Ties for the phases: tie formation and educator life phases in urban fringe and rural schools

Abstract

Friendships and close relationships within organisations and schools are integral for developing trust, engagement, and a sense of community. That said, few studies exclusively look at tie formation in friendship networks in K-12 contexts. In this study, we explore the relationship between Day and Gu’s teacher life phases and the formation of friendship ties in four schools—two elementary and two high schools—in urban fringe and rural settings. Using exponential random graph models (ERGMs), we found that across contexts there were some life phases during which educators were more likely to form friendship ties with others in the same life phase. Also, there were notable differences in the factors that contributed to friendship ties and in the odds of friendship tie formation in the elementary versus high schools. The findings have implications for those interested in supporting professional capital-building among teachers as well as those looking to establish conditions necessary to nurture and sustain teacher resilience.

Full text

Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=ctat20
Teachers and Teaching
theory and practice
ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/ctat20
Ties for the phases: tie formation and educator life
phases in urban fringe and rural schools
Jennifer R. Karnopp & Peter Bjorklund Jr
To cite this article: Jennifer R. Karnopp & Peter Bjorklund Jr (2022): Ties for the phases: tie
formation and educator life phases in urban fringe and rural schools, Teachers and Teaching, DOI:
10.1080/13540602.2022.2062723
To link to this article: https://doi.org/10.1080/13540602.2022.2062723
Published online: 12 Apr 2022.
Submit your article to this journal
View related articles
View Crossmark data

Ties for the phases: tie formation and educator life phases in
urban fringe and rural schools
Jennifer R. Karnopp
a
*
*
and Peter Bjorklund Jr
b
a
Department of Educational Leadership, San Diego State University, San Diego, CA, USA;
b
Department of
Education Studies, University of California, San Diego, CA, USA
ABSTRACT
Friendships and close relationships within organisations and
schools are integral for developing trust, engagement, and
a sense of community. That said, few studies exclusively look at
tie formation in friendship networks in K-12 contexts. In this study,
we explore the relationship between Day and Gu’s teacher life
phases and the formation of friendship ties in four schools—two
elementary and two high schools—in urban fringe and rural set-
tings. Using exponential random graph models (ERGMs), we found
that across contexts there were some life phases during which
educators were more likely to form friendship ties with others in
the same life phase. Also, there were notable dierences in the
factors that contributed to friendship ties and in the odds of friend-
ship tie formation in the elementary versus high schools. The nd-
ings have implications for those interested in supporting
professional capital-building among teachers as well as those look-
ing to establish conditions necessary to nurture and sustain teacher
resilience.
ARTICLE HISTORY
Received 15 May 2021
Accepted 1 April 2022
KEYWORDS
Professional life phases; lives
of teachers; educator
relationships; social
networks; friendship ties
Introduction
The work of educators is inherently relational in that it requires regular interactions with
students, parents, colleagues and school leadership. Over the past two decades, mounting
evidence has revealed a correlation between positive relationships among teacher collea-
gues (those characterised by trust and/or friendship) and a host of behaviours and
dispositions linked to desired outcomes for teachers and students. Such outcomes
include successful school improvement and change implementation (Bridwell-Mitchell,
2015; Coburn et al., 2013; Daly & Finnigan, 2010), the development of social capital in
support of teaching effectiveness (Pil & Leana, 2009; Spillane et al., 2015), and resilience
among educators (Day & Gu, 2010; Gu, 2013 Hopkins et al., 2019). Furthermore,
evidence suggests that educator relationships within schools can augment or diminish
CONTACT Jennifer R. Karnopp jkarnopp@sdsu.edu San Diego State University, 5500 Campanile Drive, San Diego,
CA 92182
*Jennifer R. Karnopp is now at the Department of Educational Leadership, San Diego State University.
Both authors contributed equally to this manuscript
TEACHERS AND TEACHING
https://doi.org/10.1080/13540602.2022.2062723
© 2022 Informa UK Limited, trading as Taylor & Francis Group

teacher identity, self-efficacy, motivation, and commitment (Carrillo & Flores, 2018; Le
Cornu, 2013; Day & Gu, 2010; Fox & Wilson, 2015; Frank et al., 2020; Mansfield et al.,
2014).
These beneficial behaviours and dispositions contribute to a school’s professional
capital, ‘the critical elements of what it takes to create high quality and high performance
in all professional practice—including teaching’ (Hargreaves & Fullan, 2012, p. 102).
Friendships among educators within schools are integral to creating and increasing trust,
engagement and sense of community among educators (Bryk & Schneider, 2002; Day &
Gu, 2010; Hargreaves & O’Connor, 2018). As a result, trust and friendship have increas-
ingly been used as a variable of interest in a variety of empirical studies examining the
work of educators.
Despite the importance of positive affective relationships between educators, we know
little about the social processes that contribute to the formation of these ties in schools,
and the extent to which educators engage in within-school friendships during the various
phases of their careers. As a result, we have little understanding of the social processes
that foster friendship tie formation. This paper addresses this gap through an examina-
tion of educators’ friendship networks during the various professional life phases of
teachers (Day & Gu, 2010). The teacher's life phase framework has been used to explore
teachers’ needs, concerns, struggles, and triumphs (e.g., Day & Gu, 2010; Gu, 2013), but it
has not been used to explore tie formation in social networks. Thus, we contribute to
both the social network literature and literature exploring the lives of teachers. By
including school sites in both rural and urban fringe school contexts, we gain insights
into the broader social and structural processes that facilitate the formation of critical
friendship relationships.
Educator relationships and professional life phases
Extant research identifies a variety of different phases teachers move through as they
enter the profession and develop their practice over time. Building on the seminal work
of Huberman (1989), Day and Gu (2010) identified six professional life phases of career
educators. Consistent across these frameworks is the importance of feeling a part of
a community of practice—‘a colleague among peers’ (Huberman, 1989, p. 33)—for
supporting motivation and professional engagement. Strong, trusting relationships
with colleagues have been linked with educator resilience and positive sentiments
about the teaching profession across the life stages (Day, 2012; Hopkins et al., 2019;
Lowe et al., 2019). Resilience develops in the interaction between teachers’ professional
assets and the quality of the intellectual, social and organisational environments in which
they work (Gu, 2013). Mansfield et al. (2014) note that the quality of relationships, as well
as the attitudes and conditions of the work environment, can be a resource or a challenge
to resilience and efficacy. Thus, resilience is context specific, role specific, and supports
educators through short-term challenges while also providing a longer-term equilibrium
throughout each phase of a teacher’s professional life (Gu & Day, 2013).
For early career teachers (0–7 years of experience), strong relationships are integral to
fostering psychological safety as they try to find their teacher identity and a sense of
belonging in the school (Le Cornu, 2013; Day & Gu, 2010; Fox & Wilson, 2015; Mansfield
et al., 2014). Teachers in the middle stages (8–23 years) of their careers identify collegial
2J. R. KARNOPP AND P. BJORKLUND

