1. Introduction
The last decades of the 20th century saw the beginning
of an extensive body of research on how teacher behaviour is related
to learners’ affective and cognitive learning and motivation.
Learners’ emotions, positive and negative, are thought to influence
the learning process. A number of internal factors influence these
emotions, including learners’ self-perception and beliefs (Horwitz et.
al, 1986; McIntyre, 1999), personalities (Krashen, 1982) and previous
experiences (Coleman, 2008). External factors also affect learners’
emotions, such as context-related demands (Dörnyei & Ushioda,
2021), as well as the teachers themselves and their behaviours (Guo et
al., 2022; Jiang & Dewaele, 2019; Li, 2021).
The study of teachers’ behaviours, including verbal and nonverbal acts, finds its modern roots
in the concept of immediacy, a term coined by Mehrabian (1971) to explain how people are
attracted to others. People avoid stimuli which do not provide rewards or are punishing,
favouring those which are rewarding. Creating immediacy in the classroom has been linked to
an increase in learner motivation (Allen et al., 2006; Christophel, 1990; Miller et al.,
2014; Velez & Cano, 2008; Wijaya, 2017); lower learners’ anxiety levels (Ballester,
2015); enhanced attention (Barcelos, 2020; Frymier, 1994); better attendance and
participation (Myers et al., 1998); and an enhanced perception of teachers’ credibility
(Johnson & Miller, 2002; Xie & Derakhshan, 2021).
Perceived teacher credibility itself has been correlated to improved learners’ performance
(McCroskey et al., 2004; Myers & Bryant, 2004; Pogue & Ahyun, 2006; Xie &
Derakhshan, 2021;). However, how teachers are perceived and whether this corresponds to
their actual behaviour is an area little researched. This study was set up to fill in part
of this gap by comparing actual teachers’ behaviours in the classroom and learners’
perceptions of those behaviours. In particular, we focused on nonverbal and verbal
behaviours that have been identified as improving immediacy in the classroom in past studies
(Zhang, 2006).
2. Immediacy
According to Mehrabian (1971), there are three independent dimensions to emotional states,
“pleasure-displeasure, arousal-non-arousal and dominance-submissiveness” (p. 5). In the
classroom, teacher behaviour can enhance or minimize each of these elements, thus creating
an environment more conducive to success (Frymier, 1994). Power is an inherent dimension of
the teacher but it can be manipulated through speech, such as quantity of talking
(Sorrentino & Boutillier, 1975) and also through non-verbal behaviours that help
learners feel more relaxed such as the teacher smiling or minimizing the teacher-learner
power distance by getting physically closer to the learner (Mehrabian, 1971). Arousal is
mostly conveyed through nonverbal behaviours, such as changes in body position, gesture,
facial expression or voice. While pleasure, ‘liking’, can be strengthened by using inclusive
speech, first names, humour, praise or by nonverbal behaviours such as gaze, smiles,
gestures and proximity (Richmond et al., 1987). More recent studies linking teachers’
behaviours to learners’ emotions point to strategies that engage learners (arousal) and
reduce face-threats (liking) while maintaining credibility by proving competence (Dewaele et
al., 2022). Some of these strategies include presenting clear and well-organized content at
the right level for the learners (Chesebro & McCroskey, 2001; Sidelinger, 2010;
Sidelinger & McCroskey, 1997); minimizing the face-threat of feedback to learners
(Kerssen-Griep et al., 2003) by deploying more explanatory than imperative statements, using
less formal and more complimentary language, introducing humour and self-disclosure in the
classroom discourse (Witt & Kerssen-Griepp, 2011). Self-disclosure, talking about
oneself, has also been linked to immediacy, as teachers might be perceived as being more
approachable (Miller et al., 2014).
2.1. Development of studies on immediacy in the classroom
One of the earliest studies on immediacy in the classroom correlated learners’ perceived use
of nonverbal behaviours by teachers and affective learning (Andersen, 1979). Affective
learning refers to how learners acquire skills and knowledge through emotional
experiences. Andersen found that smiling, eye contact, direct body orientation, relaxed
attitude and body movements, gestures, touching and being vocally expressive could develop
immediacy which in turn lead to higher learning scores. A number of studies found
corroborating results, in liberal art colleges (Christensen & Menzel, 1998; Richmond
et al., 1987), larger universities (Gorham, 1988) and high schools (McDowell, et al.,
1980). Nonverbal behaviours were found to be even more important in larger classes to
minimize the physical distance imposed by the environment (Gorham, 1988).
