 

 

 

 

 

 

'mm hi'. I- ‘I I '

 

I‘L Michigan State
5;) ‘1‘ Universnty
This is to certify that the
dissertation entitled

COLLECTIVE BARGAINING AND HUMAN RESOURCES
POLICIES: EFFECTS ON MENTORING AND COMMITMENT
LEVEL OF NOVICE TEACHERS

presented by

Ben Michael Pogodzinski

has been accepted towards fulﬁllment
of the requirements for the

PhD. degree in Educational Policy

PeIO’L G. W

Major Professor’s Signature
Tune Z". “.100q

 

Date

MSU is an Afﬁrmative Action/Equal Opportunity Employer

 

‘ 4- - — —.- - - — _ _
. . . . . u .-.-.--_u- —
I -I-u-I- -l-I-I---l-.-l- -u-.- -.-.-.-c--.- -.- - - - -- -

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProj/Acc&Pres/ClRC/DateDue.indd

 

COLLECTIVE BARGAINING AND HUMAN RESOURCES POLICIES: EFFECTS
ON MENT ORIN G AND CONEMITMENT LEVEL OF NOVICE TEACHERS

By

Ben Michael Pogodzinski

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Educational Policy

2009

COLLECTIVE BARGAINING AND .
ON MENTORING AND COMMITMENT LEVEL OF NOVICE TEACHERS

ACKNOWLEDGEMENTS

This dissertation stems from a multi-year study known as the Michigan Indiana
Early Career Teacher (MIECT) Study, directed by Dr. Peter Youngs and Dr. Ken Frank
at Michigan State University. I would like to thank Dr. Youngs and Dr. Frank for their
assistance throughout the dissertation process, and for making the data from the MIECT
Study accessible. I would also like to thank my fellow research assistants who worked on
the MIECT Study, Richard Holdgreve-Resendez, Nathan Jones, and Mark Low. I would
like to thank the other two dissertation committee members, Dr. Saundra Schneider and
Dr. Dale Belman, for their assistance over the course of my doctoral career. Finally, I
would like to thank my family, particularly my wife Angela, for their support through this

entire process.

The research reported in this dissertation was supported by grants from the
Carnegie Corporation of New York and the Michigan State University Intramural
Research Grants Program. The opinions expressed in the paper are those of the author

and do not necessarily reﬂect the views of the funding agencies.

iii

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................... vi
LIST OF FIGURES ............................................................................................. viii
CHAPTER 1
DISTRICT POLICY AND NEW TEACHERS ................................................... 1
Literature review ............................................................................................. 5
New teacher induction ............................................................................... 6
New teacher induction and teacher associations ........................................ 15
CHAPTER 2
CONCEPTUAL FRAMEWORK ........................................................................ 18
New institutionalism ....................................................................................... 18
Social capital theory ........................................................................................ 20
Hypotheses ...................................................................................................... 23
CHAPTER 3
METHODOLOGY AND ANALYSIS ................................................................ 28
State context .................................................................................................... 28
District sample ................................................................................................ 29
Early career teacher sample ............................................................................ 31
Mentor/Colleague sample ............................................................................... 33
District and teacher association interviews ..................................................... 34
Analysis ........................................................................................................... 34
CHAPTER 4
MENTORING POLICY ...................................................................................... 38
State context .................................................................................................... 39
Collective bargaining provisions .................................................................... 44
Formal district policy ...................................................................................... 48
Day-to-day practice ......................................................................................... 49
CHAPTER 5
EARLY CAREER TEACHER INTERACTIONS WITH THEIR MENTORS
AND COLLEAGUES .......................................................................................... 63
Frequency of interaction ................................................................................. 65
Importance of support ..................................................................................... 73
Regression analysis ......................................................................................... 78
CHAPTER 6
ORGANIZATIONAL CONTEXT AND COMMITMENT ................................ 86
Spring commitment measures ......................................................................... 89

 

99

Change-in-commitment ..................................................................................
labor-management relations ........................................................................... 105
Social network analysis ................................................................................... 107
CHAPTER 7
IMPLICATIONS FOR POLICY AND PRACTICE ........................................... 1 13
Summary of ﬁndings ....................................................................................... 115
Implications for policy and practice ................................................................ 124
Limitations of study and future research ......................................................... 127
APPENDIX .......................................................................................................... 132
REFERENCES ..................................................................................................... 137

LIST OF TABLES

3.] District demographics ........................................................................................... 30
4.1 Percentage of ECTs in the sample with mentors .................................................. 57
4.2 Variables included in analysis .............................................................................. 5 8
4.3 Likelihood of mentor assignment - Fixed effects ................................................ 5 9
4.4 Likelihood of mentor assignment ~ Random effects ............................................ 60
5.1 ECT-mentor frequency of interaction (change from fall to spring) ...................... 71
5.2 Difference between ECT interaction with mentors and colleagues ...................... 73
5.3 Correlations between organizational items and frequency of interaction ............. 77
5.4 Variance components ~ Unconditional model ...................................................... 79
5.5 Variables included in analysis ............................................................................... 80
5.6 Frequency of interaction conditional model -— Fixed effects ................................. 82

5 .7 Spearman correlations between importance of support and frequency of

interaction ..................................................................................................................... 84
6.1 Variance components —— Unconditional model ...................................................... 91
6.2 Variables used in analysis ..................................................................................... 93
6.3 Spring ECT commitment -- Fixed effects ............................................................. 96
6.4 Variance components - Conditional model .......................................................... 99
6.5 Chmge-inmmiMent from fall to spring .......................................................... 100

6.6 Variance in ECT change-in-commitment — Unconditional model ....................... 101

6.7 Change-in-commitment - Fixed effects ................................................................ 104

6.8 Variance in ECT change-inmmmiMent - Conditional model ........................... 105

vi

6.9 Spearrnan correlations —— Perceptions of labor-management relations and union
effort ............................................................................................................................. 106

6.10 Mentor and colleague inﬂuence on ECT perceptions of professional ﬁt ............ 109

vii

LIST OF FIGURES

2.1 Conceptual framework .......................................................................................... 25
5 .1 ECT interaction with their mentor ........................................................................ 66
5.2 ECT-mentor interaction by ECT years of

experience .................................................................................................................... 67
5.3 ECT-mentor interaction by organization level ...................................................... 69
5.4 ECT-mentor interaction by topic .......................................................................... 70
5 .5 ECT-colleague interaction by topic ...................................................................... 72
5.6 Importance of professional support from mentors and colleagues ....................... 74
6.1 ECT commitment to teaching the following year ................................................. 90

6.2 ECT commitment to teaching in ﬁve years .......................................................... 9O

viii

Chapter 1
District Policy and New Teachers

The experiences of teachers are directly and indirectly inﬂuenced by district- and
school-level policies aimed at ensuring that the teaching workforce is both highly
qualiﬁed and highly effective (Hess, Rotherham, & Walsh, 2004). Policies dictate the
process of teacher hiring, assignment, induction, and evaluation, all of which inﬂuence
the organizational context within which teachers perform their jobs. In many ways,
these policies also shape the social context within which teachers operate. Both the
organizational and social context has implications for certain teacher-level outcomes.

Since the 19th century, teacher associations have worked to shape educational
policy at both the national, state, and local level, thus having a hand in shaping the
organizational and social context of schooling. For example, collective bargaining
agreements can either directly or indirectly inﬂuence district human resource decisions
regarding teacher hiring, transfers, and layoffs, as well as policies related to teacher
evaluation, induction, and professional development (Fuller & Izu, 1986; Fuller,
Mitchell, & Hartman, 2000; Ballou, 2000; Koppich, 2005). Therefore, through collective
bargaining and labor-management relations, teacher associations potentially have strong
effects on teachers’ experiences because contract provisions speciﬁcally “shape and
reﬂect some of the important parameters of teachers’ work lives in technical and
normative terms” (Bascia, 1997, p. 444).

The effects that district- and school-level policies have on the experiences of
novice teachers most likely differ from the effects on more senior tenured teachers.

Most collective bargaining provisions and district policies favor tenured teachers with

regard to job protection (including order of layoffs and involuntary transfers) and
economic beneﬁts which may place novice teachers at a lower social, political, and
economic status within the organization (Eberts & Stone, 1984; Babcock & Enberg,
1999). This lower status may impact the relationships that novices have with their
colleagues and therefore impact their ability to access resources through those
relationships. In other cases, though, the novice status granted to new teachers may
increase their ability to access resources through such induction activities as common
planning time with mentor teachers or more release time for planning and observing other
teachers (Smith & Ingersoll, 2004).

What is clear though is that district policies are often designed to treat non-
tenured teachers differently than their tenured colleagues, which may not only have direct
effects on their work experiences, but also indirect effects by inﬂuencing the professional
relationships novice teachers have with their more senior colleagues. For example, how
do policies around mentoring and teacher collaboration inﬂuence the relationships that
novice teachers build with their more senior colleagues, and how are these relationships
associated with novice teacher outcomes such as beliefs and practices, as well as career
decisions? In addition to the formal structures of schools and districts, new teachers are
socialized into the profession through informal social networks that develop within a
school or district. These informal networks can also mediate the effects of formal district
policies.

This study uses both organizational theory based on the ideas of new
institutionalism as a theory of action at the organizational level, and social capital theory

as a theory of action at the individual level, to explore how district- and school-level

policies shape the mentoring and socialization of novice teachers. In particular, this
study looks at how mentoring policies are potentially shaped by collective bargaining,
how mentoring policies impact the organizational context within which novice teachers
work, and also how labor-management relations help shape the social context within
which teachers work. At the same time, individuals enact agency and through personal
interactions with colleagues have access to expertise, resources, and support; while at the
same time mentors and colleagues can enact expectations and norms onto novice
teachers. Both the organizational context and the social context have the potential to
inﬂuence important teacher outcomes, such as teacher commitment and retention.
Therefore this study aims to address three important questions: a) To what extent do
collective bargaining agreements (CBAs) inﬂuence the provision of mentoring? b) To
what extent are early career teachers (ECTs) interacting with their mentors, as well as
with other colleagues? and c) How do the organizational and social contexts within which
novices work inﬂuence their commitment to teaching, expressed through their professed
future career plans?

The ideas from new institutionalism provide a lens to examine both the formal
and informal structures within an organization. Mentoring policies are often
implemented in relation to a particular policy environment (e.g., legal requirements, new
research, CBA provisions), but the formalized mentoring policy may be “very different
from the effects generated by the networks of social behavior and relationships which
compose and surround a given organization” (Meyer & Rowan, 1977, p. 341). In other

words, there may be large gaps between the formal structure of a mentoring program, and

the actual relationship between a mentor and mentee, or between ECTs and their
colleagues (Meyer & Rowan, 1977; DiMaggio & Powell, 1983).

It is also important to consider social capital theory in addition to organizational
theory when studying mentoring policy and teacher socialization because an individual
teacher’s own social context mediates the effects of such policies and structures (Frank,
1998). The use of social capital theory provides opportunities to further explore the
supports and pressures that are associated with these types of policies, by studying how
novice teachers’ formal (i.e., mentor-mentee) and informal relationships inﬂuence
novices’ access to resources and their perceptions of the organization. For example, a
teacher’s perceptions of professional ﬁt within an organization may be partially
determined by their relationships with their mentor and other colleagues, which then may
be associated with their levels of commitment (Bryk & Schneider, 2002; Desimone et al.,
2002; Kardos et al., 2001).

The labor-management relations climate within a school and district will also
potentially affect the relationships novices have with their colleagues and administrators,
thus affecting their access to resources. Collective bargaining provisions tend to
standardize teaching work, which often encourages autonomous work and reduces
collaboration amongst teachers (Bascia, 1997; Stone, 2000). Additionally, if relations are
negative between teachers and administrators, it may reduce novice teachers’ access to
support and resources which may negatively impact their commitment. In particular,
teachers may try to transfer out of schools in which there are poor labor-management
relations. There is also reason to believe that professional relationships (amongst

teachers and between teachers and administrators) can mediate the effects of mentoring,

other induction activities, and other district policies on their practices and commitment to

teaching (Bryk & Schneider, 2002).

Literature Review

Through collective bargaining agreements (CBAs), teacher associations have
traditionally focused on increasing teachers’ salaries and beneﬁts, achieving standardized
working conditions and practices, lowering student-teacher ratios, and protecting against
job loss (Stone, 2000). Most CBAs follow the industrial union model where a premium
is placed on seniority in terms of salary, assignment, and job protection, leading some
novice teachers to view teacher associations as less relevant to their work lives (Kerohner
& Cooper, 2003). In order to help address the needs of novice teachers, some local
associations have expanded their roles into negotiating over curricular and instructional
reforms, peer assistance and review (PAR), alternative salary schedules, and induction
support and professional development for teachers (Koppich, 2005). This continues to be
the exception rather than the rule, and there has been limited research on the effects of
collective bargaining on teacher outcomes, such as commitment.

Since teacher commitment has been found to be related to teacher effectiveness
and teacher turnover, both teacher associations and districts should be concerned with
increasing teacher commitment (Weiss, 1999; Ingersoll, 2001; Ebmeier, 2003). This is
particularly important in hard-to-staff schools which serve large percentages of poor and
minority students, where teacher turnover is signiﬁcantly higher than in schools serving
higher SES white students (Boyd et al., 2002; Hanushek, Kain, & Rivkin, 2004; Smith &

Ingersoll, 2004). Teacher commitment has been found to be associated with workplace

conditions (Weiss, 1999), as well as with the collaborative social organization of a school
(Rosenholtz, 1985). This has implications for not only addressing the organizational
context of schools, but also their social context. Research, though, has not fully explored
the effects of school social organization on teacher commitment (Weiss, 1999, p. 862).

New Teacher Induction; One way schools and districts work to improve teacher
quality and retain effective teachers is through induction programs such as a) mentoring;
b) Opportunities to collaborate with colleagues; c) orientations, seminars and workshops;
and (1) additional classroom or instructional assistance (Smith & Ingersoll, 2004). In
addition to formal structures to support novices which are put into place at the district or
school level, novice teachers often receive informal support from colleagues within their
school or district. Teachers who receive formal and informal initiation into the
profession are more likely to develop norms encouraging professional growth and greater
commitment to the profession (Rosenholtz, 1989).

There is some evidence that induction programs can reduce teacher attrition rates
and provide teachers opportunities for growth (Smith & Ingersoll, 2004; Strong &
Fletcher, 2004). Induction activities related to curriculum, assessment, and teacher
evaluation can help shape the instructional expectations that are placed on new teachers
and create opportunities for novices to engage in learning activities (Grossman &
Thompson, 2004; Johnson & PNGT, 2004; Stein & D’Amico, 2002; Achinstein, Ogawa,
& Speiglman, 2004; Youngs, 2007a). At the same time, although over 75 percent of
novice teachers are participating in some form of induction activity, wide variation
remains in the quality and effectiveness of support which is provided to novice teachers

(Smith & Ingersoll, 2004). Important qualities of mentoring and other induction

activities include a) the characteristics and roles of the mentor and mentee; b) the
characteristics and roles of colleagues and administrators; c) school organizational
conditions; and d) the degree of alignment or ﬁt between novices and their mentors and
colleagues (Youngs, Qian, & Holdgreve-Resendez, in progress).

Research suggests that novices who are paired with a mentor who teaches the
same grade level, subject matter, or certiﬁcation area may receive more beneﬁts from the
relationship than novices who are paired with out-of-ﬁeld mentors. Using data from the
1999-2000 Schools and Stafﬁng Survey (SASS), Smith and Ingersoll (2004) found that
having a mentor in one’s ﬁeld reduced the likelihood of leaving teaching at the end of the
ﬁrst year by 30 percent and that being able to collaborate with colleagues on instructional
issues reduced the risk of leaving the profession by 43 percent and lowered the risk of
migration by 25 percent. A mentor or colleague who has greater knowledge of the
curricula may be able to provide more help in developing appropriate teaching strategies
and assessments.

Several qualitative studies support the ﬁndings of Smith & Ingersoll (2004),
suggesting that both mentor and mentee characteristics are important in predicting the
quality and effectiveness of a mentoring relationship (e.g., Achinstein, Ogawa, &
Speiglman, 2004; Grossman & Thompson, 2004; Youngs, 2007a). For example,
Grossman and Thompson (2004) studied the experiences of three ﬁrst-year secondary
teachers — Frank, Nancy, and Allison -— in two suburban Washington State school
districts, Prospect Harbor and Waterside. The purpose of the study was to investigate
how district and state policies concerning curriculum, induction, and professional

development affected the teachers’ opportunities to learn to teach language arts.

Differences in induction programs in each district highlighted the importance of
matching novices with mentors who have knowledge of their curriculum as well as
similar beliefs and practices. Frank’s mentor was responsible for mentoring several new
teachers which reduced the amount of time she had available to work with Frank, and she
did not have curricular expertise in all the subjects he taught, therefore the beneﬁts of the
induction program for him were limited According to Grossman and Thompson, “as a

result, she was not able to give him the kind of curricular guidance he sought” (2004, p.
292). In contrast to Frank, Allison was assigned a mentor from her department who
shared teaching strategies and materials with her because the mentor had knowledge of
the curriculum. Having a mentor in the same content area enabled her to obtain much
more assistance with curriculum and instruction than Frank.

In addition to formal mentoring assignments, the socialization of new teachers is
partially determined by the characteristics and roles of their colleagues and
administrators. Using large scale survey data of novice teachers in the Chicago Public
Schools from 2004-05, Kapadia, Cora, and Easton (2007) found that novice elementary
teachers who regulme collaborated with peers in their ﬁeld and participated in a network
of other new teachers were more likely to report having a good teaching experience and
intention to remain in teaching. Kardos and her colleagues (2001) interviewed 50 Ist-
and 2nd-year teachers from a wide variety of schools in Massachusetts, focusing part of
their inquiry on new teachers’ experiences with their school-based colleagues and

professional school cultures. They also analyzed new teachers’ accounts of the
organizational structures within which they interacted with colleagues. Using the data

collected, the researchers claimed that new teachers sought signals from their colleagues

about how to interact with students and colleagues, which instructional approaches were
acceptable, appropriate behavior in meetings, and appropriate use of time (Kardos et a1. ,
2001)

Kardos and her colleagues deﬁned professional culture as a blend of both formal
and informal norms, values, and accepted modes of professional practice that existed
among colleagues (2001). The researchers identiﬁed three types of professional school
cultures: veteran—oriented, novice-oriented, and integrated. In schools with veteran-
oriented cultures, the authors found that the school culture was deﬁned by veteran
teachers and new teachers were given no special status, which resulted in little
orientation, induction, or support. In schools labeled as having novice-oriented cultures,
the majority of the teachers were new to the profession. There were higher levels of
initial commitment and enthusiasm in these schools, but few formal induction supports
were offered, which may have impeded novice teacher instructional growth. The authors
described schools with integrated cultures as being marked by high levels of
collaboration between novice and experienced teachers. Mentors transmitted cultural
norms to novices and there were frequent opportunities for new teachers to talk with
veterans about curriculum, instruction, and assessment. The research by Kardos and
colleagues (2001) suggests, as in the case of the veteran-oriented cultures, that a lack of
collaboration or professional ﬁt in a school can decrease the level of support that new
teachers receive in handling the pressures of the profession. At the same time, where and
how novice teachers access support and expertise (i.e., from other novices vs. veterans)
can have great implications for how social capital inﬂuences novice teachers’

instructional practice and commitment to teaching. Although novices may tend to

interact with each other because they may have more in common, the help that they can
provide each other in terms of expertise may be limited compared to the potential
assistance a more veteran teacher can provide.

Principals can also play a signiﬁcant role in inﬂuencing the supportive culture of
a school, through both direct interactions with novices as well as through establishing
opportunities for novices to interact with their colleagues (Youngs, 2007b). In an
ongoing study, Desimone & Smith (2008) are investigating how school leadership shapes
new middle school math teachers’ content knowledge and instructional practices, with
focus on how the content knowledge of principals and other teachers seems related to
novices’ instructional practices. In work that stems from the study being conducted by

Desimone and Smith (200 8), McGraner and Henrick (2008) interviewed 18 middle
school principals from southern urban districts, and found that those with weak math
content knowledge viewed induction in ways that ignored certain attributes of teaching
and learning in math. Additionally, these principals did not develop organizational
supports in their schools to provide professional growth for novice teachers.

In a study of three urban Connecticut districts, Youngs (2007b) interviewed six
elementary principals, along with ﬁrst- and second-year teachers, mentors, and other
colleagues in their schools to learn more about the support provided to novices.
Principals with stronger knowledge of curricular content, assessment, and professional
development were more likely to promote instructional growth in novice teachers and
establish a collaborative organizational environment (Youngs, 2007b). Chester and

Beaudin (1996) surveyed 173 new teachers who were hired in 1989-90 to teach in

Connecticut’s nine largest school districts to learn about school organizational conditions

10

and practices. Using OLS regression, they provided evidence that “suggests that a direct
relationship exists between urban supervisors’ attention to instruction and changes in
their beginning teachers’ self—efﬁcacy beliefs” (1996, p. 246). Chester and Beaudin
added that “(i)n addition to the timing and frequency of feedback, the focus of the
feedback is an important aspect of the ﬁndings regarding supervisor observations” (1996,
p.252)

Related to the culture of support, which is partially determined by the teachers
and administrators within a school organization, is the degree of relational trust between
the adults within the school which is related to the ways that novice teachers discern the
integrity and competence of mentors, colleagues, and administrators (Ford & Youngs,
under review). Relational trust between novice teachers and their colleagues and with
their administrators derives from the set of role-relations characterizing the social
organization of schooling (Bryk & Schneider, 2002). “When relational trust is high
among the various role-sets, the school as an organizational entity is likely to exhibit
properties of its operation that are more conducive to such things as a) supporting the
practices and growth of new teachers and b) school improvement, including more
effective decision making and stronger social support for innovation and/or change”
(Youngs, Qian, Holdgreve-Resendez, in progress).

Finally, the quality and effectiveness of mentoring and induction may rely on the
level of professional ﬁt or alignment between a novice teacher and their mentor,
colleagues, and administrators (Youngs, Qian, Holdgreve-Resendez, in progress). The
degree of professional ﬁt may mediate the effects of induction policy, as well as other

district and school policies. For example, Achinstein, Ogawa, and Speiglman (2004)

11

explored the effects of state policies and local conditions on teachers’ beliefs and
practices, paying particular attention to how school professional culture inﬂuenced new
teacher socialization. Their study focused on two elementary teachers who participated
in the Santa Cruz New Teacher Center induction program in two California districts.

In reporting their ﬁndings, the authors focused primarily on two teachers in two
different districts: Liz and Sam. The differences in the two teachers’ background
characteristics and preparation routes not only inﬂuenced the teaching skills that they
acquired but also helped guide their labor market decisions, which ultimately inﬂuenced
their levels of professional ﬁt within their schools. Liz grew up in the urban district in
which she taught, attended a nearby large public university, and began teaching while she
completed her certiﬁcation. District A, in which Liz worked, recruited teachers from the
local community to reﬂect the population of the students, even though they may have
initially lacked full teaching credentials. Sam on the other hand was from an afﬂuent
northern California community and received a teaching certiﬁcate and master’s degree
from a research university. District B, in which Sam worked, recruited teachers from
research universities and looked for teachers that shared a teaching philosophy consistent
with that of the district.

The expectations placed on Liz by district policy and colleagues were to follow
routines, curriculum, and assessments designed by the state and district. The induction
program in her district also heavily revolved around implementing the literacy
curriculum. Liz was guided towards a particular way of teaching, and district policies
concerning curriculum and induction were potentially designed for someone with limited

professional preparation. In contrast to Liz, “Sam enjoyed numerous professional

12

development activities that emphasized inquiry and co-construction of knowledge”

(Achinstein, Ogawa, & Speiglman, 2004, p. 579). He beneﬁted from coaching, team
collaboration, reading academies, action research groups, membership in reform
networks, and university affiliation. The expectations in this district were for teachers to
grow as professionals with support from the district and colleagues. Sam developed
social networks that enabled him to access resources and expertise in response to
challenges and pressures that he faced. Policies in these two districts led to signiﬁcant
variations in the supports and pressures that Liz and Sam experienced, which inﬂuenced
their instructional practices in very different ways.

In addition to curricular support from the district and colleagues, both Liz and

Sam were assigned a mentor during their ﬁrst two years of teaching. Liz and her mentor
focused more on issues related to students, parents, and being rehired (she was laid off
after her ﬁrst year) than on instructional matters. Sam and his mentor met on a weekly
basis to co-plan lessons, develop literature circles and writer’s workshops, and discuss
theories related to literacy instruction. The high frequency and level of their interactions
suggest a more developed relationship and a more proper alignment of teaching
philosophies. In addition, the induction program allowed Sam to observe teachers in
other schools and districts so that he could obtain additional assistance. This study
illustrates the role that professional ﬁt within an organization has in determining the
activation of support that is available to novice teachers. This highlights that professional

ﬁt may not only determine what support is provided to novice teachers, but also to what

extent novice teachers engage with the support that is available.