relationships as a key component of job satisfaction and commitment (Day & Gu, 2010).
For veteran teachers (24 or more years), a professional life phase where many find it
difficult to sustain motivation for teaching (Day & Gu, 2010; Huberman, 1989), suppor-
tive relationships with colleagues help to maintain motivation and commitment (Carrillo
& Flores, 2018; Lowe et al., 2019). Taken together, this body of work identifies the
presence of friendships among educators within a school as important for enabling
teachers to develop and sustain a sense of efficacy and satisfaction throughout every
stage of their professional career.
Theory of social networks
The theory of social networks provides a conceptual approach to understanding relation-
ships present in a social system, such as an organisation (Wasserman & Faust, 1994).
Therefore, it is useful for examining how friendships form, persist or change over time
and how the web of social relationships in which people are embedded shape their
interactions and experiences. The people within a social network are called actors and
the relationships between actors are ties (Wasserman & Faust). Ties can be thought of as
either instrumental (a mechanism for moving resources or information) or expressive
(representing relationships or sentiments). Additionally, ties may be directed (e.g., actor
i gets advice from actor j) or undirected (e.g., actors i and j are friends). Directed ties can
be one-way (actor i seeks actor j) or reciprocal (actors i and j both seek advice from each
other) (Wasserman & Faust, 1994). Reciprocity is often an indicator of tie strength
(Borgatti & Ofem, 2010; Granovetter, 1973; Wasserman & Faust, 1994).
Network ties tend to form based on perceptions of social similarity (homophily), such
as being of the same gender, and physical proximity (propinquity), like teaching in
neighbouring classrooms (McPherson et al., 2001). Network structures also influence
tie formation (Lusher & Robins, 2013a). For example, people often reciprocate ties—if
someone confides in you about important issues you are likely to reciprocate and confide
in them (Lusher & Robins, 2013a). The expressive tie of friendship is highly correlated
with trust (McGrath & Krackhardt, 2003) and has been used as a proxy for trust in
scholarship on social networks in schools (e.g., Daly et al., 2010).
A growing body of the literature applies a social network approach to the work of
educators; e.g., the role of friendship/trust in information exchanges among educators
(Coburn et al., 2013; Daly & Finnigan, 2010; Frank et al., 2020). Geeraerts et al. (2017)
examined the relationship between educator age and four instrumental social networks—
discussing work, asking for advice, providing advice, and collaboration. The researchers
grouped teachers into three generational cohorts: young (less than 36 years of age),
middle (36–50 years of age), and old (50 years of age or more). Findings revealed distinct
patterns of behaviour between and within the different age cohorts. Furthermore, being
of the same age (age homophily) was negatively correlated with tie formation in young
and old cohorts in some of the networks examined.
Subsequent research revealed significant correlations between teachers’ age and parti-
cipation in subject-matter-specific advice networks, with the direction of the correlation
varying by subject (Geeraerts et al., 2018). Similarly, teachers’ career stage (i.e., years of
experience teaching) was associated with tie formation in advice-seeking networks with
experience being negatively associated with both seeking advice from colleagues (Spillane
TEACHERS AND TEACHING 3