Richmond et al. (1987) added the verbal dimension to immediacy studies, following Weiner,
Johnson and Mehrabian’s theories (1968). Many of the verbal immediacy behaviours would
later be associated with dimensions of credibility, such as being inclusive or enhancing
affiliation. Other verbal behaviours included the use of humour, self-disclosure,
encouraging learners to talk, addressing learners by name, providing feedback, praising,
having conversations before and after class. Of these, non-verbal behaviours were thought
to be more effective than verbal ones, according to a study by Estepp and Rogers (2015),
who asked students to report on teachers verbal and nonverbal behaviours, which
corroborated previous results by Velez (2008).
Christophel (1990) directly correlated immediacy to motivation, establishing a link between
immediacy and affective behaviours. Her results were supported by those of Frymier (1994)
who also found a positive effect on motivation but not a direct effect on student
learning. Frymier suggested that immediacy attracted learners’ attention, they then became
more involved and interested and would be looking forward to the next class. However, she
also stated that immediacy alone was not enough to motivate learners. Despite these
results, Rodríguez et al. (1996) proposed a model directly linking immediacy to affect and
affect to cognitive learning. This indirect association between immediacy and cognitive
learning was confirmed by a meta-study by Allen et al. (2006), who reported that teacher
immediacy increased student motivation and that in turn helped develop cognitive learning.
An integrated model by Zhang and Oetzel (2006) acknowledges the direct and indirect
relationships between immediacy, affect, motivation and cognition. The model states that
immediacy has a direct effect on affect and motivation, but affect also impacts
motivation, which in turn is related to cognitive learning. Guo et al. (2022) conclude
that teacher immediacy increases language enjoyment and is a contributing factor to
learners’ motivation.
Jiang and Dewaele (2019), in a study with 564 Chinese, found that English learners’ emotions,
in particular enjoyment and anxiety were specifically related to teacher’s friendliness,
joking, strictness and unpredictability (this latter behaviour only affected anxiety). A
similar study by Dewaele et al. (2022), with 360 Kuwaiti foreign languages learners, also
identified joking, predictability and use of English as behaviours affecting enjoyment and
motivation, while no effects were found on anxiety. In both studies the use of humour
emerged as one of the key behaviours enhancing learners’ enjoyment, while, at least in
Chinese learners, anxiety was more of an intrinsic factor less dependent on the teacher’s
behaviour.
2.2. Challenges to immediacy
Immediacy studies and their effect on learners do have their share of critiques, the more
valid one being that there are other behaviours taking place outside of the classroom
which are not measured, such as face to face tutorials, email responses to learners
(Allen, et al., 2006), grading practices (Hess et al., 2001) and approachability.
Komarraju et al. (2010) found that students valued outside the classroom interactions,
especially when teachers are perceived to be approachable and available as well as
respectful.
From an intercultural perspective Zhang (2006) and Lopez-Ozieblo (2015) pointed out that some
of the original items used to measure immediacy might not be relevant or appropriate in an
Asian classroom, such as touching learners or direct gaze. Mottet and Richmond (1998) had
argued that verbal immediacy strategies could not be generalised as individual differences
were too many. However, scholars in the field of politeness disagree with this statement,
as there are many examples of universal linguistic strategies deployed by speakers to
reduce psychological distance with their interlocutors (Lopez-Ozieblo, 2018).
Smythe and Hess (2006) noted that most studies report retrospective learners’ perception,
after the class has finished, not actual class observations. Only a few studies have tried
to correlate learners’ perceptions to the reality observed in the classroom. In one of
these Andersen (1979) confirmed a significant correlation between the two, having trained
observers to rate teacher immediacy based on their behaviours during classes. However, in
a study with 525 students evaluating 17 instructors, Smythe and Hess (2006) found that
learners’ perceptions of teachers’ non-verbal behaviours did not correspond to the reality
observed by their researchers.
