Teaching and Teacher Education 96 (2020) 103190

Contents lists available at ScienceDirect

Teaching and Teacher Education

journal homepage: www.elsevier.com/locate/tate

Research paper

Do teachers’ years of experience make a difference in the quality of

teaching?
Linda J. Graham a, *, Sonia L.J. White a, Kathy Cologon b, Robert C. Pianta c
a
Queensland University of Technology (QUT), Australia
b
Macquarie University, Australia
c
University of Virginia, USA

h i g h l i g h t s

Teaching quality observed for 80 teachers with 0e3, 4e5, and >5 years’ experience.

No evidence of less competence for teachers with 0e3 years’ experience.

Some evidence of decline in teaching quality in teachers with 4e5 years’ experience.

Behaviour management among highest scoring dimensions for all teachers.

Support and professional learning a priority for all teachers, not just beginners.

a r t i c l e i n f o a b s t r a c t

Article history: Extensive reforms have been made to initial teacher education (ITE) to improve “teacher quality” without
Received 17 January 2020 any evidence to support the claim that beginning teachers are less competent than experienced teachers.
Received in revised form This study adds to the evidence base by investigating associations between teachers’ years of experience
3 August 2020
and teaching quality. Results show no evidence of lower teaching quality for beginning teachers (0e3
Accepted 13 August 2020
years’ experience), but some evidence of a decline in teaching quality for teachers with 4e5 years
Available online 29 August 2020
experience. Findings suggest that the quality of teaching could be higher overall, and that targeted
support and evidence-informed professional learning would benefit all teachers.
Keywords:
Initial teacher education
© 2020 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
Beginning teachers license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Teachers’ years of experience
Teaching quality
Classroom observations

1. Introduction increased scrutiny of teaching and “teacher quality” (Scholes et al.,

2017). This shift in focus has been propelled by claims that “in
The international comparison of student performance in excess of 40 percent of the residual variance in measures of student
standardised assessments and subsequent ranking of participating performance (adjusted for students’ background and intake char-
countries over the last two decades has generated significant public acteristics) is at the class/teacher-level” (Ingvarson & Rowe, 2008,
and political anxiety, leading to an intense focus on factors believed p. 7). Increased scrutiny of teachers through the development of
to influence the academic achievement of school students (Sellar, value-added models of teacher effectiveness based on student
Thompson, & Rutkowski, 2017). In the 20 years since the OECD’s achievement in national standardised assessment systems, the
Programme for International Student Assessment (PISA) began, establishment of statutory authorities to develop and oversee
attention has shifted from in/equity in student achievement to teacher professional standards and certification, and the rein-
declines in both relative and absolute performance, leading to statement of formal inspections, has followed (Baxter, 2013; Sachs,
2016).
Since 2009, edu-tourismdwhere delegations of politicians,
principals, and teachers visit education systems at the top of the
* Corresponding author. The Centre for Inclusive Education, Faculty of Education,
Victoria Park Road, Kelvin Grove, QLD, 4059, Australia.
international league table at the hope of learning from their suc-
E-mail address: [email protected] (L.J. Graham). cessdhas swung from the “Finnish miracle” to the “Asian tigers” of

https://doi.org/10.1016/j.tate.2020.103190
0742-051X/© 2020 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/
).
2 L.J. Graham et al. / Teaching and Teacher Education 96 (2020) 103190

Singapore, Hong Kong, South Korea and Shanghai (Lingard & Sellar, will need to demonstrate “superior intelligence” in a psychometric
2019). Of interest has been the “Maths Mastery” approach to test and maintain a credit grade point average in order to register as
teaching mathematics, which was introduced and tested in England a teacher upon graduation (NSW Government Department of
through two randomised controlled trials involving 90 primary Customer Service, 2019). Other changesdincluding the increase
schools and 50 secondary schools (Jerrim & Vignoles, 2016). Also in and privileging of professional experience in ITE over theory
making the news headlines are reported differences between sys- (Gale & Parker, 2017) as well as the “stage” approach in the
tems in terms of classroom noise and disorder with the conclusion Australian Professional Standards for Teaching, which ranges from
being that teachers in lower ranked countries, like Australia Graduate, Proficient, Highly Accomplished, and Leaddare predi-
(ranked 70th of 77 countries), are poor at behaviour management cated not only on the assumption that there is a difference between
(Bolton, 2019; Carey, 2019). Blunt analyses of Teaching and Learning graduate and proficient, but that there is a positive relationship
International Survey (TALIS) data indicating higher scores for between experience and teaching quality. All these reforms, how-
experienced versus novice teachers give the impression that early ever, have occurred without the provision of any empirical evi-
career teachers are less effective, despite the caveat that beginning dence to support the claim that beginning teachers are less
teachers are disproportionately allocated to more challenging competent than teachers with more years’ experience, especially in
schools (Organisation for Economic Co-operation and Develop- behaviour management. This may be due to the limited availability
ment, 2019). of empirical evidence and because the evidence that does exist is
Scrutiny has since moved from system-level effectsdsuch as mixed.
inequities in school funding and the concentration of advantage/
disadvantage through the social segregation engineered by 1.1. Evidence of an association between experience and teaching
competitive school marketsdto the effectiveness of teacher prep- quality
aration (Gale & Parker, 2017; Mayer et al., 2017; McMahon, Forde, &
Dickson, 2015). University initial teacher education (ITE) has been At a political level and within public discourse, there is a ten-
criticised for being too theoretical and inadequately preparing dency to presume a straightforward linear relationship between
teachers for the practical realities of the classroom (Teacher teachers’ years of experience and the quality of teaching
Education Ministerial Advisory Group [TEMAG], 2014). This criti- (Brandenburg, McDonough, Burke, & White, 2016). By contrast,
cism is particularly acute with regards to behaviour management research demonstrates a complex relationship between a range of
(Joseph, 2017; O’Neill & Stephenson, 2012), which has been factors that is non-linear and cyclical, whereby experience is one of
branded as ‘ideological’ (Bolton, 2019) without any rigorous many factors influencing the quality of teaching (Brandenburg
investigation of what is being taught in ITE, whether this is really et al., 2016; Klassen & Chiu, 2010). Attempts to research this com-
where the problem lies, or whether beginning teachers are any less plex relationship has resulted in mixed findings (Chingos &
effective at behaviour management than more experienced Peterson, 2011; Rivkin, Hanushek, & Kain, 2005), which have
teachers. The effect of coupling ITE and “teacher quality,” however, been complicated by the range and incommensurability of mea-
is that it frames graduate or beginning teachers as “the problem” sures used in research, as well as inconsistency in the con-
and this, in turn, frames the development of solutions (Mockler, ceptualisation of teaching quality and the nomenclature used to
2018). In other words, a narrow focus on ITE and the graduates it describe categories of experience. For example, the terms “grad-
produces may mean that the true nature and breadth of the uate,” “beginning,” and “early career” are used interchangeably to
problems impacting school education remain undetected and un- describe teachers with any combination of experience of up to five
resolved, while others are magnified beyond their actual or prac- years (Gallant & Riley, 2014; Mockler, 2018; Sullivan, Johnson, &
tical significance. Simons, 2019). Adding further confusion, are studies that catego-
A deficit framing of ITE is particularly prominent in Australia rise teachers as “novice/expert” or “beginner/experienced” without
(Churchward & Willis, 2018), where university teacher education specifying the number of years that constitute these experience
replaced Teachers College in the 1990s and where most practising categories (Palmer, Stough, Burdenski, & Gonzales, 2005). Even in
teachers are now degree qualified. Recent solutions to “the prob- studies where experience categories are defined, there is little
lem” of quality teaching in Australia have therefore focused pre- consistency. For example, Araujo, Carneiro, Cruz-Aguayo, and
dominantly on university teacher education and the quality of ITE Schady (2016) define less experienced or “rookie” teachers as those
graduates (Gore, 2016). Australian universities are now under with 0e3 years’ experience and “experienced” teachers as those
pressure to produce “classroom ready” graduates with greater with more than three years’ experience. The groups in the Araujo
prescription of their course content through formal accreditation et al. study, however, are incommensurable with earlier research
processes (Churchward & Willis, 2018; Gale & Parker, 2017). In using the same observational measure of quality teaching, con-
2016, for example, an external assessment of the literacy and ducted by Cortina, Miller, McKenzie, and Epstein (2015). These re-
numeracy capability of teacher education students was introduced searchers describe “experienced” teachers as those who mentored
by the Australian government as a condition of graduation to student teachers, and “novice” teachers as first year graduates who
ensure “increased confidence in the skills of graduating teachers” taught the same class as their mentor teacher.
(DET, 2016 as cited in McGraw & Fish, 2018, p. 121). A further There are also differences in the conceptualisation of teaching
graduation hurdle for teacher education students, introduced in quality. In some studies, quality is a euphemism for teacher effec-
2018 and linked to the accreditation of university teacher education tiveness, which is predominantly measured indirectly using stu-
programs (Churchward & Willis, 2019), is successful completion of dent test scores. Higher test scores are presumed to indicate the
the Teacher Performance Assessment (TPA) “which uses evidence presence of quality teaching; however, these analyses are driven by
of classroom readiness determined by impact on student learning an “outcome led conception of teaching” (Flores, 2019), tend not to
and is associated with the Australian Professional Standards for define quality, are agnostic on the problem of test score manipu-
Teachers” (p. 252). lation through teaching to the test and/or excluding students with
Further, and as debate continues to rage over the sufficiency or learning difficulties (Lauen & Gaddis, 2016), anddby using aca-
otherwise of university ITE entry scores (Morgan & Aspland, 2018), demic achievement as a proxy measuredignore domain-specific
the government responsible for Australia’s largest education sys- elements of quality teaching that contribute to students’ social
tem in the state of New South Wales has mandated that graduates and emotional development (Hamre, Hatfield, Pianta, & Jamil,
 L.J. Graham et al. / Teaching and Teacher Education 96 (2020) 103190 3