et al., 2015, 2012) and providing advice (Spillane et al., 2015). This research explores the
role of experience or career stage in instrumental ties. Little is known about its relation-
ship to expressive ties.
The following research questions guide our inquiry into educators’ friendship net-
works during the various professional life phases of teachers:
●What are some of the factors that contribute to the presence of friendship ties
among educators in elementary and high schools?
●What role does educator life-phase play in tie formation in friendship networks?
●How do social structures vary between the life phases?
Data and methods
We explore educator networks in two urban fringe schools in the southwestern United
States and two remote rural schools in the mid-western United States—Urban Fringe
1
Elementary School (UES, N = 26), Urban Fringe High School (UHS, N = 43), Rural
Elementary School (RES, N = 64) and Rural High School (RHS, N = 41). Staff in each
school was surveyed in fall 2019 (see, Table 1). The range of participants who identified as
female in each school was 88% to 56%. Educator career experience averaged 21.26 years
(SD = 7.74) at UES and 11.02 years (SD = 7.80) at UHS. At each school, teachers made up
between 75% and 88% of our participants with administrators and other staff accounting
for between 25% and 12%. In both rural schools, the entire sample identified as White. At
UES and UHS, 73% and 67% identified as White, respectively. At UES, 8% identified as
Latina/o or more than one group and 4% identified as Black, Asian, and Arab/Middle
Eastern. At UHS, 12% of the sample identified as Latina/o and Asian while those who
Table 1. Sample demographics fall 2019.
UES (N = 26) RES (N = 64) UHS (N = 42) RHS (N = 41)
Female .85 .88 .76 .56
Years in Education 21.62
(7.74)
15.17
(8.82)
11.02
(7.80)
16.22
(11.70)
Role
Teacher .88 .75 .86 .83
Administrator .04 .03 .07 .05
Classified staff (Counsellor, social worker, etc.) .08 .22 .07 .12
Racial-Ethnic Group
White .73 1.00 .67 1.00
Latina/o .08 - .12 -
Asian .04 - .12 -
Black .04 - - -
Arab/Middle Eastern .04 - .02 -
More than one group .08 - .07 -
Educator Life Phases
(Day & Gu, 2010)
Phase 1 (0–3 years) .04 .12 .17 .12
Phase 2 (4–7 years) - .06 .26 .22
Phase 3 (8–15 years) .19 .39 .31 .20
Phase 4 (16–23 years) .19 .28 .19 .17
Phase 5 (24–30 years) .42 .06 .05 .12
Phase 6 (31+ years) .15 .08 .02 .17
Standard deviation of years in education is shown in parentheses.
Percentages may not add up to 100% due to rounding.
4J. R. KARNOPP AND P. BJORKLUND

identified with more than one group or as Arab/Middle Eastern made up 7% and 2% of
the samples, respectively. In terms of educator life phase, UES had more veteran
educators (57%) and most participants at RES were in the middle phases of their
professional life (67%), while the majority of participants in the high schools were in
phases one through three (64% at UHS and 54% at RHS).
We used surveys to collect social network data. Participants were asked to identify
friendships using a roster of their colleagues (Scott, 2013). As these surveys were part of
separate research projects, the rural and urban-fringe participants received slightly
different, but comparable prompts.
●Urban fringe prompt: Of your colleagues, who are those with whom you have
a close relationship? By ‘close’ we mean a person who you trust with personal
information and/or spend time with in informal settings.
●Rural prompt: Below is a list of staff at your school. Please mark the circle next to
each name indicating your close friends-those you spend time in personal activities
and/or engage in candid conversations.
We had high response rates from each site—100% at UES, 83% at UHS, 70% at RES, and
83% at RHS. We were able to retrieve demographic data for missing respondents from
publicly available sources (Table 1). Because it is preferable to impute missing socio-
metric network data than drop actors from the network (Borgatti et al., 2018), we used
the reconstruction and modal incoming network imputation options in UCINET
(Borgatti et al., 2002; Žnidaršič et al., 2012) to address missing responses. We included
only reciprocal ties (i.e., ties where both actors named each other as close friends) as
reciprocal ties are considered strong ties (Granovetter, 1973) and more accurate
(Marsden, 1990).
Data analysis and variables
We initially analysed the average normalised degree—a ratio of the number of ties for an
actor relative to the number of possible ties—between educators at the high school and
elementary school levels and then between educators in different life phases at each level.
This allowed us to compare networks of different sizes (Wasserman & Faust, 1994). We
then used exponential random graph models (ERGMs; Bjorklund & Daly, 2021; Lusher
et al., 2013) to examine tie formation within each school, as described below.
Social network data violate the assumption of independence of variables used by many
statistical tests (e.g., OLS regression; Lusher et al., 2020). As such, to explore differences
in average degree at the high school level and elementary school level, we conducted a t
test using permutation tests in UCINET (Borgatti et al., 2002, 2018). Day and Gu (2010)
categorised the six different life phases identified in their study into three broader
groupings: beginning teachers (phases one and two); middle years (phases three and
four); and veteran teachers (phases five and six). We utilised these groupings to examined
differences phase categories at the high schools and elementary schools using ANOVA
permutation tests in UCINET (Borgatti et al., 2002, 2018).
TEACHERS AND TEACHING 5

To explore tie formation in our networks we used the ergm command in the Statnet
package in R (Handcock et al., 2018; Hunter et al., 2013). ERGMs generate a large sample
of random networks with our given network characteristics and compare these to our
network to identify whether patterns of tie formation differ from random chance (Lusher
et al., 2020; Robins & Lusher, 2013a). ERGMs allow us to model network/structural
characteristics (described below) and individual actor characteristics to examine their
relationship to tie formation. Thus, ERGMs enable us to analyse a range of social
processes and personal attributes in our models (Lusher et al., 2020).
Structural effects
Structural effects represent social processes in the network (Lusher et al., 2020). We
conditioned our models on several structural effects (see, Figure 1): edges, concurrent
nodes, degree spread using geometrically weighted degree (GWD), transitivity using
geometrically weighted edgewise shared partners (GWESP), and multiple connectivity
using directed geometrically weighted dyadwise shared partners (GWDSP) as suggested
by Robins and Lusher (2013b). Each structural effect conditions our model on social
processes that occur in networks, thus supporting stronger and more ‘principled infer-
ences’ about the relationships between our actor-level effects and tie formation (Lusher &
Robins, 2013b, p. 197).
Figure 1. Structural effects included in models.
6J. R. KARNOPP AND P. BJORKLUND