In a previous meta-study of 35 studies linking immediacy with affect, motivation or cognitive
learning, Hess et al. (2001) found that 86% (30 studies) relied on learners’ perception of
teachers’ behaviours and 74% (26 studies) only used learners’ beliefs about their own
learning to evaluate learning. This methodology has been considered valid (Richmond et
al., 1987; Scott & Wheeless, 1977) although very few studies have actually observed
teachers’ behaviours and correlated them to learners’ perceptions. Hess and colleagues
(2001) accepted that immediacy “appear to be positively associated with various factors
related to affective exchange in the classroom” (2001, p. 215), however, they advocated
caution as it could be that some of the behaviours associated with immediacy could also
have a detrimental effect on learning, by focusing learners’ attention on unimportant
content (Harp & Mayer, 1998).
Lopez-Ozieblo (2015) noted that most immediacy-related knowledge is based on observations of
Western participants and that similar studies in Asian classrooms have found that out of
the classroom interactions are more important in Asian context than US ones (Zhang, 2006).
Teachers’ credibility in Asian classrooms is likely to be more important to students than
psychological closeness achieved via verbal or nonverbal behaviours. Immediacy might have
a positive emotional effect but the lack of it might not be as detrimental as in other
cultures.
Therefore, the question remains whether, in an Asian context, learners’ perceptions align
with actual teachers’ behaviours traditionally associated with the development of
immediacy. If the two are not aligned then we need to question the relevance of immediacy
studies and those that establish relationships between learners’ emotions and teachers’
behaviours. In addition, in the institution where the data was collected, teachers’
appraisals, and thus salaries, are linked to the yearly evaluations given by students of
the subject and the teacher, based on a feedback questionnaire. If students’ evaluations
are not correlated to actual behaviours, this would call into question the relevance of
such questionnaires.
3. The Study
Our first objective was to identify if perceived teachers’ behaviours could be correlated to
learners’ evaluation scores of the class/topic and the teacher, as suggested in the
literature (Andersen, 1979; Christensen & Menzel, 1998; Christophel, 1990; Gorham, 1988;
Richmond et al., 1987; Scott & Wheeless, 1977; Zhang & Oetzel, 2006). The second
objective was to identify whether learners’ perception corresponded with the reality of the
classroom, the actual teachers’ behaviours as observed via the video recordings.
This study used convenience sampling. It focused on higher education colleagues of the
researcher who volunteered to be recorded while teaching. All subjects were Language
Education related, taught at the same Hong Kong institution. Naturally occurring data was
recorded in eight hours of graduate (5 lectures) and undergraduate (3 lectures) classes,
either three or two hours long. Teachers’ speech and actions were transcribed and then
analysed. The analysis was both qualitative and quantitative to determine what verbal and
nonverbal immediacy behaviours could be observed. In addition, learners were asked to fill
in a 22-question survey to evaluate the teacher and the class and also to indicate the
perceived behaviours of the teachers. These results were then correlated to the actual
behaviours observed.
3.1. Participants
Eight full-time teachers of various Language Education topics agreed to have their sessions
recorded before COVID-19, all classes were face to face. All teachers were native English
speakers or fully bilingual. All had lived and worked in Hong Kong for over five years,
and were, therefore, familiar with the Hong Kong classroom. Further information is not
given in order not to disclose their identity. In all of the recorded sessions two cameras
were used, one at the front of the classroom and one at the back, both pointing at the
teacher in order to avoid recording learners’ faces. In all of the sessions, two members
of the research team were present to carry out the recordings and manage the surveys. Some
of the subjects were compulsory subjects, with classes of over 50 learners, while others
were electives with class sizes not exceeding 30 learners (average class size was 32
learners).
Teachers were not specifically told that their interactions with learners would be analysed
in terms of immediacy. They were aware that we would be analysing nonverbal behaviour and
gestures in particular. After the sessions, they reported having been nervous to start
with (due to the cameras) but that they had soon forgotten about them. Some of the data
recorded has also been used in a study of how teachers mitigate disagreements
(Lopez-Ozieblo, 2018, 2023).
Out of the 252 learners answering the questionnaires the majority
(90%) were from Hong Kong, the rest were from other East Asian
territories (Mainland China, Korea, Taiwan), a fifth (20%) were male
and one third (34%) undergraduates. The average age of
undergraduates was 20 and postgraduates’ ages ranged from 21 to
mid-40s. Many postgraduates were teachers themselves. Learners were
asked for their consent to record the class and to use their answers
to the 22-question survey. The surveys were filled in at the end of
the recorded session. Learners were told the objective of the study
was to identify behaviours in teachers that learners responded to
positively.
