ARTICLE IN PRESS

Teaching and Teacher Education 21 (2005) 729–741

www.elsevier.com/locate/tate

Teachers’ classroom discipline and student misbehavior in

Australia, China and Israel
Ramon Lewisa,, Shlomo Romib, Xing Quic, Yaacov J. Katzb,d
a
Institute for Education, La Trobe University, Melbourne, Victoria, Australia
b
School of Education, Bar-Ilan University, Ramat-Gan, Israel
c
Department of Psychology, Sichuan College of Education, Sichuan Province, Chengdu. 610041, PR China
d
Ministry of Education Culture & Sport and School of Education, Bar-Ilan University, Ramat-Gan 52900, Israel

Abstract

This paper reports students’ perceptions of the classroom discipline strategies utilized in Australia, China and Israel.
It examines data from 748 teachers and 5521 students to identify how teachers’ use of various disciplinary strategies,
and the extent to which these relate to student misbehavior, differ in three national settings. In general, Chinese teachers
appear less punitive and aggressive than do those in Israel or Australia and more inclusive and supportive of students’
voices. Australian classrooms are perceived as having least discussion and recognition and most punishment. In all
settings greater student misbehavior relates only to increased use of aggressive strategies. Implications are discussed.
r 2005 Elsevier Ltd. All rights reserved.

Keywords: Classroom management; Teacher–student relationship; Teacher behavior

1. Introduction additional 33 percent suggesting it was somewhat

The issue of how best to discipline students in
classrooms is of continuing interest and concern to
the community. For example, in 2002, as in nearly
all preceding years, students’ lack of discipline
ranked within the first two most serious problems
confronting the public schools in the annual Phi
Delta Kappa Polls of the public’s attitudes toward
the public schools (Lowell & Gallup, 2002). In
2002, 43 percent of respondents rated students’
lack of discipline as a very serious concern, with an
Corresponding author. Tel.: +61 3 95258482.
serious. Of all the school-related factors capable of
influencing student responsibility in classrooms,
teachers’ discipline strategies, the focus of this
paper, are among the most potent (Ingersoll, 1996;
Lewis, 1997a).
Ensuring that students behave responsibly in
classrooms is important for two independent
reasons. First, it serves as a means of preparing
students to take their place in society as respon-
sible citizens, an aim of primary importance to
schooling (Rothstein, 2000). Secondly, without
satisfactory levels of student responsibility, the
best planned and potentially most engaging