13

Although much of the previous research suggests that mentoring and
collaboration with colleagues and administrators potentially produces positive teacher
outcomes, such as improvements in instruction or increased retention rates, a recent
large-scale quantitative study challenges some of the ﬁndings of other research on
mentoring and induction Glazerman and colleagues (2008) used an experimental design
to study the effects of comprehensive induction programs developed by the Educational

Testing Service (ETS) and the New Teacher Center (NT C), compared to a control group
of districts which used their current induction strategies. Covering 17 districts in 13
states, there were 100 treatment schools and 103 control schools in 9 ETS districts, and
110 treatment schools and 105 control schools in 9 NTC districts (Glazerman et al.,
2008). Based on the ﬁrst year of data collection (2005-06), the researchers found no
statistically signiﬁcant effects between the treatment group (ETS and NTC combined)
and control group of teachers with regard to lesson implementation, lesson content,
classroom culture, teacher effectiveness, or on teacher retention (Glazerman et al., 2008,
p. xvi). The researchers note that there may be longer term effects which may be
uncovered through analysis from their second year of data collection. At the same time,
their ﬁndings suggest that induction policies may not always produce their intended
effects.

By focusing on characteristics of mentoring and other induction activities which
include a) the characteristics and roles of the mentor and mentee; b) the characteristics
and roles of colleagues and administrators; c) school organizational conditions; and d) the
degree of alignment or ﬁt between novices and their mentors and colleagues, researches

can better account for the organizational and social contexts which inﬂuence the quality

14

and effectiveness of mentoring and induction (Youngs, Qian, & Holdgreve-Resendez, in
progress). Although recent empirical research has shown no effects of induction on
teacher outcomes (Glazerman et al., 2008), many other empirical studies have found
some signiﬁcant effects on novice outcomes associated with the organizational and social
context of their school. For example, in their research from Chicago, Kapadia, Coca, and
Easton (2007) found that collaboration with colleagues and support from principals was
associated with higher levels of novice teacher commitment. It is important for
researchers to continue to explore how the organizational and social contexts inﬂuence
the work conditions of novice teachers in order to better craft mentoring and induction
policy.

New teacher induction and teacher associations. It is important to consider what
role that labor-management relations and collective bargaining have in the provision of
new teacher induction as well as other supports to novice teachers. In addition to
potentially inﬂuencing induction policy through collective bargaining, teacher
associations also affect novice teachers’ interactions with colleagues through the
deﬁnition of teachers’ work and expectations, which may either lead to more autonomous
work or potentially lead to increased teacher collaboration (Bascia, 1997). Additionally,
the climate of labor-management relations may inﬂuence the level of school-wide
professional community, thus inﬂuencing the levels of collaboration between novices and
their teacher colleagues or between novices and their administrators (Streshly &
DeMitchell, 1994).

Many induction programs were originally initiated by teacher associations as a

way to provide support to new members and help them grow as teachers (Poole, 2000).

15

Also, although many induction programs are run by district or school administrators,
when districts face budget shortfalls teacher associations often step in to help deliver
induction support and additional professional development (Bascia, 2003). Generally,
there are three types of professional development associated with teacher associations: a)
workshops, seminars, and conferences; b) formal mentoring, peer coaching, and other
induction activities; and c) informal collaboration with colleagues (Bascia, 2003). In
addition, teacher associations may encourage professional development by connecting
some elements (such as continuing college credit) to increases in salary (Bredeson, 2001 ).
Since teachers’ work days provide little time for professional development, teacher
associations often negotiate extra paid days/hours or extended contracts for summer work
in order to meet the needs of teachers (Bredeson, 2001). Additionally, some induction
programs promote collaborative planning time for academic department members or
grade-level teams, which may be more beneﬁcial in improving teacher effectiveness and
commitment (Smith & Ingersoll, 2004).

At the same time, teacher association involvement in induction and professional
development may draw their focus away from bread and butter issues. “It would be a
mistake for anyone to assume that all teacher union locals (or all districts) have embraced
education improvement and teacher quality as an essential part of their mission”

(Koppich, 2005, p. 91). Many people believe that teacher association involvement in
such non-economic areas reduces public control, limits the ﬂexibility of school
administrators, and may negatively inﬂuence district and school decisions about resource
allocations (Bredeson, 2001). Therefore, the extent to which collective bargaining shapes

new teacher induction can greatly vary depending on the dynamics of the district,

16

including the labor-management relations. Overall, teacher associations and collective
bargaining agreements not only partially deﬁne the organizational environment in which
induction policy is crafted, but may also inﬂuence the social context in which novice
teachers’ work. It is therefore important, when studying new teacher induction, to study
how labor-management relations inﬂuence new teacher outcomes such as commitment

(Kerchner & Cooper, 2003).

17

Chapter 2

Conceptual Framework

This research study is grounded in the theoretical ideas of new institutionalism
and social capital theory, utilizing both an organizational-level theory of action and an
individual-level theory of action. In particular, this study pays particular attention to the
formal and informal organizational structures which arise in response to the policy
environment regarding new teacher mentoring (Meyer & Rowan, 1977), and to the social
inﬂuence novice teachers experience within their given organizational context through
interactions with their mentors and colleagues (Meyer, Scott, & Deal, 1981; Coleman,
1988). Novice teachers make sense of their organizational context based on the formal
structures, rules, and policies of their districts and schools (organizational theory of
action), but also through their own and their colleagues’ beliefs and practices which may
vary from the formally deﬁned institutional rules (individual theory of action)
(Lounsbury, 2001; Powell & Colyvas, 2007; Frank, Krause, & Penuel, in progress).

New Institutional ism. New institutionalism stems from the ideas put forth under
“old” institutionalism, which suggests that organizations are rationally ordered to attain
goals, but the formal structures cannot eliminate non-rational dimensions of
organizational behavior, such as individuals who participate in the organization beyond
their formally deﬁned roles within the system and includes the complex informal systems
that link participants with one another (Selznick, 1948). New institutionalism pays
particular attention to how organizations relate to their environment through the

development of structural elements which often lead organizations who serve similar

18

functions to resemble each other in structure (i.e., institutionalized) (Meyer & Rowan,

1977; DiMaggio & Powell, 1983). At the same time Meyer and Rowan (1977) suggest
that because organizations fear loss of legitimacy, structural elements are decoupled from
day-to-day activities and from each other. DiMaggio and Powell ( 1983) argue that highly
structured organizational ﬁelds oﬁen lead to homogeneity in structure, culture, and
output. An organizational ﬁeld refers to “those organizations that, in the aggregate,
constitute a recognized area of institutional life: key suppliers, resource and produce
consumers, regulatory agencies, and other organizations that produce similar services and
products” (DiMaggio & Powell, 1983, p. 143). Therefore, organizations put more effort
into maintaining their institutionalized structures to maintain legitimacy rather than to
focus on the efficiency and effectiveness of activities (Meyer, Scott, & Deal, 1981).

In relation to education, Meyer, Scott, and Deal (1981) proposed, “Educational
organizations arose to bring the process of education under a socially standardized set of
institutional categories, not necessarily to rationalize the ‘production processes’ involved
in carrying out the work” (p. 46). They continued, “Organizational attention is directed
toward maintaining conformity with the socially standardized categories of the
educational system, while little effort is expended in the control and coordination of
instructional activities” (p. 49). With regards to mentoring, under this theory it would be
expected that districts and schools would respond to their environments in similar ways,
instituting mentoring programs that are in line with state regulations and common
practices. This also suggests though that districts and schools are more interested in

maintaining the structure of an acceptable mentoring system rather than its effectiveness

in producing outcomes, such as improved novice practice or teacher retention. Meyer

19

and Rowan (1977) even argue, “Institutionalized organizations seek to minimize
inspection and evaluation by both internal managers and external constituents” (p. 359).
Therefore districts and schools may view induction programs as a success based on how
they look rather than the results they produce, and may expend little effort in monitoring
their mentoring programs.

Although novice teachers, their mentors, and their colleagues operate within these
formalized structures, they still maintain personal agency. Recent research has called for
more attention to the micro-foundations of institutionalism, known as new structuralism,
to account for how individuals make sense of the organizational context (Lounsbury &
Vantresca, 2003; Powell & Colyvas, 2007). New structuralists largely focus on the
social-psychological realm of individual sense making through which individuals
negotiate rules and procedures and conﬂicting demands (Frank, Krause, & Peneul, in
progress). This sense making not only takes place within the formalized structure of the
institution, but also within the immediate social context in which individuals are located
(Granoveter, 1985; Frank, Krause, & Peneul, in progress).

Social capital theory. The attention to social context is related not only to
individuals’ sense making, but also to their ability to access support and resources within
an organization. Therefore, this research study is also grounded in the ideas of social
capital theory. Bourdieu (1986) deﬁnes social capital as “the aggregate of the actual or
potential resources which are linked to possession of a durable network of more or less
institutionalized relationships of mutual acquaintance or recognition” (p. 248). Social
capital theory, therefore, emphasizes the numerous ways that individuals are located in

networks of social relations and seeks to explicate how the nature, quality, and intensity

20

of these relations affect several important outcomes. With regard to teachers, it highlights
the resources that are available through social relations with colleagues and suggests that
different types of relations will —— to varying degrees -— mediate the effects of district and
school policies, as well as facilitate access to resources (Coleman, 1988).

Social capital is equated to investments in social relationships with expected
returns, but those returns are not necessarily guaranteed (Lin, 1999). Therefore, social
capital may not always be positive, having potentially negative consequences such as: a)
exclusion of outsiders, b) excess claims on group members, c) restrictions on individual
freedoms, and d) downward leveling norms (Portes, 1998). The returns that one sees to
their investment in the relationship may depend on many different factors, including
location within a network (Burt, 1997), bridges to other networks (Granovetter, 1973),
closure of networks (Coleman, 1988), or structural variations within a network (Lin,
1999)

This sociological view of social capital is different from that put forth by some
political scientists. For example, Putnam (1995) places social capital at the community
level with the idea of “civicness”, which facilitates action and cooperation for mutual
beneﬁt at a more aggregate level. Portes (1998) is cautious with regard to moving the
concept of social capital to the community level or even higher. In his words, “As a
property of communities and nations rather than individuals, social capital is
simultaneously a cause and an effect” (Portes, 1998, p.19). Although there are different
interpretations of social capital, in education research social capital is most often thought
to exist at the individual level, and is often deﬁned as the manifestation of supports and

pressures through social relations (Dika & Singh, 2002; Frank, 21130, & 1301111811, 2004)-

21

Therefore, this study uses the ideas of social capital as the manifestation of
supports and pressures (Bidwell, 2001; Frank, Zhao, & Barman, 2004) and variations in
levels of social capital as proposed by Lin (1999). Lin (1999) describes two types of
causation forces that are important in the analysis of inequality of social capital: 1)
structural variation, and 2) positional variation. Structural variation can be seen in the
distribution of available resources, such as teacher expertise or curricular materials,
which may vary by district, school, or subgroup. Positional variation is related to a
teacher’s social, political, and economic status within a social network which impacts
their ability to access resources and deal with pressures and expectations associated with
their work conditions. Social networks consist of relationships of individuals within a
social system, which vary in their strength and in their ability convey support or pressure.
A novice’s position within the school may greatly impact her ability to access resources
through their relationships. Given a teacher’s novice status, she may be at a lower social
and political level than more senior teachers, which limits their ability to access
resources. At the same time, if they are in a venue which supports new teachers, a
teacher’s novice status may actually help them receive extra support and resources.
Additionally, these relationships may produce pressures and expectations that impact
novice teacher commitment in negative ways. For example, novice teachers are often
expected to perform at the same level as their veteran colleagues, and the expectations
placed on them by the administration and their colleagues may add negative stress to their
professional lives.

Figure 2.1 represents the framework which I used to guide the research and

analysis. At the top, collective bargaining agreements and district policy respond to

22

environmental factors such as state regulations and mentoring research. Collective
bargaining agreements may or may not directly inﬂuence formal district mentoring
policy, represented by the dashed line between them. School policy is then at least
partially deﬁned by district policy, though school administrators often have discretion in
shaping at least parts of mentoring policy, such as assigning mentors to novice teachers
or providing release time. School policy in many ways will define the relationship
between a mentor and mentee, based on the characteristics of the assigned mentor and
based on the level of oversight that administrators have over the mentoring relationship.
School policy may also impact the level of collaboration between novices and their
colleagues, through such policies as the establishment of teacher teams or common
planning time. Through relationships with mentors and colleagues, novice teachers have
access to varying levels of resources, support, and teacher expertise. Additionally, these
relationships may also place certain pressures on novice teachers. Ultimately these
relationships may then inﬂuence novices' levels of commitment. The diagram therefore
represents both organizational-level action through deﬁning policy, but also individual-
level action where through relationships novices’ make sense of policy and are socialized
into the profession.

Hypotheses. The ideas of new institutionalism and social capital theory help
frame the experiences of novice teachers in their organizational and social contexts, as
well as how these contexts mediate the effects of district and school policies associated
mentoring. Districts and schools respond to the policy environment partially deﬁned by
state regulations, collective bargaining provisions, research, and accepted practices in

setting their induction policies. At the same time, the social contexts in which novices

23

and their colleagues work in are complex and they are not limited to their formalized
roles within the organization. Therefore mentoring policy may be set by a district or
school (organizational level), but the way in which teachers make sense of the policy and
the actual day-to-day activities regarding new teacher mentoring are often shaped by the
micro-level social context in which teachers operate (individual level). This leads to a set
of hypotheses to address my three main research questions:

Research Question I: To what extent do collective bargaining agreements inﬂuence the
provision of mentoring?

Hypothesis #1: Although there may be variation as to what extent collective bargaining
agreements deﬁne the scope of bargaining, I expect there will be little variation in the
provision of mentoring across the districts because of the eﬂects of institutional
formalization of policy structures.

Research Question 2: To what extent are early career teachers interacting with their
mentors, as well as with their colleagues?

Hypothesis #2: I expect that early career teachers are likely to interact with their close
colleagues more often than with their formally assigned mentors. This reﬂects novices ’
ability to self-select into sub-groups within an organization which may better reﬂect their
beliefs and practices regarding education, compared to those of their formally assigned
mentor.

Research Question 3: How are novices’ professional relationships associated with their
levels of commitment?

Hypothesis #3: I expect that when novice teachers feel that their professional goals are

well aligned with the goals of their colleagues and mentors, they will have greater access

24

to resources and support which will help improve their commitment. Additionally, I
expect that novice teachers who perceive better labor-management relations in their

schools will have higher levels of commitment.

Figure 2.1: Conceptual framework

 

Policy Environment
State law; research; national

 

 

 

 

 

 

 

 

 

 

 

 

programs
CBAs District Policy
scope of district """ Design of new teacher
bargaining induction
A: .‘ I
School Policy

School—level
implementation of

 

 

 

 

 

 

 

 

 

 

 

 

 

mentoring

/ ‘ ‘ ‘ak
ECT-Mentor ECT~ColleagueI
Relationship Relationship

Access to expertise, norms, Access to expertise, norms,
content & frequency of content & frequency of
interaction interaction
\Ll/
ECT Outcomes
Commitment

 

 

 

25

This study builds on existing research on teacher commitment by providing
information on how collective bargaining provisions may inﬂuence mentoring policy, the
extent to which novice teachers are interacting with their formally assigned mentors and
other key colleagues, and to what degree novices’ organizational and social context is
associated with their levels of commitment. By using such information, teacher
associations and districts can work to improve teacher commitment through re-examining
how teachers interact within a school organization, particularly paying attention to what
organizational conditions are likely to lead to more productive collaborative
environments. Additionally, this study potentially has implications for the role that labor-
management relations play in inﬂuencing the work conditions within which novices
work. This study has further implications for state and district policy conceming new
teacher induction, the physical and social organization of schools, and the extent of labor-
management relations in creating both organizational and social support for novice
teachers.

In the next chapter, I describe the research design and methodology used for
analysis of the data that was collected. In chapter 4, I review district policy and
collective bargaining provisions concerning new teacher mentoring and induction in the
sample of districts. This is largely aimed at answering research question 1. In chapter 5,

I explore the frequency and nature of early career teacher interactions with their mentors
and close colleagues. This chapter aims to address research question 2. In chapter 6, I
explore what organizational and social characteristics are associated with commitment

levels. Using hierarchical linear regressions and social network analysis I am to address

26

research question 3. In chapter 7, I provide a summary of the ﬁndings and provide

implications policy and practice and research.

27

Chapter 3

Methodology and Analysis

This dissertation research stems from a larger ongoing study of early career
teachers directed by Peter Youngs and Ken Frank in eleven urban and suburban Michigan
and Indiana districts, known as the Michigan Indiana Early Career Teacher (MIECT)
Study. In particular, this study used a mixed methods strategy, including both survey data
and structured interviews, to explore the relationship between collective bargaining,

social capital, and novice teacher outcomes such as instructional practice and
commitment. A sample of 184 novice teachers was surveyed in 2007-08 and a sub-
sample of 14 were interviewed during the same year. Additionally, the mentors and key
colleagues of the novice teachers in the study were also surveyed in 2007—08.

State context. This analysis includes data from six Michigan districts and ﬁve
Indiana districts. It is important to consider the legal context within each of the states
which may impact the scope of bargaining and legal requirements for providing new
teacher mentoring. Based on data collected by the National Center on Teacher Quality,
districts in Michigan and Indiana are both legally obligated to negotiate with their teacher
associations, and negotiation over new teacher mentoring is allowed but not required
(www.nctq.org/tr3/scope). Michigan is a union shop state, meaning teachers in public
schools are not required to join the local teacher association when they are hired into a
district, but they must pay “fair share” dues to the teacher association because they are

covered by the collective bargaining agreement. In Indiana, teachers are not required to

join the local teacher association nor are they required to pay any dues to the association

28

even though they are covered by the collective bargaining agreement. Of the 184 ECTs
in the sample, approximately 80 percent indicated that they are a member of their local
teacher association.

Michigan law states that all new teachers are to be assigned a master mentor
teacher for their ﬁrst three years in the profession, and to receive 15 days of professional
development per year (PA 289 (1995) Section 1526). Indiana law requires that all new
teachers to be assigned a mentor for their ﬁrst year of teaching, and highly recommends
that that they are assigned a mentor for their second year in the profession as well
(Indiana Administrative Code). Mentors must complete state approved training, and their
duties are outlined by recommendations from the state. The mentoring program in
Indiana is part of the Indiana Mentoring and Assessment Program (IMAP), which
requires new teachers to complete and turn in an assessment portfolio at the end of their
second year in the profession as condition for continued certiﬁcation.

District sample. The MIECT Study sought to recruit to the study medium to large
districts that varied in serving minority and/or low-income students in order to examine
whether certain induction policies and practices seemed to have different effects on new
teachers’ experiences across such districts. It was also a goal to recruit districts that had
signiﬁcant numbers of early career teachers. For example, because of declining
enrollments and tight ﬁscal budgets in many Michigan districts in 2007—08, many
districts in the state did not meet the criteria for inclusion in this study because they had
not hired many new teachers in the previous couple of years. In sum, the criteria for
selecting districts included a) variation among districts with regard to the race/ethnicity

and socio-economic status of the students sewed; and b) each district had at least 10 ﬁrst-

29

, second-, and third-year teachers in grades 1-8. Therefore the sample of districts were
purposefully recruited and selected to meet the study’s criteria.

The public school districts that were included in the study included the following:
Daus, Greenberg, Kaline, Underwood, Wagner, and Whitaker in Michigan, and Engram,
Luckrnan, Payton, Sayers, and Wilson in Indiana.1 Each of the eleven districts
participated in collective bargaining as required by state laws, and the CBAs in each
district were collected for analysis. The eleven participating districts varied signiﬁcantly
by student enrollment, percent of students eligible for free or reduced lunch, and percent

minority students.

Table 3.1: District demographics

 

 

 

Total # of Total K-12 Percent Percent Eligible
District Schools in Student Minority for Free/Reduced
District Population Lunch
Daus (MI) 36 18386 0.12 0.51
Greenberg (MI) 80 21448 0.80 0.65
Kaline (MI) 18 9139 0.50 0.42
Underwood (Ml) 43 29261 0. 1 l 0. 1 1
Wagner (MI) 12 7994 0.46 0.36
Whitaker (MI) 23 l 1645 0. 19 0.29
Englam (IN) 20 13666 0.48 0.62
Luckrnan (IN) 18 16138 0.57 0.44
Payton (IN) 13 10662 0.84 0.50
Sayers (IN) 38 21769 0.60 0.62
Wilson (IN) 19 12483 0.59 0.58

 

The three medium-sized districts in Michigan - Wagner, Kaline, and Whitaker —
served from 8,000 to 12,000 students in 2007-08 while the three larger Michigan districts
—- Daus, Greenberg, and Underwood —- served from 19,000 to 30,000 students. The

percentages of low-income students were comparable in Wagner, Kaline, and Whitaker

' Pseudonyms were used for each of the district names in order to help ensure conﬁdentiality

30

 

(29% to 42%) while the percentage of racial minority students ranged from 19% in
Whitaker to 46% and 50% in the other two medium-sized districts. While Underwood
had the highest student enrolhnent (29,803 students), it also served the lowest
percentages of minority and low-income students of the six participating Michigan
districts. The other two large districts, Daus and Greenberg, served comparable
percentages of low-income students (51% and 65%, respectively), but Greenberg served
a much higher percentage of racial minority students.”

The three medium-sized district in Indiana —- Engram, Payton, and Wilson —
served from 11,000 to 14,000 students in 2007-08 while the two larger Indiana districts —
Luckman and Sayers - served from 16,000 to 22,000 students. The percentage of low-
income students in Engram, Payton, and Wilson was comparable, ranging from (50% to
62%); the percentage of racial minority students in Engram and Wilson was around 48%,
while in Payton the percentage was around 84% minority. Luckman served around 57%
minority students, and 44% low income, while Sayers served around 60% minority
students, and 62% low income.

Early career teacher sample. Teachers who taught the core-content areas (math,
science, social studies, English/language arts, and general elementary) in grades 1-8, and
were in their ﬁrst three years of the teaching profession were invited to participate in the
study in 2007-08. Participation included the completion of a fall and spring survey and a
sample of those teachers were asked to participate in one interview. The surveys
administered were in both electronic and paper form.

In an effort to increase participation rates, a ﬁve-contact approach was used for

each survey administration (Dilhnan, 2007 ). A pre-notice letter was sent a week prior to

31

mailing the survey (which included a cover letter and consent form). A two-dollar bill
was included in the next mailing (to help increase participation) which included a link to
the online survey. Research has shown that including a token incentive (such as a two-
dollar bill) improves response rates 19-31 percent over personalized mailing alone
(Lesser et al., 1999). A reminder email with the link to the online survey was sent
approximately a week later. A thank you/reminder post card was sent near the survey
return deadline, and if a prospective study participant did not complete the survey online,
they received a paper copy of the survey. Teachers were compensated with a $20 gift
card for completing and returning a survey.

Overall, 184 early career teachers completed both the fall and spring surveys.

There was a fall response rate of 63 percent and a spring retention rate of those teachers
of 76 percent (See Appendix, Table I for breakdown of response rates by district). Of the
184 ECTs in the sample, 49 of them were ﬁrst-year teachers, 82 of them were second-
year teachers, and 5 3 were third-year teachers. The majority of them were teaching at the
elementary level, 132 elementary teachers versus 52 middle school teachers. Eighty
percent of the early career teacher sample was female, and 90 percent of the sample was
white.

Both the fall and spring surveys asked about teachers’ instructional practices; the
frequency and substance of their interactions with their mentors and colleagues; their
perceptions of relations within their schools; their work conditions; and their future career
plans. In addition, the spring 2008 survey included measures of novice teachers’
participation in teacher association activities, and their views on what teacher association

priorities should be (e. g., more say in school management, teacher evaluation, etc).

32

Items that asked about teacher background, such as degrees, certiﬁcation, and college
attended, were also asked in the spring. By collecting data on novice teachers in fall
2007 and spring 2008, I was able to measure how novice teachers’ commitment to their
grade level assignment, school, and district changed over the course of the year, as well
as how their interactions with mentors and colleagues varied from the fall to the spring.
Mentor/Colleague sample. In order to collect egocentric social network data,2 in
the fall 2007 surveys the novice teachers were asked to list their assigned mentor and
other classroom-based colleagues with whom they discussed professional issues such as
instruction and classroom management. For each novice teacher in the sample, their
mentor teacher and up to eight colleagues that they listed were invited to complete a
mentor/colleague survey. The same ﬁve-contact approach was used to help improve
participation rates from mentors and colleagues, though the mentors and colleagues
received no compensation for completing surveys. These mentors and colleagues were
then surveyed in the early winter of 2008. Data was collected from 66 teachers named as
mentors (64% response rate) and 262 teachers named as colleagues (58% response rate).
Mentors and colleagues who were named but were not regular classroom teachers or
instructional support providers were considered ineligible for participation (e.g., district
personnel) because the goal was to primarily examine the relationship between novices
and their teacher colleagues. Many of the questions asked were the same as those
directed to the novice teachers, concerning their instructional practices, their perceptions

of relations within their schools, their work conditions, and their future career plans.

2 In egocentric data, individuals name those who may be considered close colleagues, and those people are
then surveyed. In socio-centric data, everyone within an organization would be surveyed rather than those

who were named.

33

District and Teacher Association Interviews. In order to gather more extensive
information on the provision of new teacher mentoring, the effects of collective
bargaining in deﬁning the scope of induction, and the labor-management relations
context, Dr. Peter Youngs and I conducted structured interviews with the district teacher
association presidents (n=10) and the human resources (HR) directors in each district
(n=10). One district HR director and one teacher association president were not available
to complete an interview despite several attempts through emails and phone calls to
arrange an interview over the course of several months. Interview questions focused on
human resources practices concerning mentoring activities; and the labor-management
relations climate within each district and within schools in the districts. The interview
data helped inform what type of status and attention novice teachers received from both
the district and the teacher association in terms of the mentoring support they
experienced, as well as to get a better sense of the organizational climate in which new
teachers worked. Additionally, the interviews provided information on how mentors

were selected and assigned to mentees, as well as some comments on the quality of the

various mentoring programs.

