991). Preservice training and continuing professional development of teachers. Washington, D.C: Policy Studies Associates.
Berry, B., Hopkins-Thompson, T., & Hoke, M. (2002). Assessing and supporting new teachers: Lessons
from the Southeast. North Carolina: The Southeast Center for Teaching Quality at the University of
North Carolina.
Breaux, A. L., & Wong, H. K. (2003). New teacher induction: How to train, support, and retain new
teachers. Mountain View, CA: Harry K. Wong Publications, Inc.
Counts, Q. (2003). The Teacher Gap. Bethesda, MD.
Daloz, L. A. (1986). Effective teaching and mentoring: Realizing the transformational power of adult
learning experiences. San Francisco: Jossey-Bass Publishers.
Darling-Hammond, L. (1999). Solving the dilemmas of teacher supply, demand, and standards: How we can
ensure a competent, caring, and qualified teacher for every child. Washington, D.C.: National
Commission on Teaching and Americas Future.
Dillman, D. A. (1991). The design and administration of mail surveys. Annual Review of Sociology, 17,
225249.
Drago-Severson, E. (2004). Helping teachers learn: Principal leadership for adult growth and development.
Thousand Oaks, CA: Corwin Press.
Education Trust. (2003). Education watch state summaries. Washington, D.C.: Education Trust.
Evertson, C., & Smithey, M. (2000). Mentoring effects on proteges classroom practice: An experimental
field study. The Journal of Educational Research, 93(5), 294304.
Feiman-Nemser, S. (1983). Learning to teach. In L. S. Shulman & G. Sykes (Eds.), Handbook of teaching
and policy (pp. 150170). New York: Longman.
Feiman-Nemser, S. (1996). Mentoring: A critical review. Washington, D.C.: ERIC Clearinghouse on
Teaching and Teacher Education.
Feiman-Nemser, S. (2001). From preparation to practice: Designing a continuum to strengthen and sustain
teaching. Teachers College Record, 103(6), 10131055. doi:10.1111/0161-4681.00141.
Feiman-Nemser, S., & Floden, R. E. (1986). The cultures of teaching. In M. C. Witrock (Ed.), Handbook of
research on teaching (3rd ed., pp. 505526). New York: Macmillan.
Feiman-Nemser, S., & Parker, M. (1993). Mentoring in context: A comparison of two U.S. programs for
beginning teachers. International Journal of Educational Research, 19(8), 699718. doi:
10.1016/0883-0355(93)90010-H.
Feiman-Nemser, S., & Schwille, S. (2004, April). Tracing paths of influence in new teacher induction. Paper
presented at the American Educational Research Association, San Diego, CA.
Fideler, E. F., & Haselkorn, D. (1999). Learning the ropes: Urban teacher induction programs and practices
in the United States. Belmont, MA: Recruiting New Teachers, Inc.
Geschwend, L., & Moir, E. (2007). Growing together: New and veteran teachers support each other through
practices that target the needs of high school educators. Journal of Staff Development, 28(4), 2024.
Gless, J., & Moir, E. (2001). Teacher quality squared. Journal of Staff Development, 22(1), 62.
Gold, Y. (1996). Beginning teacher support: Attrition, mentoring, and induction. In J. Sikula, T. J. Buttery,
& E. Guyton (Eds.), Handbook of research on teacher education (2nd ed., pp. 548594). New York:
Simon & Schuster Macmillan.
Goodlad, J. (1984). A place called school: Prospects for the future. New York: McGraw-Hill.
Grissmer, D., & Kirby, S. N. (1997). Teacher turnover and teacher quality. Teachers College Record, 99(1),
4556.
Holloway, J. H. (2001). The benefits of mentoring. Educational Leadership, 58(8), 8586.
Huling-Austin, L. (1990). Teacher induction programs and internships. In W. R. Houston (Ed.), Handbook
of research on teacher education (pp. 535548). New York: Macmillan Publishing Company.
Humphrey, D. C., Adelman, N., Esch, C. E., Riehl, L. M., Shields, P. M., & Tiffany, J. (2000). Preparing
and supporting new teachers: A literature review. Menlo Park, CA: SRI International.
Ingersoll, R. M. (2000). Turnover among mathematics and science teachers in the U.S. National Commission on Mathematics and Science Teaching for the 21st Century.
Ingersoll, R. M. (2001). Teacher turnover and teacher shortages: An organizational analysis. American
Educational Research Journal, 38(3), 499534. doi:10.3102/00028312038003499.
Ingersoll, R. M., & Kralik, J. M. (2004). The impact of mentoring on teacher retention: What the research
says. ECS Research Review. Denver, CO: Education Commission of the States.
Ingersoll, R., & Smith, T. (2003). The wrong solution to the teacher shortage. Educational Leadership,
60(8), 3033.
Johnson, S. M. (1990). Teachers at work: Achieving success in our schools. New York: BasicBooks.