0742-051X/$ - see front matter r 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tate.2005.05.008
 ARTICLE IN PRESS
730 R. Lewis et al. / Teaching and Teacher Education 21 (2005) 729–741
lessons may fail to have the desired impact. Often
it may only require a small proportion of students
to misbehave and they become sufficiently dis-
tracting to students and frustrating to teachers
that the most carefully planned lesson fails to
promote effective learning among the students
(Barton, Coley, & Wenglinsky, 1998). This paper
examines the relationship between discipline pro-
cesses and student misbehavior in three different
national settings, namely Australia, China and
Israel.
Interest in classroom discipline relates not only
to the good it can do but also to the damage
inappropriate discipline can cause. For example,
two recent publications emphasize the potential
negative impact of particular discipline strategies.
The first conjectures that
Unnecessarily harsh and punitive disciplinary
practices against students create a climate that
contributes to school violence. This issue is little
recognized and scarcely researched. (Hyman &
Snook, 2000, p. 489).
The second publication reports the perceptions
of over 3500 school students in Australia (Lewis,
2001). This study demonstrates empirically that, in
the view of these students, their teachers are
characterized by two distinct discipline styles.
The first of these was called ‘‘coercive’’ discipline
and comprised Punishment and Aggression (yell-
ing in anger, sarcasm, group punishments, etc.).
The second style, comprising Discussion, Hints,
Recognition, Involvement and Punishment, was
called ‘‘Relationship based discipline’’. After pre-
senting a thorough data analysis the report
concludes:
Students who receive more Relationship based
discipline are less disrupted when teachers deal
with misbehavior and generally act more
responsibly in that teacher’s class. In contrast,
the impact of Coercive discipline appears to be
more student distraction from work and less
responsibility. (p. 315).
These findings appear consistent with those of
Miller, Ferguson, and Byrne (2000) who, after
examining students’ perceptions of what factors
cause classroom misbehavior, highlight the poten-
tial for teacher behavior such as shouting all the
time, unfairly blaming students, picking on kids,
and being rude, to stimulate student resistance and
subsequent misbehavior.
The present study has shown pupils to attribute
to teachers a significantly greater responsibility
for pupil misbehavior than that they attribute
to parents. (p. 93).
It is of interest to note that in an earlier Israeli
study which also examined factors seen to be
causing student misbehavior, it was reported that
although students placed their teachers’ attitude
the second most important cause of student
misbehavior, teachers ranked it as 20 out of 26
(Guttmann, 1982).
The impetus for the research discussed in this
paper came after the publication of a study of the
relationship between classroom discipline and
student responsibility and misbehavior in Austra-
lia (Lewis, 2001). As a result of that publication,
academics in a number of countries contacted the
author with requests to replicate the study in their
respective national settings. Two additional set-
tings were ultimately selected. These were China
and Israel. Australia is a typically western country,
China is a typically oriental country and Israel is
approximately half western and half oriental.
Consequently, the use of these three cultural
settings provided the opportunity for some robust
comparisons. Further, since the academics from
these settings who expressed interest in under-
taking the research were senior, very experienced,
and involved in teacher training for many years,
the issue of classroom discipline, and the research
design of the previous research appeared to have
valency in both countries.
In summary, there were two main foci for the
research. First it permitted a comparison of the
extent of usage of various discipline strategies in
three significantly different national settings.
Secondly, the relationship between student mis-
behavior and classroom discipline could be exam-
ined in each setting. In conducting this research, it
was acknowledged that there are likely to be
cultural differences associated with styles of
discipline. For example, differences of opinion
between Chinese and other Americans (Mitchel,
 ARTICLE IN PRESS
R. Lewis et al. / Teaching and Teacher Education 21 (2005) 729–741
2001) appear to focus on the relative virtues of
submission to authority versus the child’s right to
be assertive and individualistic.
The significance of classroom discipline rests not
only on its impact on students’ behavior and
learning, as outlined above. The ability of teachers
to effectively discipline students is, according to
McCormick and Shi (1999), integrally related to
teachers’ sense of professional adequacy. This
finding appears consistent with the work of God-
dard (2000), who reports the results of a study of
233 teachers’ views of their role. He notes that
‘disciplinarian’ was the third most commonly cited
metaphor provided by teachers for their work,
ranking only behind ‘leader’ and ‘knowledge
dispenser’. It is not surprising therefore that any
failure on teachers’ part to satisfactorily manage
students’ classroom misbehavior can result in
stress, and in the extreme case, burnout. Overall,
classroom discipline is a well-documented source
of teacher stress (Blase,1986; Borg, Riding, &
Falzon, 1991; DeRobbio & Iwanicki, 1996; Fried-
man, 1995; Keiper & Busselle, 1996; Kyriacou,
1980). Some results however may be tenuous as
teachers experiencing stress as a result of other
factors (for example excessive workload) may
perceive student behavior more negatively (White-
man, Young, & Fisher, 1985) and therefore inflate
its significance as a stressor. Nevertheless, disci-
pline issues rate consistently among the strongest
of teacher stressors.
Chan (1998), reporting on the stressors of over
400 teachers in Hong Kong, notes that student
behavior management rates as the second most
significant factor stressing teachers. More signifi-
cantly, perhaps, Ingersoll (2001) studied approxi-
mately 6700 teachers in the US and states that
approximately 30 percent of the 400 or so who
chose to leave the profession identify student
discipline as one of the reasons that caused them
to give up teaching. It needs to be noted that for
some teachers, the stress associated with classroom
discipline relates as much to how they are treated
by other teachers and members of the school
administration as it does to their treatment at the
hands of students (Martin, 1994).
Related research shows that it is not only the
stress arising from ineffective discipline that is a
731
cause for concern. The way teachers attempt to
cope with their concerns may add to their stress
levels rather than lower them. For example, a
recent study demonstrates that teachers who are
more worried about discipline and student mis-
behavior report greater use of six maladap-
tive coping strategies, namely, wishful thinking,
don’t tell anyone, self-blame, worry, disruption
to sleeping or eating patterns and getting sick
(Lewis, 1999). These strategies are among those
recently shown to be predictive of a range of
non-productive outcomes ranging from low self
concept to depression (Frydenberg & Lewis,
2000). Clearly, the issue of classroom discipline is
of as much significance to teachers as it is to
students.
Classroom misbehavior and discipline is a topic
that has been studied in Australia (for example,
Lewis, 2001; Lewis & Lovegrove, 1987; Oswald,
1995), China (for example, Gao, 1998; Jin &
Cortazzi, 1998; Peng, 1993) and Israel (for
example, Friedman, 1994; Kaplan, Gheen, &
Midgley, 2002; Romi & Freund, 1999). However,
this is the first attempt to systematically examine
students’ perceptions of their teachers’ disciplinary
strategies and their misbehavior in these three
national settings.
2. Measuring classroom discipline
To measure classroom discipline in three na-
tional settings was problematic. Conceptualization
of classroom discipline strategies in one setting
could not necessarily be assumed equivalent to
those in the other two. It would have been possible
to utilize exploratory factor analysis on data sets
from respective nations to obtain assessments of
discipline most appropriate to each setting. It
would even have been possible to utilize confirma-
tory factor analyses and compare goodness of fit
measures to examine the extent to which one
particular measure applied equally to all three
national settings. However, had students’ views of
classroom discipline provided different measures,
comparisons would have been prohibited. Conse-
quently, since as stated above it was the aim of this
research to undertake a replication, care was taken
 ARTICLE IN PRESS
732 R. Lewis et al. / Teaching and Teacher Education 21 (2005) 729–741
to ensure that the same measures were utilized in
each national setting.
To examine the relationship between discipline
strategies and student misbehavior, two steps were
taken. First, the 35 items on the questionnaire used
in the earlier Australian study were studied by the
colleagues in China and Israel to assess their
relevance to their respective students. As a result,
11 items were omitted and 24 were retained. These
24 items were agreed by the researchers in each
country to assess the six discipline strategies
reported in the previous Australian study and to
be of cultural relevance, albeit to differing degrees,
in their respective national settings. The strategies
measured were Punishing, Rewarding, Involve-
ment in decision-making, Hinting, Discussion and
Aggression. Examination of a number of discipline
texts (Charles, 2004; Lewis, 1997b; Tauber, 1999;
Wolfgang, 1995) indicated that one or more of
these strategies were seen as underlying most of the
available approaches to classroom discipline. For
example, Reward and Punishment related to
Interventionist approaches, Involvement to Inter-
actional approaches and Hinting and Discussion
to Non-Interventional Approaches. Although Ag-
gression was not recommended in any text, it was
seen as potentially possible in classrooms in all
settings and necessary to include given its sig-
nificance in previous research (Lewis, 2001).
The following four items were designed to assess
Punishment.

Gives out consequences to students who mis-
behave (e.g. move their seats, detention).

Increases the level of consequence if students
will not do as they are told (e.g. move seats,
detention).

Increases the level of consequence if a misbe-
having student argues.

Increases the level of consequence if a misbe-
having student stops when told, but then does it
again.
The four items developed to assess Recognition
and rewards were as follows.

Rewards individual students who behave prop-
erly.

Praises the class for good behavior.

Praises individual students for good behavior.

