Investigating How Social Elements Affect Learners with Different
Personalities
Wad Ghaban, Robert Hendley and Rowanne Fleck
School of Computer Science, University of Birmingham, B15 2TT, U.K.
Keywords:
Online Learning, Social presence, Performance, Satisfaction, Content Analysis.
Abstract:
Social presence is an essential factor in preventing learners from feeling isolation in online courses and in keep-
ing them connected. Some studies, however, point to the negative impact of the social elements in distracting
learners from concentrating on a course’s content. In this study, we investigated the influence on learners of
different personality types (using the big five model) of an optional chat added to an online learning platform.
The results show that there is a variation in the response of the learners based on their personality. However,
some personality classes spent the majority of time in the chat discussing off-topic subjects, such as fashion
or travel. Thus, although they enjoyed the features in the system, it negatively affected their knowledge gain.
We discuss the implications of our findings for adaptive online learning platforms in catering to learners with
diverse personalities.
1 INTRODUCTION
A social presence in online learning courses is be-
coming very important; it prevents learners from feel-
ing isolated and makes them feel that they belong
to the course and are connected to other learners
(Means et al., 2009). However, some researchers have
claimed that learners have different responses to so-
cial elements. Some learners find these elements en-
joyable and motivational (Kehrwald, 2008). Other re-
searchers argue that social elements are uninteresting
and cannot represent real human interaction (Cobb,
2009). Because of this difference in the perception of
social elements in online courses, we aimed to un-
derstand how different learners perceive social ele-
ments. In this research, we used personality as a
stable characteristic that can be used to describe hu-
man behaviour (Hogan and Hogan, 1989). Although
many personality theories can be considered to un-
derstand the effect of social elements, we adopted the
Big Five personality traits, a commonly used theory
to define personality. The Big Five personality traits
classifies personality into five dimensions: consci-
entiousness, extraversion, agreeableness, neuroticism
and openness to experience (Hofstee, 1994), and dif-
ferent characteristics are associated with each type of
personality.
To understand the effect of access to chat on dif-
ferent personalities, We asked two groups of learners
to use one of two versions of a learning website, either
one that included chat or one that did not. Then, we
analysed the number and the type of messages from
learners with different personalities.
We hypothesised that learners with different per-
sonalities would have varied responses to access to
chat in online learning courses. Highly extroverted
and highly agreeable learners are usually described as
social, and they like to compete and collaborate with
others (Hofstee, 1994). Therefore, we hypothesised
that these learners would send a high number of mes-
sages, which may enhance their knowledge gain and
satisfaction. Highly conscientious learners are de-
scribed as always being organised and self-triggered
to complete their tasks (Hofstee, 1994). For these
learners, we hypothesised that these learners will use
the chat in appropriate way. These learners would re-
port the same level of knowledge gain and satisfac-
tion for both versions. Further, highly neurotic learn-
ers are described as having high emotional instability.
Because of this, we hypothesised that these learners
might not be satisfied with the chat and have lower
knowledge gain in the case of online learning courses
that included chat (Judge et al., 1999).
The results confirmed our hypothesis that there
is a variation in the response to the social elements
from the different personalities. Some learners en-
joyed using the chat to talk about many unrelated top-
ics, which negatively affected their knowledge gain.
416
Ghaban, W., Hendley, R. and Fleck, R.
Investigating How Social Elements Affect Learners with Different Personalities.
DOI: 10.5220/0007732404160423
In Proceedings of the 11th International Conference on Computer Supported Education (CSEDU 2019), pages 416-423
ISBN: 978-989-758-367-4
Copyright
c
2019 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
Other learners, such as the highly conscientious ones,
used the chat in an appropriate way, which did not
affect their knowledge gain.
This kind of research is important for understand-
ing the effectiveness of online learning platforms.
Learners’ interactions with online learning platforms
need to be observed and monitored by teachers. A
good teacher will then direct the interactions in a way
that ensures that every learner is enjoying and learn-
ing from them. This is what is expected from future
technologies: looking to the learners’ characteristics
and then directing their interactions in an appropriate
way that meets the learners’ expectations. However,
to achieve this, we need a strong understanding of the
effect social components have on learners by looking
to other factors, such as the learners’ moods and their
effective state.