123

J Educ Change (2010) 11:2344

43

Johnson, S. M., & Birkeland, S. E. (2003). Pursuing a sense of success: New teachers explain their career
decisions. American Educational Research Journal, 40(3), 581617. doi:10.3102/00028312040003581.
Johnson, S. M., & The Project on the Next Generation of Teachers. (2004). Finders and keepers: Helping
new teachers survive and thrive in our schools. San Francisco: Jossey-Bass.
Kardos, S. M. (2001). New teachers in New Jersey schools and the professional cultures they experience:
A pilot study. Unpublished Special Qualifying Paper, Harvard University Graduate School of
Education, Cambridge, MA.
Kardos, S. M., & Johnson, S. M. (2007). On their own and presumed expert: New teachers experience with
their colleagues. Teachers College Record, 109(9), 20832106.
Kardos, S. M., Johnson, S. M., Peske, H. G., Kauffman, D., & Liu, E. (2001). Counting on colleagues: New
teachers encounter the professional cultures of their schools. Educational Administration Quarterly,
37(2), 250290. doi:10.1177/00131610121969316.
Kauffman, D., Johnson, S. M., Kardos, S. M., Liu, E., & Peske, H. G. (2002). Lost at sea: New teachers
experiences with curriculum and assessment. Teachers College Record, 104(2), 273300. doi:
10.1111/1467-9620.00163.
Keiley, M. K. (1996). Affect regulation in adolescents: Do males and females manage their feeling differently?. Cambridge: Harvard University Graduate School of Education.
Levy, P. S., & Lemeshow, S. (1999). Sampling of populations: Methods and applications (3rd ed.). New
York: John Wiley & Sons, Inc.
Light, R. J., Singer, J. D., & Willett, J. B. (1990). By design: Planning research on higher education.
Cambridge, MA: Harvard University Press.
Little, J. W. (1990a). The mentor phenomenon and the social organization of teaching. In C. Cazden (Ed.),
Review of research in education (Vol. 16, pp. 297351). Washington, D.C.: American Educational
Research Association.
Little, J. W. (1990b). The persistence of privacy: Autonomy and initiative in teachers professional relations.
Teachers College Record, 91(4), 509536.
Little, J. W., & McLaughlin, M. W. (Eds.). (1993). Teachers work: Individuals, colleagues, and contexts.
New York: Teachers College Press.
Liu, E., & Kardos, S. M. (2002). Hiring and professional culture in New Jersey schools. Cambridge: Project
on the Next Generation of Teachers at the Harvard Graduate School of Education.
Lortie, D. C. (1975). Schoolteacher: A sociological study. Chicago: University of Chicago Press.
McCabe, M. (2006). State of the states, Quality counts at 10: A decade of standards-based education.
Education Week, 25(17), 74.
McDonald, F. J. (1980). The problems of beginning teachers: A crisis in training. Vol. I. Study of induction
programs for beginning teachers. Princeton, NJ: Educational Testing Service.
McDonald, F. J., & Elias, P. (1983). The transition into teaching: The problems of beginning teachers and
programs to solve them. Summary Report. Princeton, NJ: Educational Testing Services.
McLaughlin, M. W. (1993). What matters most in teachers workplace context? In J. W. Little &
M. W. McLaughlin (Eds.), Teachers work: Individuals, colleagues, and contexts. New York: Teachers
College Press.
Meyer, J. W., & Rowan, B. C. (1978). The structure of educational organizations. In M. Meyer and
Associates (Eds.), Environments and organizations: Theoretical and empirical perspectives. San
Francisco: Jossey-Bass.
Moir, E., Gless, J., & Baron, W. (1999). A support program with heart: The Santa Cruz Project. In
M. Scherer (Ed.), A better beginning: Supporting and mentoring new teachers. Alexandria, VA:
Association for Supervision and Curriculum Development.
Murnane, R. J., Singer, J. D., Willett, J. B., Kemple, J. J., & Olsen, R. J. (1991). Who will teach?: Policies
that matter. Cambridge, MA: Harvard University Press.
National Commission on Teaching and Americas Future. (2003). No dream denied: A pledge to Americas
children summary report. Washington, D.C.: National Commission on Teaching and Americas Future.
Powell, A. G., Farrar, E., & Cohen, D. K. (1985). The shopping mall high school: Winners and losers in the
educational marketplace. Boston: Houghton-Mifflin.
Rea, L. M., & Parker, R. A. (1997). Designing and conducting survey research (2nd ed.). San Francisco:
Jossey-Bass.
Rosenholtz, S. J. (1989). Teachers workplace: The social organization of schools. New York: Longman.
Rowan, B. (1990). Commitment and control: Alternative strategies for the organizational design of schools.
In C. Cazden (Ed.), Review of research in education (Vol. 16). Washington, D.C.: American Educational Research Association.
Rust, F. O. (1994). The first year of teaching: Its not what they expected. Teaching and Teacher Education,
10(2), 205217. doi:10.1016/0742-051X(94)90013-2.