Rewards the class when students behave well.
The following two sets of items determine the
extent to which teachers attempted to include
students’ voice in the decision-making related to
discipline.
The first four relate to an emphasis on the class
as the determiner of the discipline process.

Organizes the class to work out the rules for
good behavior.

Decides with the class what should happen to
students who misbehave.

Makes students leave the room until they decide
to behave properly.

Lets students know that the way they are
behaving is not how the class expects them to.
The next four items provide for the voice of
individual students.

Discusses students’ behavior with them to allow
them to figure out a better way to behave in
future.

Lets students talk about their side of things so
that it can be clearly understood.

Gets students to understand why their behavior is
a problem for others by discussing it with them.

Gets students to change the way they behave by
helping them understand how their behavior
affects others.
The next set of four items relates to a strategy
that usually precedes more formal intervention by
the teacher, and provides students with awareness
that in the teacher’s eyes, all is not as it should be.
It probably also communicates some level of trust
that students will self-regulate their behavior. The
following four items were designed to assess this
process which was called Hinting.

Describes what students are doing wrong, and
expect them to stop.

Asks students questions like ‘‘What are you
doing?’’ to get them to think about how to
behave better.
 ARTICLE IN PRESS
R. Lewis et al. / Teaching and Teacher Education 21 (2005) 729–741

Reminds misbehaving students about the class
rules.

Describes how students are misbehaving to
make them decide whether to stop or not.
The final four items were written to permit
measurement of teacher Aggression. This strategy
was defined as the use of strategies which, while
legal, may in some settings negate the student’s
sense of well-being and possession of natural
rights.

Yells angrily at students who misbehave.

Deliberately embarrasses students who misbe-
have.

Keeps the class in because some students
misbehave.

Makes sarcastic comments to students who
misbehave.
In summary therefore, the six classroom disci-
pline strategies examined in the previous study
(Lewis, 2001) were assessed by a total of 24
questionnaire items.
To enable the collection of data without
identifying any individual teacher by name, ques-
tionnaires specified one of six subject areas taught
(for example Science, Mathematics, English (Chi-
nese, Hebrew), Social Studies). Students were then
requested to concentrate on that one class and the
teacher who teaches it when completing the
questionnaire.
To measure the extent of student misbehavior,
students were asked to indicate ‘‘How often do
you misbehave in this teacher’s class?’’. To
respond they chose from the alternatives, Almost
never, Only a little, Sometimes and Often, which
were coded 1–4, respectively.
The questionnaire was translated into both
Chinese and Hebrew and in each case back-
translated into English to ensure accuracy. To
respond to the discipline items, students indicated
on a 6-point scale how frequently the teacher acted
as described in the statement ‘‘when trying to deal
with misbehavior’’. The response alternatives
provided, namely, Nearly always, Most of the
time, A lot of the time, Some of the time, Hardly
ever and Never were coded 6–1, respectively.
733
Students at grade levels 7–12 completed these
items documenting their perception of their
teacher’s use of each classroom discipline stra-
tegy. The introduction to the questionnaire was
brief and indicated that the questions to follow
focused on ‘‘classroom discipline and how you feel
about it’’. There was no indication as to the
research questions being addressed. Written ex-
planation of response formats were provided as
required.
3. Sample
The three purposive samples utilized in this
study, and described below, were restricted to
students attending years 7–12 at coeducational
schools. Although representative sampling was not
attempted, care was taken when selecting partici-
pating schools to ensure that the sample included
both larger and smaller schools, situated in a range
of socioeconomic and geographic areas. In addi-
tion, schools which appeared ‘atypical’ were not
included, for example extremely large, small or
isolated schools, or schools which were selective in
intake. In Australia, all secondary schools (years
7–12) in the north eastern region of Victoria and a
small number in the Melbourne metropolitan
region were invited to participate in the study.
The response rate of 70 percent reflects the
importance attributed to the topic of classroom
discipline in secondary schools.
In Israel, a sample of four high schools (years
10–12) and eight junior high schools (years 7–9) in
the geographic center of Israel were invited to
participate in the study. All accepted.
In China, the sample of teachers and students
was drawn from eight schools in Chengdu region,
Sichuan Province. Two of these were lower
secondary schools (years 7–9) and the remainder
comprised years 7–12. In each Chinese and Israeli
school, a random sample of classes at all year
levels were selected. As a research assistant
administered questionnaires to these classes their
teachers completed their questionnaires.
Table 1 records the number of teachers, and
students at year levels 7–8, 9–10 and 11–12 for
Australia, Israel and China, respectively.
 ARTICLE IN PRESS