Analysis

I performed qualitative and quantitative analysis in multiple stages to align with
the research questions. To address the ﬁrst research question regarding the provision of
mentoring, I utilized both qualitative and quantitative data. For the qualitative data, text
from collective bargaining agreements from each of the districts and notes associated

with the review of the agreements were organized by using N‘Vivo7 software. This

34

software allowed me to easily compare mentoring language in the collective bargaining
agreements across districts. Additionally, the interviews with HR directors and teacher
association presidents were audio recorded and then transcribed to learn more about
formal district policy regarding mentoring (e.g., mentor selection, mentor assignment,
requirements for meeting). The transcriptions were then coded using the same categories
with which the CBAs were coded.

In the ﬁrst stage of analysis, I used NVivo7 to look for systemic patterns of
variation in formal district policy related to the provision of mentoring, and how these
policies seemed related to collective bargaining agreements. For the second stage of
analysis, I drew on interview data from the HR directors and teacher association
presidents to discern the ways in which these policies were implemented across the
districts, particularly paying attention to indications of gaps between the formal policy
and the day-to-day mentoring activities.

In addition to using the qualitative data to learn about the provision of mentoring
and the role that collective bargaining plays in setting induction policy, I used HLM
software to run an ordinal logistic regression to examine which organizational factors
may be associated with the probability that an early career teacher indicated that they had
been assigned a mentor. In particular, I looked at how the presence of CBA language
pertaining to mentoring was associated with the probability of being assigned a mentor.

To address the second research question regarding the extent to which early career
teachers were interacting with their mentors and colleagues, I ﬁrst used SAS software to
produce overall frequencies of novice interaction with their mentors, as well as to look

across interactions around speciﬁc topics, such as curriculum, teaching strategies, or

35

student behavior. I also compared the ﬁequency of interactions between mentors and
mentees by novices’ years of experience and level at which they taught (elementary
versus middle school). In addition, I compared the frequency and substance of ECTs’
interactions with mentors to their interactions with their close colleagues, as well as the
reported level of importance of support provided by the mentors and colleagues. I then
produced correlations to examine how frequency of interactions with mentors and
colleagues may be associated with early career teachers’ perceptions of their
organizational context. Finally, I used SAS software to run a multilevel linear
regression to examine whether or not mentor-mentee match and importance of mentor
support were associated with more frequent interactions with mentors.

To address the third research question, I ﬁrst used SAS software to run a series of
hierarchical linear regressions to examine the organizational and social factors that may
be associated with novices’ commitment to their grade level assignment, school, and
district. In particular, attention was paid to early career teachers’ perceptions of labor-
management relations within their schools, their perceptions of professional ﬁt, and the
level of importance they placed on the support they received from their mentors and
colleagues. I then used hierarchical linear regression to examine the organizational and
social factors that may be associated with novices’ change-in-commitment from fall to
spring. Finally, I used social network analysis to analyze survey data gathered from

novice teachers and their mentors and colleagues. In particular, I examined how novice
teachers’ perceptions of organizational context were inﬂuenced by the perceptions of

their mentors and colleagues.

36

By utilizing both qualitative and quantitative data I aimed to thoroughly address
the three key research questions. The use of qualitative data, particularly the interview
data, provides greater context for understanding the formal structures of mentoring within
7 the districts, but it also provides an opportunity to learn more about how the
implementation actually takes place. This partially addresses the concerns put forth in
new institutionalism regarding the gap between formal structures and informal activities
within organizations. The survey data gathered from the early career teachers, their
mentors, and their colleagues provides even further information on the informal
structures within the schools. The data provides valuable information on the frequency
and substance of ECTs’ interactions with mentors and colleagues, as well as ECTs’
perceptions of the organizational context. Finally, this analysis allows for better

understanding of how the organizational and social context is associated with novice

teachers’ commitment levels.

37

Chapter 4

Mentoring Policy

This chapter addresses the ﬁrst guiding research question; to what extent do
collective bargaining agreements inﬂuence the provision of mentoring? Using the ideas
from new institutionalism as the guiding framework, I aimed to explore both the formal
and informal structures which guide mentoring policy within the sample of districts.
New institutionalists are concerned with how organizations become institutionalized as
they relate to their environments, leading similar organizations (e.g., school districts) to

develop the same or similar formalized structures (Meyer & Rowan, 1977; DiMaggio &
Powell, 1983). In the case of designing and implementing a mentoring program, school
districts respond to state law, research and commercial induction programs, as well as
historical factors within their districts (DiMaggio & Powell, 1983).

Under this framework, one would expect districts to work towards implementing
mentoring programs that ﬁt the mold of “acceptable” to maintain legitimacy, with less
concern about the effectiveness or efficiency in providing support to novice teachers and
improving teacher and student outcomes (Meyer, Scott, & Deal, 1981). This lead to my
ﬁrst hypothesis: Although there may be variation as to what extent collective bargaining
agreements deﬁne the scope of bargaining, I expect there will be little variation in the
provision of mentoring because of the eﬂects of institutional formalization of policy
structures. At the same time, although the formalized structure of mentoring may not

signiﬁcantly vary across the districts, especially within a given state, the informal

38

structures in which mentoring operates may lead to variations in effectiveness, given
individuals and school contexts.

I ﬁrst provide some information on the legal context and requirements for
providing mentoring to new teachers in Michigan and Indiana. Next, using information
ﬁom the coded collective bargaining agreements from each of the eleven districts, and
drawing from interviews with the human resources directors and teacher association
presidents, I outline the formal provision of mentoring in each of the districts. In
particular, I am interested in the extent to which the collective bargaining agreements and
district policy deﬁne the scope of mentoring. I am also interested in how the day-to-day
implementation of mentoring may be different than what is formally deﬁned, so I drew
from the interviews to get a sense of any potential gaps between what policy dictates, and
how mentoring is actually provided. I then present descriptive data on the provision of
mentoring across the eleven districts. Finally, I present output from a multi-level logistic
regression to see if there was a statistically signiﬁcant effect on the probability of having
a mentor assigned to a teacher based on the presence of collective bargaining provisions
concerning new teacher mentoring.

State context. Both the states of Michigan and Indiana require that new teachers
are assigned a mentor, though they do vary in the extent that state policy may guide 1003]
Policy. Michigan law states that all new teachers are to be assigned a master mentor
teacher for their ﬁrst three years in the profession (PA 289 Section 1526; I995). The
Michigan State Board of Education (SBE) made a statement regarding the provision of
mentoring:

The SBE believes that the New Teacher Induction/Teacher .
Mentoring process is a cooperative arrangement between peers m

39

which new members of the teaching profession are provided
ongoing assistance and support by one or more skilled and
experienced teachers. This relationship should be collegial in
nature, and all experiences should be directed toward the
development and reﬁnement of the knowledge, skills and
dispositions necessary for effective learning. This process is
expected to be mutually beneﬁcial for all parties involved and to
result in improved instructional practice and professional
performance (www.michigan.gov/mde).

In order to try to reach the goals laid out in the state mandated mentoring program, the

state board also made recommendations for mentor qualiﬁcations, but deferred actual

mentor selection criteria, qualiﬁcations, and training to local education agencies. There

recommendations include:

Criteria for selection of mentors: It is strongly recommended that
teacher mentors be selected by a district/school committee using
operational guidelines developed at the local level. It is also
recommended that the teacher mentor volunteer willingly for their
responsibility (www.michigan.gov/mde).

Inrportant characteristics of mentors: Demonstrated excellence in
teaching, participation in professional development activities, same
certiﬁcation or specialty area as the new teacher and be located in
same building (if possible), active and open listener
(www.michigan.gov/mde).

The statements posted by the State Board of Education lay out the general goals

of mentoring from the perspective of the state. Generally, the state policy aims to ensure

that novices have district/school support in developing their subject matter and

instructional knowledge and skills, and that the mentoring relationship is collegial in

nature as opposed to evaluative. This is to help promote a relationship based on trust and

“gimme, rather than one in which the mentor may have power over a mentee.

Additionally, the state acknowledges the potential importance of matching a novice With

40

a mentor who has content knowledge in the subject that the novice is teaching, and the
importance of being in the same building to promote more frequent and meaningful
interaction. At the same time, the law and policy gives a lot of discretion to the local
level to design and implement the mentoring policy. This could potentially lead the
greater variation in program design, as well as in program effectiveness.

Similar to Michigan, Indiana law requires that all new teachers to be assigned a
mentor, but only for their ﬁrst year of teaching. The state, though, highly recommends
that new teachers are assigned a mentor for their second year in the profession as well.
As part of a two—year continuing certiﬁcation process (Indiana Mentoring and
Assessment Program - IMAP), which includes the submission of a portfolio at the end of
a teacher’s second year, Indiana state law speciﬁes the provision of mentoring to a
greater extent than in Michigan. Although the provision of mentoring may be more
regulated at the state level in Indiana, the message given by the two states are similar.
The State of Indiana indicates:

Student learning is directly related to teacher knowledge and

abilities and that teachers, like students, should be continual

learners. As a result, beginning teachers will receive guidance and

support by trained mentors, as well as professional development

activities provided by their schools. Such guidance and support

within a collaborative environment should provide practical

strategies to enhance the capabilities of beginning teachers to

increase student learning

(http://www.doe.in. gov/dps/IMAP_online.html).

In contrast to Michigan regulations, Indiana state regulations set criteria for who can
serve as a mentor, particularly requiring them to complete state approved training:

A certiﬁed mentor is an accomplished teacher who has been

selected by the local school district and has completed state-

approved mentor training. This individual is assigned to support a
beginning teacher during their ﬁrst year of participation in the

41

IMAP. The individual mentor is trained to provide instructional

support to the beginning teacher as well as identify additional

resources needed to ensure the beginning teacher develops

competency in his/her teaching. After successfully completing

state—approved mentor training, the participant is qualiﬁed to serve

as a mentor. The ﬁrst priority for the assignment of mentor

teachers must be given to individuals who complete a mentor

training program if available.

The Indiana state policy also acknowledges the potential importance of creating a
collegial relationship between mentors and mentees, and stresses the importance of such
a relationship in developing novices’ knowledge and skills in the classroom. In Indiana
though, they require the mentors to complete state approved training, and because it is
linked to the [MAP certiﬁcation process for new teachers, the mentoring which is done at
the local level is potentially more constrained to meet the requirements of the IMAP. At
the same time, it is left to the districts to identify teachers to complete the mentoring
program, and therefore considerable discretion remains at the district level.

In addition to the state requirements regarding the provision of mentoring, state
regulations concerning collective bargaining are important to consider, particularly the
legal guidelines for the scope of collective bargaining in both Indiana and Michigan. In
Michigan and in Indiana, negotiation over the provision of new teacher induction is
permissive but not mandatory (www.nctq.org/tr3/scope). Districts may seek to avoid
putting language pertaining to induction within a collective bargaining agreement
because it then becomes enforceable by a neutral third party. At the same time, teacher
associations may seek to include provisions concerning induction in collective bargaining

agreements to ensure that novice teachers are supported. Not all local teacher

associations would be interested in negotiating over a permissive issue such as new

42

teacher induction, especially if they are having difficulty negotiating over mandatory
issues such as salary and beneﬁts (Bredenson, 2001; Koppich, 2005).

Other external environmental factors may inﬂuence the provision of mentoring,
particularly economic factors. During the time of data collection, several districts across
Michigan and Indiana faced tight ﬁscal budgets due to the poor economic climate in both
states. For example, both the Daus and Whitaker teacher association presidents indicated
that due to budget cuts in their districts cuts were made to their mentoring programs,
which resulted in the elimination of full-time released mentors. Additionally, in three of
the six Michigan districts in the sample there were layoffs during the year of data
collection (though most were called back), and in several of the districts in the sample
from Indiana they ﬁnally experienced stagnant growth in student population after years of
steady increases. Standardized data on the ﬁscal situation within the districts was not
collected though.

In summary, both states require mentoring for new teachers, in Michigan for the
ﬁrst three years and in Indiana for the ﬁrst year. Both states stress the importance of
collaboration between teachers as a key to professional growth and its relation to student
outcomes. Additionally, both states indicate that selection of mentors should occur at the
local level, but in Indiana teachers are required to complete state approved training before
they can serve as a mentor. State law also allows for the negotiation between teacher
associations and districts concerning the provision of mentoring, but it is not a
requirement. These state regulations are a key part of the environment in which districts
respond to in developing and implementing their mentoring policies. Since district

mentoring policy is at least partially shaped by state regulations, the structure of district

43

mentoring policy may show several similarities between districts. At the same time,
quite a bit of discretion is given to districts in designing mentoring programs, the
selection of teachers who become mentors, and importantly the match between mentor
and mentee. Other factors though should be considered when thinking about the
provision of mentoring, such as the economic situation that districts face.

Collective bargaining provisions. Collective bargaining agreements can be
thought of as responding to both external forces such as state regulations, and internal
forces such as labor-management relations and the needs of teachers. Negotiation of
mentoring policy is allowed but not required under state law in both Michigan and
Indiana, but only four of the 11 collective bargaining agreements reviewed contained any
language pertaining to the mentoring of new teachers. This may reﬂect the fact that
many teacher associations and school districts simply focus on negotiating over
mandatory issues such as salary, beneﬁts, and work conditions rather than on efforts to
improve instructional quality (Johnson & Kardos, 2000; Koppich, 2005). It is possible
though that having language in a collective bargaining agreement pertaining to mentoring
ensures that mentoring policy is fully implemented and enforced, and may signal to
teachers that both the district and teacher association are promoting new teacher
development. At the same time, goals and intentions expressed through collective
bargaining may not ultimately lead to changes in action where institutions encounter
loose coupling among bargaining units (Scott, 1998). In other words, changes in a
collective bargaining agreement may not ultimately lead to changes in behavior on the

ground level if structural elements within the district (e.g., schools, departments,

44

bargaining units, etc) are loosely coupled and in reality are less interrelated and closely
coordinated as organizational charts would show.
It is possible that within the four districts which have collective bargaining

agreements with language pertaining to new teacher mentoring, there is a better
relationship between the administration and the teacher association. This could possibly
allow them to not only successfully negotiate mandatory issues such as salary to a
satisfactory degree for both parties, but also to negotiate over permissive issue areas such
as new teacher mentoring. Or it could be that they are able to negotiate over a permissive
issue area such as mentoring to the degree that it builds common goals and promotes a
collaborative bargaining style in order to move towards negotiating over more difﬁcult
mandatory issues. The data collected in this study can not directly attest to this issue, but
it does point out the need for further research.

The four districts which contained language pertaining to mentoring in the CBA
include Greenberg, Kaline, Wagner, and Whitaker, all of which are in Michigan. The
Kaline and Whitaker CBAs simply stated that the mentoring relationship was to be
conﬁdential and should not be included in evaluation of mentor teacher or mentee. This
is aligned with the state speciﬁcation that the relationship should be collegial in nature, in
order to promote a better professional relationship not based on formal authority. The
Greenberg CBA stated that when possible, mentors and mentees were to be matched by

program and school. Additionally, it also indicated that the mentoring relationship was to
be conﬁdential and not to be used for evaluative purposes. In addition, it suggested that
school administrators provide release time for mentors and mentees to work together. At

the same time, this was a suggestion and discretion was left to the building

45

administrators. In order to possibly recruit teachers to become mentors, or at least
provide extra incentive, the CBA said that at the principals’ discretion, mentors could be
compensated $400 for mentoring a lst year teacher, $200 for mentoring a 2nd year
teacher, and $100 for mentoring a 3rd year teacher. It is not clear that principals ensured
that mentors were compensated, and it’s not clear how much ﬁnancial incentive was
necessary to attract and retain an adequate number of high quality mentors.

In Wagner, the CBA indicated that when possible, mentors and mentees were to
be matched by certiﬁcation and school. Additionally, the mentoring relationship was to
be conﬁdential and not to be used for evaluative purposes. Similar to Greenberg, the
CBA said that school administrators were to provide release time for mentors and
mentees to work together, and suggested that the mentor and mentee be assigned
common preparation time. Again, this seemed to be more of a suggestion, and discretion
was given to the building administrator to manage their building given the needs of their
teachers and students. The CBA also guaranteed that mentors were compensated 2
percent of BA base salary (about $750) for mentoring a lst year teacher, and 1 percent of
BA base for mentoring a 2nd or 3rd year teacher.

Echoing the language that stems from the state regulations, language in the
Greenberg and Wagner CBAs stressed the importance of mentor-mentee match in
promoting a valuable mentoring experience. Additionally, both suggest the provision of
structural elements, such as release time, to promote more frequent and substantive
interactions. At the same time, the matches are made at the school level by the building
administrators, and it is up to the building administrators to schedule release time or

common planning time. It is not clear as to what extent the district monitored the

46

provision of mentoring in individual school buildings. Having language in the CBAs

though may provide recourse to new teachers and mentors who feel that the formal
mentoring policy is not being implemented properly at the school level. Additionally,
building level principals may be more inclined to follow policy deﬁned by a CBA rather
than policy not deﬁned by a CBA to avoid any formalized teacher grievance. Yet, the
language within the collective bargaining agreements was relatively soft, in the sense that
they use words such as “suggest” and “when possible” or “at the principal’s discretion”.
This may reﬂect the fact that because negotiating over new teacher mentoring is not
mandatory, when it is contained within the collective bargaining agreements it is an area
that pays lip service to promoting new teacher induction rather than acting as an
enforceable code of policy.

Overall, the majority of collective bargaining agreements reviewed contained no
language pertaining to mentoring, and two of the four that did was quite limited in
deﬁning the scope of the mentoring program. The two collective bargaining agreements
that contained more language (Greenberg and Wagner) may provide greater opportunities
for mentors and mentees to interact because it calls for release time, and also may make it
more likely that a mentor and mentee are matched by certiﬁcation or grade level. Yet,
formal district policy across all the districts may not vary greatly regardless of provisions
in the collective bargaining agreements because the collective bargaining provisions did
not really set further guidelines than what the states did. Additionally, quite a bit of

discretion was given to building level administrators, therefore although policy might
look uniform across districts and within districts, the quality and substance of mentoring

may vary greatly on the ground level.

47

Formal district policy. In conjunction with or in the absence of collective
bargaining provisions which guide mentoring policy, all eleven districts had some form
of formal mentoring policy. In accordance with state regulations, all six Michigan
districts in the sample had an induction policy which assigned a mentor to a teacher for
their ﬁrst three years in the profession. Additionally, in all six districts mentor

assignments were made at the building level, the mentors had to have tenure, and in ﬁve
of the six Michigan districts mentors received some form of compensation.

In Underwood, mentors were required to attend an orientation, and to keep a log
of their interactions with their mentees. In Kaline, mentors and mentees received training
on how their relationship should develop. In both districts, the HR directors indicated
that it was policy to try to match mentor and mentee by grade level or subject matter
when possible. In Wagner, mentors were offered training but it was voluntary, and they
also were required to keep track of their interactions with their mentees through
documentation In Daus, mentors and mentees were supposed to attend occasional after-
school meetings which were directed by two master teachers, and they were also required
to write in mentoring journals about their interactions. In the words of the Daus HR
director, “It’s one thing to have a casual relationship with your mentor, it’s another one to
actually develop a formal mentoring journal.”

Because Indiana’s mentoring requirements were a little more centralized at the
state level, there were some differences compared to the districts in Michigan,
particularly due to the required state approved training for mentors. For example, in
Engram the human resources director had a direct role in assigning mentors in

collaboration with building principals, because she had a centralized list of which

43

teachers in her district had completed the required training to become a mentor.

Similarly, the Payton HR director said that mentors were assigned by the district, but

most of the time they were in the same building as the new teacher to whom they were
assigned. The Wilson HR director said that their mentoring program was part of a two-
year professional development program in addition to what building level principals were
doing to meet their school speciﬁed goals. When asked about the mentoring program in
his district, the Luckman HR director said, “That is more state-directed than it is the
district. I think we do more than the state minimum.” As with Luckman, the Sayers I-IR
director indicated that the mentoring program was more guided by state policy rather than
anything speciﬁc at the district level.

Differences in the state policy environment did produce some differences in the
provision of mentoring across districts in Michigan and Indiana. In Michigan, more
discretion was given to districts in assigning and training mentors, who in turn gave more
discretion to building principals to make the assignment decisions. In Indiana, because
the state regulations were more prescriptive in what training a teacher needed to be a
mentor, and its connection to the two-year portfolio assessment for new teachers, the
mentoring programs seemed to be more centralized at the district level. It could be
expected that the differences in these formalized structures may lead to differences in the
actual provision of mentoring, but ultimately the effectiveness of the mentoring programs
relied on the quality of the relationships that were developed between the mentor and
mentee at the school-level.

Day-to-day practice. In line with the framework, it is important to consider the

potential gaps between actual practice and the formalized structures which make up and

49

give legitimacy to mentoring programs. The framework suggests that organizations are
more concerned with institutionalized legitimacy rather than ensuring that programs are
necessarily effective, often decoupling practice from evaluation and supervision (Meyer
& Rowan, 1977). Interviews with both the HR directors and teacher association
presidents shed some light on this matter.

The interviews with the human resources directors and the teacher association
presidents indicated their view of the importance of a good mentor-mentee match. For
example, the Underwood HR director stated, “The principals try to match those pairings
based on subject area, based on proximity, you know location in the building, grade level.
And they try to have a good ﬁt in terms of who the mentor is.” This is inline with the
individual level theory of action derived from social capital theory, which stresses the
importance of accessing resources and support through relationships. If a novice teacher
is paired with a teacher who has extensive knowledge of the curriculum for which they
are using because they teach the same grade or subject matter, it may be more likely that
the mentor will therefore be able to provide higher levels of assistance in using the
curriculum, developing teaching strategies, or assessing students.

When asked about how important the mentors are to the new teachers she
responded:

“I think they ﬁnd it beneﬁcial. Especially where we have good

matches. The new teachers’ response I get is that they deﬁnitely

appreciate what they get from their mentor. If it isn’t a good ﬁt,

what I ﬁnd is that the mentees seek someone else as a mentor.

Whether it’s their formal mentor or an informal mentor, new

teachers deﬁnitely beneﬁt from having a mentor in the building
they’re working in.”

50

Her comment regarding the “informal mentor” indicates that support for new teachers
may not only come through the formal structures put in place by district policy. If the
ECTs do not ﬁnd the support from the formal mentor beneﬁcial, many seek out other
colleagues to provide expertise, resources, and other supports. This was echoed by the
Payton HR director, “We talk to them about their assigned teacher mentor, and we also
suggest that they ﬁnd a senior teacher that they have some comfort level with along with
their mentor to augment mentors inﬂuence in their early years.” These statements
indicate that they acknowledge that mentoring alone may not provide the necessary
support new teachers need, especially if the mentor-mentee match fails to provide
novices access to the necessary resources and support they need.

The Kaline HR director also suggested that the mentor-mentee match is important
in increasing the likelihood that the mentoring relationship will be beneﬁcial.
Additionally, he felt that principals were best suited to assign mentors.

“I try to have the principals pick them because they obviously have

the most experience with those new staff and they know their own

staff. They have to be tenured teachers, the mentors are, so that

means they’ve been with the district four years. They’ve all been

trained, so you can’t be a mentor without being trained. And they

try to pick them personality wise and subject matter wise. So we

leave it to the building principals to pick the appropriate mentor.”
He clearly laid out the goals through the eyes of the district administration, but beyond
setting minimum criteria (mentors have to be tenured and receive some sort of training),
the mentoring relationship was largely made at the building level. Again, therefore

regardless of formal district mentoring policy, principal discretion plays a large role in

promoting eﬁective mentoring.

51

In contrast to the Kaline HR director, the Engram HR director discussed the
difﬁculty in making appropriate matches, particularly because mentors in Indiana needed
to complete state required training. She stated:

“It’s been a disaster for the last three years. The reason for that is
that Indiana requires us to use trained mentors, mentors that have

been through the 40 hours of a mentor training program. It has
taken us three years to get up to speed to have anywhere near

enough mentors.”
This speaks to the difficulty in recruiting highly effective mentors. Districts and
principals often are limited in which mentors they have because they are volunteers, and
they have limited resources to compensate them. They are often forced to match mentors
and mentees based on who is available given their personnel constraints, rather than who
they believe (and profess through policy) to be a desirable match.

Not only does the match by grade level or subject matter seem to matter, but also
matches by personality. That is why many of those interviewed indicated that it was best
for principals to assign the mentors because they were in a better position to know the
personalities of their staff. But the Engram HR director pointed out the difﬁculty in this
as well, “We have discussed with the mentors that it’s impossible to match them
personality-wise.” She felt it was too difficult to predict who would click in terms of
personalities when the district or school assigns a mentor. These concerns indicate that
when considering matches between mentors and mentees, the intangibles (such as
personality) may be as important if not more important than subject matter/ grade level
match in determining the quality of the relationship. The social aspect of mentoring,
although acknowledged within policy, is often difficult to formulize in practice when an

administrator has to make a one-to—one professional match.