123

44

J Educ Change (2010) 11:2344

Ryan, K. (Ed.). (1970). Dont smile until Christmas: Accounts of the first year of teaching. Chicago: The
University of Chicago Press.
Ryan, K., Newman, K. K., Mager, G., Applegate, J., Lasley, T., Flora, R., et al. (1980). Biting the apple:
Accounts of first year teachers. New York: Longman Inc.
Sizer, T. R. (1984). Horaces compromise: The dilemma of the American high school. Boston: HoughtonMifflin.
Smith, T. M., & Ingersoll, R. M. (2004). What are the effects of induction and mentoring on beginning
teacher turnover? American Educational Research Journal, 41(3), 681714.
Stansbury, K., & Zimmerman, J. (2000). Lifelines to the classroom: Designing support for beginning
teachers. San Francisco: WestEd.
Troen, V., & Boles, K. C. (2003). Whos teaching your children? Why the teacher crisis is worse than you
think and what can be done about it. New Haven, CT: Yale University Press.
Useem, B. (2001, April). New teacher staffing and comprehensive school reform: Philadelphias experience.
Paper presented at the American Educational Research Association, Seattle, WA.
Veenman, S. (1984). Perceived problems of beginning teachers. Review of Educational Research, 54(2),
143178.
Villani, S. (2002). Mentoring programs for new teachers: Models of induction and support. Thousand Oaks,
CA: Corwin Press.
Wideen, M., Mayer-Smith, J., & Moon, B. (1998). A critical analysis of the research on learning to teach:
Making the case for an ecological perspective on inquiry. Review of Educational Research, 68(2),
130178.
Wilson, S., Darling-Hammond, L., & Berry, B. (2001). A case of successful teaching policy: Connecticuts
long-term efforts to improve teaching and learning. Seattle, WA: Center for the Study of Teaching and
Policy.
Youngs, P. (2002). State and district policy related to mentoring and new teacher induction in Connecticut.
National Commission on Teaching and Americas Future.
Zeichner, K. (1979). Teacher induction practices in the United States and Great Britain. Madison, WI:
University of Wisconsin, Department of Curriculum and Instruction.

123

Copyright of Journal of Educational Change is the property of Springer Science & Business Media B.V. and its
content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.