734 R. Lewis et al. / Teaching and Teacher Education 21 (2005) 729–741

Table 1 edged that the data could only provide a general

Study sample indication of the extent of the issue.
Country Respondents
The second question asked ‘‘To what extent is
the issue of classroom discipline and student
Teachers Students misbehavior an issue of concern to you?’’ To
answer, teachers selected either Of major concern,
Year 7–8 Year 9–10 Year 11–12 Total
Of moderate concern, Of minor concern, Of
Australia 491 1713 1624 846 4183 almost no concern or Of no concern. These
Israel 98 261 334 241 836 alternatives were coded 5–1, respectively. Tables
China 159 159 147 196 502
2 and 3 report the relevant data for teachers from
the three national settings.
Inspection of the w2 figures for Table 2
(w2 ð16Þ ¼ 20:191, p ¼ :212) shows there are no
Of the total of 5521 students, 48 percent are statistically significant differences in the perceived
males, although the percentage of males varies levels of classroom misbehavior in Australia, Israel
from 38 to 60 percent depending on year level and China.
within country. The gender distribution of the 748 In contrast, initial inspection of the data in
teachers is less evenly divided. There were 11, 42 Table 3 would seem to show that teachers in Israel
and 46 percent of males, respectively in Israel, report significantly more concern about classroom
Australia and china. It is of interest to note the misbehavior than do teachers from either Austra-
small proportion of men in the Israeli sample. This lia or China. However, the substantial differences
is not atypical however, as noted in previous were due to another cause. Although great care
research (Romi & Katz, 2003), and reflects the was taken to ensure that translation of all terms on
predominance of women in the teaching force in the questionnaire were accurate, an error occurred
non-religious schools in Israel. for this question. In the Hebrew version of the
questionnaire the word ‘‘concern’’ was translated
as ‘‘importance’’. This discrepancy was not noted
after back-translation to English. Therefore,
4. Findings
Table 2
Prior to comparing the extent to which the Level of perceived classroom misbehavior by setting
various discipline strategies are utilized in different
national settings and examining the relationship Country Level of perceived misbehavior
between various types of discipline perceived by Nearly all Most Some Hardly any None
students and classroom misbehavior, it was
considered helpful to document how significant Australia 1 2 33 54 10
Israel 1 2 54 41 2
an issue classroom discipline is for the teachers of
China 0 2 42 50 6
the students responding to the survey. To gain
such information, a 10 percent sample of the
teachers at the selected schools answered two Table 3
questions as part of the survey. The first required Level of concern by setting
them to state ‘‘How many students misbehave in
the first class you would normally be teaching next Country Level of concern
Monday?’’ Five alternative responses were pro- Major Moderate Minor Almost none None
vided, namely Nearly all, Most, Some, Hardly any
and None. These responses were coded 5–1, Australia 12 27 25 24 12
Israel 90 10 0 0 0
respectively. Since what constituted ‘‘misbehavior’’
China 11 26 30 25 8
was not defined or exemplified, it was acknowl-
 ARTICLE IN PRESS

R. Lewis et al. / Teaching and Teacher Education 21 (2005) 729–741 735

rather than reporting to what extent the issues of
classroom discipline and student misbehavior was
an issue of concern to them, the Israeli teachers
indicated the extent it was an issue of importance.
The comparison of the Australian and equivalent
Chinese data showed no significant differences
(w2 ð4Þ ¼ 2:774, p ¼ :596).
It is of interest to note that in a recent study 294
Australian teachers were asked to indicate their
levels of concern related to any inability to
discipline classrooms as they would prefer. In
reporting these data Lewis (1999, p. 162) states ‘‘it
can be argued that the gap between best and
current discipline practice was, on average, of only
moderate concern to these teachers.’’ It would
appear therefore that the 2 sets of data, collected
over 3 years apart, from independent samples of
teachers, are very consistent.
As stated above, the data in Tables 2 and 3,
which provide a general framework for the
subsequent discussion on classroom discipline
strategies used in classrooms, show that a sub-
stantial group of teachers in two independent
national settings report at least moderate levels of
concern over student misbehavior in class. In
addition, the Israeli teachers sampled clearly
highlight the significance of the topic. It is likely
that part of the teachers’ interest in classroom
discipline is based on their perception that, on
average, a little fewer than ‘‘some’’ but more than
‘‘hardly any’’ students in the classes they expect to
teach are likely to misbehave.
To document the reliability of the measures
of teachers’ discipline strategies, the students’
responses to the sets of four items were considered
for internal consistency in each national setting,
Cronbach a coefficients were computed for each
set. If the deletion of any item from a set increased
the magnitude of the respective a, that item was to
be removed. The reliance on a did not reflect a
belief that high values for internal consistency of
responses within a set of items was essential to
justify the value of a scale. It was assumed, given
the diversity of ways in which some strategies may
be operationalized in different national settings,
that increased likelihood of one behavior in a set
of four may only be a weak predictor of the
utilization of those comprising the remainder of
the set. Nevertheless, items would be treated as a
scale only if each item added shared explanatory
variance to the measure.
The only scale for which any items were
excluded was Involvement. In both Australia and
Israel, the removal of the item ‘‘Makes students
leave the room until they decide to behave
properly’’ improved the magnitude of the respec-
tive a coefficient. On this basis it was removed
from the scale. Consequently there were only three
items remaining in the Involvement scale.
Table 4 reports the scale means (M), average
item means for the items comprising respective
scales (AIM), standard deviations (SD) and
Cronbach a coefficients for each of the six scales
measuring classroom discipline strategies.
Inspection of the a coefficients in Table 4 shows
that most a’s are moderate. Some are clearly low
and reflect a loose (although statistically signifi-
cant) association between the behaviors related to