As noted above, we use geometrically weighted effects for degree, transitivity, and
multiple connectivity in our models. Geometrically weighted effects use a weighting
parameter that diminishes the impact of high numbers of triads or ties that may
inhibit model convergence (Snijders, 2011; Snijders et al., 2006). We tested multiple
weighting parameters until we found the parameters that gave us the best fit for our
models.
Homophily effects
Ties tend to form based on homophily in gender, ethno-racial group, and similar job
tasks (McPherson et al., 2020; Spillane et al., 2015, 2012). We controlled for homo-
phily related to gender, role (i.e., teacher, principal, counsellor), and racial-ethnic
identity in our models. We omitted homophily for race/ethnicity in the rural schools
where all educators identified as White. We also explored homophily based on life
phase in two different models for each network. The first set of models (Table 2)
incorporated a single variable that examined the general likelihood of tie formation
based on life phase homophily. Our second set (Table 3) includes a binary effect for
each life phase to see if there are specific life phases with more homophily than others.
We omitted some life phases in our second models due to perfect correlation (i.e., all
ties were between actors in the same life phase) or because there were no educators or
one educator in the category.
Table 2. ERGM models with categorical life-phase variable.
Urban fringe ES OR Rural ES OR urban fringe HS OR Rural HS OR
Structural Effects
Edges −4.84***
(1.37)
−2.48***
(.84)
−4.22***
(.81)
−4.01***
(.41)
Concurrent Nodes .16
(1.03)
−.78
(.64)
a−.05
(.68)
Degree (GWD) 1.37
(1.23)
−1.10
(.68)
.31
(.76)
−.08
(.61)
Closed Triads (GWESP) 1.08***
(.28)
.71***
(.14)
1.34***
(.24)
.48**
(.17)
Open Triads (GWDSP) −.09
(.11)
−.14
+
(.08)
−.07
(.05)
.10***
(.03)
Homophily & Heterophily Effects
Same gender −.02
(.29)
.33
(.22)
.38**
(.14)
1.46 .47*
(.23)
1.60
Same role 1.02
+
(.62)
2.77 .47*
(.18)
1.60 .21
(.15)
.02
(.18)
Same racial-ethnic group .64*
(.29)
1.90 a.24
+
(.14)
1.27 a
D&G Life phases .36
(.29)
−.07
(.21)
.71***
(.28)
2.03 .64*
(.28)
1.91
***p< .001
, **
p < .01, *p < .05,
+
p < .1
Note: ‘a’ indicates omission from the model due to perfect correlation
TEACHERS AND TEACHING 7

Results
Figure 2 shows each school network. We initially wanted to see if there was a significant
difference between elementary and high-school educators’ friendship ties. A t test using
a permutation test (Borgatti et al., 2018) found a significant difference (p < .001) between
the two groups. High-school teachers reported roughly 7% more friendship ties than
their elementary school counterparts (see, Table 4).
The ANOVA test examined the differences between teachers in the beginning, middle,
and veteran life phases (Table 5) and revealed significant differences between the life
phases in both high schools (F = 3.26, p < .05) and elementary schools (F = 7.94, p < .01).
Being in one of the three life phases accounted for 7% of the variance in average degree
for high-school educators (η
2
= .07) and 17% of the variance for elementary school
educators (η
2
= .17). Interestingly, high-school educators in the veteran phases reported
a lower average degree (.09) than their peers in the beginning (.17) and middle (.18)
phases, whereas elementary school teachers in the veteran phase reported higher average
degree (.16) relative to their peers in the beginning (.05) and middle (.08) phases. In other
words, veteran high-school educators reported fewer reciprocal friendship ties relative to
their peers, while veteran elementary school educators reported more.
Table 3. ERGM models with individual binary life-phase variables.
UES OR RES OR UHS OR RHS OR
Structural Effects
Edges −4.98***
(1.34)
−2.57***
(.86)
−4.18***
(.78)
−4.15***
(.38)
Concurrent Nodes .16
(1.09)
−.79
(.64)
a−.09
(.72)
Degree (GWD) 1.43
(1.24)
−1.07
(.68)
.69
(.93)
−.08
(.65)
Closed Triads (GWESP) 1.08***
(.28)
.71***
(.14)
1.39***
(.23)
1.38***
(.23)
Open Triads (GWDSP) −.09
(.11)
−.14
+
(.08)
−.08
+
(.04)
.11***
(.02)
Homophily & Heterophily Effects
Same gender −.02
(.29)
.36
(.22)
.36*
(.14)
1.43 .51*
(.24)
1.64
Same role 1.14*
(.56)
3.12 .52*
(.19)
1.68 .19
(.14)
.02
(.14)
Same racial-ethnic group .62*
(.29)
1.86 a.20
(.13)
a
D&G Life Phases
Phase 1 c−.63
(.68)
.96*
(.38)
2.61 a
Phase 2 c−.71
(1.04)
.74**
(.23)
2.09 .79*
(.36)
2.20
Phase 3 .36
(1.12)
−.02
(.22)
.71***
(.21)
2.03 .84*
(.40)
2.31
Phase 4 1.22*
(.61)
3.39 −.54
(.50)
.58
+
(.34)
1.79 .55
(.60)
Phase 5 .20
(.32)
1.48
(.95)
a.25
(.89)
Phase 6 .92
(.71)
1.29*
(.60)
3.63 c.54
(.66)
***p< .001
, **
p < .01, *p < .05,
+
p < .1
‘a’ indicates omission from the model due to perfect correlation
‘c’ indicates omission from the model due to only one or zero participants in the category
8J. R. KARNOPP AND P. BJORKLUND