52

In addition to difficulties in making favorable matches, some of the HR directors
and teacher association presidents questioned the effectiveness of their programs.
Although the Daus CBA contains no language regarding mentoring, the teacher
association president stated, “We do follow the minimum state guidelines of assigning a
mentor and we do track professional development.” He continued, “We’re following the
basic requirements of the law, but we’re not doing anybody a service.” He indicated that
in previous years they had full-time mentors who were released from teaching, and that
they were better able to provide directed assistance to new teachers. Because of budget
constraints they changed policies. The lack of resources and trained mentors, in his view,
left a shell of a mentoring program which was legitimated because it met the
requirements of the law but was not effective. Although data was not collected
pertaining the level of resources that each district spent on new teacher mentoring, the
Daus HR director’s sentiments about declining resources was echoed by several of the
other HR directors and teacher association presidents in both Michigan and Indiana.

The Kaline teacher association president also suggested that district mentoring
policy was not very effective, even though there was language in the CBA pertaining to
mentoring:

“There needs to be a mentorship process teacher-to-teacher to help

walk them through that whole thing during the course of the year

so that if you run into a bump, or you see something that you don’t

understand you have someone you can go to who can talk you

through. That is something we have contractually. I wish that the

association and district were better able to carry it through, because
it would avoid a lot of issues that do come up.”

He was concerned about the gap between policy and contractual requirements and what

actually takes place in the schools. So, in his district, although there is formal mentoring

53

policy, partially deﬁned by collective bargaining, there were few formal mechanisms to

evaluate and supervise the implementation of mentoring policy.

The Payton teacher association president indicated that the quality of mentoring
could vary by building depending on who the principal was:

“In the great buildings with the best administrators there is a
process that the administrator interviews the teacher and then
chooses within their mind the best one for the start. If there is a
problem that administrator is very comfortable with changing. In
some of our lesser strong, our least strong buildings, we have a
couple of principals who chose who your mentor is and that’s who
they are for the entire time. If you don’t like it, you can leave.”

This highlights the discretion given to building administrators, and also the ways in
which formal policy may look similar across districts but the way that it is implemented
on the ground varies signiﬁcantly. This can have a great impact on the quality and
effectiveness of a mentoring program, but since the mentoring program appears to be

legitimate by state and professional standards the quality of the program is not fully

assessed.

The Wagner HR director went even further in describing the lack of effectiveness

of their mentoring program:

“That you know I’m afraid in our district is not terribly proud of
this, but I think that is an area that is not systematic enough. In
other words, what we have a problem with is, and I do provide
mentor training to our mentors, but it’s voluntary. So I get about
1/3 of the mentors to show up, and then what we end up with is the
quality of the mentoring that takes place is all over the map. It’s
not very systematic. And we also try to hold them accountable in
terms of making them turn in reports three times a year, how many
hours they spent together. But nobody really knows that they
really spent that amount of time, and we don’t know what they are
talking about really. So, the mentoring program I think, though it’ 5
provided to new teachers, is probably an area of the greatest need
for us to try to build in some quality assurance. Because some of

54

our mentees come out of that and say ‘that is the greatest thing that

happened during my early career; I had the greatest mentor, and it

made this difference in my career’. And then I have some mentees

who didn’t want to tell on their mentor at the time. They’ll say,

‘You know that was the biggest waste of your money you ever

used.’ And we haven’t ﬁgured out how to get after that yet.”

These two quotes illustrate that regardless of state-level or district-level policy, mentoring
can look very different depending on the speciﬁc individuals involved and depending on
which building the novice teachers are placed. This illustrates the loose coupling
between regulators, districts, schools, and individuals which often leads to differences
between policy and practice. Additionally, since there is a role for the principal in
assigning mentors in most of the districts, the quality of principal leadership may have
important consequences in leading productive mentoring. Also, the supervision of
mentoring may need to take place at the school-level rather than just completing logs to
meet district or state requirements. This quote also speaks to the investment that mentors
need to make in the professional relationship. Are they entering into the mentoring
relationship because their principal asked them to (or told them to), or because they feel
that it is important for the development of their novice colleagues?

The interviews with the HR directors and teacher association presidents show that
the actual day-to-day practice, at least in some cases, varied from the formalized structure
0f mentoring mandated by the districts and states. Because of staffing restraints, it often
proved difﬁcult to ensure that new teachers were paired with someone who taught the
same grade level or subject matter as they did, and as pointed out, at times it was very

difficult to accurately match teachers based on their personalities. Also, besides mentor-

mentee logs there seemed to be very few mechanisms to evaluate the quality and

55

effectiveness of the mentoring that was taking place. For example, the Greenberg HR
director indicated that many of the mentors and mentees were only meeting once a
month, and that was not enough in her opinion to develop a productive relationship which
would allow novices to receive guidance in instruction.

In addition to reviewing the collective bargaining agreements and interviewing
the district HR directors and teacher association presidents, I wanted to use the survey
data to see if there was an association between having collective bargaining language
pertaining to mentoring and the likelihood that a novice teacher indicated that they have a
mentor. As previously stated, having language regarding the assignment of a mentor in. a
CBA may be a mechanism for ensuring that policy is fully implemented. Table 4.1
shows the percentage of ECTs who indicated that they had been assigned a mentor for the
2007-08 school year. Sixty-six percent of the ECTs in this sample indicated that they had
a mentor, and a higher percentage of ﬁrst year teachers and elementary teachers indicated
that they had a mentor than either more experienced novices or middle school teachers.
Overall, 86 percent of those required by law to be assigned a mentor indicated that they
did have a mentor. This indicates that a substantial number of teachers were not
receiving the formal support which the law stipulates is an essential part of teacher
development. For example, an ECT who was in her third year of teaching in Michigan,
indicated in an interview that she had been assigned a mentor during her ﬁrst year of
teaching but had not interacted with her during the current school year. This further

highlights the decoupling of formalized policy from day-to-day practice.

56

Table 4.1: Percent of ECTs in sample with mentors

 

 

 

 

 

 

Mean Std. Dev.
ECT indicated had mentor 0,66 0.47
lst yr ECT indicated had mentor 0.94 0.24
2nd yr ECT indicated had mentor 0.68 0.47
3rd yr ECT indicated had mentor 0.38 0.49
Elementary school ECT indicated had mentor 0.7 l 0.45
Middle school ECT indicated had mentor 0.54 0.50 l

 

 

 

In order to explore the relationship between formalized policy and the likelihood
of being assigned a mentor, I ran a multi-level logistic regression with “having a mentor”
(0 = no; 1 = yes) as the outcome variable. Assuming a Bernoulli distribution, I ran the
following model to particularly explore whether or not having language in a collective
bargaining agreement was associated with the probability of an ECT indicating that they

had a mentor (Table 4.1 describes the selected variables used in the analysis):

'lijk = 7000 + 7001 (CBA mentor provision)k + 7002 (MI district)k + 7003 (district Silelk +
7010 (percent white)J-k + 7020 (percent free/reduced lunch)J-k + 7100 (mid sch teacher)ijk +
7200(white teacher),-J-k + 7300 (2nd yr teacher)ijk + 7400 (3rd yr teacher )ijk + 7500 (female),'jk

+ r01k + uoor
Using a logit 1in function, this model explores the log-odds Ilijk that teacher ‘i’ nested

Within school ‘j’ nested within district ‘k’ indicated that they had a mentor based on the
Presence of collective bargaining language pertaining to mentoring, district size, percent

0f Students within a school who are eligible for free or reduced priced lunch, percent 0f

57

students within a school who are white, level of school taught, race, years of experience,

gender, and what state a district was located in. Person level controls for race and gender

were included to help control for personal factors even though the vast majority of the

sample was white and female.

Table 4.2: Variables included in analysis

 

 

 

 

 

 

Variable Name Description Mean
T eacher-Ievel
have a mentor Outcome variable; a dummy variable (1 == had a 0.67
mentor; 0 = did not have a mentor)
mid sch teacher A dummy variable (1 = ECT teaches middle 0.26
school; 0 = ECT teaches elementary school)
white teacher A dummy variable (1 = ECT is white; 0 = ECT is 0.92
non-white)
2nd yr teacher A dummy variable ( 1 = ECT was in 2nd year of 0.46
teaching; 0 == ECT was not in 2nd year of
teaching)
3rd yr teacher A dummy variable (1 = ECT was in 3rd year of 0.29
teaching; 0 == ECT was not in 3rd year of
teaching)
female A dummy variable (I = ECT was female; 0 = 0.83
ECT was male)
School-level
percent white Percent of students in school who were white 0.52
(sd = 0.30)
percent Percent of students in school who were eligible 0.53
free/reduced lunch for free or reduced priced lunch (sd = 0.24)
District-level
CBA mentor A dummy variable (1 = CBA contained language 0.36
provision pertaining to mentoring; 0 = CBA contained NO
language pertaining to mentoring)
MI district A dummy variable (1 = district is in Michigan; 0 0.34
= district is in Indiana)
district size Total prekindergarten through twelfth grade 15.72
L__ student population in the district divided by 1,000 (sd = 6.56)

 

 

58

The distribution is assumed to be a Bernoulli distribution, with level-I sampling
Yijkltpijk ~ B(mijk, (pijk), where Yijk is the binomial distribution with milk “trials” and
probability of mentor assignment per trial (Pijkv The level-2 and level-3 variance
distribution is r0jk ~ N(O, Tu) and l~l00k ~ N(0, TB) respectively. Table 4.3 shows the

ﬁxed effects results from running the conditional model, and Table 4.4 shows the random

effects.

Table 4.3: Likelihood of mentor assignment -— Fixed effects

 

 

 

- _ Coefﬁcient
Fixed Effect (11 -— 171) lt-statj
Intercept, 7000 300*
( l .94)
District-level
CBA mentor provision, 7001 40.17
(0.16)
W district, 7002 1.68
(1.76)
district size, 7003 0.002
(0.3 1)
School-level
percent white, 7010 -0.57
(0.50)
percent free/reduced lunch, 7020 0.21
(0.17)
T eacher-level
mid school teacher, 7100 -0.77*
( 1 .70)
white, Yzoo 0. 18
(0.26)
2nd yr teacher, 7300 -2.25*"
(2.84)
3rd yr teacher, 7400 -3.60***
(4.45)
female, 7500 -O.27
(0.48)

 

statistically signiﬁcant: *** 0.01; ** 0.05; * 0.10

 

 

The coefﬁcients for both 2nd year teacher and 3rd year teacher were both
statistically signiﬁcantly and negative. This corresponded to the probability of having an
assigned mentor of 0.10 and 0.03 respectively relative to a ﬁrst-year teacher, ceteris
paribus. This is also not surprising considering that second and third year teachers in
Indiana were not required by law to be assigned a mentor. I did include a state dummy
variable for whether or not the district was in Michigan, and although the coefﬁcient was
positive it was not statistically signiﬁcant. Therefore I was not able to reject the null
hypothesis that teachers in Michigan were more or less likely to be assigned a mentor
compared to teachers in Indiana. Additionally, middle school teachers were less likely to

have indicated that they had a mentor, holding all other variables constant.

Table 4.4: Likelihood of mentor assignment - Random effects

 

 

Random Standard Variance df Chi-square p—value
effect deviation component

lev-I & lev-2

Intl, R0 0.38 0.15 82 184.46 0.000
lev-3

Intl/Int2, 0.14 0.02 7 3.63 >.500
U00

 

 

This chapter aimed to address the following question: To what extent do CBAs
inﬂuence the provision of mentoring? I hypothesized that they would have minimal
effects, because district policy would vary little in response to state regulations.
Additionally, because negotiating over the provision of mentoring is permissible but not
mandatory in both Michigan and Indiana, it suggests that districts would seek to avoid
having binding language concerning mentoring within a CBA. This chapter shows that

the institutionalized nature of mentoring policy is likely a result of districts responding to

60

their external environment (state law), and in line with the ﬁamework of new
institutionalism the realities of the internal environment created gaps between policy and
practice. Further, having language in the collective bargaining agreement did not seem to
be related to the actual day-to-day provision of mentoring. Although four of the CBAs
contained language pertaining to the provision of mentoring, those CBAs did not require
anything beyond what was being implemented by many of the other districts in the
sample. It is not clear though whether the teacher association inﬂuenced mentoring
policy through informal channels not represented in the collective bargaining agreements.
Additionally, formalized policy across the districts did not vary greatly regardless of
CBA language, in part because district-level policy was responding to the requirements
of state regulations and did not necessarily go beyond that.

Additionally, there was variation in the potential quality of mentoring at the
school and individual level. A substantial amount of discretion was given to building
level administrators in overseeing mentoring programs, such as in making mentor-mentee
assignments and ensuring compliance. The interviews suggested though that the quality
of mentoring in their buildings varied in part based on the quality of building level
administrators. This is in line with previous research which found that principals’
conception of mentoring and subject matter knowledge was associated with the quality of
mentoring in the building (Youngs, 2007b; McGraner & Henrick 2008). The HR
directors and teacher association presidents also felt that mentor-mentee match by grade
level, subject matter, and personality was an important factor in making it a worthwhile
relationship. This is supported by research that suggests proper alignment between

mentor and mentee is associated with lower rates of teacher attrition and greater

61

opportunities for learning (Smith & Ingersoll, 2004; Grossman & Thompson, 2004).

This match may be related to the frequency of interaction between mentor and mentee, as
well as what subjects they talk about. But, as summed up by the Wagner HR director,
“Because some of our mentees come out of that and say ‘that is the greatest thing that
happened during my early career; I had the greatest mentor, and it made this difference in
my career’. And then I have some mentees who didn’t want to tell on their mentor at the
time. They’ll say, ‘You know that was the biggest waste of your money you ever used.
Therefore, although districts may try to legitimate their mentoring programs by creating
institutionalized structures, the quality of the mentoring often depends on individual

action.

62

Chapter 5

Early Career Interactions with their Mentors and Colleagues

This chapter addresses the second guiding research question: to what extent are
early career teachers (ECTs) interacting with their mentors, as well as with their
colleagues? Continuing to use the ideas from new institutionalism as the guiding
framework, I aimed to explore the extent to which novices were interacting with their
formal mentors, and the extent that they were interacting with other teacher colleagues.
Early career teachers make sense of their organizational context through formal
mechanisms such as policy, but also through interactions with their colleagues, which
may lead to different outcomes (Lounsbury, 2001; Powell & Colyvas, 2007; Frank,
Krause, & Penuel, in progress). Here, ideas from social capital theory are also important
to consider, particularly how professional relationships mediate the effects of district and
school policies, as well as facilitate access to resources (Coleman, 1988).

Under this framework, one would expect that novice teachers will invest in
relationships that help them more clearly understand their organizational context, as well
as provide them access to resources and support Findings from recent research suggest
that novice teachers access support and resources through both formal and informal
channels, and that the degree to which they access these supports is partially based on
policy and partially based on personal needs (e.g., Achinstein, Ogawa, & Speiglman,

2004; Grossman & Thompson, 2004; Kardos et al., 2001; Coburn & Russell, 2008). As
suggested by the conceptual framework, novices operate both within the formal

institutionalized structure of the organization (e.g., in a mentoring program), but they also

63

 

‘3,

9 I

It};

W

in
K

.1“.

he

function within the informal structure (e. g., interacting with colleagues). The extent to
which the formal and informal mechanisms of socialization differ can have important
consequences for how new teachers make sense of their organizational context. This
lead to my second hypothesis: I expect that early career teachers are likely to interact
with their close colleagues more often than with their formally assigned mentors. This
reflects novices ' ability to self-select into sub-groups within an organization which may
better reﬂect their beliefs and practices regarding education, compared to those of the ir
formally assigned mentors. Continuing from ﬁndings from the ﬁrst chapter, it could be
expected that because mentoring policy is often institutionalized in order to be
legitimated, it is less concerned with being effective in assisting teachers. Therefore
novice teachers may seek out other individuals to obtain resources and supports, rather
than go to their formally assigned mentor with whom they may have few shared interests
or practices, which may ultimately affect commitment levels.

In this chapter, I ﬁrst present frequency charts indicating the frequency and
substance of ECT interaction with their mentors and colleagues. In particular, I was
interested in exploring the differences in the levels and substance of interaction. Next, I
present data on the ECTs’ indication of the level of importance of support received ﬁom
the mentor and colleagues. I then present some correlations between frequency of
interaction with mentors and various organizational factors. Finally, I present ﬁndings
from running a multi-level linear regression which aims to examine the relationship
between mentor-mentee frequency of interaction and organizational and social factors,
particularly mentor-mentee match and ECTs’ reported importance of support from their

mentors.

64

u.

,u

Frequency of interaction. The early career teachers who indicated that they had a
mentor were asked to indicate their overall frequency of interaction with their mentor in
both the fall and spring surveys, ranging from “never”, “less than once a month”, “1-3
times per month”, “1-2 times per week”, “3-4 times per week”, and “every day”. Chart
5.1 represents the ECTs’ responses, comparing their responses in the fall to their
responses in the spring (see Appendix Table II for percent responses). Sixty-six percent

of the ECTs indicated that they were interacting with their mentors at least once a week
in the fall, but only 59 percent reported that they were interacting with their mentors at
least once a week in the spring. Additionally, a small percentage of ECTs indicated that
they had never or had rarely (i.e., less than once a month) interacted with their mentors;
16 percent in the fall and 19 percent in the spring.

Overall, the data indicates that the ECTs were interacting with their mentors
slightly less frequently in the spring than they were in the fall, though the difference was
not statistically signiﬁcant. It is possible that some ECTs interact with their mentors less
frequently in the spring than in the fall because they have less need for support; such as
the mentor served as a bridge to other support in the beginning months but needed to play
a smaller role in the spring once the ECTs were more established. At the same time, it is
possible that for some ECTs, the support they received from their mentors in the fall was
not highly valued and therefore they sought less assistance from their mentors as the year
progressed. The large percentage of novices who indicated low frequency of interactions

(less than once a month/never) could be cause for concern. It at least could be an
indication that they do not ﬁnd the mentoring relationship valuable, or they do not ﬁnd

sufficient time to actually meet with their mentOr.

65

‘1'
‘l'
it

,1“

\i

Figure 5.1: ECT interaction with their mentor

 

I Fall
I Spring

    

 

Never Less than 1-3 times 1-2 times 34 times Every day
once a per month per week per week
month

 

 

In addition to looking at the overall level of interaction between the ECTs and
their mentors, I explored differences by years of experience and level of teaching. Chart
5.2 shows the responses by ECT years of experience for their frequency of interactions
with their mentors in the spring (see Appendix Table III for percent responses).
Interestingly, no ﬁrst-year ECTs reported that they never interacted with their mentor and
a lower percentage of them indicated that they interacted with their mentor less than once
a month as compared to second- and third-year teachers. This may suggest that ﬁrst-year
teachers were interacting more frequently with their mentors because they were in need
of more assistance than their slightly more experienced novice colleagues. There could
have also been more pressure placed on mentors of ﬁrst-year teachers by building or

district administrators to actively initiate interactions with the brand new teachers. A
larger sample size for each of the cells (i.e., years of experience) may provide more

information on the distribution of frequency of interaction.

66

Figure 5.2: ECT-mentor interaction by ECT years of experience

 

ECT-Mentor Interaction by Yrs Exp

mum
UIOOIS

4‘
DUI

  

Percent Response
M
O

 

 

001

Never Less than 1-3 times 1-2 times 34 times Every day
once a per month per week per week
month

Frequency of Interaction j

 

 

 

The relatively high percentage of third-year teachers who reported interacting
with their mentors every day was not expected, but this may in part be due to the small
sample size when comparing teachers by years of experience (i.e., only 20 third-year
teachers indicated that they had a mentor). Because of the small sample size and due to
the categorical nature of the ECT-mentor interaction variable, I ran a Wilcoxon rank-sum
test (equivalent to the Mann-Whitney U—test) to compare ﬁrst-year teacher responses to
the second- and third-year teacher responses (Cody & Smith, 1997). Although the ﬁrst-
year teachers reported slightly more frequent interactions with their mentors, the

difference was not statistically signiﬁcant (Two-Sided Pr >= |S - Mean} 0.14). If a
second- or third-year teacher had developed a good working relationship with their

mentor teacher, it might be expected that they would continue to meet on a regular basis

67

with them throughout their early career. Additionally, for the second-year teachers in
Indiana, they may have required more assistance with their certiﬁcation portfolio
assessment throughout their second year.

In addition to looking across the ECTs’ years of experience, I also compared their
frequency of interaction with their mentors based on whether they taught elementary
school or middle school. The organizational structure of elementary schools and middle
schools are quite distinct, with middle school teachers often teaching ﬁve or six sections,
and over 100 students, while elementary teachers are primarily responsible for a much
smaller core group of students. Additionally, middle schools are often organized by
departments which may impact the interactions of teachers differently. Chart 5.3 shows
the responses of the ECTs based on whether they taught at the elementary or middle
school level (see Appendix Table IV for percent responses).

The rank-sum test indicates that for this sample of teachers, middle school ECTs
were interacting with their mentors less frequently than their elementary counterparts
(Two-Sided Pr >= [S - Meant 0.07). This may reﬂect the organizational differences
between elementary and middle schools. Elementary schools tend to be in smaller
buildings, which may make the proximity of an ECT and mentor within a building closer,
and the structure of classes and classroom activities may make mentors more accessible.
Additionally, there may be a more likely chance that the mentor at the elementary school
will have knowledge of the curriculum that the elementary ECT is using; they are also
most likely to both hold a general elementary teacher certiﬁcation. At the middle school

level, there may be a greater chance that an ECT will be matched with an out-of-ﬁeld

mentor, not in their department, and in a different location within the building.

68

Figure 5.3: ECT-mentor interaction by organization level

 

    

 

i I I I
Never Less than 1-3times 1-2 times 3-4times Everyday

once a per month perweek perweek
month

 

In addition to being asked about their overall frequency of interaction with their
mentors, the ECTs were asked about the frequency of their interactions with their
mentors regarding particular topics. Their responses were collapsed into four categories
from the six; “less than once a month”, “less than once a week” (at least once a month),
“1-2 times per week”, “3 or more times per week”. Chart 5 .4 shows the ECTs’ responses
for their frequency of interaction with their mentors in the spring regarding curriculum,
teaching strategies, student behavior, classroom assessments, standardized testing, and
psychological support.

As indicated by the chart, a greater proportion of ECTs indicated that they were
interacting with their mentors at least weekly with regards to curriculum, teaching
strategies, and student behavior. The ECTs reported much less interaction with their
mentors regarding standardized testing. There were also relatively high proportions of
ECTs who reported that they were interacting with their mentors less than once a month

across all topics, especially around standardized testing and psychological support. To

69

the extent that is important for novices to receive assistance in regards to these topics, a
signiﬁcant proportion of the ECTs were rarely interacting with their mentors. Again, it
could be a sign that a) the novices do not feel that the relationship provides support in
these areas, b) many of the mentors may not have expertise in these areas, or c) there may
be structural barriers to frequent interaction (e. g., no release time, overburdened

schedules).

Figure 5.4: ECT-mentor interaction by topic

 

 

   

 

 

 

ECT-Mentor Interaction Spring
5 Psychological Support
‘5' ' ' ﬁ—h
g Standardized Testing l Less than once a month
it Assessments I Less than once a week
g Student Behavior D 1-2 times perweek
3:. Teaching Strategies Cl 3 or more times per week'
o ,
"' Cum'culum 7 ‘
0% 20% 40% 60% 80% 100%
Proportion of Response

 

Taking the difference between the ECT spring responses and fall responses, I ran
a series of Wilcoxon signed rank tests to examine whether or not there was a statistically
signiﬁcant change in frequency of interaction over the course of the year. Table 5.1
indicates that for nearly all of the categories the ECTs were interacting with their mentors

less frequently in the spring than in the fall, and the mean differences (except for

70

psychological support and overall interaction) were statistically signiﬁcant at least at the
0.10 level. This could be because a) the ECTs needed less support as the year
progressed, b) novices and their mentors became more busy over the course of the year,
or possibly because c) they did not ﬁnd the assistance from their mentors as valuable as

other possible supports.

Table 5.1: ECT-mentor frequency of interaction (change from fall to spring)

 

 

ECT-Mentor Inte tion Std.

(spring-fall) me Me” Dev. Pr >= '5‘
Overall Interaction -0.09 1.30 0.3961
Psychological Support 0.00 1.25 0.9818
Standardized Testing -0.21** 1.10 0.0288
Assessment -0. 19'” 0.94 0.0192
Student Behavior -0.26* 1.21 0.0576
Teaching Strategies -0.23** 1.15 0.0406
Curriculum -O.19* 1.17 0.0688

 

 

 

Statistically signiﬁcant: *** 0.01; '"' 0.05; * 0.10

In addition to getting support from formal mentors, novice teachers interact with
their colleagues. In the fall ECT survey, they were asked to list up to eight school-based
colleagues who were responsible for instruction and with whom they engaged in
professional interactions (e.g., interactions about curriculum, instruction, students, school
policies, parents, etc.). They were then asked to report on their frequency of interaction
around the same topics with those colleagues, again the responses were collapsed to the
following: “less than once a month”, “less than once a week” (at least once a month), “ 1-
2 times per week”, “3 or more times per week”. Chart 5.5 shows the ECTs’ responses for

their frequency of interaction with their colleagues in the spring regarding curriculum,

71

teaching strategies, student behavior, assessments, standardized testing, and

psychological support.