Table 4
Discipline strategy usage by country

Scale Country

China Israel Australia

M AIM SD a M AIM SD a M AIM SD a

Punishment (n ¼ 4) 11.98 (3.00) 3.94 .62 13.02 (3.26) 4.28 .75 14.54 (3.64) 4.75 .79
Discussion (n ¼ 4) 17.45 (4.35) 4.30 .72 13.45 (3.36) 4.77 .77 13.10 (3.28) 4.46 .73
Recognition (n ¼ 4) 16.32 (4.08) 4.36 .66 13.16 (3.29) 4.92 .79 12.51 (3.13) 5.32 .83
Aggression (n ¼ 4) 9.22 (2.30) 4.03 .69 11.65 (2.92) 4.22 .62 11.02 (2.76) 4.88 .73
Involvement (n ¼ 3) 11.80 (3.93) 3.37 .49 8.49 (2.83) 3.11 .48 8.49 (2.83) 3.14 .53
Hinting (n ¼ 4) 16.04 (4.01) 3.91 .58 14.62 (3.65) 4.03 .62 14.48 (3.62) 4.04 .66
 ARTICLE IN PRESS
736 R. Lewis et al. / Teaching and Teacher Education 21 (2005) 729–741
one strategy. For example, the Involvement scale’s
a of approximately .5 reflects correlations of .17,
.26 and .40 between the three respective pairs of
items. The likelihood of teachers letting students
know that ‘‘the way they are behaving is not how
the class expects them to’’ is not highly associated
with the likelihood of them organizing the class to
work out the rules for good behavior and what
should happen to students who misbehave, even
though the latter two behaviors are more likely to
co-occur. Another explanation for the low a
coefficients is the difficulty inherent in transferring
strategies conceptualized in one culture to another.
Nevertheless, despite some concern about the low
internal consistency of responses to items in the
Involvement and Hinting scales, it was determined
to use all six scales for purposes of replication,
acknowledging the tenuousness of findings relating
to the less-reliable measures.
5. Classroom discipline strategies in Australia,
Israel and China
Before examining the relationship between
classroom discipline and misbehavior, considera-
tion was given to the comparison of the frequency
of usage of different strategies in differing national
settings and the correlations between strategies.
Inspection of the AIM, also reported in Table 4,
indicates that the pattern of usage of the various
classroom discipline strategies appears relatively
similar in Australia and Israel. In both countries,
teachers commonly react to misbehavior by letting
students know that there is a problem in the hope
that they will improve their behavior. In addition,
they are more than sometimes likely to Punish
misbehaving students and discuss with them the
impact their misbehavior has on others in a bid to
have them determine a better way to behave. They
recognize appropriate behavior more than some-
times, to increase the likelihood of its reoccur-
rence, The two strategies utilized less frequently
than sometimes are aggression and involvement of
the class in setting rules and consequences. The
pattern in China is a little different in that students
report greater use of all strategies except Aggres-
sion and Punishment. Although the relative usage
of strategies also varies by Country the only
strategy to vary rank within a country by more
than two ranks is Punishment, which ranks as the
most common strategy in Australia, and the fourth
and fifth most commonly used strategy in Israel
and China, respectively.
In interpreting these data, it is of value to note
that there are a number of substantial correlations
between strategies. In each national setting, when
correlations greater than .4 are considered, four
strategies intercorrelate. These are Discussion,
Reward, Hint and Involve. All appear to demon-
strate a positive attitude toward students. In
addition, in each setting, aggression correlates
with punishment, forming a potentially more
negative response to misbehavior, although in
Australia punishment also correlates with hint.
These patterns of relationships are very similar in
different national settings, suggesting that teachers
may be seen by students in terms of discipline
‘‘styles’’ comprising a number of strategies.
In order to more systematically compare na-
tional differences in the extent of usage of different
discipline strategies, a 5-way MANOVA was
performed where Country, level of student mis-
behavior in class, gender of the student, the gender
of the teacher and year level (year 7/8, 9/10 or 11/
12) were utilized as independent variables and the
six discipline strategies were the dependent vari-
ables. This enabled the consideration of the
students’ age and sex and the sex of the teacher
to be considered in the analysis in addition to
national setting and level of misbehavior. Since
level of student misbehavior was one of the
independent variables, the analyses also permitted
the relationship between discipline and misbeha-
vior to be examined in the different settings.
The analysis indicated a total of seven statisti-
cally significant effects. Since the analysis investi-
gated 31 predictors, a conservative probability
level of .01 was utilized for statistical significance.
Table 5 reports the Mutivariate F values (Pillai’s
trace) for the significant effects.
Inspection of the data in Table 5 shows that
three main effects and four interaction effects were
statistically significant. Consideration of the rele-
vant univariate tests and Sheffe tests for post hoc
comparisons (po:05) for year level indicate that