The questionnaire was an adaption of the institution’s “Student Evaluation Questionnaire”
(already used by learners to evaluate the teacher and the subject) and Zhang’s
questionnaire (2006) which identifies specific verbal and non-verbal behaviours that are
associated with the creation of immediacy in Asian classrooms (see Appendix 1). The first five
questions related to the evaluation of the topic and the session, including: learner’s
attitude towards the topic, how much they understood and perception of the contents in
terms of how interesting, relevant and informative they were and how much the session
stimulated learners’ interest in the topic. The following eight questions related to the
evaluation of the teacher’s competence and performance in terms of making learners feel
comfortable. The last seven questions targeted specific teacher behaviours, use of humour,
movement, hand gestures, smiles, eye-contact, first names and praise. Participants were
asked to agree with the statements provided on a 1 to 5 Likert scale (1 = I totally
disagree). An additional two questions, not related to this study, were added to identify
whether learners preferred holistic teachers (whose role was also that of a friend, a
parent, a teacher), to confirm the results of a previous study (Lopez-Ozieblo, 2015).
Although we were only recording a snap-shot of the whole course, we believe it provided a
valid example of the teacher’s practices. By carrying out the recordings and
questionnaires after week 8 of the academic semester, we ensured that learners were
familiar with teachers and so their responses to the questionnaire would have been based
on the session recorded but also on previous interactions with the teacher (Abelson,
1985).
To minimise the impact of the external observation, a member of
the research team was selected partly for her knowledge of the
teachers observed, having been their student in previous years.
This, partly emic, knowledge gave us more confidence in our analysis
of certain behaviours, although all cases were discussed by all
three team members.
3.2. Data collection and analysis
The behaviours observed were: teachers use of humour, moving around the classroom and
gesturing. In addition, we also noted the amount of time spent on self-disclosure.
Learners were asked to evaluate their perception of how often they observed these
behaviours in their teachers. Each one of these behaviours was linked to a survey
question. For ‘the teacher is funny’ we had counted the number of times learners laughed;
for ‘moves around’ the number of times the teacher moved from one spot to another; for
‘gestures’ the time spent gesturing and proportion of time spent gesturing. Talley and
Temple (2015) report that positive hand and arm positions are positively correlated to
immediacy, while negative ones are not. They described positive hand and arm position as
having the hands in front of the body fingers together, either clasped or in a steepling
position and palms up or outwards. While negative positions are hands in pockets, arms
crossed or behind the back. It would seem that the lack of hand/arm movements is overall
negative, therefore we noted whether there was hand/arm movement or not. As many of the
gestures included long holds (the gesture froze mid-air), we measured the overall time
spent gesturing (rather than the more common measurement of frequency of gestures). This
was calculated as a percentage of the total time the teacher was speaking.
Teachers’ smiles and eye contact with students, were not included
in the final analysis as in many instances the quality of the
recording did not allow us to confirm these behaviours. Similarly,
praise and instances where teachers’ used students’ names were also
not analysed as they were only observed in two teachers.
The speech from the recordings of each session was transcribed by student helpers, using
Praat, a voice transcription software, and checked by two research
assistants and the author. The transcriptions were then imported into
ELAN a free software for multimedia analysis, where instances of the
teacher moving and hand gesturing, were annotated by the research assistants and the
author (checked by each other obtaining intercoder reliability of 95% of the data after
discussion).
Teachers’ speech was divided according to its content: relating to the subject matter;
exercises learners had to complete in-class (including time spent explaining these and
time debriefing); teachers’ self-disclosures: personal asides and personal examples (e.g.
“when I was an English teacher in Japan”); and student related asides (e.g. referring to
work that would need to be submitted in the future) or examples (e.g. “when do you
code-switch?”). This was done by viewing the recordings and manually categorizing chunks
of speech. All findings were normalised as a percentage of the total length of each
lecture.
Next, the responses to the surveys were consolidated and analysed by section: class
evaluation, teacher evaluation and perception of teacher behaviour. A number of analyses
were carried out to test:
The correlation between class/topic evaluation and teacher
evaluation scores. This was done to confirm results from
existing studies correlating the two.