Figure 5.5: ECT-colleague interaction by topic

 

ECT-Colleague Interaction Spring

 

Psychological Support
Standardized Testing
Assessments

Studert Behavior
Teaching Strategies -

    
   
    

I Less than once a month
I Less than once a week

   

B 1-2 times per week

 

Topic of Interaction

I] 3 or more times per week

CUMCUlU“ ,1»

 

V

0% 20% 40% 60% 80% 100%

Proportion of Response

 

 

Similar to their reported interactions with their mentors, ECTs were interacting
with their colleagues more frequently around curriculum, teaching strategies, and student
behavior. This may indicate that these areas are the topics which were of greatest
concern for the novice teachers. I also ran Wilcoxon signed rank tests on the difference
between fall and spring interaction with colleagues, and none of the differences were
statistically signiﬁcant; this indicates that overall ECTs were maintaining their support
system of colleagues over the course of the year, while they tended to interact less

frequently with their mentors as the year progressed.

72

Comparing Chart 5.5 to Chart 5.4, ECTs reported more frequent interaction with
their colleagues than with their mentors across all categories, but especially around
curriculum, teaching strategies, and student behavior. To test this relationship, I took the
difference between the ECTs’ reported interactions with their colleagues and with their
mentors, and then ran a series of Wilcoxon signed rank tests. Table 5.2 indicates that the
differences in reported interaction with colleagues compared to mentors were relatively

large, and all were statistically signiﬁcant at the 0.01 level.

Table 5.2: Difference between ECT interactions with mentors and colleagues

 

 

 

 

Mentor vs. Colleague Interaction Mean 3::- Pr >= lS]
Curriculum 103*" 1.68 <0.0001
Teaching Strategies 1.17“” 1.46 <0.0001
Student Behavior 093*" 1.69 <0.0001
Assessment 0. 86*“ 1 .29 <0.0001
Standardized Testing 067*” 1.11 <0.0001
Psychological Support 076*" 1.42 <0.0001

 

Statistically signiﬁcant: "W 0.01; ** 0.05; * 0.10

Importance of support. Again, the differences in frequency of interaction may be
in part because of issues of access, and/or in part due to the value ECTs found in their
relationships with their mentors and colleagues. Additionally, since the ECTs were
allowed to self—select into different sub-groups of peers, they may have been more
aligned by personality or beliefs about teaching and learning. In both the fall and spring
surveys, the ECTs were asked, “How important is the professional support that you
receive from your mentor and school-based colleagues?” They were asked to indicate

either “not at all important”, “somewhat important”, “very important”, or “extremely

73

important”. Chart 5.6 shows a comparison of the ECTs’ spring responses for the
importance of the professional support from their mentors and colleagues (see Appendix

Table V for frequencies).

Figure 5.6: Importance of professional support from mentors and colleagues

 

Importance of professional support from mentor and

 

 

 

 

 

 

 

 

colleagues

I
0
E
8.
3 I Mentor
E IColleagues
§
0
n- 0 , ‘ , it ‘ .

Notat all Somewhat Very important Extremely

important important important

 

 

Chart 5.6 indicates that generally ECTs found the support from their colleagues
more important than the support that they received from their formally assigned mentor
(mean difference 0.15, standard deviation 0.86, P > [Si 0.07). Additionally, none of the
ECTs indicated that the support that they received from their colleagues was not
important at all, while slightly over 10 percent of ECTs indicated that the support they
received from their mentor was not important at all. This led me to consider what
organizational and policy factors may be associated with the frequency of interaction

with both mentors and colleagues, particularly mentors.

74

I calculated correlations between the ECTs’ frequency of interaction with their
mentors and measures of their perceptions of the organization. I also calculated
correlations between the ECTs’ frequency of interaction with their colleagues and
measures of their perceptions of the organization. I used the Speannan correlation
coefﬁcient because of the categorical nature of the frequency of interaction with mentors
and because of the nature of the other key variables of interest (Cody & Smith, 1997).
The measures of the perceptions of the organization include perceptions of professional
ﬁt, relational trust amongst teachers, labor-management relations in the school, degree of
collective responsibility of the teachers. I also calculated the correlations with the
reported level of importance of support they received from their mentors and colleagues.

For perceptions of professional ﬁt, the ECTs were asked to indicate their level of
agreement (1 = “strongly disagree”, 2 = “disagree”, 3 == “agree”, 4 = “strongly agree”)
with the following survey items: a) my approach to teaching ﬁts in throughout this
school; b) my professional interests are the same as those of other teachers throughout
this school; c) I identify with other teachers throughout this school; (1) my professional
goals are the same as those of other teachers throughout this school; C) I matter to other
teachers throughout this school; and 0 other teachers throughout this school matter to me
(Bryk & Schneider, 2002). I computed the correlation alpha between the items (or =
0.90), and then created a composite variable by taking the average response across all the
items (mean = 3.31; standard deviation = 0.51).

For perceptions of relational trust among teachers in their school, the ECTs were
asked to indicate their level of agreement with the following statements: a) it’s OK in this

school to discuss feelings; b) teachers in this school trust each other; e) teachers in this

~75

school respect other teachers who take the lead in school; and d) teachers in this school
respect those colleagues who are experts in their craft (Bryk & Schneider, 2002). I
computed the correlation alpha between the items (a = 0.89), and then created a
composite variable by taking the average response across all the items (mean = 3.12;
standard deviation = 0.64).

For perceptions of the degree of collective responsibility shared by the teachers in
the school, the ECTs were asked to indicate the proportion of teachers (1 = “none”; 2 =
“less than half”; 3 = “about half”; 4 = “most”; and 5 = “all”) who do the following: a)
help maintain discipline in the entire school, not just their classrooms; b) take
responsibility for helping one another do well; 0) take responsibility for improving the
overall quality of teaching in the school; d) feel responsible for helping students develop
self-control; e) set high expectations for academic work; and f) feel responsible for
ensuring that all students learn (Bryk & Schneider, 2002). I computed the correlation
alpha between the items (a = 0.91 ), and then created a composite variable by taking the
average response across all of the items (mean = 3.60; standard deviation = 0.73). For
perceptions of labor-management relations in their school, they were asked to rate
relations between teachers and administrators in their school: 1 = poor; 2 = fair; 3 = good;
and 4 = excellent (mean = 2.78; standard deviation 0.95). I was also interested in how
their indication of importance of support from their mentors and colleagues was

correlated with their frequencies of interaction.

Table 5 .3 shows the Spearman correlation coefficients for the organizational
items and the overall frequency of interactions with the ECTs’ mentors, and with the

average overall frequency of interaction with the colleagues whom they listed in the fall

76

c\.

.I t.

survey. The largest correlation coefficient was for the reported importance of support
provided by the mentor or colleagues, both statistically signiﬁcant at the 0.01 level. This
makes intuitive sense considering that if novices ﬁnd value in the relationship they would
spend more time investing in that relationship. This seemed to be particularly important
when looking at the frequency of interaction with the mentors. At the same time, it may
be that through more frequent interaction more valuable support is provided. Regardless,
to the extent that frequency of interaction and the value of support are correlated, district
and school administrators could work towards providing opportunities for more frequent,
substantive interaction between mentors and mentees through such things as common
planning time or extra release time. This though may often be difficult considering the

resource constraints that plague many districts and schools.

Table 5.3: Correlations between organizational items and frequency of interaction

 

 

 

 

Variable Mentor Colleagues
Professional ﬁt 0.05 024*“
(0.56) (0.002)
Relational trust 0.04 0.18M
(0.64) (0.02)
Collective responsibility 0.25 *** 019*"
(0.0005) (0.01)
Labor-management relations 0.13 020*“
(0.14) (0.008)
Importance of support 059*" 042*”
(<0.0001) (<0.0001)

 

statistically signiﬁcant: *** 0.01; ** 0.05; * 0.10

Overall though, with the exception of the reported importance of support
received, the correlation coefﬁcients are relatively small, though all in the expected

direction. All of the correlation coeﬂicients between the key variables and average

77

frequency of interaction with colleagues were statistically signiﬁcant, at least at the 0.05
level. So although there may not be large correlations, they are distinguishable from
zero. In addition, the ECTs’ perception of profession ﬁt with their colleagues has the
largest correlation coefficient (with the exception of importance of support) with regards
to interactions with colleagues. Again, this makes intuitive sense, if an ECT has shared
professional goals and beliefs with her colleagues she may tend to interact with them
more often.

Generally speaking, the correlation coefﬁcients were larger and overall more
likely to be statistically signiﬁcant with regards to interactions with colleagues, with the
exception of perceptions of collective responsibility. This may reﬂect the fact that when
ECTs perceived higher levels of collective responsibility shared by the teachers in the
school, it was an indication that the mentor actually was more willing to provide regular
support to their mentee. For example, the Whitaker teacher association president
indicated that often times the new teachers did not want to initiate contact with their
mentors because they did not want to appear to be incompetent, and therefore it was very
important for mentors to initiate interactions.

Regression analysis. To the extent that more frequent interactions with formal
mentors lead to opportunities for novice teacher learning and growth, I explored whether
mentor-mentee match or the degree to which ECTs value the support from their mentor
was associated with their frequency of interaction. I ﬁrst ran a regression with various
control variables (see Table 5.5 for description of variables used in the analysis) to
review the distribution of the error term of the outcome variable “frequency of interaction

with mentor”. Through nmning PROC UNIVARIATE in SAS, the statistical tests for

78

normality indicated that I can reject the null hypothesis that the error terms are not

normally distributed (i.e., Shapiro-Wilk statistic of 0.99, p-value pr < W 0.511).

Therefore, it is assumed the variance at level-1 is normally distributed, eijk ~ N(0, 02),

and it is also assumed that the level-2 and level-3 error variance is normally distributed,

TOjk ~ N(0, T“) and H001: ~ N(0, T B). Table 5.4 shows the distribution of the variance
components from running an unconditional model, Yijk = TOjk + 1’0jk + I100k + eijk where
Yijk is the frequency of interaction of individual ‘i’ and their mentor in school ‘j’ within

district ‘k’, and 703k is the average frequency of interaction.

Table 5.4: Variance components - unconditional model

 

Covariance parameter Unit Estimate
@tandard error)

 

 

Intercept School-level 040*
(0.3 l )

Intercept District-level 0.27
(0.23)

Residual Teacher-level 171*"

(0.3g

 

 

statistically signiﬁcant: *** 0.01; ** 0.05; * 0.10

Table 5.4 indicates that the majority of the variation lies between individuals
(approximately 72 percent), though roughly 16 percent of the variance is between
schools. To help explain some of the variance I ran the following conditional model,
particularly to explore the association between mentor-mentee match and the importance

of mentor support with frequency of interaction:

79

Yijk = 7000 + 7001 (CBA mentor provision)k + 7002 (district size)k + 7010 (percent white)jk
+ 7020 (percent free/reduced lunch)jk + 7100 (mid sch teacher)ijk + y200(white teacher)ijk +
7300 (lst yr teacherhjk + 7400 (femalehjk + 7500 (matchhjk + 7600 (SUPPOfOijk + YOjk + “0%

+ eijk

Table 5.5 describes the variables used in the analysis. I performed a test of variance

inﬂation, and only the variables for percent white students and percent students eligible

for free/reduced lunch had variance inﬂation factors over 2 (2.33 and 2.07 respectively).

Table 5.5: Variables included in analysis

 

 

 

Variable Name Description Mean

T eacher—Ievel

mid sch teacher A dummy variable (1 = ECT teaches middle 0.23
school; 0 = ECT teaches elementary school)

white teacher A dummy variable (1 = ECT is white; 0 = ECT is 0.91
non-white)

lst yr teacher A dummy variable (1 = ECT was in lst year of 0.38
teaching; 0 == ECT was not inlst year of teaching)

female A dummy variable (I = ECT was female; 0 = ECT 0.82
was male)

match A dummy variable (1 = ECT and mentor teach 0.51
same grade-level in elementary school or subject
matter in middle school; 0 = ECT and mentor DO
NOT teach same grade-level in elementary school
or subject matter in middle school)

support A dummy variable (1 = ECT rated importance of 0.67
mentor support as “very important” or “extremely
important”; 0 = ECT rated importance of mentor
support as “somewhat important” or “not
important at all”)

School-level

percent white Percent of students in school who were white 0.48

(sd = 0.29)
percent Percent of students in school who were eligible for 0.55
Lfrge/reduced lunch free or reduced priced lunch (sd = 0.22)

 

80

 

Table 5.5 (cont’d)

 

 

 

 

 

amiable Name Description

 

District-level

CBA mentor A dummy variable (1 = CBA contained language 0.25

provision pertaining to mentoring; 0 = CBA contained NO
language pertaining to mentoring)

district size Total pre-kindergarten through twelfth grade 15.71

student population in the district divided by 1,000 (sd = 6.01)

 

 

 

 

 

 

 

 

Table 5.6 shows the results from running the multi-level conditional linear
regression with the variables from Table 5.5. The coefﬁcient for CBA mentoring
language was not statistically signiﬁcant, therefore I cannot reject the null hypothesis that
having CBA language requiring mentoring is not associated with more frequent
interaction between mentor and mentee. It should be noted that there are only eleven
districts in the sample, and only four of those districts had CBAs which contained
language pertaining to mentoring; therefore, the lack of variation and the small sample of
districts make it difficult to model any district level effects. This is reinforced by
ﬁndings from the previous chapter which suggested that collective bargaining agreement
provisions concerning mentoring had no statistically signiﬁcant association with the
assignment of mentors, and the qualitative data also suggests that the quality of
mentoring often had more to do with person-level relationships rather than structural
forces.

The variable “mentor-mentee” match is also not statistically signiﬁcant, indicating
that I can not reject the null hypothesis that the association between match and ﬁ-equency
of interaction is zero. I had expected that the mentor-mentee match coefficient would be

Statistically signiﬁcant and positive, because based on the theory of action at the

81

individual-level one could expect that when properly matched, mentors would have
greater expertise in the content and teaching strategies in the ECTs’ given areas, and
therefore the ECTs may interact with their well aligned mentors more frequently. At the
same time, the measure of match for this sample is somewhat crude; no information on
matches by certiﬁcation area was available for this larger sample, mentors may teach
similar grade-levels or subject matters to the ECTs which also gave them expertise; and
match by personality may be more important. The qualitative data received from
interviewing the district HR directors and teacher association presidents does indicate
that the personalities of the mentors and mentees play a signiﬁcant role in determining

the quality of the relationship.

Table 5.6: Frequency of interaction conditional model — ﬁxed effects

 

 

 

 

. _ Coefficient
Fixed Effect (n —- 97) [t-statl
Intercept, 7000 3.28"
(2.67)
District-level
CBA mentoring provision, 7001 0.20
(0.35)
district size, 7002 0.09"
(2.01)
School-level
percent white, '0-47
7010 (0.49)
percent free/reduced lunch, 7020 '0-30
(0.28)
T eacher-Ievel
mid school teacher, '0-04
"00 (0.12)
white, 7200 4'22“".
(2.65) A

 

 

82

 

 

73‘

Table 5.6 (cont’d)

 

 

 

 

- ._ Coefﬁcient
F rxed Effect (11 -— 97) [t-statj
lst yr teacher, 7300 0.47“
(1.77)
female, 7400 -0.49
(1.42)
mentor-mentee match, 7500 0.26
(0.98)
support, 7600 159*"
' (5.86)

 

statistically signiﬁcant: *** 0.01; ** 0.05; * 0.10

The coefﬁcient for importance of mentor support is positive and statistically
signiﬁcant at the 0.01 level. This suggests that when an ECT values the support received
from their mentor they are likely to interact with them more frequently, ceteris paribus.
This makes intuitive sense; it is expected that an individual would seek out the assistance
of a mentor more frequently when they can provide them with support, resources, and
teaching strategies. At the same time, highly effective mentors may seek out their
mentees to provide them with resources and not wait for the ECT to come to them.
Although this represents individual action, it also has implications for policy, particularly
concerning the placement of ECTs within a building which can grant them more access to
their mentors, as well as providing additional release time for ECTs and mentors in order
to promote more frequent substantive interactions.

By examining the frequencies regarding the topics the ECTs and mentors
generally interact around, it would be expected that interactions around curriculum,

teaching strategies, and student behavior were viewed as being important. Table 5.7

83

.\m

uIA

‘-m

shows the Spearman correlations between reported importance of mentor support and

frequency of interaction around certain topics.

Table 5.7: Spearman correlations between importance of support and frequency of
interaction

 

 

 

Importance of mentor support Correlation coefﬁcient
Curriculum 0.48***
(<0.0001)
Teaching Strategies 056*"
(<0.0001)
Student Behavior 045*“:
(<0.0001)
Assessment 0.45":
(<0.0001)
Standardized Testing 03 Sue
(<0.0001)
Psychological Support 046*"
(<0.0001)

 

 

Statistically signiﬁcant: "* 0.01; '"' 0.05; "' 0.10

Although the correlations are moderate they are not that different from each other,
and more information would need to be gathered concerning exactly what type of support
is provided. For example, do novice teachers ﬁnd support more valuable when their
mentors provide them with curricular resources, help them develop assessments, practice
teaching strategies, or help them develop strategies for dealing with disruptive students?
The data available from the surveys did not allow for the exploration of this question, but
it does suggest that more research needs to be conducted.

This chapter addressed the research question, “To what extent are early career
teachers (ECTs) interacting with their mentors, as well as with their colleagues?” As I
had hypothesized, the ECTs interacted more often with their colleagues than with their

mentors, although they tended to interact with both their mentors and colleagues more

84

.7"-

r
'o
‘1

it
Ii -

L,.‘
I5

 

often around issues related to curriculum, teaching strategies, and student behavior as
opposed to standardized testing or psychological support. The data shows that although
nearly 60 percent of ECTs were interacting with their mentors at least once a week in the
spring, a signiﬁcant percentage were meeting less than once a month. In addition to
meeting less frequently with their mentors, generally the ECTs indicated that the support
that they received from their mentors was less important than the support that they
received from their colleagues.

Novices on average were interacting with their formal mentors less frequently,
and found the support less important compared to the support from colleagues, possibly
because they often did not have a choice in who was assigned to them, thus they may not
have shared professional goals and beliefs or have complimentary personalities. They
may have been interacting with their colleagues more often because they were more
accessible (such as within their department or in classrooms close to their own), and/or
because they had the knowledge and resources that the ECTs needed access to. This may
be particularly true if the formal mentors lacked knowledge and expertise in the grade-
level or subject matter in which the ECTs were teaching. Additionally, only 51 percent
were matched with their mentor by either grade level or subject matter, which may
indicate that several of the mentors lacked the expertise in the curriculum that the novice
teachers were using. If the mentor and mentee were not teaching the same grade level or
subject matter, the novice may just had preferred to consult with colleagues who were
teaching the same subject matter and grade level. Additionally, the ECTs may have been

interacting with their colleagues more frequently because they had a better professional

85

F6"

ﬁt or alignment with them regarding their beliefs and practices. Future research and
analysis can provide more information on this subject.

As the framework of new institutionalism suggests, there may be variations in the
quality of mentoring even though the formalized structures of mentoring may not
signiﬁcantly vary across the districts. Each district had instituted a mentoring program as
required by law, but as the data shows the frequency of interaction with mentors and the
value that novices attribute to their mentoring relationships greatly varied. The social and
organizational structure of individual schools potentially plays a big role in determining
the quality of mentoring which takes place, regardless of formalized policy at the state or
district level. Additionally, in line with social capital theory, it is then important to
consider to the extent to which to which individual action through mentoring
relationships and relationships with colleagues are associated with various novice teacher
outcomes. If novice teachers are receiving support from an array of colleagues, the
importance of formalized one-to-one mentoring may not be as important in producing
more effective, committed teachers. Yet overall, the relationships with both mentors and

colleagues may be related to larger social and organizational factors.

86

Chapter 6

Organizational Context and Commitment

This chapter aims to address the third guiding research questions: “How do the
organizational and social contexts within which novices work inﬂuence their
commitment to teaching, expressed through their professed future career plans?” In
particular, I am interested in exploring the association between their perceptions of the
labor-management relations within their school, professional ﬁt, and the support they
received from their mentors and colleagues. As the previous two chapters suggest, there
were gaps between formalized mentoring policy and the actual day-to—day practice of
mentoring within many of the schools in the samme. Beyond the formalized policy, I
contend that organizational factors inﬂuence the quality of the mentoring novices receive,
as well as the quality of support that they receive from their colleagues. This is where the
concept of social capital theory may continue to help address how novices’ professional
relationships (both formal and informal) may be associated with their commitment levels.

As expressed earlier, previous research has found that teachers’ perceptions of
the organization (e.g., ﬁt within the school, leadership, practice), partially determined by
their relationships with their mentor and other colleagues, are associated with their levels
of commitment (Bryk & Schneider, 2002; Desimone et al., 2002; Kardos et al., 2001).
Additionally, it is through relationships that novice teachers can access resources and
support, while at the same time it is through these relationships that novices’ are
socialized into the organization (Coleman, 1988; Kardos et al, 2001). Therefore, for the

third research question, I hypothesized: I expect that when novice teachers feel that their

87

professional goals are well aligned with the goals of their colleagues and mentors, they
will have greater access to resources and support which will help improve their
commitment. Additionally, I expect that novice teachers who perceive better labor-
management relations in their schools will have higher levels of commitment.

For this study, commitment is being measured as the novices’ indication of their
future career plans, such as their desire to continue to work in a particular school the
following year or for the next ﬁve years. Career choice is certainly not the only possible
measure of commitment, but research has found that commitment is related to teachers’
career choices (Weiss, 1999). Additionally, due to the study’s design, data collection in
2008-09 will allow me to track whether or not the novices were in a different teaching
assignment compared to 2007-08. If they transferred to a different building but remained
in the district, a section in the fall 2008 survey asked them the reasons for the move. If
they left the district they were sent a follow up survey asking them about the reasons they
left the district. Therefore there will be retention data and I will be able to relate it to
their previous commitment measures.

In this chapter, I ﬁrst provide ﬁndings from running a series of multi-level linear
regressions, examining the relationship between spring commitment measures and the
ECTs’ perceptions of the organization. I then provide ﬁndings from running multi-level
linear regressions, but with “change-in-commitment” from fall to spring as the outcome
variable. This allows me to further control for the effects of the ECTs’ prior career plans.
Finally, I provide ﬁndings from performing social network analysis, particularly
examining the influence mentors’ and colleagues’ beliefs and perceptions have on ECTs’

perceptions of professional ﬁt within their school.

88

NI

ml

‘1 I

w‘

5‘

w.

 

.3» .‘ -
r l-

Spring commitment measure. I hypothesized that novices who perceive better
relationships amongst teachers and between teachers and administrators will have higher
levels of commitment. To test this hypothesis I ran a series of hierarchical linear
regressions, with commitment measures from the spring 2008 survey as the outcome
variables, and perceptions of the organization as the key independent variables of interest.
The outcome variables include the novices’ indication of commitment to their grade level
assignment, school, and district.

For commitment to their grade level assignment, school, and district, novices
were asked to indicate their level of agreement (1 = “strongly disagree”, 2 = “disagree”, 3
== “agree”, 4 = “strongly agree”) as to whether they preferred to teach in the same grade-
level / school / district the next year, and whether they could see themselves teaching in
the same grade-level / school / district in ﬁve years (Smith & Ingersoll, 2004). On
average, the novice teachers indicated high levels of commitment across all of the
categories. Chart 6.1 shows the frequency of responses to the commitment questions for
their preferences for the next year, and Chart 6.2 shows the responses to whether they
could see themselves in the same assignment in ﬁve years. As the charts show, overall
the ECTs were quite committed as the distribution of responses is heavily skewed
towards “agree/strongly agree”. They were slightly more committed to remaining in their
assignment the following year as compared to ﬁve years down the road, but even those
responses were heavily skewed towards “agree/strongly agree”. The ECTs were also
slightly more committed to teaching in their current school and district as compared to

their grade-level assignment.

89

Figure 6.1: ECT commitment to teaching the following year

 

 

 

  
  

.4
o

70
8 60
i 50
40 IGrede
g 30 ,, lSchool
2o 5,. _CIDistrlct
i

 

- ~* ‘ I

strongly disagree agree strongly
disagree agree
Response

 

 

 

Figure 6.2: ECT commitment to teaching in five years

strongly disagree agree strongly
disagree agree
Ramona

 

 

I Grade
I School
El District

 

 

 

 

 

 

Because the responses to the commitment measures for the next year and within
ﬁve years were on average very close, I took the average response for each category to

make a single measure of commitment to grade level, school, and district. In addition, I

90

made a composite variable by taking the average response to each of the six items (a =
0.87). I checked the distribution of the error terms for each of the outcome variables, and
although the statistical tests for normality did not allow for the rejection of the null
hypothesis that they are not normally distributed, a review of the histograms indicated
that they approached a normal distribution Under the central limit theorem, with a larger
sample it would be expected that the distribution of the error terms would approach

normality. Therefore, for the series of three-level models the assumption is that the error

terms are normally distributed for all three levels: eijk ~ N(O, 02), TOjk ~ N(O, Tu), and

ﬂoor ~ N(0, TB)-
1 ﬁrst ran a series of unconditional models to examine the distribution of variance

between all three levels (see Table 6.1), Yijk = YOjk + rojk + “00k + eijk: where Yijk is the

measure commitment and YOjk is the intercept. The majority of variation in commitment

levels lies among individual teachers, though for each of the categories there is a
statistically signiﬁcant amount of variation among schools. For example, for
commitment to school, approximately 76 percent of the variance in commitment is
among teachers, and 14 percent is among schools.
I then ran a series of multi-level models to explore the relationship between
novice teachers’ perceptions of their organization and their levels of commitment. The
model controls for district size, student SES in each school, student race in each school,
level taught, and ECT years of experience. Key variables of interest included the level of

importance of support received from mentors and colleagues. I also included

91

independent variables that represent the novices’ perceptions of their school organization

as well as their work conditions.