R. Lewis et al. / Teaching and Teacher Education 21 (2005) 729–741
Table 5
Significant predictors of classroom discipline strategies
Independent variable F Hyp Error
value df df
Year level 2.28 12 9022
Country 30.10 12 9022
Misbehavior 2.92 18 13536
Country  Year level 3.25 24 18052
Country  Teacher Sex 2.23 12 9022
Country  Year  Misb’r 1.49 66 27090
Country  Tchr sex  Year 3.65 24 18052
there is significantly more Discussion (F ¼ 7:04,
p ¼ :001), Recognition (F ¼ 5:96, p ¼ :003) and
Involvement (F ¼ 6:72, p ¼ :001) in years 7 and 8
than in years 9 and 10, and 11 and 12 (which did
not differ significantly). In addition, there is more
Hinting perceived in years 7 and 8, and 9 and 10,
than in years 11 and 12. These data may reflect a
greater need for intervention with younger stu-
dents who are new to secondary schooling and in
the process of adaption. However, if that were the
case it is surprising that there is not significantly
more Punishment at this level also. In summary,
however, it may be argued that teachers are willing
to use more positive strategies, ones which imply
trust, with younger students.
The main effect for Country applied to all
six classroom discipline techniques and showed
that students in China, compared to those in
Australia or Israel, report less usage of Punish-
ment (F ¼ 15:71, p ¼ :000) and Aggression
(F ¼ 11:92, p ¼ :000), and greater use of Recogni-
tion (F ¼ 56:65, p ¼ :000), Discussion (F ¼ 71:92,
p ¼ :000), Hinting (F ¼ 14:66, p ¼ :000) and
Involvement (F ¼ 100:15, p ¼ :000). Whereas stu-
dents in Australia and Israel do not differ in
reported levels of Involvement, Aggression and
Hinting, Australian students report less use of
Discussion, Recognition and more Punishment
than do those in Israel.
The final main effect indicated that students
who reported greater levels of misbehavior were
more likely to perceive Aggression by teachers
(F ¼ 7:87, p ¼ :000), although no other strategy
produced statistically significant univariate results.
ARTICLE IN PRESS
737
Some elaboration of these main effects comes
about when interaction effects are considered. The
Significance
only interaction effect which failed to produce a
level statistically significant univariate effect (po:01)
was Country by Year by Misbehavior. The other
.007 four therefore required explanation.
.000
Firstly, the Country by Teacher Sex interaction
.000
.000 showed that even though, as reported above, there
.000 is less Aggression reported in China, it is primarily
.006 due to it being seen as used less by men (F ¼ 7:69,
.000 p ¼ :000). Secondly, the Country by Year level
interaction indicated that there were three strate-
gies with statistically significant differences,
namely Punishment (F ¼ 6:38, p ¼ :000), Aggres-
sion (F ¼ 5:73, p ¼ :000) and Discussion
(F ¼ 6:60, p ¼ :000). The first two of these show
that whereas year 11 and 12 students in Australia
and Israel are least likely to report the use of
Punishment and Aggression, those in China report
such strategies more frequently than do younger
students. The last Country by Year level interac-
tion effect, for Discussion, shows that in Israel the
variation in usage by year level is four times that
found in either China or Australia.
The final significant interaction (Country by
Teacher sex by Year level) sheds a little more light
on some of the earlier findings. Four univariate
effects were noted as significant, namely for
Punishment (F ¼ 5:54, p ¼ :000), Aggression
(F ¼ 3:68, p ¼ :005), Involvement (F ¼ 7:05,
p ¼ :000) and Hinting (F ¼ 3:55, p ¼ :007). The
first two results indicate that the main reason older
students in China report more Punishment and
Aggression (see above) is because it is seen as
coming primarily from women teachers (as is
Punishment in Israel). In Australia, male and
female teachers display similar, low levels of
Aggression and Punishment with older students.
The level of involvement of students in class-
room management decision-making and the
amount of hinting at the inappropriateness of
student behavior by women teachers in both
Australia and China decreases with increasing
student age. For women teachers in Israel, most
student involvement occurs with students at years
11 and 12, although the differences between year
levels are very small. For men, the picture is
similar in that most Hinting and Involvement
 ARTICLE IN PRESS
738 R. Lewis et al. / Teaching and Teacher Education 21 (2005) 729–741
occurs for students in years 7 and 8. The least is
offered to years 9 and 10 in both Israel and China,
and 11 and 12 in Australia. In China the
differences are very small for Hinting.
In Israel, there is more Hinting at, and Involving
of, students in years 7 and 8 by men teachers than
by women. At years 9 and 10 the opposite is the
case. In contrast, in Australia, women teachers are
more likely than men to Involve and Hint at
students in all year levels, although at years 9 and
10 the difference is very small. In China, men
teachers are more likely to hint at, and involve,
students in years 11 and 12. At years 7 and 8,
women are more likely than men to Hint at their
students.
6. Discussion
The broad pattern of results indicates that
teachers sampled from China appear more inclu-
sive and supportive of students voices when it
comes to classroom discipline, and are less
authoritarian (punitive and aggressive) than those
in Israel or Australia. In contrast, the Australian
classrooms are perceived as having least discussion
and recognition and most punishment. Israeli
teachers are situated between these two positions,
even though they exhibit the most aggression.
(Inspection of relevant means shows that their
aggression is primarily in the form of yelling in
anger at their students.)
Cultural factors may be posited as part of the
reason for these patterns. In China, teachers are
held in very high esteem (Li, Xie, & Wang, 1998)
and as argued by Jin and Cortazzi (1998), Chinese
students would ‘‘follow what teachers say out of
respect’’. In a recent study for example, which
compared Chinese and American students’ percep-
tions of the aims of schooling, Lau et al. note that
the former were more likely to report that learning
to respect authority was a significant outcome of
education. Consequently, even though Chinese
teachers report as much ‘‘misbehavior’’ in class
as do Australian and Israeli teachers, the nature of
the misbehavior in Chinese classrooms may be less
extreme. These assumptions are supported by the
work of Aldridge, Fraser, and Huang (1999) in
Taiwan. After visiting classrooms in both coun-
tries they report
Classroom observations and interviews suggest
that there could be differences in the ways in
which students regard their teachers; students in
Taiwan had more respect for teachers than did
students in Australia.
Although the teachers in both countries com-
plained about discipline problems with stu-
dents, we noted that there was more evidence of
disruptive behavior in science classes in Aus-
tralia (described in the first story as answering
back and chatting between friends) than in
Taiwan. (p. 58).
An assumption of less provocative forms of
misbehavior in Chinese classrooms could explain