2014). Other studies use direct measures that articulate a specified observation. For example, in North America, Stuhlman and Pianta
range of quality teaching practices and then observe for them. In (2009) investigated the relationship between years of teaching
these studies, teaching quality is conceptualised as a multidimen- experience and teaching quality in 820 Grade 1 classrooms from
sional construct that is theoretically derived from empirical 700 schools across 32 states using the Classroom Observation
research identifying a range of teaching practices that make a System for First Grade (COS-1; NICHD ECCRN, cited in Stuhlman &
positive contribution to students’ emotional, behavioural, and ac- Pianta, 2009); an early prototype of the Classroom Assessment
ademic outcomes (Pianta and Hamre, 2009). Scoring System (CLASS). Teachers’ years of experience in this study
ranged from less than one to 41 years. Four classroom types were
1.1.1. Evidence from studies using indirect measures identified: 1) positive emotional climate/lower academic demand,
The most common indirect measure of teaching quality is stu- 2) high overall quality, 3) mediocre quality, and 4) low overall
dent performance in standardised assessments and findings from quality. No significant differences between the four classroom types
this research are mixed. There is some evidence that teachers’ years were found in relation to teachers’ years of experience.
of experience have little or no impact on student outcomes (Rivkin In another North American study, Schachter, Spear, Piasta,
et al., 2005; Stuhlman & Pianta, 2009), however, there is also evi- Justice, and Logan (2016) investigated the effect of teaching expe-
dence to the contrary (Araujo et al., 2016; Podolsky, Kini, & Darling- rience on language and literacy instruction with 222 early child-
Hammond, 2019). Again though, this evidence is not clear cut, hood educators who had zero to 36 years teaching experience. This
suggesting that there is some effect for experience but that this is team used the Individualizing Student Instruction (ISI) classroom
limited and not cumulative. For example, some research across observation system (Connor, Piasta et al. cited in Schachter et al.,
national and international contexts provides evidence of an early 2016), which was adapted to focus on the language and literacy
effect for experience whereby beginning teachers quickly improve, instruction provided in early childhood education settings. Teacher
but this association declines after their initial adjustment to the years of experience was found to be negatively associated with the
field (Chingos & Peterson, 2011; Klassen & Chiu, 2010; Rivkin et al., amount of instruction, with less experienced teachers achieving
2005; Rockoff, 2004). Similarly, in a large North American study significantly higher scores. A positive relationship between quan-
drawing on data made available by the Florida Department of Ed- tity of instruction, and teaching quality/student learning outcomes
ucation involving 84,031 teachers across Grades 4 to 8, Chingos and was assumed and not investigated in this study.
Peterson (2011) found that teachers generally become more effec- An effect for teachers’ years of experience on student learning
tive in the areas of mathematics and reading after their initial outcomes was found in a large study funded by the Inter-American
transition into the field (particularly after the first year), but the Development Bank involving 24,000 Kindergarten children in
return to experience plateaus or declines after four to five years. In Ecuador (Araujo et al., 2016). Teaching quality was assessed using
another North American study using panel data on 300 teachers the Classroom Assessment Scoring System (CLASS; La Paro, Pianta,
and standardised test scores for 10,000 students, Rockoff (2004) Stuhlman, & Hamre, 2002), in conjunction with assessments of
found a small but positive impact of teacher experience in the children’s math, language, and executive function, conducted at the
first two years of teaching for math computation, and in the first 10 beginning and end of the Kindergarten year. Participating teachers
years of experience for vocabulary and reading comprehension. In a were grouped into two experience categories: (1) inexperienced or
similar study in North America involving data from more than half “rookie” teachers with three years’ experience or less, and (2)
a million students in over 3000 schools across Grades 3 to 7, Rivkin “experienced” teachers, conceptualised as those with more than
et al. (2005) found a positive effect for teaching experience on three years teaching experience. The authors report that Kinder-
student outcomes in mathematics and reading for teachers in their garten children taught by “rookie” teachers achieved test scores
first year of teaching. However, these effects did not continue that were 0.17 of a standard deviation lower than experienced
beyond the initial period of transition into the profession. teachers. However, the study did not compare the CLASS scores of
The evidence for associations between teachers’ years of expe- these two groups and there are anomalies in the report. For
rience and teaching quality from studies using indirect measures example, it appears the authors tried to optimise the variance
(e.g., attitudes, student performance) is by and large mixed and explained in learning outcomes using CLASS and experience. Re-
does not suggest a clear set of conclusions regarding the impor- gressions were run with different definitions of “rookies” (0 years,
tance of experience as a predictor of quality. The literature also has 0e1 years, 0e2 years … 0-N years) and the maximum variance was
addressed questions concerning the returns to teacher experience 0e3 years of experience, which represented a small six percent of
by examining more direct indicators of teaching quality, specifically the sample. In their footnote, the authors report that “the returns to
teachers’ observed classroom behaviours in areas of classroom experience in Ecuador rise sharply in the first three years, and then
management, social support for students, and instruction (Reyes, flatten out” (p. 11). An alternative interpretation of the iterative
Brackett, Rivers, White, & Salovey, 2012; Rucinski, Brown, & regressions, using years of experience, could be that there were
Downer, 2018). Given the oft-cited reports that novice teachers some extreme CLASS scores for teachers with three years of expe-
describe challenges in classroom management (Voss, Wagner, rience, driving the finding with a small sample of “rookie” teachers.
Klusmann, Trautwein, & Kunter, 2017) and research indicating Analyses investigating teacher transience also found that “rookie”
that increasing years of experience can lead to improved perfor- teachers were more likely to move schools, however, no analyses
mance in this domain (Ladd & Sorensen, 2017), it stands to reason were conducted to assess the potential contribution of transience
that examining teachers’ actual classroom interactions could be to lower test scores achieved by the students of those teachers.
useful in attempts to understand the links between experience and Mean CLASS scores were also significantly lower for teachers in the
performance. Moreover, given that interactions with students are Ecuadorian study (3.7) than the cited US sample (4.5) (Araujo et al.,
the most direct and proximal indicator of teachers’ influence on 2016), suggesting that findings from this study should be inter-
students, observations of these interactions are an important in- preted with caution.
dicator of teaching quality. Cortina et al. (2015) compared observation ratings, also using
the CLASS, to the eye movement data of 24 teachers (12 experi-
1.1.2. Evidence from studies using direct observation enced or “supervising” teachers and 12 “novices” with less than one
Only a small number of studies have investigated the link be- years’ experience), teaching the same classes in Grades 2 to 8 in
tween teaching experience and teaching quality using direct North America. The participating teachers wore mobile eye
4 L.J. Graham et al. / Teaching and Teacher Education 96 (2020) 103190