Table 2 shows the ERGM results using educator life phase as a single categorical
variable and Table 3 shows the results when we examined homophily in tie formation at
each life phase. Our ERGM models present results in log odds. For ease of interpretation,
Figure 2. Networks in each school.
Table 4. T-test examining difference between the average degree between high school and
elementary school educators (10,000 permutations).
High School Elementary School Absolute Difference
Average Degree Normalised .16
(.13)
.09
(.10)
.07***
***p < .001
Standard errors in parentheses
Table 5. ANOVA test examining difference between the average degree between educator life phases
at the high school and elementary school level (10,000 permutations).
High School Elementary School
Beginning
Phases
Middle
Phases
Veteran
Phases
Beginning
Phases
Middle
Phases
Veteran
Phases
Average Degree
Normalised
.17 .18 .09 .05 .08 .16
F-Statistic 3.26* 7.94**
Eta-squared .07 .16
**p < .01, *p < .05
TEACHERS AND TEACHING 9

we transformed the log odds of significant actor-level effects into odds ratios. A positive
and significant coefficient for structural effects indicates that the network structure is
found in the observed network more frequently than would occur by chance alone.
For actor-level effects, we examined homophily in terms of gender, same role, same
racial ethnic group, and Day and Gu’s (2010) educator life phases. We found that being in
the same role significantly increased the odds of tie formation at UES (exp [1.02] = 2.27)
and RES (exp [.47] = 1.60), meaning that being in the same role more than doubled the
odds of a tie at UES, and increased the odds by 60% at RES. This effect was not significant
at UHS or RHS. Conversely, being of the same gender was positively correlated with tie
formation at UHS (exp [.38] = 1.46) and RHS (exp [.47] = 1.60). Being of the same race/
ethnicity increased the odds of tie formation at UES (exp [.64] = 1.90) and UHS (exp
[.24] = 1.27). Interestingly, we found that being of the same educator life phase was not
a significant predictor of tie formation at either UES or RES, but it was one of the
strongest predictors at UHS (exp [.71] = 2.03) and RHS (exp [.64] = 1.91). Being of the
same educator life phase at UHS doubled the odds of a friendship tie and it was 91% more
likely than we would expect by chance alone at RHS. Broadly speaking, these results
indicate that being of the same educator life phase increases the odds of friendship tie
formation.
Parsing out the different life phases revealed new results. Table 3 (below) details our
findings related to homophily at each educator life phase. At UES, being in life phase 4
increased the odds of forming a friendship tie with another educators in the same life
phase three times over (exp [1.22] = 3.39). Similarly, at RES, educators within life phase
six were more than three and a half times likely to form ties with each other then would
be expected by chance alone (exp [1.29] = 3.36). At UHS, there were high levels of
homophily between educators in life phase 1 (exp [.96] = 2.61), phase 2 (exp [.74] = 2.09),
phase 3 (exp [.71] = 2.03), phase 4 (exp [.58] = 1.79), and phase 5 (perfect correlation). At
RHS, there were high levels of homophily for life phase 1 (perfect correlation), phase 2
(exp [.79] = 2.20), and phase 3 (exp [.84] = 2.31).
Discussion and conclusion
The purpose of this study was to explore the structures of educators’ within-school
friendship networks and examine the social processes that facilitate tie formation, with
specific attention to the professional life phases of educators (Day & Gu, 2010). Findings
revealed that across all four sites, there were specific life phases during which educators
were more likely to form friendship ties with others in the same phase, thus confirming
that within-school friendships can be an important source of professional support for
educators at different life phases. Prior research indicates that teachers who feel emo-
tionally supported and connected to their colleagues build relational resilience (Jordan,
2005) and are more likely to remain in teaching (Gu, 2013; Gu & Day, 2013). Positive
relationships between educators are strongly related to teacher well-being and feelings of
belonging (Roffey, 2012). Therefore, understanding that a friendship tie is more likely to
form between educators of the same professional life phase can help school leaders to
employ practices that will foster these relationships as a strategy for increasing educator
engagement and well-being. That said, it is important that school leaders recognise that
attitudes and conditions of the work environment mediate resilience and efficacy
10 J. R. KARNOPP AND P. BJORKLUND