Table 6.1: Variance components - unconditional model

 

 

Grade School District Overall

(estimate / (estimate / (estimate / (estimate /

std. error) std. error) std. error) std. error)
Teacher level variance 049*" 056*“ 044*" 027*"

(0.07) (0.08) (0.06) (0.04)
School level variance 0.07* 018*" 017*“ 012*"

(0.05) (0.07) (0.06) (0.04)
District level variance 0 O 0.05 0.01

(0) ﬂ) £905) (002)

 

 

Statistically signiﬁcant: *** 0.01; ** 0.05; * 0.10

For perceptions of labor-management relations in their school, they were asked to
rate relations between teachers and administrators in their school: 1 = poor; 2 = fair; 3 =
good; and 4 = excellent. I collapsed the variable into three dummy variables, (poor
perceptions of labor management relations; fair/ good perceptions of labor-management
relations; and excellent perception of labor-management relations); the fair/ good
perception variable was left out of the model as the reference category.

Other key independent variables of interest included the novices’ perceptions of
professional ﬁt within the organization, as well as their level of stress and burnout. The
measures of perceptions of professional ﬁt and stress/burnout were based on composite
variables created from taking the average of responses across a series of items in which
the novices were asked to indicate their level of agreement with each response: 1 =

“strongly disagree”, 2 = “disagree”, 3 = “agree”, 4 = “strongly agree”. Please see

Appendix Table VI and Table VII for the listing of all items used to create the composite

92

variables of perceptions of professional ﬁt and stress/bumout. Table 6.2 provides the

description of all of the variables used in the analysis. I also tested for multicollinearity

by checking for variance inﬂation. Only the variables for “percent of students who are

white”, “mentor”, and “mentor support” had variance inflation statistics that were just

over 2. For “mentor” and “mentor support” this makes sense since the “mentor” variable

serves as a ﬂag in the models in order to increase the sample size by including those in

the analysis who indicated that they did not have a mentor.

Table 6.2: Variables used in analysis

 

Variable Name

 

 

A dummy variable (1 = ECT indicated they currently
had a mentor; 0 = ECT indicated they currently DID

NOT have a mentor)

93

Description Mean
T eacher-Ievel
fall commitment to Avg commitment to grade level in the fall 3.51 (sd 0.70)
grade
spring commitment to Avg commitment to grade level in spring 3.40 (sd 0.75)
grade
fall commitment to Avg commitment to school in fall 3.31 (sd 0.88)
school
spring commitment to Avg commitment to school in spring 3.26 (sd 0.87)
school
fall commitment to Avg commitment to district in fall 3.46 (sd 0.80)
district
spring commitment to Avg commitment to district in spring 3.39 (sd 0.81)
district
fall overall Avg overall commitment in fall 3.48 (sd 0.66)
commitrnent
spring overall Avg overall commitment in spring 3.39 (sd 0.64)
commitment
mid sch teacher A dummy variable (1 = ECT teaches middle school; 0 0.28
= ECT teaches elementary school)
white teacher A dummy variable (1 = ECT is white; 0 = ECT is non- 0.90
white)
lst yr teacher A dummy variable (1 == ECT was in lst year of 0.27
teaching; 0 = ECT was not in lst year of teaching)
female A dummy variable (1 = ECT was female; 0 = ECT was 0.83
male)
mentor 0.66

 

Table 6.2 (cont’d)

 

 

 

 

 

 

Variable Name Description Mean
mentor support A dummy variable (1 = ECT rated importance of 0.43
mentor support as ‘Wery important” or “extremely
important”; 0 = ECT rated importance of mentor
support as “somewhat important” or “not important at
all”)
colleague support A dummy variable (1 = ECT rated importance of 0.89
colleague support as “very important” or “extremely
important”; 0 == ECT rated importance of colleague
support as “somewhat important” or “not important at
all”)
poor labor relations A dummy variable (1 = ECT indicated that there were 0.13
poor labor-management relations within their school; 0
= ECT indicated labor-management relations as
fair/good/excellent
excellent labor A dummy variable (1 = ECT indicated that there were 0.23
relations excellent labor-management relations within their
school; 0 = ECT indicated labor-management relations
as poor/fair/good
professional ﬁt Composite measure of ECTs’ perceptions of their 3.31 (sd 0.51)
professional ﬁt with their colleagues ((1 = 0.90)
stress/bumout Composite measure of ECTs’ levels of stress/bmnout 2.26 (sd 0.61)
(a = 0.93)
School-level
percent white Percent of students in school who were white 0.49 (sd =
0.28)
percent free/reduced Percent of students in school who were eligible for free 0.55 (sd =
lunch or reduced priced lunch 0.22)
District-level
district size Total prekindergarten through twelfth grade student 15.62 (sd =
population in the district divided by 1,000 5.69)

 

I ﬁrst ran the hierarchical linear models with all of the variables from Table 6.2

included (see Appendix Table VIII for the output), and then reduced the models to only

contain variables of interest that were at least statistically signiﬁcant at the 0.10 level, as

well as keeping district, school, and teacher characteristic variables as controls. Table 6.3

reports the results from running the regressions. In addition to reporting the un-

standardized estimates, the standardized estimates are also shown to make comparison

94

 

 

l
a‘t'
it.“

p)!
l.’
“m

«(an

H .:

Li: a

between binary variables and numeric variables more interpretable. The binary variables

were standardized to have a mean of 0, and the numeric variables were standardized by

subtracting the mean and then dividing by two standard deviations (Gelman, 2007).

Being a middle school teacher was also associated with lower levels of

commitment to grade-level, district, and overall. This may be related to the differences in
the organization of middle schools and elementary schools, where teachers at the middle
school level may teach a wider range of grade levels in a given school year, as well as
different levels of content (such as honors or remedial).

Key independent variables of interest are the ECTs’ reported level of importance
of support from their mentors and colleagues, their perceptions of labor-management
relations, and perceptions of professional ﬁt within their schools. The coefﬁcients for
importance of mentor support was only statistically signiﬁcant in the model for
commitment to school, yet the coefficient for “have a mentor” was also statistically
signiﬁcant but negative. This would suggest that when an ECT indicated that the support
from their mentor was important that they had higher levels of commitment to their
school, but they had lower levels of commitment when they indicated that they had a
mentor. Because these two variables have a relatively large correlation, the estimates
may be mis-speciﬁed. The coefﬁcients for reported level of importance of support from
colleagues were not statistically signiﬁcant in any of the models. Therefore I cannot

reject the null hypothesis that the association between commitment and importance of

colleague support is zero.

95

if!!!
I’ll!

 

 

 

 

 

 

 

 

 

l 8mm A 00W .2 02
. - . 000 @000 80$
8 0 :MMMWW 8.0- tea-.0- m _ .0- :Megw 000- ....Ngow Begum
. . 28 0.:
3 o .33 o 2 0 0.3.3.0 : .o 3.330 E Enameomem
A32
end- ..i... :5- 8338 «wee-eons doom
2 2 o ”$an teen—8 8:88 .«o 35:00::
. m 2 .o- 0.”..me 3:58 0 03:
3.0- 5.00 . :20 Ass 5.:
00.0- No 0- 80- 000- 0_ .0. 00.0 2 .0 2050"—
2 .0- A02 . c . 60.00 A0000 5.00
2 .0. 00 0- 2 .0- 0000- .00- 000 00.0 323
. A2 00 £0.00 €0.00 80.:
_0 0 ~00 ~00 3.0 000. 00.0- 000- 2 0.. 2002 a .2
. ads 30.: an: an:
2 0- :20. E .0- .000- 02 .0- 2 .0- m 2 .0- ..000- .282 .800... 2003
. $20 30.3 20: £0.00
S. 0 :50 00.0 2.300 2.0 58.0 2.0 500.0 305558 .3
22a assess
. A0000 80.00 A: .3 30.00
S 0 00.0 ~00- 00.0. 00 0 :0 000.0 000 082 0800880 eased
$0.00 8000 E00 @000
00.0 2 .0 2 .0 00.0 .00 2 .0 30.0 20.0 333 aeooam
00$ 800-0
@000 G2 00 $0.00 5.00
~00 800 20.0 «00.0 80- 0000- 000- 000.0- one area
20$ “0.00-,3
$0.2 $2.0 A0. .00 00000
.33.“ aaaomﬁ 3504 «aims..— 309.35
0:80am 003-02 cassam 02$: easem- eawé usaem sag-02 .
£03m \oﬁaemm Evaceﬁm 3008sz 0032855 \oaasmem Eﬁeagm \oaaaemm . . . . . macho”
.335 How-5&0 Bosow .954 035 a .mo 0 a... .8 o a: . .

 

 

acute BE -. 3252558 son.— uarem ”no use...

96

Another key variable of interest was the novice teachers’ perceptions of
professional ﬁt. This variable tries to capture an ECTs’ overall perception of their ﬁt
within the organization, in such areas as shared beliefs and goals with teachers within the
school. For the models for commitment to grade-level, school, and overall commitment
the coefﬁcients for perceptions of professional ﬁt were positive and statistically
signiﬁcant at the 0.01 level. Although the standardized estimates are not as large as those
for stress/burnout, or in the case of commitment to school as large as the estimate for
poor perceptions of labor-management relations, it suggests that when novice teachers
have higher perceptions of professional ﬁt within the organization they are more likely to
have higher levels of commitment, holding all other variables constant.

The coefﬁcient for poor perceptions of relations between teachers and
administrators in the school was negative and statistically signiﬁcant at the 0.01 level in
the model for commitment to school. For commitment to school, the standardized
estimate for poor perceptions of labor-management relations was the largest with the
exception of prior commitment. The labor-management relations climate may not only
directly affect an individual ECT, but also the whole school as an organization which
may impact the relations among teachers themselves.

The ﬁndings on poor perceptions of labor-management relations make intuitive
sense. Principals play a signiﬁcant role in shaping the organizational environment in
which novices work, and can also play a direct role in providing instructional support.
Principals are responsible for making school-level job assignments, evaluating teachers,
handling student and parent problems, as well as providing direct support to teachers and

indirect support through creating collaborative work environments (Youngs, 2007b;

97

Desimone & Smith, 2008; Chester & Beaudin, 1996). Therefore if relations between
administrators and teachers is poor it may not only effect the one-to—one relationship
between a teacher and an administrator, but it may also impact the overall professional
climate within a school which may effect teacher commitment.

Overall, these hierarchical linear models suggest that there are some key variables
that may be associated with novice teachers’ future career plans, and need to be explored
further. In particular, the variables for “poor perceptions of labor-management relations”,
“perceptions of professional ﬁt”, and “stress/burnout” were at least in some of the models
statistically signiﬁcant at the 0.01 level. Colleagues and administrators can be excellent
sources of guidance and support, and can impact organizational climate of the
school/district within which the teachers are working. Additionally, levels of
stress/burnout may not only be related to personal characteristics, such as personal skills
to perform the teaching tasks, but also the level of support that novices are receiving
within their organization. Also, the coefﬁcients for “professional ﬁt” in the commitment
suggest that when teachers feel that they are part of their professional community they are
more likely to want to remain in that assignment.

Table 6.4 shows the variance components from the restricted model. The
variance components at the teacher level were all statistically signiﬁcant at the 0.01 level.
Although these models explained a substantial amount of the variation at the teacher and
school level, there is more to be accounted for. For example, the conditional model for
commitment to school accounted for approximately 63 percent of the variation among
teachers. The model for commitment to district though only accounted for approximately

27 percent of the variation among teachers. There are likely several factors, including

98

 

personal factors such as family inﬂuences, which are not accounted for within these

models, even though a prior level of commitment is included.

Table 6.4: Variance components - conditional model

 

 

Grade School District Overall

(estimate / (estimate / (estimate / (estimate /

std. error) std. error) std. error) std. error)
Teacher level variance 0.27*** 0.21*** 0.32*** 0.12***

(0.05) (0.04) (0.05) (0.02)
School level variance 0.01 0.03 0.03 0.04”

(0.03) (0.03) (0.04) (0.02)
District level variance 0.004 0 0.009 0

(9.009) @103)

 

 

 

Statistically signiﬁcant: *** 0.01; ** 0.05; * 0.10

Change- in- commitment. The regression analysis used spring commitment
measures as the outcome variables and controlled for prior fall commitment, but it is also
important to consider modeling change-in-commitment over time. To further explore
organizational factors’ inﬂuence on novice teacher commitment, I created composite
measures of change-in-commitment, and used hierarchical linear models to see what
relationship the key variables of interest have with changes in commitment from fall to
spring.

For both fall and spring measures of commitment, I ﬁrst combined the measures
by category (grade-level, school, and district) and took the average of the responses to
commitment to the following year and to ﬁve years. I also calculated an overall measure
of commitment; ﬁrst I calculated the correlation alpha between the six items (a = 0.87),
and then created a composite variable from the average response across all six items.

Finally, I subtracted the fall composite measures from the spring composite measures to

99

form the change-incommitment measures. Table 6.5 shows the mean responses for fall,
spring, and change. It should be noted that the ECTs reported very little change-in-
comrnitrnent from the fall to the spring, and only the difference for commitment to grade-
level was statistically signiﬁcant based on a paired t-test. Yet, it should be stressed that
the mean change was negative, suggesting that the sample of teachers were less

committed in the spring compared to the fall.

Table 6.5: Change-in-commitment from fall to spring

 

 

 

Variable 11:21“ Std. dev. 13mg Std. dev. ﬁfe Std. dev
Grade-level 3.51 0.70 3.40 0.74 -011" 0.67
School 3.31 0.88 3.26 0.87 -005 0.58
District 3.45 0.80 3.39 0.81 -0.06 0.69
Overall 3.48 0.66 3.39 0.64 -007 0.49

 

 

Statistically signiﬁcant: "" 0.01; ** 0.05; "‘ 0.10

I checked the distribution of the error terms for each of the change-in-
commitrnent outcome variables, a review of the histograms of the distributions indicated
that they approached a normal distribution. Under the central limit theorem, with a larger
sample it would be expected that the distribution of the error terms would approach

normality. Therefore, for the series of three-level models the assumption is that the error

terms are normally distributed for all three levels: Cijk ~ N(0, 02), TOjk ~ N(0, Tu), and

000k ~ N(0, T5).

To explore how the variance in the novices’ change-in-commitment levels was

distributed among individual teachers, among schools, and among districts the following

unconditional three-level model was run for each of the commitment measures: Yijk =

100

 

YOjk + YOjk + #00k + eijk, where Yijk is the measure change-in-commitment, York is the
intercept, roik is the error term at the district level, nook is the error term at the school level,
and eijk is the error term at the individual level. As Table 6.6 shows, the distribution of
variance in change—in-commitment overwhelmingly was among individuals, as opposed
to among schools or districts. Part of the reason is that there was relatively little change
in commitment from fall to spring, and several person level characteristics and

circumstances inﬂuence an individual’s perceptions of their future career plans, such as

plans to raise a family or the career choices of a spouse.

Table 6.6: Variance in ECT change-in-commitment — Unconditional model

 

 

 

Grade Level School District Overall
T h l 1 . 043*" 027*“ 045*" O.19"'M
83" e' M “name (0.06) (0.06) (0.09) (0.05)
. 0.03 0.009 0.03
School level varlance 0 (0.05) (0.06) (0.05.)
. . . 0.008 0.02 0.04
Dlstnct level vanance L002) (002) 40-05) 0

 

After running the baseline unconditional models, I then ran conditional models
with all of the spring independent variables which were used in the previous hierarchical
linear regression models (see Appendix Table IX for the output). Using data from the
output, I reﬁt each of the models to only include independent variables which had
coefﬁcients which were statistically signiﬁcant at least at the 0.10 level (keeping controls
for district and school demographics, as well as teacher level characteristics), and then re-
ran the conditional models (see Table 6.7). In addition to the parameter estimates, the

table also shows standardized estimates. I standardized the variables and re-ran the

101

models in order to more easily make comparisons across the different variables,
particularly because many of the variables of interest are binary variables. For the binary
variables, they were standardized to a mean of zero. For the numeric variables, I
subtracted the mean and divided by two times the standard deviation in order to make the
coefﬁcients directly comparable to the binary variables (Gelman, 2007).

For change-in-commitrnent to grade-level assignment, ECTs who reported higher
levels of stress had more change in commitment from the fall to the spring. A standard
deviation increase in stress/burnout levels was associated with 0.22 standard unit negative
change-in—commitment, ceteris paribus. This makes intuitive sense as well, if an ECT is
stressed throughout the year, especially in the spring, it is likely to have an effect on her
commitment. For change-in-commitment to school and change-in-commitment to the
district, there were no statistically Signiﬁcant variables that remained in the ﬁnal
conditional models. This may be in part because there was little change in commitment
from the fall to spring, and the change was not statistically Signiﬁcant. Therefore it is
hard to predict what would cause that change, especially given the data that was
available.

The coefﬁcient for “importance of support from colleagues” was statistically
signiﬁcant and negative in the model for overall change-in-commitment. The data
indicates that when a novice teacher found the support from their colleagues important,
on average they had 0.41 standard unit change-in-commitment, holding all other
variables constant. This was not expected considering I would expect that if novice
teachers value the support they are receiving from their close colleagues they would be

more willing to invest time and energy in those relationships, thus making them more

102

committed One possible explanation is that this variable is not a good measure, since 89
percent of the teachers in the sample indicated that the support they received from their
colleagues was ‘Wery important” or “extremely important”.

To further explore the relationship between “importance of colleague support”
and change in commitment, I ran an analysis of covariance (ANCOVA) model with
spring overall commitment as the outcome variable. The key independent variables of
interest were the fall overall commitment measure, the spring “importance of colleague
support” measure, and an interaction variable (“fall overall commitment” x “importance
of colleague support”). The model controlled for the same district, school, and teacher-
level characteristics as the previous models.

The coefﬁcient for “importance of colleague support” was negative and
statistically signiﬁcant at the 0.01 level (standardized coefﬁcient of -1.58; S.e. 0.49), but
the coefﬁcient for the interaction term was positive and statistically signiﬁcant at the 0.01
level (standardized coefficient of 1.10; SC. 0.36). This indicates that the effect of
“importance of colleague support” does depend on the previous commitment level of the
ECTs. The negative association between level of importance of colleague support and
spring overall commitment was on average lower when ECTS had higher levels of overall
commitment in the fall. It is possible that those with higher levels of commitment in the
fall had established a more beneﬁcial support network early on and therefore had smaller

changes in commitment from fall to spring.

103

 

00 .0 . J8.0 : :00 2: ”2800000 £3300an

 

 

 

 

104

 

 

2: NE _0_ «.2
$0.00
NNdr 3.3.?th Egan
3000 am. 20
:3- .030. 3.0. and. oamuozoo .00 35:00::
m H .o. A?” "W .5808 a 005:
£0.00 $0.00 90.00 20.00
000.0 000 0 00.0 00.0 00.0 8.0 00.0 00.0 20500
£0.00 $0.00 £0.00 $0.00
2.0- t .0. 00 .0- z .0. 0000.0- 80.0- 00.0 00.0 323
5.00 90.00 80.00 A00. .0
00.0- 00.0. 80.0. 00.0. 00.0- 00.0- 2 .0- m: .0. 0082 a E
60.00 :0. 00 80.00 30.00
:0- 00 .0. 3 .0- 2 .0. 00.0- 00.0. 00.0- 00.0. .2080 .802. 0.0002
333 uﬁﬁkﬁw
5 .00 § .00 G0. 00 £0.00
~00 3 0 8.0 00.0 m. _ .0 00.0 00.0 m _ .0 025. 08002000 0080.0
3.00 30.00 A: .00 80.00
00.0 : 0 3 .0 00.0 8.0- 3.0. 00.0 E .0 223 008.000
0200 08.5.
As .00 @000 8000 G 0 . _0
8.0- 80.0. 8.0 80.0 0.0.0- 0.00.0. _ 0.0- 00.0- 800 6:05
003 0385
:0. 00 5 .00 5 .00 9.0.00
3. 0 00.0 00.0- £0 00855
820000 020?: 03:50.0 003030 seesaw 0020?: 20800.0 0029.:
000030030 \oaaﬁm 030030050 inseam 03030050 3380000 08003003 3020000
ggm BﬁEdnam hogan 030585—
.355 .055 .8000 .33
0006

 

$98.00 tour.— .. «acﬁumﬁﬁectamrowaagu "5.0 03 ah.

 

I also wanted to see to what extent these models accounted for the variation which
was established by the unconditional model. Table 6.8 shows the variance components
from the four multi—level conditional models shown in Table 6.5. From these I calculated
the percentage of variance explained by the models. For change-in-commitment to
grade-level and change-in-commitment to school, the conditional model only accounted
for approximately 7 percent of the variance among teachers.

The conditional model for change-in-commitment to school did not account for
any of the variation among individual teachers. For change in overall commitment, there
was actually more variation at the teacher level in the conditional model, though less
variation among schools. Overall, these models explained very little of the variance

among teachers, but because the mean change from fall to spring was so small this was

not surprising.

Table 6.8: Variance in ECT change-in-eommitment - Conditional model

 

 

Grade Level School District Overall
T h l l . 040*“ 025*" 046*" 022*"
63° 6’ eve “"3““ (0.05) (0.05) (0.08) (0.03)
. 008* 0.001
School level variance 0 (0. 0 6) (0.05) 0
. . . 0.006 0.02 0.02
DlStl’lCt level variance £0.05) (0.03) £10 4) 0

 

 

Labor-management relations. In this study I was particularly interested in the
role of labor-management relations played in inﬂuencing the organizational context of
schooling for teachers. Findings from the hierarchical linear models with spring

commitment measures indicate that ECTS’ perceptions of labor-management relations

105

within their school, particularly poor perceptions of labor-management relations, were
associated with their commitment levels. The Spring survey asked the ECTS to indicate
the extent of the role that they believed their teacher association should play in a
improving the work lives of teachers. Speciﬁcally, they were asked on a series of items,
“How much effort do you think your local union should put into each of the following
areas?” They were asked to specify: 1 = No effort, 2 = A little effort, 3 = Don’t know, 4
= Some effort, and 5 = A lot of effort. I calculated Spearman correlations between these
measures and the measure of their perception of labor-management relations within their

school, this time coded 1 = excellent, 2 = good, 3 = fair, and 4 = poor (see Table 6.9).

Table 6.9: Spearman correlations - Perceptions of labor-management relations and union
effort

 

 

 

 

Variable Correlation

. . 0.05
Gettlng better frlnge beneﬁts (055)

. . . 0.11
Improvrng jOb security (0.14)

. . . . 0.15"
Getting teachers more say ln how they do then Jobs (0 04)
Negotiating over standards for teacher evaluation and how the 014*
evaluations are used (0.06)
Helping to make teaching more interesting $030)

*
Negotiating how No Child Left Behind affects teachers 3930)
Giving members a say in how the union is run 250289)
Getting teachers a say in how the administration and school 026*"
board run the school system (0001)
Getting teachers more say in their teaching assignments or 014*
transfers (0.07)
. . . . 0 07
T ll '
e lng members what the unlon ls dorng (0.34)
. . 021*"
H
andllng member grievances {0.003)

 

Statistically Signiﬁcant: *** 0.01, ** 0.05; * 0.10

106

 

 

Overall, the coefﬁcients were relatively small, but were all positive indicating that
the lower the perceptions of labor-management relations within the school, the more
effort they wanted the teacher association to put into a given area. Correlations that were
statistically signiﬁcant primarily relate to the management of teaching. For example, the
largest correlation in the table is for “getting teachers a say in how the administration and.
school board run the school system”, and is statistically signiﬁcant at the 0.01 level. This
suggests that in schools with lower perceptions of labor-management relations, the lower
the satisfaction the novice teachers had with the management of the school and their
teaching, which ultimately may impact their commitment to their school. If teacher
associations are able to work to promote more teacher voice in the management of
individual schools or the district as a whole, it may work to increase teacher commitment.
It should also be noted that there was no statistically signiﬁcant difference in perceptions
of labor-management relations when comparing those who said they were a member of a
teacher association and those who indicated that they were not part of a teacher
association. 4

Social network analysis. In addition to exploring what possible variables may be
associated with outcomes of commitment, I was interested in exploring how novices’
perceptions of the school as an organization were possibly shaped by the perceptions of
their formally assigned mentors and key colleagues. Novices’ perceptions are not formed
within a vacuum, and may in part be inﬂuenced by those with whom they interact with in
the school. To explore this possibility, information about the perceptions of the ECTS’
mentors and key colleagues who they listed in the fall survey was gathered through

surveying the mentors and colleagues.