tracking glasses, and their eye movement data indicated the dis- were drawn from seven government primary schools in Queens-
tribution of visual attention across the class. Cortina et al. (2015) land, Australia, with enrolments ranging from approximately
conceptualised teacher quality in classroom monitoring as being 210e760 children. In 2014, the schools had Index of Community
equal distribution across all students, rather than focused attention Socio-Economic Advantage (ICSEA) values between 884 and 977;
on just a few students. According to this criterion, the experienced all below the national mean of 1000.2 The teachers in the 80
teachers scored better in classroom monitoring when compared to classrooms observed were all degree qualified with an average of
the novice teachers, who typically had more focused attention on 9.23 years (SD ¼ 9.27) teaching experience. Teachers’ years of
individual students rather than equal distribution across the whole experience ranged from three weeks through to 38 years. The study
class. However, when correlating the eye movement patterns commenced in 2014 upon receipt of Queensland University of
against CLASS observation ratings, a more complex relationship Technology Human Research Ethics Committee and Queensland
was evident, particularly for feedback quality. Novice teachers who Department of Education approvals.
had high feedback quality on CLASS ratings also demonstrated
greater focused attention on individual students. The novice 2.2. Data collection
teachers who had low feedback quality had a more even distribu-
tion of visual attention and were considered to have good class- 2.2.1. Classroom observations
room monitoring, whereas the experienced teachers maintained Classroom interactions were observed annually using the
their evenly distributed classroom visual monitoring irrespective of Classroom Assessment Scoring System (CLASS), a standardised
high or low feedback quality rating. This finding indicates the observational tool designed to capture and evaluate classroom
complexity of quality teaching and the individual differences in quality from pre-Kindergarten through to 12th Grade (Hamre et al.,
approaching feedback. For example, when beginning as a teacher 2007). It is used in 30-min observation cycles (20 min observation
and providing high feedback quality, there may be a cost to overall plus 10 min scoring) with a minimum of four observational cycles
classroom monitoringdas measured by eye movementsdbut this per classroom. The measure has demonstrated adequate criterion
does not necessarily mean a reduction in overall teaching quality. and predictive validity in studies involving more than 4000 class-
rooms in the United States (Hamre et al., 2007), mainly due to
reliability and objectivity sustained by the mandatory training of
1.2. Summary of evidence across studies
observers (Cortina et al., 2015). In this study, classroom observa-
tions were conducted using the CLASS K-3 measure, incorporating
Across studies there is very limited support for the claim that
Prep to Grade 3. Each of the 80 classrooms was observed for a
beginning teachers are less competent than teachers with more
minimum of 2 h (at least four cycles) by certified observers and the
years of experience. The extant research provides mixed evidence,
first author, a certified CLASS K-3 trainer. In each year of the study,
with some correlations between teaching quality and teacher
the first 10 observation cycles were conducted by all observers and
experience, however these might be conceptualised. However,
moderated to ensure consistency. All observations were completed
these findings are confined to a subset of areas of impact and
within a two-month period in the third term of each school year. In
windows of time. Other research provides no evidence that expe-
the CLASS K-3, classroom interactions (teacher-student, student-
rience makes a difference. Findings from the four studies using
teacher, student-student) are grouped into three domains and 10
direct measures suggest that more experience does not necessarily
dimensions (Table 1).
result in higher quality classroom environments, however, research
For each of the 10 dimensions, ratings based on the observed
using direct measures is costly and is therefore rare. The present
presence or absence of indicators of quality teaching and learning
study aims to contribute to the literature by drawing on stand-
are made on a 7-point scale, with 1e2 representing scores in the
ardised classroom observations of 80 primary school classrooms (K-
low range, 3e5 representing scores in the mid-range, and 6e7
6) and data on years of experience from 80 classroom teachers to
representing scores in the high range. Mean scores are calculated
investigate whether there is any association between experience
for the 10 dimensions and then aggregated to determine mean
and the quality of classroom processes that are important for
scores for each of the three domains.
children’s academic, social, and behavioural outcomes (Hamre,
The CLASS is used annually by Head Start in the United States to
Pianta, Mashburn, & Downer, 2007; Pianta & Hamre, 2009).
measure classroom quality in grantee programs. In 2016, the
domain mean scores for 319 Head Start grantees were 6.00 for
2. Methodology and methods Emotional Support, 5.73 for Classroom Organisation, and 2.83 for
Instructional Support (Head Start Early Childhood Learning &
2.1. Participants Knowledge Center, 2016). Large scale studies have shown that
mean scores tend to be higher in Emotional Support and Classroom
This study analyses standardised classroom observation data Organisation, than Instructional Support (Pianta & Hamre, 2009).
from 18 Preparatory Year teachers, 28 Grade 1 teachers, 20 Grade 2 This same research indicates that a threshold of 5.00 and above is
teachers, and 14 Grade 3 teachers (n ¼ 80).1 Participating teachers needed in both the Emotional Support and Classroom Organisation
domains “to promote positive social development and reduce
problem behaviours,” while a threshold of 3.00 and above in the
1
The teachers reported here are those who participated in both classroom ob- Instructional Support domain is necessary “to achieve greater gains
servations and individual interviews. In the event of a teacher participating mul- in early academic and language skills” (The National Center on
tiple times in the longitudinal study, we used only the first year of data collected for
Quality Teaching and Learning, 2011).
that teacher, to avoid replicating individual scores that may skew analyses, either
positively or negatively. In this way we have prioritised a range of teachers, rather
than repeating individual teacher scores. 2.2.2. Teacher years of experience
2
The ICSEA values are derived from child-level data, which includes parent Participating teachers were asked how long they had been
occupation and education, and community-level data, which includes remoteness teaching and these data were recorded in the main statistical
and percent Indigenous enrolment. With a national mean of 1000 (±100), ICSEA
values range from approximately 500 (schools with children with extremely
dataset. We conducted analyses in two stages guided by the
educationally disadvantaged backgrounds) to approximately 1300 (schools with research literature investigating teaching quality and returns on
children with very educationally advantaged backgrounds). experience, which suggests two critical timepoints of increase,
 L.J. Graham et al. / Teaching and Teacher Education 96 (2020) 103190 5

Table 1
The classroom assessment scoring system (CLASS) domains and dimensions.