The correlation between learners’ perceptions of specific behaviours, based on their
survey answers, and the overall evaluation of the class and the teacher. This
addressed our first objective.
The correlation between learners’ perceptions of teacher’s gesture, use of humour,
positive language and movement and the actual observed behaviour for each of the eight
teachers. This addressed our second objective.
The correlation between class evaluation and teacher evaluation scores and the actual
time spent on exercises, self-disclosure, content and student related issues. This was
calculated to confirm existing results indicating that the objective of teachers’
speech (in particular self-disclosure) affects learners’ evaluation of the
teacher.
We combined the answers from all the participants. As the data
was not normally distributed, there were very few cases of scores of
1 or 2 in the evaluations, the correlations were investigated using
Spearman’s coefficient, a moderate Spearman’s rho was considered to
be between 0.3 and 0.7. All correlations were run at a significance
level of 95%.
4. Results
4.1. Learners’ evaluations of class/content and teacher
The four correlations we investigated are detailed below with figures and with an explanation
of the results of each analysis. The first analysis investigated the correlation between
learners’ evaluation of the class/topic and of the teacher. Figure 1 shows the average learners’ evaluation of each class/topic
(in blue, dashed line) and of the teacher (in orange, full line). Overall, the evaluation
of the class was always lower than the evaluation of the teacher (the maximum score was 5,
shown in the y-axis), showing that the teacher’s knowledge and competence to deliver the
content was rated higher than the content presented. This supports current research on the
emotional impact of the teacher in the overall learning process (Dewaele et al., 2022).
Although we did not test (and are not confirming) causality, it would seem that how the
teacher is perceived might be dictating the evaluation of the class/topic.
Figura 1. Learners’ evaluations of the topic and the class and the teachers’ performance
![]()
A Spearman’s correlation coefficient confirmed a strong positive correlation between the two:
Spearman’s rho = 0.708, p = <0.001 Confidence
intervals (CI) = [0.765, 0.641]. This confirms the findings of previous
studies that correlate perceived teacher’s caring and competence with better students’
performance, also associated with a higher interest and better understanding of the topic
(McCrosky et al., 2004; Myers & Bryant, 2004; Pogue & Ayhun, 2006). Perceived
teacher’s credibility is a mix of perceived competence, trustworthiness and caring and
somewhat correlated to improved learners’ performance and students’ learning (Finn et al.,
2009).
4.2. Correlating learners’ perceptions of teachers’ behaviours and evaluations
In the second analysis we ran two sets of correlations between learners’ perceptions of
teachers’ behaviours (the teacher used my name, smiled, looked at me, was funny, moved,
gestured, praised me) and (1) class/topic evaluations and (2) teacher evaluations. Figure 2 shows the Spearman’s rho
for the significantly correlated class/topic evaluations (in blue, dashed line) and
teacher evaluations (in orange, full line), p values <0.001, vs. the
perceived behaviour for each teacher. The graph gives the correlation coefficients,
Spearman’s rho as the Fishers’ z effect size in brackets
(see Appendix 2).
Every behaviour is positively correlated with the evaluation
score of the class and the teacher, although the correlations
between the perceived teachers’ behaviours and the evaluation of the
teacher are stronger than with the class/topic evaluations.
Figura 2. Spearman’s correlation coefficients between perceived teacher behaviour and class/topic
evaluations and teacher evaluations
![]()
Note: The y-axis refers to the Spearman’s correlation coefficient
rho. A medium correlation is between 0.3 and 0.7 (American
Psychological Association, 2010). The letter in brackets refers to
the Fisher’s effect size z, H = high; M = moderate and L = low. A
medium effect is between 0.3 and 0.5 (Lipsey & Wilson,
2001).
The highest correlations with the strongest effects were found between the evaluation of the
teachers and with the teacher perceived as ‘being funny’, ‘smiling’ and ‘looking at me’.