(Mansfield et al., 2014), as illustrated by the work of Meredith et al. (2020) on burnout
contagion. Thus, leaders should attend to the quality of relationships among educators
they encourage.
Across both rural and urban fringe contexts, high-school educators were significantly
more likely to form friendship ties with others in the same professional life phase. This
finding indicates that the needs met by these ties transcend context (Le Cornu, 2013; Fox
& Wilson, 2015) and aligns with prior research indicating that trusting relationships are
important for supporting early-career teachers in the face of challenges (Frank et al.,
2020) to sustain them in the profession (Le Cornu, 2013; Fox & Wilson, 2015; Mansfield
et al., 2014). At the other end of the professional life-phase spectrum, veteran teachers
identify colleagues as an important support group that positively impact their sense of
efficacy, professional growth, and resilience (Carrillo & Flores, 2018). Past research has
shown that teachers with more experience are less likely to seek advice from peers
(Spillane et al., 2015) and that older teachers are less likely to be sought out for advice
regarding novel ideas and technology (Geeraerts et al., 2018). As such, our results
indicate that they may seek friendship and support from peers in the same professio-
nallife phase. These results seem to run counter to Geeraerts et al.’s (2017) finding that
age cohort homophily was negatively correlated with advice seeking. One explanation
may be that these educators share similar responsibilities or they face similar challenges
(Spillane et al., 2015). Positive collegial relationships sustain commitment in the face of
challenges (Day & Gu, 2010).
While all schools in our sample demonstrated some correlation between friendship
ties and specific life phases, there were notable differences in the factors that contributed
to friendship ties in the elementary schools versus the high schools. Being in the same life
phase was one of the strongest predictors of tie formation in both high schools, particu-
larly among early-career educators, but was not significant at either elementary school.
One exception was that veteran elementary school teachers were significantly more likely
to form ties with one another. Interestingly, the opposite was true at the high school,
where there was no significant correlation between friendship and being a veteran
teacher. This may be explained by the distinct organisational structures of high schools
versus elementary schools (e.g., self-contained classrooms versus departments), thus
aligning our finding with prior research indicating that organisational structures influ-
ence educators’ personal support networks (Coburn et al., 2013; Le Cornu, 2013; Fox &
Wilson, 2015). Leadership in particular has been found to have a mediating role among
veteran teachers (Lowe et al., 2019).
This study identified additional characteristics that made a significant contribution to
friendship tie formation among building colleagues. Overall, high-school educators were
more likely to form friendship ties than elementary school educators. While this may be
related to the distribution of educators within each professional life phase at each school
(e.g., there were no UES educators in phase 2 of their career), a compelling explanation
for these findings is that the routines and structures of secondary schools facilitate more
educator interactions than is common in elementary schools. Furthermore, gender
homophily was important to tie formation in both high schools, but being in the same
professional role was not while, within both elementary schools, being in the same
professional role was important to friendship tie formation, but gender was not. This
aligns with Yavuzkurt and Kıral (2012) who identified that secondary school teachers
TEACHERS AND TEACHING 11

perceived high opportunities for friendship within their work contexts with opportunity
perception being significantly higher among female educators than among males. The
authors proposed that this may be attributed to gender differences around constructs of
friendship.
Finally, it is important to note that in the schools with a racially diverse staff, race was
a significant predictor of friendship tie formation. This aligns with Bridwell-Mitchell
(2015) who identified that faculty of colour was not a part of the ‘mega-click’ of White
educators in their school. Subsequently, they were left out of the reform-related informa-
tion and advice exchanges that occurred informally at lunch and outside of school. Such
findings have important implications for the inclusion of diverse perspectives in the
development of school culture and reform implementation, particularly as school orga-
nisations work towards anti-bias and culturally conscious practices.
In regards to practice, these findings suggest that by strategically fostering these
positive affective ties, school leaders can develop organisational capital (Hargreaves &
Fullan, 2012), enhance a sense of collective efficacy and organisational commitment (Day
& Gu, 2010), and establish the conditions necessary to nurture and sustain teacher
resilience (Le Cornu, 2013; Fox & Wilson, 2015; Gu, 2013). Leaders can support educator
resilience by helping beginning teachers to become more familiar with the schools’
systems and help them build supportive relationships with staff (Fox & Wilson, 2015;
Mansfield et al., 2014). These supportive relationships can promote feelings of relational
trust which is postively related to retention of early career teachers (Hopkins et al., 2019).
At the elementary level, traditional organisational structures may create barriers to the
organic development of friendship ties, therefore leaders should consider intentionally
creating opportunities for friendship tie development through routines and practices
(e.g., scheduled meetings, peer mentoring structures, use of shared space). Furthermore,
potential ethnic or racial social divisions among staff may be bridged through a culture of
equitable engagement enhanced by ensuring diverse voices are assigned to collaborative
tasks that encourage social interactions.
Friendship relationships are integral to the inherently emotional work of teaching
(Roffey, 2012). More research is needed exploring the mechanisms and social processes
that facilitate these ties in other school contexts. Future research exploring tie formation
using mixed methods or qualitative network approaches (e.g., Froelich et al., 2020).; Fox
& Wilson, 2015; Thomas et al., 2019) can provide a more nuanced understanding of why
these ties form as well as focus on the quality of these ties and the nature of the resources,
support, and emotional connections being passed through them.
Note
1. Classifications based on National Center for Education Statistics’ Rural Educational
Achievement Programme guidelines
Disclosure statement
No potential conflict of interest was reported by the author(s).
12 J. R. KARNOPP AND P. BJORKLUND