Emotional Positive Climate Polite language and listening are aspects of respect within a class. Teachers demonstrate positive regard by smiling, making
Support eye contact, and using children’s names to acknowledge and encourage effort.
Teacher Sensitivity Addressing both academic and emotional needs, teacher sensitivity measures the teacher’s moment-to-moment response to
students cues for support and help.
Negative Climatea Presence of negative attitudes and behaviours that can include sarcasm, criticism, humiliation, hostility, harsh tone, threats,
shouting and physical aggression.
Regard for Student Teachers’ interactions with students and classroom activities should place an emphasis on students’ interests, points of view
Perspectives and independence to encourage student responsibility and autonomy.
Classroom Behaviour Proactive strategies are implemented effectively together with clear expectations, consistency, and clarity of rules with
Organisation Management careful monitoring to keep students on task and to anticipate or redirect and address misbehaviour.
Productivity Efficient management of instructional time, materials and transitions through prior organisation, appropriate pacing, clear
instructions, and use of clear and consistent routines.
Instructional Learning The “hooks” teachers use to engage students, such as a variety of materials and modalities, which allow students to access and
Formats attend to the instructional opportunities given.
Instructional Quality of Feedback Teachers drive learning by intelligently responding to students’ questions, comments and actions, and advance student
Support learning with follow-up questions.
Concept Development Measures the teachers’ use of instructional discussions and activities to promote students’ higher-order thinking skills and
cognition, as well as the teacher’s focus on understanding rather than on rote instruction.
Language Modelling Where teachers intentionally model more advanced, more complex language for students by, for example, introducing,
explaining and using new words or synonyms.
a
Negative climate is reverse scored.

followed by plateau or decline: 0e3 years and 3e5 years, 3. Results

respectively.
In the first stage of analysis, we used two groups to categorise Overall, the 80 participating teachers scored highest in the do-
participating teachers’ experience (beginning: 0e3 years; experi- mains of Emotional Support (M ¼ 5.24, SD ¼ 0.94,
enced: more than 3 years). There were 25 teachers in the 0e3 Range ¼ 2.13e6.81) and Classroom Organisation (M ¼ 4.90,
group, and 55 teachers in the more than 3 years group. SD ¼ 1.24, Range ¼ 1.83e6.75), and lowest in Instructional Support
In the second stage of analysis, we used three groupings: (M ¼ 3.60, SD ¼ 1.11, Range ¼ 1.42e5.92). The distribution of CLASS
beginning: 0e3 years; transitioning: 4e5 years; and experienced: scores is represented in Fig. 1.
more than 5 years. By distinguishing 0e3 years and 4e5 years, our
second analysis reflected two important stages within the 0e5 year
period. The 0e3 year category more accurately reflects beginning or
novice teacher status than would a 0e5 year period and takes into
account research that has identified (i) initial gains for experience
within the first few years post-graduation, and (ii) a plateau or
3.1. Beginning teachers (0e3 years) and experienced teachers (þ3
decline in teacher effectiveness after three years (Chingos &
years)
Peterson, 2011; Klassen & Chiu, 2010; Rivkin et al., 2005; Rockoff,
2004). The 4e5 years category captures a transitional period in
In the ANCOVA models adjusting for grade level taught, there
which this plateau/decline may commence, and when the majority
were no significant differences between the two teacher experi-
of early career teacher attrition is said to occur (Chingos & Peterson,
ence categories (0e3 years and more than 3 years) for any of the
2011; Rivkin et al., 2005). There were 25 teachers in the 0e3 group,
three domains (Emotional Support, Classroom Organisation, and
11 teachers in the 4e5 group, and 44 teachers in the more than 5
Instructional Support) (ps > .05; Table 2). Similarly, when investi-
years group.
gating the individual CLASS dimensions, there were no significant
differences between the beginning and experienced teacher cate-
gories (ps > .05; Fig. 2).

2.3. Data analysis

All statistical analyses were completed using SPSS (Ver 25.0,

www.ibm.com/spss). Descriptive statistics were used to describe
CLASS domains (Emotional Support, Classroom Organisation, and
Instructional Support) and subsequently the dimensions that
comprise each domain. Univariate analysis of covariance was con-
ducted to examine the differences between the CLASS domain
scores for the teacher experience groupings. To adjust for the po-
tential impact of the grade level taught on CLASS scores, grade level
taught was included as a covariate in all analyses. A subsequent
univariate analysis of covariance examined the effects at a dimen-
sion level. These analyses we first conducted with two groups
(beginning: 0e3 years; experienced: more than 3 years), then three
groups (beginning: 0e3 years; transitioning: 4e5 years; and
experienced: more than 5 years). For analysis with the three
groups, pairwise comparisons used Bonferroni correction for mul-
tiple comparisons. Fig. 1. Distribution of CLASS Scores for each domain for participating teachers (n ¼ 80).
6 L.J. Graham et al. / Teaching and Teacher Education 96 (2020) 103190

Table 2
CLASS domains for two group analysis: Beginning and experienced.

CLASS Domain Teacher Experience Category F(1, 77) p

Beginning (0e3 years) n ¼ 25 Experienced (þ3 years) n ¼ 55

M ± SD M ± SD

Emotional Support 5.33 ± 0.97 5.20 ± .94 .123 .727

Classroom Organisation 4.75 ± 1.36 4.96 ± 1.18 2.063 .155
Instructional Support 3.39 ± 1.19 3.71 ± 1.06 2.489 .119

Fig. 2. Comparing mean CLASS scores for Beginning (0e3 years) and Experienced (þ3 Fig. 3. Comparing mean CLASS scores for Beginning (0e3 years), Transitioning (4e5
years) teacher groups for each dimension. There were no statistically significant group years) and Experienced (þ5 years) teacher groups for each dimension. # Negative
differences for any dimension. # Negative Climate is reverse scored. Climate is reverse scored. *p < .05. **p < .01.