the lesser likelihood of the more stringent forms of
disciplinary interventions such as Punishment and
Aggression. The Chinese teachers’ greater usage of
hinting, discussion and inclusion may reflect
confidence on their part that the children will
listen to them and to their peers, as argued by Jin
and Cortazzi (1998). Chinese teachers can also rely
on parents to support them in their attempts to
make students self-disciplined (Gao, 1998; Peng,
1993). With regard to the reported greater use of
Aggression and Punishment by women teachers in
China (in higher year levels), and greater involve-
ment of older students by male teachers, it may be
that more traditional respect is accorded to males
within the Chinese culture, particularly by older
students. Consequently male teachers are less
likely than females to need to resort to coercive
strategies to reorient misbehaving students. This
explanation could also account for why women
teachers in China are also less likely to try and
involve older students in determining rules, and
consequences for misbehavior. An alternative
explanation may be related to the reported greater
stress levels experienced by female teachers in
China (Dong, 2001; Zhou & Zhao, 1998). Such
stress is likely to be more pronounced in higher
year levels because of the college entrance exam-
ination, and the low entrance rate of students to
universities.
It may be argued that the Australian and
Israeli teachers’ relative unwillingness to empower
 ARTICLE IN PRESS
R. Lewis et al. / Teaching and Teacher Education 21 (2005) 729–741
students in the decision-making surrounding class-
room discipline is related to the lower levels of
unconditional respect they are likely to receive
from students, and the reduced levels of support
parents provide. The fact that the Australian
teachers report less use of discussion and more
punishment than do those in Israel may signify
that the former have relatively less ‘‘legitimate’’
power, and need to rely on more ‘‘coercive’’ power
(Tauber, 1999) to manage their classrooms. The
greater likelihood for women teachers of years 8
and 9 in Australia and Israel to be inclusive could
signify a lesser concern about their legitimate
power, although as reported, in Israel men
teachers were more inclusive for years 7 and 8.
This explanation based on power may also
account for the greater inclusion of younger
students in all three countries (although the effect
is most pronounced in Israel). That is, students in
years 7 and 8 are more likely to accord teachers
unconditional respect than those in later years.
The results for teachers’ usage of recognition
and reward for good behavior is difficult to
interpret in a way consistent with the above
analysis. One might have assumed that teachers
who have less legitimate power may try to use
more ‘‘Reward’’ power (Tauber, 1999) to compen-
sate. The data however show greatest use of
recognition and reward by the Chinese teachers,
followed by those in Israel, with the Australian
sample of teachers using least. These findings
suggest an additional explanation for some of the
findings that is consistent with the reported
associations between students’ self-reported mis-
behavior and teacher aggression.
As stated earlier, in all three settings, students
more prone to misbehavior report greater levels of
aggressive teacher disciplinary behavior. It may be
assumed that the more provocative students
stimulate more anger in teachers and consequently
more aggression. Angry or upset teachers, as
argued by Glasser (1997), may not be interested
in being reasonable toward unreasonable and
disrespectful students. They therefore will find it
unpalatable to recognize very difficult students
when they act appropriately. Rewarding ‘Nean-
derthals’ for being normal may not come natu-
rally. Similarly, they may find it unpleasant and
739
unproductive to spend time letting such students
tell their side of events, trying to get them to
acknowledge that their behavior is unfair and
needs to change.
Having argued why it is understandable for
teachers to react to provocative and possibly
confronting student behavior by becoming more
aggressive, this does not excuse such a response. It
is recognized that
Children who have significant emotional and
behavioral problems respond less positively to
others and this elicits fewer positive responses
and more negative responses from others (Pace,
Mullins, Beesely, Hill, & Carson, 1999, p. 151).
However, teachers are professionals who need
to respond in the best interests of their clients, the
students. As argued by Roeser, Eccles, and
Sameroff (2000), ‘‘Teachers need to protect
adolescents from situations they perceive as
threatening to their self yor threatening to their
social image’’ (p. 454). If not, then ‘‘adolescents
will feel less motivated to learn and more unhappy
and will be more likely to manifest academic or
social problems’’(p. 454). In summary, teachers
cannot allow themselves and difficult students to
be locked into ‘‘a vicious cycle of reciprocal
causation’’ (Pace et al., 1999, p. 151).
In making a recommendation that teachers need
to work harder to foster quality relationships with
difficult students, we are aware how difficult that
may be. However, a clear starting point would be
to minimize the usage of aggressive disciplinary
strategies while increasing the frequency with
which teachers recognize students’ responsible
behavior, however rare it may be. Secondly, they
need to make the opportunity to let students talk
about their side of things so that it can be clearly
understood, to get them to understand why their
behavior is a problem for others and to obtain a
plan for a better future. In many schools in
Australia, Israel and China, this conversation is
conducted by someone other than the classroom
teacher, as a result of a ‘referral’. It is the year
level coordinator (class teacher), school counselor
or a senior teacher who gets the advantage of
the relationship power that arises from such
 ARTICLE IN PRESS
740 R. Lewis et al. / Teaching and Teacher Education 21 (2005) 729–741
conversations (Tauber, 1999), whereas it is the
classroom teacher who needs it.
Encouraging teachers to build rather than
destroy good will with students who are more
provocative is a challenging request. It will not be
easy. In the experience of one of the authors who is
working in schools to achieve this aim it can take
many years of persistent effort accompanied by
considerable support. If teachers are to act in the
best interests of students in an area as emotive as
classroom discipline, then
creating professional work environments where
teachers feel supported by other professionals
and school leaders in relation to their own
needs for competence, autonomy, and quality
relationships is essential (Roeser et al., 2000,
p. 466).
Acknowledgments
The research project reported in this paper was
in part sponsored by the Institute for Community
Education and Research, School of Educatiion,
Bar-Ilan University, Israel.
References
Aldridge, J. M., Fraser, B. J., & Huang, T.-C. I. (1999).
Investigating classroom environments in Taiwan and
Australia with multiple research methods. The Journal of
Educational Research, 93(1), 48–61.