Notes on contributors
Jennifer Karnopp, Ph.D. is an assistant professor in the department of Educational Leadership at
San Diego State University. Her research explores the intersection of leadership practice and
organizational change for more equitable and inclusive schooling, with specific attention to the
role of formal and informal relationships in organizational functioning.
Peter Bjorklund Jr., Ph.D. is a postdoctral researcher in the Education Studies Department and a
Lecturer in the Human Developmental Sciences Programme at University of California San Diego.
His research focuses on social network analysis, belonging, trust, and identity in K-16 contexts.
ORCID
Jennifer R. Karnopp http://orcid.org/0000-0002-4671-2279
References
Bjorklund Jr., P., & Daly, A.J. (2021). The ties that belong: Tie formation in preservice teacher
identification networks. Teaching and Teacher Education, 97, 103223. https://doi.org/10.1016/j.
tate.2020.103223
Borgatti, S. P., Everett, M. G., & Freeman, L. C. (2002). Ucinet 6 for Windows: Software for social
network analysis [Computer software]. Analytic Technologies.
Borgatti, S. P., Everett, M. G., & Johnson, J. C. (2018). Analyzing social networks. Sage.
Borgatti, S. P., & Ofem, B. (2010). Overview: Social network theory and analysis. In A. J. Daly (Ed.),
Social network theory and educational change (pp. 17–30). Harvard Education Press.
Bridwell-Mitchell, E. N. (2015). Theorizing teacher agency and reform: How institutionalized
instructional practices change and persist. Sociology of Education, 88(2), 140–159. https://doi.
org/10.1177/0038040715575559
Bryk, A., & Schneider, B. (2002). Trust in schools: A core resource for improvement. Russell Sage
Foundation.
Carrillo, C., & Flores, M. A. (2018). Veteran teachers’ identity: What does the research literature
tell us? Cambridge Journal of Education, 48(5), 639–656. https://doi.org/10.1080/0305764X.
2017.1394982
Coburn, C. E., Mata, W. S., & Choi, L. (2013). The embeddedness of teachers’ social networks:
Evidence from a study of mathematics reform. Sociology of Education, 86(4), 311–342. https://
doi.org/10.1177/0038040713501147
Daly, A. J., & Finnigan, K. S. (2012). Exploring the space between social networks, trust, and urban
school district leaders. Journal of School Leadership, 22(May 2012), 493–530. https://doi.org/10.
1177/105268461202200304
Daly, A. J., Moolenaar, N. M., Bolivar, J. M., & Burke, P. (2010). Relationships in reform: The role
of teachers’ social networks. Journal of Educational Administration, 48(3), 359–391. https://doi.
org/10.1108/09578231011041062
Daly, A. J., & Finnigan, K. S. (2010). A bridge between worlds: understanding network structure to
understand change strategy. Journal of Educational Change, 11(2), 111–138. 10.1007/s10833-
009-9102-5
Day, C. (2012). New lives of teachers. Teacher Education Quarterly, 39(1), 7–26. https://files.eric.
ed.gov/fulltext/EJ977354.pdf
Day, C., & Gu, Q. (2010). The new lives of teachers. Routledge.
Fox, A. R., & Wilson, E. G. (2015). Networking and the development of professionals: Beginning
teachers building social capital. Teaching and Teacher Education, 47, 93–107. https://doi.org/10.
1016/j.tate.2014.12.004
TEACHERS AND TEACHING 13

Frank, K. A., Kim, J., Salloum, S. J., Bieda, K. N., & Youngs, P. (2020). From interpretation to
instructional practice: A network study of early-career teachers’ sensemaking in the era of
accountability pressures and Common Core state standards. American Educational Research
Journal, 57(6), 2293–2338. https://doi.org/10.3102/0002831220911065
Froehlich, D.E., Van Waes, S., & Schäfer, H. (2020). Linking quantitative and qualitative network
approaches: A review of mixed methods social network analysis in education research. Review of
Research in Education, 44(1), 244–268. https://doi.org/10.3102/0091732X20903311
Geeraerts, K., Van den Bossche, P., Vanhoof, J., & Moolenaar, N. (2017). Intergenerational
professional relationships in elementary school teams: A social network approach. Frontline
Learning Research, 5(2), 78–98. https://doi.org/10.14786/flr.v5i2.293
Geeraerts, K., Vanhoof, J., & Van den Bossche, P. (2018). Teachers’ intergenerational advice and
information seeking: Content matters! Journal of Professional Capital and Community, 3(4),
256–271. https://doi.org/10.1108/JPCC-11-2017-0026
Granovetter, M. (1973). The strength of weak ties. American Journal of Sociology, 78(6),
1360–1380. https://doi.org/10.1086/225469
Gu, Q. (2014). The role of relational resilience in teachers’ career-long commitment and effective-
ness. Teachers and Teaching: Theory and Practice, 20(5), 502–529. https://doi.org/10.1080/
13540602.2014.937961
Gu, Q., & Day, C. (2013). Challenges to teacher resilience: Conditions count. British Educational
Research Journal, 39(1), 22–44. https://doi.org/10.1080/01411926.2011.623152
Handcock, M., Hunter, D., Butts, C., Goodreau, S., Krivitsky, P., & Morris, M. (2018). ergm: Fit,
simulate and diagnose exponential-family models for networks [The Statnet Project R package
version 3.9.4.]. https://CRAN.R-project.org/package=ergm
Hargreaves, A., & Fullan, M. (2012). Professional capital: Transforming teaching in every school.
Teachers College Press.
Hargreaves, A., & O’Connor, M. T. (2018). Solidarity with solidity: The case for collaborative
professionalism. Phi Delta Kappan, 100(1), 20–24. https://doi.org/10.1177/
0031721718797116
Hopkins, M., Bjorklund Jr., P. & Spillane, J.P. (2019). The social side of teacher turnover: Closeness
and trust among general and special education teachers in the United States. International
Journal of Educational Research, 98 292–302. 10.1016/j.ijer.2019.08.020
Huberman, M. (1989). The professional life cycle of teachers. Teachers College Record: The Voice of
Scholarship in Education, 91(1), 31–57. https://doi.org/10.1177/016146818909100107
Hunter, D. R., Handcock, M. S., Butts, C. T., Goodreau, S. M., & Morris, M. (2008). ergm:
A package to fit, simulate and diagnose exponential-family models for networks. Journal of
Statistical Software, 24(3), 1–29. https://doi.org/10.18637/jss.v024.i03
Jordan, J.V (2005) Relational resilience in girls. Handbook of resilience in children. (Eds.)
Goldstein, S., Brooks, R.B. Springer. (pp. 79–90).
Le Cornu, R. (2013). Building early career teacher resilience: The role of relationships. Australian
Journal of Teacher Education (Online), 38(4), 1–16. https://ro.ecu.edu.au/cgi/viewcontent.cgi?
referer=&httpsredir=1&article=1992&context=ajte
Lowe, G., Gray, C., Prout, P., Jefferson, S., & Shaw, T. (2019). Still keen and committed: Piloting an
instrument for identifying positive veteran teachers. Teachers and Teaching, 25(4), 418–433.
https://doi.org/10.1080/13540602.2019.1621829
Lusher, D., Koskinen, J., & Robins, G. (2013). Exponential random graph models for social network
analysis: Theory, methods, and applications. Cambridge University Press.
Lusher, D., & Robins, G. (2013a). Formation of social network structure. In D. Lusher, J. Koskinen,
& G. Robins (Eds.), Exponential random graph models for social network analysis: Theory,
methods, and applications (pp. 16–28). Cambridge University Press.
Lusher, D., & Robins, G. (2013b). Personal attitudes, perceived attitudes and social structures:
A social selection model. In D. Lusher, J. Koskinen, & G. Robins (Eds.), Exponential random
graph models for social network analysis: Theory, methods, and applications (pp. 189–201).
Cambridge University Press.
14 J. R. KARNOPP AND P. BJORKLUND