107

The following model represents the statistical technique used to measure the

inﬂuence of the mentors’lcolleagues’ perceptions and beliefs on the perceptions and

beliefs of the ECTs in the sample: Perceptionsrj = BOj + pj [Xii’j Yitj] + Perceptionsijt-1 +
ﬂag ij + eij , where Perceptionsij represents the perceptions of ECT “i” in school “j” in
the spring, 130,- represents the intercept, p,- represents the inﬂuence parameter and [Xii’j

Yi’j] represents the ECTS’ exposure to inﬂuence (mean exposure) based on the

relationship X between ECT “i” and mentor/colleagues “i’” in school “j” and the

perceptions of those mentors/colleagues (Yi’j-l) at the time 1, Perceptionsin represents
the perceptions of ECT “i” in school “j” in the fall, ﬂag ij is a ﬂag (0/1) indicating

whether or not there was data from a mentor or colleagues for a given ECT, and eij

represents the error term. I used this social network inﬂuence model to explore what
inﬂuence mentors and colleagues have on ECTS’ perceptions of their professional ﬁt
within the school. I ran a model for the mentors separately from the model for colleagues
(see Table 6.10). The error term is assumed to be normally distributed with mean 0 and
variance 02. A check for normality was completed, and the Shapiro-Wilk (Pr < W
0.3460) and a review of the distribution of the residuals allowed me to reject the null
hypothesis that the distribution is not normally distributed. The parameter estimates and
standardized estimates are both reported. For these models, the variables were
standardized to a mean of zero and standard deviation of 1.

The colleague inﬂuence coefﬁcient for “professional ﬁt” is statistically

signiﬁcant, even after controlling for prior levels. A standard deviation in total inﬂuence

108

for professional ﬁt was associated with 0.28 standard unit higher ECT perceptions of
professional ﬁt. Even when controlling for ECTs’ prior perceptions and beliefs, the
colleagues’ beliefs and perceptions appear to have at least a small inﬂuence on the ECT’S
perceptions and beliefs concerning their ﬁt. It could be that an ECTS’ perception of
professional ﬁt was developed through interacting with colleagues, and therefore they
were shaped through exposure to their colleagues beliefs and practices. Alternatively, an
ECTS’ perceptions of professional ﬁt may be a function of the difference between their

beliefs and practices with that of their colleagues.

Table 6.10: Mentor and colleague inﬂuence on ECT perceptions of professional ﬁt

 

 

 

 

Mentor Colleagues
Parfme‘e‘ Standardized Parfmete’ Standardized
Estrmate/ Estimate Estrmate/ Estimate
Jt-valuel jt-valuej

Intercept 1.25"“ 1.12***
(4.99) (5.06)

total inﬂuence -0.0l -0. 14 002*“ 0.28
(1.05) (2.68) ,

fall measure 066*“ 0.60 0.59““ 0.54
(9.65) (8.63)

ﬂag -0.07 -0.07 0.24“ 0.19
(0.51) 41.85)

11 180 180

R-square 0.35 0.37

 

 

Statistically signiﬁcant: *** 0.01; ** 0.05; "‘ 0.10

The coefﬁcient for the inﬂuence term for mentors is not statistically signiﬁcant,
so I can not reject the null hypothesis that the perceptions of mentors has no signiﬁcant
relationship with the perceptions of the ECTS. Even though the coefﬁcient is not
statistically signiﬁcant, I expected the coefﬁcient to be positive, which it is not. This may

reﬂect the fact that often times the ECTS had little say in which teacher was assigned to

109

them as their mentor, and therefore may have different beliefs and practices regarding
teaching. For example, previous data reported suggests that a substantial number of
ECTS are rarely interacting with their mentors, and approximately 11 percent ﬁnd the
support from their mentors not at all important.

This chapter set out to address the following research question, “How do the
organizational and social contexts within which novices work inﬂuence their
commitment to teaching expressed through their professed future career plans?” In
particular, I was interested in exploring the association between the support they received
from their mentors and colleagues, the perceptions of labor-management relations in their
school, and their perceptions of professional ﬁt. Following the ideas from social capital
theory, I hypothesized that support received from mentors and colleagues would be
associated with their levels of commitment, and that novices’ perceptions of their
organizational context would also be associated with their levels of commitment.

In line with previous research, this research found that ECTS’ perceptions of
professional ﬁt was positively associated with their spring measures of commitment to
grade-level, school, and the overall measure of commitment. When novice teachers share
similar beliefs and practices as their colleagues, they are more likely to believe that they
are a member of the professional community and are more likely to have access to
resources and support (Bryk & Schneider, 2002; Desimone et al., 2002; Kardos et al.,
2001). They may be more likely to invest their time into the profession and in particular
into their organization and professional relationships. At the same time, measures of
perceptions of professional ﬁt were not statistically signiﬁcant when the outcome

variable was “change-in-commitment”.

110

The ECTs’ perception of labor-management relations in their schools was
associated with their spring commitment to school and change-in—commitment to school.
In particular, when an ECT had poor perceptions of relations between teachers and the
school administration, the ECTs’ commitment was signiﬁcantly lower. This indicates the
importance that administrators play in shaping the school climate within which teachers
operate. Principals’ content knowledge has been found to be associated with teachers’
instructional practice (Desimone & Smith, 2008), and they also provide opportunities for
teachers to interact and receive support (Youngs, 2007b). Principals are also directly
responsible for the evaluation of teachers, teacher assignment, and often the
implementation of district and state policy within the school. Therefore school
administrators partially deﬁne the work conditions as well as the organizational context
within which teachers work.

When the work conditions are poor and the organizational climate is poor, it is
expected that novice teachers (and all teachers for that matter) would have lower levels of
commitment and would be more likely to try to transfer out of that particular school. In
interviews with several district HR directors and teacher association presidents, they
indicated that the quality of principal leadership can vary greatly within a given district,
especially in their ability to provide support to teachers. The Greenberg teacher
association president even indicated that there were a few buildings in his district which
had very poor quality principals, which he believed led to high rates of teacher transfer
from those buildings, as those who were able to leave, did so.

I also explored the possible inﬂuence mentors and colleagues had on the

perceptions and beliefs of the ECTs. The social network analysis indicated that

111

colleagues’, more so than mentors’, perceptions and beliefs potentially inﬂuence the

perceptions and beliefs of the ECTS with whom they interact. This may also reﬂect the
fact that the ECTS in this sample, on average, were interacting much more frequently
with their key colleagues compared to their formally assigned mentors, and also rated the
support that they received from their colleagues higher. It should be noted though that
inﬂuence can at times have negative effects, such as exacting undue pressure on novices
to conform (Kardos et al, 2001; Portes, 1998). Therefore it is important that novices have
the opportunities to interact with colleagues who are well aligned with them (e.g., share
professional goals and beliefs) and also have high levels of expertise and access to
resources.

Overall, this chapter conﬁrms the hypothesis that novice teachers’ professional

relationships with colleagues and administrators matter in regards to their commitment.
It is these relationships that often deﬁne the conditions within which they work, including
determining what resources and support the ECTS have access to. As the conceptual
framework suggests, novice teachers make sense of their organization not only through
responding to institutional rules and policies, but also through the informal mechanisms
of teacher socialization. In addition, there are often gaps between the institutional ﬂames
and the informal organizational make-up of an organization, which differentially shape a
novice teachers’ experience. Therefore the professional relationships novice teachers

have within the institutional frame can have important consequences for their beliefs and

practices.

112

 

Chapter 7

Implications for Policy and Practice

Teacher mentoring can be part of a comprehensive induction program that
provides novice teachers opportunities to improve their effectiveness in teaching and to
become socialized into the profession through a collaborative support system. Research
has shown that mentoring can provide novice teachers opportunities to learn and improve

their effectiveness, as well as reduce the chances that they migrate out of a particular

 

school or district, or leave the profession all together (Smith & Ingersoll, 2004; Grossman
& Thompson, 2004; Johnson & PNGT, 2004; Stein & D’Amico, 2002; Achinstein,
Ogawa, & Speiglman, 2004; Youngs, 2007a). Recent research by Glazermarl and
colleagues (2008) questions these claims, but more so than disproving previous claims
their research raises questions about the formation and implementation of mentoring and
induction policies.

Using ideas from new institutionalism and social capital theory I aimed to look at
how both formal and informal organizational structures impact the mentoring of new
teachers. First, I was interested in exploring the role that teacher associations play in
Shaping mentoring policy through the collective bargaining process. Teacher
associations are often thought to constrain the practices of districts (see Hoxby, 1996;
Ballou, 2000; Moe, 2005), though in some cases teacher associations work with districts
to support the professional development of their members (Bascia, 2003). Teacher
associations therefore may play an important part in shaping the formalized structures

within which teachers work and mentoring programs operate.

113

Secondly, I was concerned about the gap between formalized policy and actual
day-to—day practice. The framework suggests that institutionalized structures will look
similar across schools and districts as they respond to the same or similar environmental
cues (Meyer & Rowan, 1977; DiMaggio & Powell, 1983). Additionally, the
effectiveness of a mentoring program is often measured by its formalized structure which
meets the standards of state regulation or accepted practice, rather than the results that it
produces. Yet, the effectiveness of mentoring is largely based on the individuals within
an organization who are involved with the process, because it is largely social in nature.
Also, the role that informal professional networks play in the socialization of teachers is
often not considered in relation to formal one-on-one mentoring, and policymakers may
miss opportunities to identify how novice teachers are actually being socialized into the
profession.

Finally, I was concerned with what organizational elements were associated with
novice teacher commitment. Novice teachers seek cues from their colleagues as to
acceptable behavior and practice within the formalized structure of the organization, and
it is through these cues that novices shape their perceptions of the organization
Gaounsbury & Vantresca, 2003; Kardos et a1, 2001; Powell & Colyvas, 2007). The
novices’ perceptions of the organization, such as the labor-management relations climate
or their own professional ﬁt with their colleagues, have the potential to inﬂuence their
decision to remain in the organization. Additionally, through their professional
relationships, novice teachers access resources and support, which in turn may relate to

their ability to improve their instruction and their levels of commitment (Coleman, 1988).

114

In this chapter, I ﬁrst brieﬂy summarize the ﬁndings which were drawn from
addressing the three guiding research questions: a) To what extent do collective
bargaining agreements inﬂuence the provision of mentoring? b) To what extent are
novice teachers interacting with their mentors, as well as their colleagues? and c) How
are novices’ professional relationships associated with their levels of commitment? I
then use the ﬁndings to draw some implications for both policy and practice. Finally, I
offer some thoughts on the limitations of this study and some implications for future
research on new teacher mentoring and induction.

Summary of ﬁndings. In chapter 4, I addressed Research Question 1: “To what
extent do collective bargaining agreements inﬂuence the provision of mentoring?” In
particular, I was interested in exploring how collective bargaining may inﬂuence formal
district policy concerning new teacher mentoring, and also how actual day-to-day
practice may vary from what is formally speciﬁed. Following the framework, I proposed
that all of the districts in the samme, both in Indiana and Michigan, responded to their
state policy contexts regarding the requirements for mentoring new teachers. New
teachers in Michigan were required to be assigned a mentor by the district for their ﬁrst
three years in the profession, but beyond that the design and implementation of mentoring
policy was left to the local level. In Indiana, new teachers were required to be assigned a
mentor for their ﬁrst year in the profession as part of a two-year portfolio certiﬁcation
process. Additionally, mentors in Indiana were required to complete state approved
training.

To address this question I ﬁrst reviewed collective bargaining agreements (CBAS)

from all 11 districts in the sample, looking for language pertaining to new teacher

115

mentoring, induction, and professional development. Only four of the 11 CBAS actually
contained language pertaining to mentoring, and although two of the CBAS were more
speciﬁc in outlining the mentoring policy, they did not go beyond what was already
common district policy. For example, the Greenberg CBA indicated that principals
should match mentors and mentees by program area, but this was standard policy in the
other districts as well. At the same time, having language in the CBA may give teachers
cause for ﬁling a grievance if they feel that the mentoring practice within their building is
not following the speciﬁcations of the agreement. This though, was unlikely because in
the Greenberg CBA it was more of a suggestion rather than a requirement. Koppich
(2005) indicated that it would be a mistake to assume that it is the goal of all teacher
associations to work to improve teachers’ instructional quality. They may feel that
mentoring is an area for management to be concerned with, and they may put their focus
more on issues of job protection and money matters.

Although this was a small sample of districts, and only four of the CBAS
addressed mentoring, I argue that overall collective bargaining played a minimal role in
deﬁning the provision of mentoring. Even if a local teacher association wanted to work
towards inﬂuencing teacher mentoring they might face resistance from the administration
because it may be seen as trying to add restrictions into the management of schools and
human resources (Bredenson, 2001). Mentoring is a permissive topic of bargaining in
Michigan and Indiana, not a mandatory topic of bargaining. Therefore, overall district
policy in this sample was largely formed in the absence of inﬂuence from collective
bargaining. It is not clear though whether teacher associations guide mentoring policies

in other ways.

116

Through interviews with the district human resources (HR) directors and teacher
association presidents in the 11 districts, I was able to learn more about the formal district
mentoring policy, as well as potential gaps between policy and practice. The interviews
indicated that across most of the districts, policy suggested that when possible, mentors
and mentees should be matched by grade-level, subject matter, or certiﬁcation area.
Previous research has indicated that having a good match by grade-level/area is a key part
of encouraging substantive and frequent mentor-mentee interaction (Grossman &
Thompson, 2004; Achinstein, Ogawa, & Speiglman, 2004). Additionally, building level
administrators were given quite a bit of discretion in making mentor-mentee matches. An
exception was in a couple of Indiana districts where the district administration had some
hand in assigning mentors. In Engram for example, the human resources director had a
direct role in assigning mentors in collaboration with building principals, because she had
a centralized list of which teachers in her district had completed the required training to
become a mentor. In addition to assigning mentors, across the districts there was some
formal or Speciﬁed way of keeping track of mentor-mentee meetings, such as in Daus, the
mentees were required to keep a mentoring journal. Additionally, in Indiana the process
was a little more formalized around the two-year portfolio requirement.

Yet, although the HR directors and teacher association presidents laid out the
formalized policy, they also Spoke about gaps in the implementation of the policy, and
issues concerning the effectiveness of mentoring in their districts. For example, the Daus
teacher association president indicated that his district was following the letter of the law,
but in reality was not doing anyone a service. The Engram HR director reported that

although they try to make good matches between mentors and mentees, it was very

117

difﬁcult to ﬁnd enough teachers who had received training as well as to ensure that their
personalities would align which ultimately impacts the effectiveness of the relationship.

This highlights the fact that in each of these districts mentors were volunteers, so
the administration was limited by their available human resources. For example,
although it was district policy to match mentors and mentees by program area, only 51
percent of the early career teachers in this study had a mentor in their grade level or
subject matter. Further, the Greenberg HR director indicated that many of the mentors
and mentees were rarely meeting (once a month or less) which was ineffective for
providing support and resources to novice teachers.

The evidence in this chapter suggests that, although CBAS have the potential to
shape mentoring policy, in this sample of districts they did not appear to do so.
Additionally, as the framework suggests, formal mentoring policy did not vary that much
from district to district reﬂecting institutionalized frames, especially within a given state,
and also there were gaps between policy and practice (DiMaggio & Powell, 1983; Meyer
& Rowan, 1977). Several of the pe0ple interviewed indicated that the quality of
mentoring not only varied by individual mentoring pairs, but also by building based on
the quality of building leadership in assigning mentors and overseeing the mentoring
relationship. Previous research has indicated that the principal plays a key direct and
indirect role in providing support, and their conception of mentoring can impact the
quality of support novices receive (Youngs, 2007b; Desimone & Smith, 2008).
Additionally, as indicated by the Wagner HR director, often times both the district and
building level administration had no way of knowing how a mentoring relationship was

progressing, such as how often they met or what they were even talking about.

118

Therefore, there was great variation in the actual implementation of mentoring as well as
its quality and effectiveness.

In Chapter 5, I addressed Research Question 2: “To what extent are novice
teachers interacting with their mentors, as well as their colleagues?” To address this
question I utilized data from the early career teacher (ECT) fall and spring surveys
regarding the frequency and substance of the novice teachers’ interactions with their
mentors and colleagues. I was interested in comparing the interactions between ECTS
and their mentors and the interactions between ECTS and their other colleagues who they
named as people they interacted with professionally. The framework suggests that both
formal and informal mechanisms can shape the experience of an ECT, but possibly in
different ways. Additionally, it is through these professional relationships that the novice
teachers can access resources and support, but are also socialized into the profession,
learning the norms and expectations within the organization (Coleman, 1988; Portes,
1998)

The data indicated that ECTS in this sample were interacting with their mentors
and colleagues most frequently around curriculum, teaching strategies, and student
behavior, compared to such topic areas as assessments, standardized testing, or
psychological support. Overall though, the ECTS indicated that they were interacting
with their colleagues more frequently than with their formally assigned mentors. A
signiﬁcant percentage of ECTS were rarely interacting with their mentors at all; in the
spring approximately 16 percent indicated that they were meeting with their mentor less
than once a month and 3 percent indicated that they never interacted with their mentor.

Additionally, from the fall to the Spring, the ECTS tended to interact with their mentors

119

less frequently, while their interactions with their colleagues continued at the same
frequency, suggesting that the relationships with the colleagues were more beneﬁcial, if
not professionally at least socially.

The ECTS were asked about the level of importance of the support they received
from mentors and colleagues. On average, the ECTS found the support from their
colleagues more important than the support they received from their mentors.
Additionally, approximately 11 percent of the ECTS indicated that the support they
received from their mentor was not important at all. So not only were the ECTS typically
interacting with their colleagues more frequently, they valued the support they received
from them more than the support from their mentors.

To the extent that frequent interaction with mentors is important for the
socialization of teachers, I was interested in seeing what possible policy elements might
be associated with frequency of interaction. In particular, policy across the districts
stressed the importance of mentor-mentee match. With the data available, I was able to
determine if the ECT and their assigned mentor were at least matched by grade level
(elementary school) or subject matter (middle school). Iran a multi-level linear
regression with frequency of interaction as the outcome variable, and “match” as one of
the key independent variable of interest and ECt reported level of importance of mentor
support as the other (controlling for district and school demographics, as well as ECT
years of experience and level taught). The coefﬁcient for “match” was not Statistically
signiﬁcant, which was unexpected. At the same time, because the measure of match was
somewhat crude it could be a weak measure. For example, mentors who had previously

taught the grade level or teach a similar subject matter may have as much expertise as

120

someone who is matched exactly by grade-level or subject matter. Additionally, the
personalities of the mentors and mentees may be more associated with frequency of
interaction. Therefore, although alignment between mentor and mentee seems to be an
important element of mentoring policy, it may not always indicate that it will lead to
more frequent interaction.

When the ECTS reported that the support from their mentors was “very
important” or “extremely important”, there was a statistically signiﬁcant and positive
association with frequency of interaction. This compliments previous research which
suggests that when ECTS are receiving valuable support they are more likely to continue
to seek out those individuals for assistance (Youngs, 2007a; Grossman & Thompson,
2004; Achinstein, Ogawa, & Speiglman, 2004). Additionally, this may be an indication
that the mentors are actively seeking out the mentees to provide support and not waiting
for the novices to come to them.

In Chapter 6, I addressed Research Question 3: “How are novices’ professional
relationships associated with their levels of commitment?” I was particularly interested
in exploring how the novices’ perceptions of the support they received from their formal
mentors and colleagues, perceptions of professional ﬁt, and perceptions of labor-
management relations in their school were associated with their commitment levels,
expressed by their indication of their future career plans. Novices’ perceptions of their
organization are in part based on how they relate to the formal structures, but also based
on their relationships within the organization, both with their fellow teachers as well as

with their administrators. Additionally, it is through these relationships that they gain

121

access to resources and support, but also feel pressure and experience expectations, which
may have negative consequences on their commitment levels.

To explore this research question, I ﬁrst ran a series of hierarchical linear
regressions with commitment to teaching in their grade-level / school / district as the
outcome variables (constructed from spring responses). The key independent variables of
interest included a measure of the importance of support the mentors and colleagues
provided to the ECT, perceptions of labor-management relations in their school, and
perceptions of their professional ﬁt within the school. I also controlled for district and
school demographics, prior levels of commitment, as well as ECT years of experience
and level taught.

The regression results indicated that perceptions of professional ﬁt were
positively associated with overall commitment, commitment to grade-level, and
commitment to school. This makes intuitive sense, given that if one shares professional
goals and beliefs with colleagues they are more likely to invest themselves in the
organization. Previous research also provided evidence that when a teacher has similar
beliefs and practices as their colleagues, they are more likely to have access to resources
and support and have higher levels of commitment (Bryk & Schneider, 2002; Desimone
etal., 2002; Kardos et al., 2001).

When the ECTS had poor perceptions of relations between teachers and
administrators in their school, it was associated with lower commitment levels to their
school. Principals not only provide direct support to new teachers, but they also partially
deﬁne the professional climate within which they work (Youngs, 2007b; Desimone &

Smith, 2008; Chester & Beaudin, 1996). Principals are most often directly responsible

122

for teacher assignment within a building, teacher evaluation, the implementation and
oversight of policy within a building, and also responsible for handling student and parent
issues. This potentially makes them a very critical component in the socialization of
novice teachers.

In addition to spring commitment measures as outcomes, I ran hierarchical linear
regression models with “change-in-commitment” from fall to spring as the outcome. The
coefficients for professional ﬁt and poor perceptions of labor-management relations were
not statistically signiﬁcant in any of the four models. This may reﬂect the fact that there
was very little change-in-commitment from fall to spring, which makes it hard to model
that change. Additionally, there was no information regarding the novices’ personal lives
(e. g, changes in spouses career, illness, family plans, etc) that may be better predictors of
career change. Surprisingly, the more coefﬁcient for colleague support (valued as “very
important” or “extremely important”) was negative and statistically signiﬁcant for change
in overall commitment. This was not expected, but because there was very little change
from fall to spring, and approximately 90 percent of ECTS rated support from colleagues
as “very important” or “extremely important”, the estimate may be biased.

Finally, I conducted social network analysis using data from the ECT surveys and
surveys of ECTS’ mentors and colleagues. The aim was to see how the perceptions of the
mentors and colleagues pertaining to professional ﬁt were related to the perceptions and
beliefs of the ECTS. The data indicated that the colleagues’, more so that the mentors’,
perceptions of professional ﬁt were positively related to the levels of perceptions of
professional ﬁt of the ECTS. Previous research indicates that individuals make sense of

their organization not only through their own response to institutional rules and norms,

123

but also through their interactions with individuals within the organization (Lounsbury,
2001; Powell & Colyvas, 2007; Frank, Krause, & Penuel, in progress). This study further
provides evidence that how novices make sense of their organization is partially
determined through their interactions with their colleagues.

Implications for policy and practice. Based on the ﬁndings, I propose three
general implications for policy and practice. The ﬁrst is for policy makers to instill more
oversight of mentoring program implementation in order to monitor and evaluate the
effectiveness of mentoring programs. The interviews with several of the human
resources directors and teacher association presidents indicated that there were few
formal mechanisms to accurately monitor and evaluate the effectiveness of their
programs, and therefore the quality of mentoring could vary greatly within a district and
even within a building. Many of the districts required the new teachers and their mentors
to keep a log, but as the Wagner human resources director stated, the district was still
unable to really know whether or not they were actually meeting, and knew even less
about what they were talking about. The data from the surveys also indicates that many
of the ECTS met with their mentors rarely or never.

Part of the reason why some mentors and mentees may not have been interacting
regularly was because they were not teaching the same grade-level or subject matter (as
indicated by the data). Although it was often district policy to match mentors and
mentees by grade-level, subject matter, or certiﬁcation (as well as by personality), the
assignment was left to principals, and as indicated by the interviews some principals did a
better job of matching mentors and mentees than others. I am not suggesting that the

central administration take on the role of assigning mentors directly (although in some of

124

the Indiana districts they did have a larger role), but to hold principals accountable for the
quality of the mentoring that takes place in their building. If there is more district
oversight over principals in the administration of mentoring policy, and more oversight
by principals in ensuring the mentoring relationship is beneﬁcial, then there will likely be
a reduction in the variation in quality (raising the bar), as well as a reduction in the gap
between the formalized policy and actual practice. This may be an area where teacher
associations can play a larger role; if they are dedicated to promoting teacher quality and
support, they can negotiate for more oversight of mentoring programs to help ensure
some quality control. Even outside of collective bargaining, teacher associations can help
promote greater support for novice teachers through promoting greater collaboration
among their members.

This of course takes time and resources, which are often in short supply within a
school and district. Additionally, regardless of policy and oversight, the crux of a
mentoring program is the relationship that develops between the mentor and mentee.
Administrators can put elements of the program in place which can facilitate more
productive interaction (e.g., proper matches, release time), but ultimately the value that
early career teachers receive from the mentor comes down to the relationship that they
build. As shown, often times it is difﬁcult to make matches, and even if mentors and
mentees have release time it does not mean that they will necessarily use the time
productively.

Therefore, this leads to a second general policy implication; instead of focusing
on a one-to-one mentoring program, instill a network of collegial support based in part on

the direction of the novice teacher. The novices in this sample, on average, interacted

125

more frequently with key colleagues and found the support they provided more important
that the support offered by their mentors. Several schools have developed instructional
teams based on grade-level assignment or subject matter which can work in similar ways
to a one-to-one mentoring relationship in providing resources and support to new
teachers. By having a network of support, which is still in some ways formally assigned,
novice teachers may have increased access to resources, but will also have more choice to
gravitate to someone within the network to whom they can relate professionally and
socially. Additionally, different colleagues are more likely more equipped to provide
different valuable resources.