3.2. Beginning (0e3 years), transitioning (4e5 years) and Pairwise comparisons for Behaviour Management and Instruc-
experienced teachers (þ5 years) tional Learning Formats dimensions were inconsistent. Pairwise
comparisons for the Behaviour Management dimension were not
3.2.1. CLASS domains significant. Pairwise comparisons for the Instructional Learning
In the ANCOVA models adjusting for grade level taught, there Formats dimension indicated that only the Experienced teacher
was a significant difference between the teacher experience cate- group had scores that were significantly higher than the Tran-
gories for the Classroom Organisation domain (p < .05; Table 3), but sitioning group (p < .04). There were no statistically significant
not the Emotional Support and Instructional Support domains. In differences with the Beginning teacher group for pairwise com-
each domain, regardless of statistical significance, the Beginning parisons for Instructional Learning Formats.
and Experienced teacher groups scored higher than the Tran-
sitioning group.
4. Discussion
3.2.2. CLASS dimensions
The Beginning teacher group had very similar mean scores to This research investigated whether there is any association be-
the Experienced teacher group (þ5 years) in nearly all dimensions tween teachers’ years of experience and teaching quality using
of the CLASS, with the Transitioning group (4e5 years) being CLASS scores from observations of 80 Prep to Grade 3 classroom
relatively lower (Fig. 3). Significant group differences were present teachers. Teachers’ years of experience were initially categorised
in Negative Climate (F (2, 76) ¼ 6.29, p ¼ .003, h2p ¼ 0.14), Behaviour into two groups: Beginning (0e3 years) and Experienced (more
Management (F (2, 76) ¼ 4.51, p ¼ .014, h2p ¼ 0.11), and Instructional than 3 years). Our first analysis found no significant differences
Learning Formats (F (2, 76) ¼ 4.48, p ¼ .014, h2p ¼ 0.11). between Beginning and Experienced teachers on any of the three
Pairwise comparisons for Negative Climate indicated that the domains or 10 dimensions measured by the CLASS. These findings
significant difference was in reference to the Transitioning group of are consistent with previous research using the CLASS that found
teachers. The Transitioning group demonstrated more negative no differences between beginning/novice and experienced teachers
classroom climate behaviours than both the Beginning and Expe- (Stuhlman & Pianta, 2009). However, to account for the develop-
rienced teacher groups (ps  .02). There was no significant differ- mental differences between beginning (or graduate) and early
ence between the Beginning and Experienced teacher groups for career teachers with more years of experience, we divided our
negative classroom climate. sample into three groups: Beginning (0e3 years), Transitioning

Table 3
CLASS domains for three group analysis: Beginning, transitioning and experienced.

CLASS Domain Teacher Experience Category F(2, 76) p h2p

Beginning (0e3 years) n ¼ 25 Transitioning (4e5 years) n ¼ 11 Experienced (þ5 years) n ¼ 44
M ± SD M ± SD M ± SD

Emotional Support 5.34 ± 0.97 4.72 ± 1.39 5.32 ± .76 1.604 .208 .04
Classroom Organisation 4.76 ± 1.36 4.26 ± 1.40 5.14 ± 1.07 3.296 .042 .08
Instructional Support 3.39 ± 1.19 3.18 ± 1.34 3.84 ± .96 2.60 .081 .06
 L.J. Graham et al. / Teaching and Teacher Education 96 (2020) 103190 7

(4e5 years), and Experienced (more than 5 years). In this second CLASS K-3, and the intellectual demand of the Australian Curricu-
analysis, a significant main effect of group was found for Classroom lum in the primary school grades may contribute to higher scores in
Organisation. The Classroom Organisation domain difference re- the Instructional Support domain.
flected differences at the dimension level with Beginning and A similar pattern was evident at the dimension level with no
Experienced teachers outperforming teachers in the Transitional group achieving high range mean scores in any of the 10 CLASS
group in three dimensions: Negative Climate, Behaviour Manage- dimensions. Mean scores for all three groups (Beginning, Tran-
ment, and Instructional Learning Formats.3 sitioning, Experienced) fell below the minimum threshold of 5.00
Neither level of analysis provides evidence to support the claim in two dimensions of Emotional Support (Teacher Sensitivity, Re-
that beginning teachers (with 0e3 years’ experience) are inade- gard for Student Perspectives) and one dimension of Classroom
quately prepared for the classroom and may instead provide Organisation (Instructional Learning Formats). Further, while our
further evidence of a decline in the quality of teaching after the first Experienced (more than 5 years) group scored above the threshold
three years for some teachers (Chingos & Peterson, 2011; Klassen & in Productivity, the Beginning and Transitioning groups each scored
Chiu, 2010; Rivkin et al., 2005; Rockoff, 2004). The removal of initial below. The Transitioning group also scored more poorly in Negative
support structures, together with an increase in workload and re- Climate than the other two groups and was the only group to score
sponsibilities, and emergence of emotional burnout may contribute below the threshold in Behaviour Management. One possibility
to this decline (Buchanan et al., 2013; Weldon, 2018). Previous that should not be overlooked and which is worthy of further
research investigating associations between teaching quality and research is whether cultural differences play a role in the lack of
years of experience has also questioned whether a previously high range mean scores achieved by this sample of Australian
detected post-five-year improvement is associated with early teachers, compared to patterns established in the United States.
career teacher attrition and the departure of less effective early Interestingly, the “enthusiasm” of American teachers and the pos-
career teachers (Chingos & Peterson, 2011; Clotfelter, Ladd, & sibility that some CLASS dimensions, particularly those feeding into
Vigdor, 2006). Taken together, these findings suggest that the the Emotional Support domain, may be “culturally bound” has been
assessment of beginning teachers’ performance could be negatively raised by European researchers (Bihler, Agache, Kohl, Willard, &
impacted by the use of broad experience categories that extend Leyendecker, 2018, p. 3). The possibility of a similar effect of cul-
beyond the initial 0e3 year phase to potentially include what ture should be examined here.
several studies, including this one, have indicated may be a tran- Despite claims that Australian teachers are poor at managing
sitional period and one that may be difficult for some early career student behaviour, Behaviour Management was the highest scoring
teachers. Further research with a larger number of transitional- dimension for experienced teachers and the second highest scoring
stage teachers from a range of different cultural, school phase, dimension, after Positive Climate, for both Beginning and Tran-
and socioeconomic contexts is required before generalising the sitioning teachers with only the Transitioning group scoring below
findings reported here. the threshold of 5. This finding suggests that calls for more pro-
In the meantime, however, our findings do suggest that placing fessional development in behaviour management may be mis-
emphasis on the accreditation, content, and availability of high- directed and is unlikely, by itself, to improve the overall quality of
quality mentoring and ongoing professional learning for all teach- teaching. We say this because teaching is complex and multidi-
ers may be more appropriate than further experiments in ITE mensional. The CLASS observational measure has been designed to
pathways and the destabilisation of university teacher education. take this complexity into account and, as such, there is some
This observation is based on the mean scores achieved by our full interrelatedness between dimensions. For example, low Produc-
sample which, while not representative of all Australian teachers tivity can play into Behaviour Management (Graham, 2018), as lack
and including only primary school teachers, is comparable with of lesson preparation and the absence of smooth transitions can
research conducted using the CLASS internationally. This research lead to students sitting idle which can prompt talking, fidgeting,
points to an established pattern of higher scores in the Emotional and calling out. Some teachers deal with this situation quickly and
Support and Classroom Organisation domains, together with lower effectively to move back into teaching. When observed using the
scores in the Instructional Support domain (Hamre & Pianta, 2009). CLASS, these teachers would score well in Behaviour Management
The distribution of scores in our sample for each of these domains is but lower in Productivity.
also similar to that of Head Start grantees in the United States (Head However, some teachers can get distracted by these emergent
Start Early Childhood Learning & Knowledge Center, 2016), behaviours and respond reactively or ineffectively (Graham, 2018),
although there is greater spread in scores in this study indicating thereby fuelling the instructional vacuum created by low produc-
more individual variability in teaching quality. At the same time, tivity. These teachers would score poorly in both Productivity and
however, our results differ to those of Head Start grantees in that Behaviour Management and, depending on how they attempt to
there were no mean scores in the high range (6e7) in any of the regain control, possibly in Negative Climate as well. Any future
three CLASS domains. Further, mean scores in the Classroom initiatives to improve the quality of teaching should match the
Organisation domain typically fell below the minimum threshold complexity and multidimensionality of this problem and seek to
score (5.00) considered necessary for positive social, academic and address behaviour through the implementation of Positive Behav-
behavioural development (The National Center on Quality Teaching iour Intervention and Supports (PBIS) in schools at a systemic level
and Learning, 2011). By contrast, the mean score of teachers in our as part of a comprehensive Multi-Tier Systems of Support (MTSS)
sample surpassed the minimum threshold of 3.00 for Instructional framework that also supports academic and social emotional
Support, which again differs from the Head Start grantee pattern learning (Bohanon, Gilman, Parker, Amell, & Sortino, 2016; Yeung
where the mean score was 2.83. A possible explanation for this et al., 2016). This is not something that can be achieved through
difference is that Head Start uses the CLASS Pre-K, rather than the changes to ITE alone, although mandating inclusion of at least one
stand-alone unit within the ITE curriculum to enable coverage of
the principles of MTSS and PBIS, as well as evidence-based class-
3
room management practices would be a welcome change that
The Behaviour Management and Instructional Learning Format dimensions
measure teachers’ proactive use of different strategies, modalities and materials to
might prevent this important area of practice being embedded or
engage students and to keep them on task. Negative Climate increases when lost to a crowded curriculum.
teachers’ use punitive methods to achieve control, such as threats or sarcasm. These recommendations are supported by our results which, in
8 L.J. Graham et al. / Teaching and Teacher Education 96 (2020) 103190