Lusher, D., Wang, P., Brenneke, J., Brailly, J., Faye, M., & Gallagher, C. (2020). Advances in
exponential random graph models. In R. Light & J. Moody (Eds.), The Oxford Handbook of
social networks (pp. 234–253). Oxford University Press.
Mansfield, C., Beltman, S., & Price, A. (2014). ‘I’m coming back again!’ The resilience process of
early career teachers. Teachers and Teaching, 20(5), 547–567. https://doi.org/10.1080/13540602.
2014.937958
Marsden, P. V. (1990). Network data and measurement. Annual Review of Sociology, 16(1),
435–463. https://doi.org/10.1146/annurev.so.16.080190.002251
McGrath, C., & Krackhardt, D. (2003). Network conditions for organizational change. The Journal
of Applied Behavioral Science, 39(3), 324–336. https://doi.org/10.1177/0021886303258267
McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: Homophily in social
networks. Annual Review of Sociology, 27(1), 415–444. https://doi.org/10.1146/annurev.soc.27.
1.415
Meredith, C., Schaufeli, W., Struyve, C., Vandecandelaere, M., Gielen, S., & Kyndt, E. (2020).
“Burnout contagion” among teachers: A social network approach. Journal of Occupational and
Organizational Psychology, 93(2), 328–352. https://doi.org/10.1111/joop.12296
Pil, F. K., & Leana, C. (2009). Applying organizational research to public school reform. Academy
of Management Journal, 52(6), 1024–1101. https://doi.org/10.5465/AMJ.2009.47084647
Robins, G., & Lusher, D. (2013a). A simplified account of an exponential random graph model as
a statistical model. In D. Lusher, J. Koskinen, & G. Robins (Eds.), Exponential random graph
models for social network analysis: Theory, methods, and applications (pp. 29–36). Cambridge
University Press.
Robins, G., & Lusher, D. (2013b). Illustrations: Simulation, estimation, and goodness of fit. In
D. Lusher, J. Koskinen, & G. Robins (Eds.), Exponential random graph models for social network
analysis: Theory, methods, and applications (pp. 167–185). Cambridge University Press.
Roffey, S. (2012). Developing positive relationships in schools. In S. Roffey (Ed.), Positive relation-
ships: Evidence based practice across the world (pp. 145–162). Springer.
Scott, J. (2013). Social network analysis (2nd ed.). Sage.
Snijders, T. A. (2011). Statistical models for social networks. Annual Review of Sociology, 37(1),
131–153. https://doi.org/10.1146/annurev.soc.012809.102709
Snijders, T. A., Pattison, P. E., Robins, G. L., & Handcock, M. S. (2006). New specifications for
exponential random graph models. Sociological Methodology, 36(1), 99–153. https://doi.org/10.
1111/j.1467-9531.2006.00176.x
Spillane, J. P., Hopkins, M., & Sweet, T. M. (2015). Intra- and interschool interactions about
instruction: Exploring the conditions for social capital development. American Journal of
Education, 122(1), 71–110. https://doi.org/10.1086/683292
Spillane, J. P., Kim, C. M., & Frank, K. A. (2012). Instructional advice and information providing
and receiving behavior in elementary schools: Exploring tie formation as a building block in
social capital development. American Educational Research Journal, 49(6), 1112–1145. https://
doi.org/10.3102/0002831212459339
Thomas, L., Tuytens, M., Devos, G., Kelchtermans, G., & Vanderlinde, R. (2019). Beginning
teachers’ professional support: A mixed methods social network study. Teaching and Teacher
Education, 83, 134–147. https://doi.org/10.1016/j.tate.2019.04.008
Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and applications (Vol. 8).
Cambridge University Press.
Yavuzkurt, T., & Kıral, E. (2020). The relationship between workplace friendship and job satisfac-
tion in educational organizations. International Journal of Progressive Education, 16(5), 404–
425. https://doi.org/10.29329/ijpe.2020.277.25
Žnidaršič, A., Doreian, P., & Ferligoj, A. (2012). Absent ties in social networks, their treatments,
and blockmodeling outcomes. Advances in Methodology & Statistics/Metodoloski Zvezki, 9(2),
119–138. https://doi.org/10.51936/redp2838
TEACHERS AND TEACHING 15