Barton, P. E., Coley, R. J., & Wenglinsky, H. (1998). Order in
the classroom: Violence, discipline and student achievement.
Princeton, NJ: Policy Information Center, Educational
testing service.
Blase, J. J. (1986). A qualitative analysis of sources of teacher
stress: Consequences for performance. American Educa-
tional Research Journal, 23, 23–40.
Borg, M. G., Riding, R. J., & Falzon, J. M. (1991). Stress in
teaching: A study of occupational stress and its determi-
nants, job satisfaction and career commitment among
primary schoolteachers. Educational Psychology, 11(1),
59–75.
Chan, D. W. (1998). Stress, coping strategies, and psychological
distress among secondary teachers in Hong Kong. American
Educational Research Journal, 35(1), 145–163.
Charles, C. (2004). Building classroom discipline: From models to
practise (6th ed.). New York: Longman.
DeRobbio, R. A., & Iwanicki, E. F. (1996). Factors accounting
for burnout among secondary school teachers. Paper pre-
sented at the annual conference of the American Educa-
tional Research Association. New York.
Dong, H. (2001). A psychological investigation of teachers’
over strees and anxiety. Journal of Urumqi Adult Education
College, 2, 55–58.
Friedman, I. A. (1994). Conceptualising and measuring teacher-
perceived student behaviors: Disrespect, sociability and
attentiveness. Educational and Psychological Measurement,
54(4), 949–958.
Friedman, I. A. (1995). Student behavior patterns contributing
to teacher burnout. The Journal of Educational Research,
88(5), 281–289.
Frydenberg, E., & Lewis, R. (2000). The coping scale for adults:
Construct validity and what the instrument tells us. ERIC
TM031533 ED443 888.:12
Gao, L. (1998). Cultural context of school science teaching and
learning in the People’s Republic of China. Science
Education, 82, 1–13.
Glasser, W. (1997). A new look at school failure and school
success. Phi Delta Kappan, 597–602.
Goddard, J. T. (2000). Teaching in turbulent times: Teachers’
perceptions of the effects of external factors on their
professional lives. Alberta Journal of Educational Research,
46(4), 293–310.
Guttmann, J. (1982). Pupils’, teachers’ and parents’ causal
attributions for problem behavior at school. Journal of
Educational Research, 76(1), 14–21.
Hyman, I. A., & Snook, P. A. (2000). Phi Delta Kappan, 81(7),
489–501.
Ingersoll, R. M. (1996). Teachers’ decision-making power and
school conflict. Sociology of Education, 68(2), 159–176.
Ingersoll, R. M. (2001). Teacher turnover and teacher
shortages: An organizational analysis. American Educa-
tional Research Journal, 38(3), 499–534.
Jin, L., & Cortazzi, M. (1998). Dimensions of dialogue: Large
classes in China. International Journal of Educational
Research, 29, 739–761.
Kaplan, A., Gheen, M., & Midgley, C. (2002). Classroom goal
structure and student disruptive behavior. British Journal of
Educational Psychology, 72(2), 191–212.
Keiper, R. W., & Busselle, K. (1996). The rural educator and
stress. Rural educator, 17(2), 18–21.
Kyriacou, C. (1980). Stress, health, and school-teachers: A
comparison with other professions. Cambridge Journal of
Education, 10, 154–159.
Lau, S., Nicholls, J. G., Thorkilsen, T. A., & Patashnick, M.
(2000). Chinese and American adolescents’ perceptions
of the purposes of education and beliefs about the
world of work. Social Behavior and personality, 28(1),
73–89.
Lewis, R. (1997a). Discipline in schools. In L. J. Saha (Ed.),
International encyclopedia of the sociology in education
(pp. 404–411). Oxford, UK: Permagon.
Lewis, R. (1997b). The discipline dilemma (2nd ed.). Melbourne:
The Australian Council for Educational Research.
 ARTICLE IN PRESS
R. Lewis et al. / Teaching and Teacher Education 21 (2005) 729–741
Lewis, R. (1999). Teachers coping with the stress of classroom
discipline. Social Psychology of Education, 3, 1–17.
Lewis, R. (2001). Classroom discipline and student responsi-
bility: The students’ view. Teaching and Teacher Education,
17(3), 307–319.
Lewis, R., & Lovegrove, M. N. (1987). What students think of
teacher’s classroom control techniques: Results from four
studies. In J. Hastings, J. Schwieso (Eds.), New directions in
educational psychology, Vol. 2: Behaviour and motivation
(pp. 93–113). The Falmer Press.
Li, X. Y., Xie, J. S., & Wang, Z. H. (1998). Reform and
development of Chinese higher teacher education. Jiangxi,
China: Jiangxi Higher Education Publishing House.
Lowell, C., & Gallup, A. M. (2002). The 34th annual Phi Delta
Kappa/Gallup poll of the public’s attitude towards the
public schools. Phi Delta Kappan, 84(1), 41–56.
Martin, S. C. (1994). A preliminary evaluation of the adoption
and implementation of assertive discipline at Robinton high
school. School Organisation, 14(3), 321–330.
McCormick, J., & Shi, G. (1999). Teachers’ attributions for
responsibility for their occupational stress in the People’s
Republic of China and Australia. British Journal of
Educational Psychology, 69, 393–407.
Miller, A., Ferguson, E., & Byrne, I. (2000). Pupils’ causal
attributions for difficult classroom behavior. British Journal
of Educational Psychology, 70, 85–96.
Mitchel, K. (2001). Education for democratic citizenship:
Transnationalism, multiculturalism, and the limits of
liberalism. Harvard Educational Review, 71(1), 51–78.
Oswald, U. (1995). Difficult to manage students: A survey of
children who fail to respond to student discipline strategies
in government schools. Educational Studies, 21, 265–276.
Pace, T. M., Mullins, L. L., Beesely, D., Hill, J. S., & Carson,
K. (1999). The relationship between children’s emotional
741
and behavioral problems and the social responses of
elementary school teachers. Contemporary Educational
Psychology, 24, 140–155.
Peng, S. S. (1993). Fostering student discipline and effort:
Approaches used in Chinese schools. Paper presented at the
Annual Meeting of the American Educational Research
Association (April 12–16, 1993).
Roeser, R. W., Eccles, J. S., & Sameroff, A. J. (2000). School as
a context of early adolescents’ academic and social-
emotional development: A summary of the research
findings. The Elementary School Teacher, 11(5), 443–471.
Romi, S., & Freund, M. (1999). Teachers’, students’ and
parents’ attitudes towards disruptive behaviour problems in
high school; a case study. Educational Psychology, 19(1),
53–70.
Romi, S., & Katz, Y. (2003). Affective education: The nature and
characteristics of teachers’ and students’ attitudes toward
school in Israel. Unpublished paper, School of Education,
Bar Ilan University. Israel.
Rothstein, R. (2000). Towards a composite index of school
performance. The Elementary School Teacher, 100(5),
409–441.
Tauber, R. T. (1999). Classroom management: Sound theory and
effective practice. Pennsylvania: Bergin & Garvey.
Whiteman, J. L., Young, J. C., & Fisher, L. (1985). Teacher
burnout and the perception of student behavior. Education,
105, 299–305.
Wolfgang, C. H. (1995). Solving discipline problems: Strategies
for classroom teachers (3rd ed.). Massachusetts: Allyn and
Bacon.
Zhou, Z., & Zhao, D. (1998). A study on the symptoms of
anxiety and related factors of primary and secondary
school teachers. Journal of China Civilian Medicine, 10(5),
305–306.