finding no additional return to experience and rather some evi- than teachers with more years of experience at either the domain
dence of a decline after three years, align with the bulk of previous or dimension level measured by the CLASS. However, differences
empirical research investigating teachers’ years of experience and were detected for teachers in the Transitioning group (4e5 years’
the quality of teaching. To date, however, this empirical evidence experience) who achieved significantly lower scores in the Class-
appears not to have been consulted in the construction of policies room Organisation domain. At the dimension level, the Tran-
attempting to “vet” potential entrants to the teaching profession sitioning group also scored significantly lower in Negative Climate,
through changes to and greater prescription of university ITE pro- Behaviour Management, and Instructional Learning Format.
grams. Together, these policies and their accompanying rhetoric Contrary to reports suggesting Australian teachers are poor at
have positioned beginning teachers as “the problem” affecting managing behaviour, Behaviour Management was among the
student engagement, achievement and behaviour, and reform of highest scoring dimensions for all three groups, behind only Posi-
university teacher education as “the solution”. However, as there tive Climate for Beginning and Transitioning teachers. The distri-
are now multiple studies using both direct and indirect measures bution of scores for teachers in the full sample was not unlike those
that have found no significant difference in the quality of teaching reported in the international research literature, with higher scores
between beginning teachers and teachers with many more years of in Emotional Support and Classroom Organisation, than Instruc-
experience, continuation of this framing suggests that political tional Support. Unlike some international samples, however, no
calculation takes higher priority in the formulation of education groups achieved mean scores in the high range for any domain or
policy and workforce development than empirical evidence. While dimension of the CLASS. While teachers in this sample typically
attracting, educating and retaining high quality candidates into the achieved scores above the minimum threshold considered neces-
teaching profession is of critical importance, our data suggest that sary for students to make positive gains in the social, emotional,
focusing purely on beginning teachers and their preparation is not and academic domains of learning, the results were not
the only “solution” to teaching quality, and that any future initia- outstanding.
tives should aim to improve the capacity for quality teaching of all The findings from this research suggest that Beginning teachers
teachers, especially those who may be experiencing emotional are doing as well or better than teachers with more years of
burnout. experience, but that the overall quality of teaching could be higher.
As such, better support and professional learning is necessary to
4.1. Limitations improve the quality of teaching for all teachers, not just those at the
beginning of their career. Rather than participate in “formal, one-
A potential limitation in this analysis is the number of teachers size-fits-all ‘in-service’ or conference-style” (Mockler, 2020, p. 4)
(n ¼ 11) in the Transitioning group (4e5 years’ experience). With professional development, it may be that beyond the first years of
the exception of Negative Climate, which is reverse scored and was teaching (0e3 years), and particularly during the 4e5 year transi-
significantly higher for the Transitioning group, the Transitioning tion period, that professional support needs to become more
group typically had lower and more variable CLASS scores than the individualised. For example, identifying those teachers with low to
Beginning and Experienced teachers. Further, as presented in mid-range CLASS scores (4) for targeted professional support that
Table 3, the Transitioning group means typically had the largest is relevant to the individual and to the grade level they teach.
standard deviations, which may indicate individual outliers. This Strengthening this transition may result in greater teacher satis-
warrants further investigation with a larger number of teachers faction and retention, and higher levels of teaching quality as these
with 4e5 years’ experience. teachers develop greater expertise and confidence in their
Another potential limitation is that the CLASS measure was teaching.
developed and normed in the United States. The possibility that To help produce a more responsive, targeted and effective pro-
some CLASS dimensions may be “culturally bound” has been raised fessional learning and support culture for all teachers, not just
in research conducted in Europe where, for example, German those at the beginning of their careers, members of the teaching
teachers scored lower as a group in some dimensions, than Amer- profession should explain to those policy actors with the power to
ican teachers. As a group, Australian teachers may be more make decisions that: (i) teaching is a complex and multidimen-
emotionally reserved than their American counterparts, however, sional activity underpinned by ongoing professional learning and
many individual teachers in this sample also did achieve scores in support, (ii) more years of experience do not by themselves
the high range suggesting that high levels of positivity are possible. translate into higher quality practice, (iii) beginning teachers can
Finally, this study included only primary school teachers using and do engage in quality teaching at the level of more experienced
the CLASS K-3 measure. Recognising that developmental and cur- teachers, and that (iv) some beginning teachers have difficulty
riculum differences have an influence on classroom pedagogy, two transitioning past the initial induction/support period, and that this
other versions of the CLASS were developed for the Upper needs to be identified and addressed sensitively and comprehen-
Elementary (Grades 5 and 6), and Secondary (Grades 7e12) school sively with individually targeted support.
phases. Further research is necessary to assess whether there is a
relationship between teachers’ years of experience and the quality Author contribution
of teaching in the middle and secondary school years.
Linda J. Graham: Conceptualisation, Funding acquisition, Project
5. Conclusion administration, data collection, analysis, writing e review and
original draft preparation. Sonia L. J. White: Data curation, quanti-
This study used standardised classroom observations from 80 tative data analysis, writing e review and original draft prepara-
Queensland primary school teachers teaching Prep to Grade 3 to tion. Kathy Cologon: Writing e review and original draft
examine associations between teachers’ years of experience and preparation. Robert C. Pianta: Analytical strategy, writing and
the quality of teaching. Teaching quality was assessed using the original draft preparation.
Classroom Assessment Scoring System (CLASS) and results were
compared across two, and then three, experience categories based Acknowledgements
on the extant literature. Our analyses found no evidence that
Beginning teachers (0e3 years’ experience) were less competent The first two years of this six-year longitudinal study were
 L.J. Graham et al. / Teaching and Teacher Education 96 (2020) 103190 9