Humour and smiling are immediacy behaviours that can help develop liking (Mehrabian,
1971). Teachers use of humour has been linked to enhanced students’ learning and a
decrease of learner’s anxiety overall (Aylor & Oppliger, 2003; Gorham &
Christophel, 1990). Humour varies by culture and Chinese are thought to use it in a more
constrained manner than other cultures (Nevo et al., 2001). In the Chinese classroom,
teacher’s humour has been found to be positively correlated with a classroom communication
apprehension (Zhang, 2005). Zhang believes this might be related to how humour focuses on
the individual which in collectivist societies, like the Chinese, might be a cause of
anxiety (Hackman & Barthel-Hackman, 1993). However, in this study, participants
reacted positively to the use of humour by the teacher. One possible explanation is that
culturally, Hong Kong might be regarded as a more Westernized society than Mainland China
(where Zhang’s study took place) and this is reflected in a more relaxed interpretation of
humour. Smiling and looking at learners on the other hand are behaviours that transmit
care. Chinese students value holistic teachers who go beyond the transmission of knowledge
(Jin & Cortazzi, 1998), a trait shared by Hong Kong students (Lopez-Ozieblo, 2015); it
is likely that smiling was thus interpreted as a positive and valued trait.
The lowest correlation was found with the behaviour ‘the teacher uses my name’. Originally,
immediacy studies focused on white North American males, these results might not be
relevant to the Asian context (Zhang & Oetzel, 2007). In particular those relating to
the power distance between the teacher and the students. Chinese culture has a larger
power distance than North America (Hofstede, 1991), which means that addressing students
by their first names might not be seen as appropriate (teachers involved in the study used
students’ first names, rather than last names, when addressing them). We speculate that
the power distance difference might also explain why praise does not have a large effect.
In a large-power distance culture, praise is not expected and might not be processed as
positively as in a low-power distance one, where praise is common.
Moving and gesturing, another two common immediacy behaviours in
North American classrooms (Andersen, 1979), might not be so relevant
in the Hong Kong classroom (Lopez-Ozieblo, 2015). Moving might be a
behaviour to close the physical distance between the teacher and the
learner but in the Asian classroom this action might increase
learners’ anxiety as it focuses the attention on one individual.
Despite Hong Kong’s
Western culture, it still shares some of the traits of a
collectivist society (Hofstede, 1991). Although gestures have been
correlated with enhanced learning in a number of studies (Macrine
& Fugate, 2022), the value of gestures in the classroom is a
topic that requires further research as their benefit varies
depending on a number of factors, including learners’ personal
traits (Lopez-Ozieblo et al., under review).
4.3. Observed and perceived teachers’ behaviours
The third analysis compared the actual behaviours of the teachers, based on the video
recordings, with those perceived by the learners. As it is quite likely that teachers’
behaviours will be different when explaining content rather than when engaging with
learners doing exercises, we focused on time spent delivering content to calculate what
percentage of time was spent gesturing. As teachers will also be funny and move about the
room when talking about themselves and during teacher-learner interactions, for the
analysis of these behaviours we focused on the full session (rather than just on the time
spent on content).
For the perceptions and behaviours relating to gesturing, Figure 3 shows in blue (dashed) the learners’ perception scores when evaluating
the statement ‘the teacher gestures with his/her hands when talking’ (maximum score = 5,
left y-axis) and in orange (full line) the observed percentage of time each teacher spent
talking about content-related issues and gesturing at the same time (maximum score = 100%,
right y-axis).
Figura 3. Time spent gesturing while delivering content and learner’s perception of teacher
gesturing
![]()
As Figure 3 shows learners’ perceptions of gestures
seem to be close to the actual observations. The Spearman correlation coefficient rho
indicated a moderate-high positive correlation between the perceptions and the actual
behaviours, however it was not significant: Spearman’s rho = 0.683 p =
0.062, CI [0.937, —0.042], Fisher’s z = 0.835. The lack
of statistical significance means that the results are not conclusive.
We also ran a correlation between the actual percentage of time spent gesturing and the
teachers’ evaluation scores given by learners and found a positive significant
correlation, Spearman’s rho = 0.81 p = 0.022, CI [0.964,
0.244], Fisher’s z = 1.126. These results indicate a strong positive
correlation between gesturing and a positive evaluation by the learners, with a
statistically significant result and a high level of confidence. Test 2 had found a
moderate-low correlation between gesture perception and the learner’s evaluation of the
teacher but test 3 suggests that gestures might have a stronger effect in appealing to
learners’ emotions than they might be aware of.
For the behaviours regarding addressing learners by name, moving around the classroom and
being funny, we noted the number of instances the teacher used names, walked around the
room (not just a few steps right-left or front —back but using the full length/width of
the room) and how many times learners laughed. In most cases there were only a few
instances of these behaviours.