The third area with policy and practice implications concerns the labor-
management relations climate within the schools. When an ECT had poor perceptions of
relations between the teachers and administrators in a school, they were more likely to
have lower levels of commitment. As stated, several of the human resources directors
and teacher association presidents indicated that there were several schools where labor-
management relations were poor, and where the quality of the administrative leadership
was poor. Principals play an important role in providing assistance to new teachers as
well as in establishing a climate of collaboration amongst all the teachers. In these
circumstances, districts need to improve their ability to identify sub-par principals in
order to provide them assistance or remove them, and to identify schools where the
relationship between the administration and the teachers is suffering in order to
implement some type of intervention. Teacher associations in these circumstances can

also play an important role in working to improve collaboration between teachers and

126

administrators in these schools, with the ultimate goal of ensuring that teachers have the
resources and support they need to properly perform their jobs.

Limitations of study and future research. There are several limitations of this
study which can be addressed by future research. First, while this study was able to
explore the frequency and some of the substance of interaction between ECTS and their
mentors and colleagues, this study was not able to gauge exactly what type of support the
mentors and colleagues were providing. For this sample of teachers, the novice teachers
were most often interacting with their formal mentors and their colleagues around
curriculum, teaching strategies, and student behavior. This may be an indication that
these are the areas with which novice teachers need the most assistance, but I was not
able to gauge what type of support the mentors and colleagues were providing in these
areas. For example, were they sharing tricks-of-the-trade in dealing with common
student problem behaviors, or were they discussing a particular problem student they
both taught to discuss and evaluate interventions? Were they discussing general teaching
strategies such as how to pr0perly use group work, or were they sitting down to co-plan a
unit or lesson? By better understanding the type of support that is provided and actually
valued by the novices, program designers can provide training to mentors or teams as to
what type of support to provide.

This study also did not collect detailed information as to the context within which
the interactions with mentors and colleagues were taking place. The survey did ask them
to indicate if they met with their mentor before school (73 percent said yes), after school
(58 percent said yes), during lunch (41 percent said yes), and during their planning period

(43 percent said yes), but the survey did not collect information about the context of

127

interactions with colleagues. Also the surveys did not ask about the duration of their
interactions. Were the novices interacting with their mentors and colleagues at regularly
scheduled times, or were they available for quick pop-in questions? Were little bits of
continuous help valuable, or was it more of programmed professional development
through interactions that had a larger impact? This has implications for the design of
support systems of novice teachers; for example, it may be more beneﬁcial for a novice
teacher to be located within a building surrounded by those who have knowledge of her
curriculum and students, who she can quickly turn to between classes or at lunch, as
opposed to having regularly scheduled meetings between a mentor and a mentee.

This study also did not ask the ECTS about their interactions with building-level
administrators, nor was data collected from building-level administrators. Previous
research and this study suggest that principals can play a signiﬁcant role in the
socialization of new teachers through direct interactions with novice teachers as well as
through implementing mentoring programs and creating a collaborative work
environment. The interviews with human resources directors and teacher association
presidents indicated that the effectiveness in mentoring varied by building in part because
of the variation in building leadership. It would be beneﬁcial to know what processes
principals use to recruit teachers to become mentors, what processes they use to assign
mentors, and how they monitor the effectiveness of the mentoring program.

Additionally, how in general do building-level administrators work to promote a
collaborative work environment and maintain beneﬁcial relationships with their staff? If
principals are assigning mentors to mentees just to meet the requirements of the district

and the state, they are not as likely to be systematic in how they go about it, and this may

128

lead to relationships that are not as productive as they could be. Additionally, if their
involvement in the mentoring program ends after they make the mentor assignment they
have no true way of measuring the effectiveness of the program. Principals not only can
directly inﬂuence the support new teachers have, but by creating a collaborative work
environment they can ensure a wider network of support for new teachers, and for all
teachers. This may also work to improve the relationship between the administration and
the teachers, which may further work to improve the commitment of teachers.

There were also some technical limitations with this study, particularly
concerning the generalizeability of the ﬁndings. The ECT response rate was relatively
low, with a 63 percent fall survey response rate and a 76 percent spring retention rate.
Part of the concern is that novice teachers, eSpecially ﬁrst-year teachers, can be very
overwhelmed by their work and participating in a study will not be their top priority.
Additionally, because the study is longitudinal it is a challenge to retain teachers in the
study. The teachers who initially completed the fall survey and those who remain in the
study in the spring may have been flmdamentally different from the eligible sample in the
eleven districts. If those ECTS who were more overwhelmed by their busy schedules
were more likely not to participate, then this study misses a key group of teachers that we
should be concerned about, and the ﬁndings may be generalizeable even the new teachers
in the eleven districts in the sample.

Additionally, the novice teachers were asked to name their mentor and key
colleagues in the fall survey. It is more than likely that several of those who did not
complete the fall survey did so at least in part because they were not comfortable naming

names and involving their colleagues with the study. Even though the ECTS were

129

 

informed that the mentors/colleagues would not be informed as to who listed them,
knowing that their colleagues would be contacted may have prompted some ECTS to not
participate. Again, if that sample of teachers were fundamentally different from the
sample that completed the surveys then it challenges the ability to generalize the ﬁndings.

Concerning the collection of social network data, this study collected egocentric
social network data as opposed to socio—centric data. The study relied on the ECTS to
name individuals with whom they interacted with, which may not cover the extent of the
network of support that novice teachers are exposed to. Collecting socio-centric data
involves surveying all teachers/professionals within a school, and allows researchers to
get better school-wide measures. The study collected egocentric data due to time, human,
and monetary resource constraints. Collecting socio-centric data, at least in a sample of
schools, may allow more detailed analysis of the extent of novice teachers’ access to
resources and support, as well as provide opportunities to better understand the culture
and norms of the whole school organization.

Although there are limitations with this study, the ﬁndings from this research do
add to previous research on new teacher induction by further exploring the role that
collective bargaining plays in the provision of mentoring, exploring the socialization of
new teachers through one-to-one mentoring and collaboration with colleagues, and by
exploring how novices’ perceptions of their organizational context is associated with
their commitment levels. In this sample, there was a very limited role that collective
bargaining played in deﬁning the scope of mentoring, but there could be a greater role for
teacher associations to help ensure that mentoring policy is reformed and to provide

greater oversight in the implementation of mentoring. Because novice teachers on

130

average valued the support from their colleagues as opposed to their mentors, districts
and schools should consider developing networks of support as opposed to or in addition
to one-to-one mentoring programs. Finally, districts need to evaluate the role principals
play in providing support to novice teachers and their role in developing a collaborative
work environment, because they potentially play a pivotal role in the socialization of new

teachers into the profession.

131

APPENDIX

Table I: ECT response rates by district (fall & spring completers)*

 

 

 

 

# of # of Middle
maxim 1316:]ng BREE?” Middle Middle School Overall
District ECTS ECTS in Response Schools School ECT Resp.
eligible study Rate ECTS ECTS 1n respons Rate
elmgble studL e rate
Daus (MI) 26 7 26.9 20 4 20.0 23.9
Greenberg (MI) 20 8 40.0 5 2 40.0 40.0
Kaline (MI) 14 9 64.3 9 5 55.6 60.9
Underwood (MI) 23 8 34.8 7 5 71.4 43.3
Wagner (MI) 8 7 87.5 0 0 NA 87.5
Whitaker (MI) 4 2 50.0 8 6 75.0 66.7
Englam (IN) 53 30 56.6 19 12 63.2 58.3
Luckman (TN) 14 8 57.1 5 3 60.0 57.9
Payton (IN) 27 18 66.7 5 2 40.0 62.5
Sayers (IN) 29 20 69.0 13 5 38.5 59.5
Wilson (IN) 31 15 48.4 28 8 28.6 39.0
Total 249 132 0.53 119 52 0.44 0.50

 

* 63 percent response rate in the fall, and a 76 percent retention rate in the spring

Table II: ECT interaction with mentor - Percent ECT response

 

 

 

‘ Frequency Fall Spring
Never 7.5 3.3
Less than once a month 8.5 15.8
1-3 times per month 17.9 21.7
l-2 times per week 24.5 17.5
3-4 times per week 13.2 15.8
Everyday 28.3 25.8 [‘5

 

 

132

 

Table III: ECT interaction with mentor by years experience - Percent ECT response

 

 

 

 

Frequency lst yr 2nd yr 3rd yr
Never 0 5.6 5.3
Less than once a month 12.8 16.7 21.1
1-3 times per month 23.4 24.1 10.5
1-2 times per week 14.9 20.4 15.8
3—4 times per week 21.3 13 10.5
Every day 27.7 20.4 36.8

 

Table IV: ECT interaction with mentor by level — Percent ECT response

 

 

Frequency Elem Mid Sch
Never 2.2 7.1
Less than once a month 15.2 17.9
1-3 times per month 20.7 25
1-2 times per week 15.2 25
3-4 times per week 16.3 14-3
Every day 30.4 10.7

 

 

 

133

 

Table V: Importance of mentor and colleague support (frequency of response)

 

 

 

 

 

 

Response Mentor Support Colleaﬁgg: Support
Not at all important 12 0
Somewhat important 25 16
Very important 28 64
Extremely important 48 86

 

 

Table VI: Perceptions of professional fit items

Indicate your level of agreement or disagreement with each of the following statements.

 

 

 

 

Strongly Strongly Not
Dar/ten one circle on each line disarm Disagree Agree ﬂ sure"
My approach to teaching ﬁts in throughout
this school 0 O O O 0
My professional interests are the same as
those of other teachers throughout this school 0 O O O O
I identify with other teachers throughout this
school 0 O O O 0
My professional goals are the same as those of
other teachers throughout this school 0 O O O 0
I matter to other teachers throughout this 0 O O O O
Other teachers throughout this school matter _ _ _ _ _
to me ‘ L) U U U U

 

" Not Sure was recoded to “missing”

Table VII: Perceptions of stress/burnout items

Indicate your level of agreement or disagreement with each of the following statements.

Strongly Strongly N03
disagree Disagree Agree agree sure

 

Darker: one circle on emh line

 

I feel emotionally drained ﬁ'om my work 0 O O 8 (CD)

I feel used up at the end of the workday O O O

I feel fatigued when I have to get up in the

morning and face another day on the job 0 O 0 g g

I feel blurred out ﬁom my work 0 O 8 O Q

I feel ﬁ'ustrated by my work 0 O O o O

I feel I’m working too hard on my job 0 O O O O

I feel fed up with my work 0 0

I’ve become more callous toward people since I 0

took this job 0 O O O o
o O 0 O

 

.. I wogy that this job is hardening me emotionally

‘ Not Sure was recoded to “missing”

134

 

000 .. m00.0 a... :00 4...... 0080000 00800030

 

 

 

 

02 02 03 02 a

mmd 0:00th —m.N sen-Nd.- mmN 0:;th 3mm canon-.61 unogemmobm
0.40 1.3.0 5.0 8.0 000-. 3.00.0 00d 00000.0 00 302000.020
00.0 3 .0 0N4 2.0 00.0 E .0 $0 8.0 00000002 0008-0000. Beam
0: S .0- 2.. 060- ohm 0:00.0- m0._ 2 .0 00200—2 0050-300— .500
00.0 3.0. 00._ 3.0- 02 8.0 5.0 000.0. 0.5003 .9800 .0..an—
30 00 .0 00.0 2 .0 2L 0080 00 .0 00.0.. ton—0:0 0358 .«o 0.50::
00.0 00.0 5.0 E .0 -.~ :50- 0: 00.0 .3002
3.0 00.0- 3.0 v0.0- 00.0 00.0- 00.0 00 .0 2080.0
00.0 3.0. 00.0 00 .0. 00 .0 8.0 3.0 00.0 823
000 3.0- 2 .0 ~00- 00.0 800.0- mm; 00 .0- 8:003 .S 02
$2 0. .0- 9..— 0N.0- 1: 2.0.. 00.0 .30- 5508 60:00 0:022
3.0 anemmd 26 9.....de and aeahnd Nwd4 aaaamd Eon—“goo =0&
«0.0 ~00 $0 000- 00.— 3.0 00 .0 8.0 :25— 0330500 E3000
00.0 3 .0 5.0 320 00,— $0 0N0 00.0- 823 003.60
3.0 000.0 3.0 000.0 m: 8.0- 3.0 800- one 8305
EN :00; 3.0 second V00 000.0 3am 0:00; 0009—2:—

_o=_0>-0_ 036E Wag-0— ouaemﬁm 7029-0— Baﬁsmm Fag-g 308000.”—

888200 868.20.— 888800 0308800

.385 8505 Begum 0264

0005

 

 

.000:— EEEGEB I Haﬁz—=58 hum mutam ”:5 030,—.

135

00 .0 .. M3.0 r. 00.0 4...... 05000000 0.800.080

 

 

 

 

 

MS 02 34— 02 0

~00 00.0- N0.0 00.0- 00.0 00.0- m0.~ 1.3.0- Beggar-chm
3.0 m0 .0 0N0 00.0 00.0 00.0 7: 2.0 2.0 000208.080
040.0 00 .0 34.0 and 00.0 :0 0m .0 00.0 8000002 ﬁes-500— .oxm
0m .0 00.0- on .0 00.0- %: «~0- 30 00 .0 00000.2 0008-0000— .500
0N.N 2.00.0- 8." 2.3.0- 30 00.0.. 0_ .0 mad- :00020 28.00 0000::
00.0 00.0 00.0 00 .0 02 3.0 00.0 0000- 00.0003 80:08 .00 €095
00 .0 00.0 00.0 00.0 00.— 1.3.0- E4 E0 .8052
00.0 000.0 00.0 3.0 and 3.0 00.0 00.0 0.080...“
S .0 3.0- 8.0 0~.0- N00 0 0 .0 mm .0 00.0- 833
0: S .0- 00.0 00.0- 000 2 .0- 00; 03.0. 8:002 .5 02
00.0 :0- 004 00 .0- 0nd 8.0- 00.0 00 .0- 8:080 .0208 0:022
3.0 2 .0 2.10 2.0 m: 30.0 $0 mm .0 00::— 03300500 000200
3.0 00.0 00.0 00.0 8.0 00.0 00 .0 :0 323 00800.0
0m .0 30.0 00.0 000.0 9.0 000.0- 000 000.0- 030 8505
«N0 3 .0 3.0 00.0 00.0 0N0. wmd 0nd- E3025

Wag-0— 30800m $3.834 3080a Egg-07 8083M egg-00 8085mm

0808800 8860.30 0088800 88885

5:95 «om-005 Begum 00.54

0.005

 

 

.25.: 3:03—0:00 l pecans—Bec-E-owaaac H.00— "E 030,—.

136

REFERENCES

Achinstein, B., Ogawa, R., & Speiglman, A. (2004). Are we creating separate and
unequal tracks of teachers? The impact of state policy, local conditions, and
teacher characteristics on new teacher socialization. American Educational
Research Journal, 41(3), 557-603.

Babcock, L., & Enberg, J. (1999). Bargaining unit composition and returns to education
and tenure. Industrial and Labor Relations Review (52).

Ballou, D. (2000). Teacher contracts in Massachusetts. White Paper No. 12. Boston:
Pioneer Institute for Policy Research.

Bascia, N. (1997). Teacher unions and teacher professionalism in the US: Reconsidering
a Familiar Dichotomy. In El . Biddle (Ed). Fundamental Change: International

Handbook of Teachers and Teaching. (pp. 225-245). Netherlands: Kluwer
Academic Publishers.

Bascia, N. (2003). Triage or tapestry? Teacher unions’ work toward improving teacher

quality in an era of systemic reform. Report for Center for the Study of Teaching
and Policy.

Boyd, D., Lankford, H., Loeb, S., & Wycoff, J. (2002). Initial matches, transfers, and
quits: The role of teachers’ career decisions in the disparities in the average
teacher qualiﬁcations across schools. Working Paper.

Bourdieu, P. (1985). The forms of capital. in JG Richardson (Ed), International

Handbook of Theory & Research for Sociology of Education (pp. 241-258). New
York: Greenwood.

Bredeson, P.V. (2001). Negotiated learning: Union contracts and teacher professional
development. Education Policy Analysis Archives, 9: retrieved from
http://epaa.asu.edu/epaa/v9n26.htrnl.

Bryk, A., & Schneider, B. (2002). Trust in schools: A core resource for improvement.
New York: Russell Sage Foundation.

Burt, R. (2004). Structural holes and good ideas. American Journal of Sociology, 110(2),
241-270.

Chester, M.D., & Beaudin, B.Q. (1996). Efﬁcacy beliefs of newly hired teachers in urban
schools. American Educational Research Journal, 33(1), 233-257.

Cody, R., & Smith, J. (1997). Applied statistics and the SAS programming language.
Upper Saddle River, New Jersey: Prentice Hall.

137

Coleman, J .S. (1988). Social capital in the creation of human capital. American Journal
of Sociology, 94, (Supplement 1988): S95-S120.

Desimone, L.M., & Smith, TM. (2008). Improving middle-school math instruction and
student achievement. Paper presented at the annual meeting of the American
Educational Research Association, New York.

Desimone, L., Porter, A.C., Garet, M., Suk Yoon, K., & Birman, B. (2002).
Effects of professional development on teachers’ instruction: Results from a
three-year study. Educational Evaluation and Policy Analysis, 24(2), 81-1 12.

Dika, S., & Singh, K. (2002). Applications of social capital in educational literature: A
critical synthesis. Review of Educational Research, 72(1), 31-60.

Dillman, DA. (2007). Mail and internet surveys: The tailored design method. Hoboken,
NJ: John Wiley & Sons, Inc.

DiMaggio, P., & Powell, W. (1983). The iron cage revisited: Institutional isomorphism

and collective rationality in organizational ﬁelds. American Sociological Review,
48, 147-160.

Eberts, Randall, & Stone, Joe. 1984. Unions and public schools. Lexington, KY:
Lexington Books.

Ebmeier, H. (2003). How supervision inﬂuences teacher efﬁcacy and commitment: An

investigation of a path model. Journal of Curriculum and Supervision, 18(2),
1 10-141.

Ford, T., & Youngs, P. (under review). How policy context shapes trust in schools:

Understanding how " Success for All" promotes trust in a high-performing urban
Midwestern district.

Frank, K.A. (1998). The social context of schooling; Quantitative methods. Review of
Research in Education, 23, 171-216.

Frank, K.A., Krause, A, & Penuel, W. (in progress). Knowledge ﬂow and organizational
change.

Frank, K.A., Zhao, Y., & Borman, K. (2004). Social capital and the diffusion of
innovations within organizations: The case of computer technology in schools.
Sociology of Education, 77(2), 148-171.

Fuller, B., & Izu, J .A. (1986). Explaining school cohesion: What shapes the
organizational beliefs of teachers? American Journal of Education (94): 501 -535.

138

Fuller, H., Mitchell, G., & Hartman, M (2000). Collective bargaining in Milwaukee
Public Schools. in Tom Loveless (Ed) Conﬂicting Missions? Teachers Unions
and Educational Reform Washington, DC: Brookings Institution Press.

Gelman, A. (2008). Scaling regression inputs by dividing by two standard deviations.
Statistics in Medicine (2 7) :2865-2873.

Glazerman, S., Dolﬁn, S., Bleeker, M., Johnson, A., Isenberg, E., Hugo-Gil, J. Gruder,
M., Britton, E., Britton, E., & Ali, M. (2008). Impacts of comprehensive teacher
induction: Results from the ﬁrst year of a randomized controlled study.
Washington, DC: US. Department of Education, Institute of Education Sciences.

Granovetter, M. (1973). The strength of weak ties. American Journal of Sociology, 78(6),
1360-1380.

\

Granovetter, M. (1985). Economic action and social structure: The problem of
embeddedness. American Journal of Sociology, 91(3), 481-510.

Grossman, P.L., & Thompson, C. (2004). District policy and beginning teachers: A lens
on teacher learning. Educational Evaluation and Policy Analysis, 26(4), 281-301.

Hanushek, E, Kain, J ., & Rivkin, S. (2004) winter. The revolving door: A path-breaking
study of teachers in Texas reveals that working conditions matter more than
salary. Education Next, winter.

Hess, F .M., Rotherham, A.J., & Walsh, K. (2004). A qualified teacher in every
classroom? Cambridge, MA: Harvard University Press.

Hoxby, C. (1996). How teachers’ unions affect education production. The Quarterly
Journal of Economics 11 1 : 671-718.

Ingersoll, RM. (2001). Teacher turnover and teacher shortages: An organizational
analysis. American Educational Research Journal, 38(3), 499-535.

Johnson, S.M., & Kardos, S. (2000). Reforming Bargaining and Its Promise for School
Improvement. In Tom Loveless (Ed). Conﬂicting Missions? Teachers Unions and
Educational Reform. (pp. 746). Washington, DC: Brookings Institution Press.

Johnson, S.M., & The Project on the Next Generation of Teachers. (2004). Finders and
keepers. San Francisco: J ossey-Bass.

Kapadia, K., Coca, V., & Easton, J .Q. (2007). Keeping new teachers: A ﬁrst look at the

inﬂuences of induction in the Chicago Public Schools. Chicago: Consortium on
Chicago School Research, University of Chicago.

139

Kardos, S.M., Johnson, S.M., Peske, HG., Kauffman, D., & Liu, E. (2001). Counting on
colleagues: New teachers encounter the professional cultures of their schools.
Educational Administration Quarterly, 3 7(2), 250-290.

Kerchner, C., & Cooper, B. (2003). ’Ravening tigers’ Under Seige: Teacher union
legitimacy and institutional turmoil. In William Lowe Boyd and Debra Miretzky
(Eds). American educational governance on trial: Change and challenges.
Chicago: The University of Chicago Press.

Koppich, J. (2005 ). Addressing teacher quality through induction, professional
compensation, and evaluation: The effects on labor-management relations.
Educational Policy, 19, 90-11 1.

Lin, N. (1999). Building a network theory of social capital. Connections, 22, 28-51.

Lounsbury, M. (2001). Institutional sources of practice variation: Staffing college and
university recycling programs. Administrative Science Quarterly, 46, 29-56.

Lounsbury, M. & Ventresca, M. (2003). The New structuralism in organizational theory.
Organization, 10, 457-480.

McGraner, K.L., & Henrick, E. (2008). Effective principal leadership for teacher
induction: DevelOping a conception of the principal’s role in advancing teacher
development. Paper presented at the annual meeting of the American Educational
Research Association, New York.

Meyer, J ., & Rowan, B. (1977). Institutionalized organizations: Formal structure as myth
and ceremony.” American Journal of Sociology, 30, 431—450.

Meyer, J ., Scott, R., & Deal, T. (1981). Institutional and technical sources of
organizational structure: Explaining the structure of educational organizations. In
Herman D. Stein (Ed). Organization and the human services. (pp. 151-178).
Temple University Press.

Poole, W. (2000). The construction of teachers’ paradoxical interests by teacher union
leaders. American Educational Research Journal, 3 7, 93-119.

Portes, A. (1998). Social capital: Its origins and applications in modern sociology. Annual
Review of Sociology, 24, 1-24.

Powell, W. & Colyvas, J. (2007). Microfoundations of institutional theory. In R.
Greendwood (Ed). Handbook of Organizational Institutional ism. New York: Sage
publications.

Putnam, RD. (1995). Bowling Alone: America’s declining social capital. Journal of
Democracy, 6: 65 -7 8.

140

Rosenholz, SJ. (1985 ). Political myths about educational reform: Lessons from research
on teaching. Phi Delta Kappan, 66, 349-355.

Rosenholtz, SJ. (1989). Teachers' workplace: The social organization of schools. New
York: Longman.

Scott, W.R. (1998). Organizations: Rational, natural, and open systems. Upper Saddle
River, New Jersey: Prentice Hall.

Selznick, P. (1948). Foundations of the theory of organization. American Sociological
Review, 13, 25-35.

Smith, T.M., & Ingersoll, RM. (2004). Reducing teacher turnover: What are the
components of effective induction? American Educational Research Journal,
41(3), 681-714.

Stein, M.K., & D’Amico, L. (2002). Inquiry at the crossroads of policy and learning: A
study of a district-wide literacy initiative. Teachers College Record, 104(7), 1313-
1344.

Stone, J. (2000). Collective bargaining and public schools. In Tom Loveless (Ed).
Conﬂicting missions? Teachers unions and educational reform. (pp. 47-
68).Washington, DC: Brookings Institution Press.

Streshly, W., & DeMitchell, T. (1994). Teacher Unions and TQE. Thousand Oaks, CA:
Corwin Press, Inc.

Strong, M., & Fletcher, S. (2004). An investigation of the effects of teacher experience
and teacher preparedness on the performance of Latino students in California.
Report to the Latino Policy Center, Berkeley, CA.

Weiss, EM. (1999). Perceived workplace conditions and ﬁrst-year teachers’ morale,
career choice commitment, and planned retention: A secondary analysis.
Teaching and Teacher Education, I 5, 861-879.

Wooldridge, J. (2000). Introductory econometrics: A modern approach. South-Westem
College Publishing.

Youngs, P. (2007a). District induction policy and new teachers’ experiences: An
examination of local policy implementation in Connecticut. Teachers College
Record, 109(3), 797-837.

Youngs, P. (2007b). How elementary principals’ beliefs and actions inﬂuence new

teachers’ induction experiences. Educational Administration Quarterly, 44(2),
101-137.

141

Youngs, R, Qian, H., & Holdgreve-Resendez. (in progress). Teacher induction for
diverse, urban contexts.

142

       

    

ATE UNIVE

”'Cli'iiiii‘llun um 1111311111111“

1293 030