funded by the Australian Financial Markets Foundation for Children Head Start Early Childhood Learning & Knowledge Center. (2016). A national over-
view of grantee CLASS® scores in 2016. Retrieved from https://eclkc.ohs.acf.hhs.
(2013-030). The final four years of the study were supported by the
gov/sites/default/files/pdf/national-class-2016-data.pdf.
Australian Government through the Australian Research Council’s Ingvarson, L., & Rowe, K. (2008). Conceptualising and evaluating teacher quality:
Discovery Projects funding scheme (DP160100319). The views Substantive and methodological issues. Australian Journal of Education, 52(1),
expressed herein are those of the authors and are not necessarily 5e35. https://doi.org/10.1177/000494410805200102
Jerrim, J., & Vignoles, A. (2016). The link between East Asian ‘mastery’ teaching
those of the Australian Government or the Australian Research methods and English children’s mathematics skills. Economics of Education
Council. Review, 50, 29e44. https://doi.org/10.1016/j.econedurev.2015.11.003
Joseph, B. (2017). Getting the most out of Gonski 2.0: The evidence base for school
investments. Sydney, NSW: Centre for Independent Studies. Retrieved from
References https://www.cis.org.au/app/uploads/2017/10/rr31-e.pdf.
Klassen, R. M., & Chiu, M. M. (2010). Effects on teachers’ self-efficacy and job
Araujo, M. C., Carneiro, P., Cruz-Aguayo, Y., & Schady, N. (2016). Teacher quality and satisfaction: Teacher, gender, years of experience and job stress. Journal of
learning outcomes in kindergarten. Quarterly Journal of Economics, 131(3), Educational Psychology, 102(3), 741e756. https://doi.org/10.1037/a0019237
1415e1453. https://doi.org/10.1920/wp.ifs.2016.0916 La Paro, K. M., Pianta, R. C., Stuhlman, M., & Hamre, B. (2002). Classroom assessment
Baxter, J. A. (2013). Professional inspector or inspecting professional? Teachers as scoring system (CLASS). Charlottesville, VA: National Center for Early Develop-
inspectors in a new regulatory regime for education in England. Cambridge ment and Learning. University of Virginia.
Journal of Education, 43(4), 467e485. https://doi.org/10.1080/ Ladd, H. F., & Sorensen, L. C. (2017). Returns to teacher experience: Student
0305764X.2013.819069 achievement and motivation in middle school. Education Finance and Policy,
Bihler, L. M., Agache, A., Kohl, K., Willard, J. A., & Leyendecker, B. (2018). Factor 12(2), 241e279. https://doi.org/10.1162/EDFP_a_00194
analysis of the classroom assessment scoring system replicates the three- Lauen, D. L., & Gaddis, S. M. (2016). Accountability pressure, academic standards,
domain structure and reveals no support for the bifactor model in German and educational triage. Educational Evaluation and Policy Analysis, 38(1),
preschools. Frontiers in Psychology, 9, 1232. https://doi.org/10.3389/ 127e147. https://doi.org/10.3102/0162373715598577
fpsyg.2018.01232 Lingard, B., & Sellar, S. (2019). Schooling reform in Australia: Legitimation through
Bohanon, H., Gilman, C., Parker, B., Amell, C., & Sortino, G. (2016). Using school “projections” onto Shanghai and east asian schooling systems. In F. Waldow, &
improvement and implementation science to integrate multi-tiered systems of G. Steiner-Khamsi (Eds.), Understanding PISA’s attractiveness: Critical analyses in
support in secondary schools. Australasian Journal of Special Education, 40, comparative policy studies (pp. 49e64). London, UK: Bloomsbury Publishing.
99e116. https://doi.org/10.1017/jse.2016.8 Mayer, D., Dixon, M., Kline, J., Kostogriz, A., Moss, J., Rowan, L., & White, S. (2017).
Bolton, R. (2019, July 25). Ideology crushes teachers’ ability to control classes. Studying the effectiveness of teacher education. In D. Mayer, M. Dixon, J. Kline,
Australian Financial Review. Retrieved from https://www.afr.com/policy/health- A. Kostogriz, J. Moss, & L. Rowan (Eds.), Studying the effectiveness of teacher
and-education/ideology-crushes-teachers-ability-to-control-classes-20190724- education (pp. 13e26). Singapore: Springer. https://doi.org/10.1007/978-981-
p52a6s. 10-3929-4_2.
Brandenburg, R., McDonough, S., Burke, J., & White, S. (Eds.). (2016). Teacher edu- McGraw, A., & Fish, T. (2018). Selection and rejection in teacher education: Qualities
cation: Innovation, intervention and impact. Springer. https://doi.org/10.1007/ of character crucial in selecting and developing teacher education students.
978-981-10-0785-9. Asia-Pacific Journal of Teacher Education, 46(2), 120e132. https://doi.org/
Buchanan, J., Prescott, A., Schuck, S., Aubusson, P., Burke, P., & Louviere, J. (2013). 10.1080/1359866x.2017.1355048
Teacher retention and attrition: Views of early career teachers. Australian McMahon, M., Forde, C., & Dickson, B. (2015). Reshaping teacher education through
Journal of Teacher Education, 38, 112e129. https://doi.org/10.14221/ the professional continuum. Educational Review, 67(2), 158e178. https://
ajte.2013v38n3.9 doi.org/10.1080/00131911.2013.846298
Carey, A. (2019, December 4). Australian students ‘among the worst in the world’ for Mockler, N. (2018). Early career teachers in Australia: A critical policy historiogra-
class discipline. The Sydney Morning Herald. Retrieved from https://www.smh. phy. Journal of Education Policy, 33(2), 262e278. https://doi.org/10.1080/
com.au/politics/federal/australian-students-among-the-worst-in-the-world- 02680939.2017.1332785
for-class-discipline-20191204-p53gwk.html. Mockler, N. (2020). Teacher professional learning under audit: Reconfiguring practice
Chingos, M. M., & Peterson, P. E. (2011). It’s easier to pick a good teacher than to in an age of standards. Professional Development in Education. https://doi.org/
train one: Familiar and new results on the correlates of teacher effectiveness. 10.1080/19415257.2020.1720779
Economics of Education Review, 30(3), 449e465. https://doi.org/10.1016/ Morgan, A., & Aspland, T. (2018, September 8). Viewpoints: Should universities raise
j.econedurev.2010.12.010 the ATAR required for entrance into teaching degrees? The conversation. Retrieved
Churchward, P., & Willis, J. (2019). The pursuit of teacher quality: Identifying some from https://theconversation.com/viewpoints-should-universities-raise-the-
of the multiple discourses of quality that impact the work of teacher educators. atar-required-for-entrance-into-teaching-degrees-102841.
Asia-Pacific Journal of Teacher Education, 47(3), 251e264. https://doi.org/ Nsw Government Department of Customer Service. (2019, September 3). Ensuring
10.1080/1359866X.2018.1555792 excellence new standards for graduate teachers. Retrieved from https://www.
Clotfelter, T., Ladd, H. F., & Vigdor, J. L. (2006). Teacher-student matching and the nsw.gov.au/your-government/the-premier/media-releases-from-the-premier/
assessment of teacher effectiveness. Cambridge, MA: National Bureau of Eco- ensuring-excellence-new-standards-for-graduate-teachers/.
nomic Research Working Paper No. 11936. https://doi.org/10.3386/w11936 Organisation for Economic Co-operation and Development. (2019). September 10).
Cortina, K. S., Miller, K. F., McKenzie, R., & Epstein, A. (2015). Where low and high Review education policies: Learning time and disciplinary climate. Retrieved from
inference data converge: Validation of CLASS assessment of mathematics in- https://gpseducation.oecd.org/revieweducationpolicies/#!node¼41718&amp;
struction using mobile eye tracking with expert and novice teachers. Interna- filter¼all.
tional Journal of Science and Mathematics Education, 13(2), 389e403. https:// O’Neill, S., & Stephenson, J. (2012). Does classroom management coursework in-
doi.org/10.1007/s10763-014-9610-5 fluence pre-service teachers’ perceived preparedness or confidence? Teaching
Flores, M. A. (2019). Unpacking teacher quality: Key issues for early career teachers. and Teacher Education, 28(8), 1131e1143. https://doi.org/10.1016/
In A. Sullivan, B. Johnson, & M. Simons (Eds.), Attracting and keeping the best j.tate.2012.06.008
teachers (pp. 15e38). Singapore: Springer. Palmer, D. J., Stough, L. M., Burdenski, T. K., Jr., & Gonzales, M. (2005). Identifying
Gale, T., & Parker, S. (2017). The prevailing logic of teacher education: Privileging the teacher expertise: An examination of researchers’ decision making. Educational
practical in Australia, England and Scotland. In M. A. Peters, B. Cowie, & Psychologist, 40(1), 13e25.
I. Menter (Eds.), A companion to research in teacher education (pp. 521e535). Pianta, R. C., & Hamre, B. K. (2009). Conceptualization, measurement, and
Singapore: Springer. https://doi.org/10.1007/978-981-10-4075-7_35. improvement of classroom processes: Standardized observation can leverage
Gallant, A., & Riley, P. (2014). Early career teacher attrition: New thoughts on an capacity. Educational Researcher, 38(2), 109e119. https://doi.org/10.3102/
intractable problem. Teacher Development, 18(4), 562e580. https://doi.org/ 0013189X09332374
10.1080/13664530.2014.945129 Podolsky, A., Kini, T., & Darling-Hammond, L. (2019). Does teaching experience
Gore, J. M. (2016). Reform and the reconceptualisation of teacher education in increase teacher effectiveness? A review of US research. Journal of Professional
Australia. In R. Brandenburg, S. McDonough, J. Burke, & S. White (Eds.), Teacher Capital and Community, 4(4), 286e308. https://doi.org/10.1108/JPCC-12-2018-
education (pp. 15e34). Singapore: Springer. https://doi.org/10.1007/978-981- 0032
10-0785-9_2. Reyes, M. R., Brackett, M. A., Rivers, S. E., White, M., & Salovey, P. (2012). Classroom
Graham, L. J. (2018). Student compliance will not mean “all teachers can teach”: a emotional climate, student engagement, and academic achievement. Journal of
critical analysis of the rationale for “no excuses” discipline. International Journal Educational Psychology, 104(3), 700. https://doi.org/10.1037/a0027268
of Inclusive Education, 22(11), 1242e1256. https://doi.org/10.1080/ Rivkin, S. G., Hanushek, E. A., & Kain, J. F. (2005). Teachers, schools, and academic
13603116.2017.1420254 achievement. Econometrica, 73(2), 417e458.
Hamre, B. K., Hatfield, E, Pianta, R, & Jamil, F (2014). Evidence for general and Rockoff, J. E. (2004). The impact of individual teachers on student achievement:
domain-specific elements of teacherechild interactions: Associations with Evidence from panel data. The American Economic Review, 94(2), 247e252.
preschool children’s development. Child Development, 85(3), 1257e1274. https://doi.org/10.1257/0002828041302244
https://doi.org/10.1111/cdev.12184 Rucinski, C. L., Brown, J. L., & Downer, J. T. (2018). Teacherechild relationships,
Hamre, B. K., Pianta, R. C., Mashburn, A. J., & Downer, J. T. (2007). Building a science of classroom climate, and children’s social-emotional and academic development.
classrooms: Application of the CLASS framework in over 4,000 US early childhood Journal of Educational Psychology, 110(7), 992. https://doi.org/10.1037/
and elementary classrooms. New York: Foundation for Child Development. edu0000240
10 L.J. Graham et al. / Teaching and Teacher Education 96 (2020) 103190