As can be seen from Figure 4, learners’ perceptions
seem to be quite closed to the observed reality in respect to the use of humour for five
out of the eight teachers. We found a positive and significant correlation between the
actual use of humour and the teachers’ evaluation scores given by learners, Spearman’s
rho = 0.846 p = 0.008, CI [0.971,
0.35], Fisher’s z = 1.242. Although there was a moderate positive
correlation between observed instances of humour and learners’ evaluation of the teachers,
this was not significant, Spearman’s rho = 0.589 p =
0.124, CI [0.914, —0.197 ], Fisher’s z = 0.676, making
the results inconclusive. This is likely to be related to the size of the sample and the
few data points observed.
Figura 4. Perceived and actual use of humour
![]()
Figura 5. Perceived and actual use of classroom space.
![]()
Learners’ perceptions also seem to be close to the observed reality in respect to how the
teachers use the space (Figure 5). We found a
moderate positive correlation, not significant, between the actual use of space and the
teachers’ evaluation scores given by learners, Spearman’s rho = 0.518
p = 0.188, CI [0.896, —0.294], Fisher’s
z = 0.574. There was no significant correlation between observed
instances of walking around and teachers’ evaluation Spearman’s rho =
0.391 p = 0.338, CI [0.859, —0.433], Fisher’s
z = 0.413. These results could be due to a lack of data (there were few
instances of the teachers using all the classroom space) but should be investigated
further.
4.4. Type of teachers’ activity and learners’ evaluation
The fourth analysis focused on the breakdown of the class by type of speech and activity:
time spent on exercises, on self-disclosure, on content and on the learners (see Table 1). We were not able to find any significant
correlations between the time spent on any of these activities and the evaluation of the
class/topic or the teacher (see Appendix 3). Teachers spent a relatively low percentage of class time talking
about themselves, or giving student related examples or information, perhaps because of
this, there is no significant correlation between this type of content and the evaluations
of the class or the teacher. Some of the teachers spent most of the class time on
exercises while others were mostly delivering content.
For all teachers the proportion of time spent on self-disclosure was very low. This is a
behaviour associated with immediacy in North American classrooms but they might not be
appropriate in the Chinese classroom as it breaks down the expected power-distance
dynamics (Myers et al., 1998; Zhang, 2007), making learners uncomfortable. The
correlations of time spent on the learner and the evaluations of the class and teacher
were the only ones to have a moderate Spearman’s correlation coefficient
rho = 0.599 and 0.527
respectively, but in neither case was it significant, although the effect size was also high,
Fisher’s z = 0.691 and 0.586. Perhaps the same factors that mitigate the
effect of getting close to learners in the Chinese classroom are at play here, and some
students do not feel comfortable being the object of attention.
Tabla 1. Observed proportion of time spent on different activities
| Teacher |
Class evaluation |
Teacher evaluation |
Time on exercises |
Time on selfdisclosure |
Time on content |
Time on learner |
| 1 |
4.38 |
4.47 |
79% |
2% |
14% |
8% |
| 2 |
4.19 |
4.39 |
35% |
7% |
37% |
17% |
| 3 |
4.32 |
4.62 |
48% |
0% |
18% |
30% |
| 4 |
3.89 |
4.03 |
65% |
0% |
17% |
1% |
| 5 |
4.16 |
4.53 |
32% |
0% |
56% |
2% |
| 6 |
4.15 |
4.26 |
23% |
3% |
58% |
5% |
| 7 |
4.06 |
4.13 |
4% |
4% |
56% |
9% |
| 8 |
4.1 |
4.14 |
63% |
0% |
24% |
1% |
5. Conclusions
This study sought to identify the accuracy of learners’ perceptions by comparing them with
researchers’ observations of actual teachers’ behaviours and to test whether there is a
correlation between teachers’ behaviours and learners’ evaluations of the class/topic and
the teacher.
In the first instance our results confirm those from other studies that have observed a
positive correlation between learners’ interest in the topic (based on the questions
addressing class/topic evaluation) and perceived teachers’ credibility and care for them
(based on the questions addressing teachers’ evaluations). In this study, learners’
evaluations of teachers were always higher that the evaluations of the class/topic. It is
likely that there is a causality effect, teachers who are well-regarded lead to high
evaluations of the class, however that was not the focus of the study and, although
suspected, cannot be confirmed.
The answers to the second question, whether learners’ evaluations
of the class and the teacher can be correlated to those behaviours the
literature identifies as developing immediacy, indicate that
classrooms in the Hong Kong context are more similar to those in
Mainland China than North American ones. Learners’ evaluations of the
class/topic and the teacher were positively and significantly
correlated with the perceived teachers’ verbal and non-verbal
immediacy-creating behaviours tested. In the participants tested, the
three highest correlations were found with teacher smiling, being
funny and looking at the students, behaviours that can be associated
with caring and relaxed teachers. Teachers gesturing and moving around
the classroom, were noted to also have positive correlations but not
as strong. In the case of gestures, further research is called for as
their effect on learning has been noted by a number of studies, but
also seems to be affected by a number of factors, including personal
differences (Lopez-Ozieblo et al., under review).
Overall, our results suggest that learners’ perceptions are not
that far from the reality of the classroom. When observing time spent
gesturing, use of humour and teachers’ use of space (moving in the
classroom), our results show high and moderate-high positive
correlations between the actual behaviours and learners’ perceptions
(although the correlation on use of space was not significant). The
first two actual behaviours were also found to have a positive
significant correlation with learners’ evaluations of the teacher.
We are well aware that there are a number of limitations to our
study, but it certainly highlights the effects of teachers’ behaviours
in the creation of affect and an emotionally positive context for
learners. The main limitation of the study is that there are many
variables and these are difficult to separate. We have tried to
identify some, such as the content of the teachers’ speech, but many
others are not addressed, due to the size of the sample. Our sample
population was small (with just eight teachers from one specific
discipline) with additional factors that might also be variables, such
as gender, nationality and experience. Other variables that need to be
identified and, if possible, isolated are learners’ gender, size of
class, postgraduate vs. undergraduate and even issues related to
context, such as time of day, location, number of learners, core or
elective subjects and content. The behaviours we analysed were the
ones we could more readily observe. However, there are others related
to immediacy that we were not able to report on, such as smiling.
5.1. Pedagogical implications
Our results corroborate Andersen’s (1979) original relationship between perception and
observed behaviour, contradicting those of Smythe and Hess (2006). Our results support
findings about the emotional impact of the teacher in the learning process (Dewaele et
al., 2022), even if learners’ awareness of this varies. Learner’s perception as to the
teachers’ use of humour corresponds quite closely to the reality observed, indicating that
learners are attuned and aware of this behaviour. However, learners might not be
consciously aware of teacher’s gestures and movements in the classroom and so they might
not value them as highly as they might value humour. As our results indicate that teachers
who gesture more when explaining content were rated higher than those who gestured less,
teachers might want to think about their pedagogical gestures. Teachers’ perception of
gestures is usually positive (see Nathan et al., 2019 for a survey of American teachers),
believing that gestures can help to attract students’ attention and to convey complex
ideas more clearly.
In the Hong Kong context, learners’ perceptions of gestures are often shaped by popular ideas
on body language and the extend to which gestures express emotions, rather than content
(ongoing study with Hong Kong students). To enhance the benefit of teacher’s gestures it
might be necessary to first raise students’ awareness as to the value of gestures, not as
tools to express confidence or hide anxiety but as conveyors of meaning. This awareness
raising could be done by asking learners to gesture themselves, copying teachers’ gestures
or producing their own, which will have the added benefit of enhancing learning and recall
(Macedonia, 2019).
Our study indicates that students’ evaluation of the course and content under study is partly
determined by the behaviour of the teacher. These observations are not new, but teacher
training is still mostly focused on methods and strategies to convey content with less
attention paid to behavioural and emotional factors. We believe that students’ classroom
anxiety and enjoyment is affected by teacher’s behaviours and emotional state, a point
that every teacher should be aware of. In the language classroom, teachers need to be
aware that their behaviour can help to produce an affective context which might enhance
learning and the overall evaluation of the subject. We call for further research in this
area as students’ emotional state, prior knowledge and extraneous factors are likely to be
affecting their perceptions. Learners are more likely to note unusual behaviours, rather
than recurrent ones, the former being more likely to heighten emotions in students. Future
studies might want to test this and also analyse each behaviour independently and identify
whether specific ones hold greater affective weight with students.