Sachs, J. (2016). Accountability, standards and activism: A challenge or opportunity Teacher Education Ministerial Advisory Group. (2014). Action now: Classroom ready
for teacher education. In C. Day, & J. C. Lee (Eds.), Quality and change in teacher teachers. Canberra, Australia: Australian Government, Department of Education.
education: Western and Chinese perspectives (pp. 251e262). Switzerland: Retrieved from https://docs.education.gov.au/system/files/doc/other/action_
Springer International Publishing. https://doi.org/10.1007/978-3-319-24139-5_ now_classroom_ready_teachers_print.pdf.
15. The National Center on Quality Teaching and Learning. (2011). Understanding and
Schachter, R. E., Spear, C. F., Piasta, S. B., Justice, L. M., & Logan, J. A. (2016). Early using the CLASS for program improvement. Retrieved from http://prekkid.org/
childhood educators’ knowledge, beliefs, education, experiences, and children’s head-start/agency/training-and-tech-assistance/Class%20Overview%
language- and literacy-learning opportunities: What is the connection? Early 20Materials/NCQL%20Understanding%20and%20Using%20the%20CLASS%20%
Childhood Research Quarterly, 36, 281e294. https://doi.org/10.1016/ 5BHappy%20Tsugawa-Banta%5D.pdf.
j.ecresq.2016.01.008 Voss, T., Wagner, W., Klusmann, U., Trautwein, U., & Kunter, M. (2017). Changes in
Scholes, L., Lampert, J., Burnett, B., Comber, B. M., Hoff, L., & Ferguson, A. (2017). The beginning teachers’ classroom management knowledge and emotional
politics of quality teacher discourses: Implications for pre-service teachers in exhaustion during the induction phase. Contemporary Educational Psychology,
high poverty schools. Australian Journal of Teacher Education, 42(4), 19e43. 51, 170e184. https://doi.org/10.1016/j.cedpsych.2017.08.002
https://doi.org/10.14221/ajte.2017v42n4.3 Weldon, P. (2018). Early career teacher attrition in Australia: Evidence, definition,
Sellar, S., Thompson, G., & Rutkowski, D. (2017). The global education race: Taking the classification and measurement. Australian Journal of Education, 62(1), 61e78.
measure of PISA and international testing. Edmonton, Alberta: Brush Education. https://doi.org/10.1177/0004944117752478
Stuhlman, M. W., & Pianta, R. C. (2009). Profiles of educational quality in first grade. Yeung, A. S., Craven, R. G., Mooney, M., Tracey, D., Barker, K., Power, A., & Lewis, T. J.
The Elementary School Journal, 109(4), 323e342. https://doi.org/10.1086/593936 (2016). Positive behavior interventions: The issue of sustainability of positive
Sullivan, A., Johnson, B., & Simons, M. (Eds.). (2019). Attracting and keeping the best effects. Educational Psychology Review, 28(1), 145e170. https://doi.org/10.1007/
teachers: Issues and opportunities. Singapore: Springer. s10648-015-9305-7