2 BACKGROUND
Online learning was defined by (Richardson and
Swan, 2003) as any course that offers an entire cur-
riculum online and gives learners the ability to ac-
cess the materials anytime from anywhere (Anderson,
2008). (Richardson and Swan, 2003) noted that in on-
line learning, learners and teachers no longer have to
meet physically in order to share knowledge. These
courses are free from the constraint of time and space
(Ally, 2004). However, learners in online courses lack
face-to-face interaction and often miss the feeling of
belonging to a class, which may result in feelings of
isolation. Much research has suggested adding social
components to prevent learners from feeling isolated
and to keep them connected to one another (Means
et al., 2009).
2.1 Social Presence
(Garrison, 2007) noted that three elements must be
present for success in learning: teacher presence, cog-
nitive presence and social presence.
Social presence in online courses can be defined
as the degree to which participants in computer-
mediated communication can effectively connect
(Swan and Shih, 2005). (Tu and McIsaac, 2002) de-
fined social presence as the degree to which partici-
pants are aware of others in real interaction.
(Lowenthal, 2010) categorised social presence
into three categories: a) effective responses contain-
ing personal expressions of emotions and feelings, b)
cohesion responses containing expressions related to
building and supporting relationships, such as greet-
ings and c) interactive responses containing agree-
ments and disagreements. (Richardson and Swan,
2003) found that learners usually begin interactions
with a large number of cohesion responses. How-
ever, over time, the number of interactive responses
increased.
(Tu and McIsaac, 2002) noted that social presence
elements are an essential factor of learners’ success
in online courses. Social presence can occur in on-
line courses in many forms, such as chats, discussion
boards and emails (Aragon, 2003). Interactions be-
tween learners can contain texts, voice messages and
emojis (DeSchryver et al., 2009). Research shows
a strong correlation between a high social presence
and the learners’ outcomes (Richardson and Swan,
2003). (Swan and Shih, 2005) pointed out that learn-
ers who felt more comfortable using social elements
scored higher in their perceived learning. (Swan and
Shih, 2005) examined the effect of social elements on
learners’ satisfaction and achievement by asking 91
students to enroll in one of two versions of a course:
one with social elements and the other without. At
the end of the experiment, they measured how social
the learners were (high, medium or low) using a spe-
cial questionnaire. They also measured learners’ out-
comes and satisfaction levels. They found that learn-
ers who were more social were more satisfied and had
better perceived outcomes. These results were sup-
ported by (Richardson and Swan, 2003), who indi-
cated that the high use of social elements is a predictor
for learners’ satisfaction levels and outcomes. (McIn-
nerney and Roberts, 2004) pointed out that these ele-
ments can prevent learners from feeling isolated.
(Tu and McIsaac, 2002) showed that learners in
online courses need more information about other
learners’ profiles in order to have a more effective
interaction. However, (Garrison, 2007) argued that
some problematic issues exist pertaining to social ele-
ments in online courses, as text-based communication
cannot replace face-to-face expression and body lan-
guage. In addition, delayed responses from the learn-
ers may annoy some learners, whereas other learners
may feel insecure about computer-to-computer inter-
action. (Swan and Shih, 2005) interviewed learners
and asked them about their opinion of the social ele-
ments. Some learners described text communication
as ’cold’ and unlike physical interaction. Other learn-
ers stated that social elements motivate them to com-
plete the course. However, other learners mentioned
that social elements are not challenging or interesting.
Therefore, (Cobb, 2009) explained that the responses
of different learners towards social elements in online
courses varied. Thus, it is essential to understand how
learners with diverse personalities respond to social
elements in online courses.
Investigating How Social Elements Affect Learners with Different Personalities
417
2.2 Personality
Personality can be described as a set of character-
istics that describe how individuals think and feel
(Hofstee, 1994). (Hogan and Hogan, 1989) stated
that personality is consistent, and it may develop
over time. There are different theories used to
describe personality. For example, there are three
common theories of personality: Eysenck’s theory of
personality, the Myers-Briggs Type Indicator and the
Big Five personality traits (Claridge, 1977). In this
research, the focus will be on the Big Five, which is
widely used in similar research (Wiggins, 1996).
Big Five Model. This theory is one of the most
popular theories used to explain and understand
personality (Wiggins, 1996). This theory classifies
learners’ personalities into five dimensions (types):
conscientiousness, extraversion, agreeableness,
neuroticism, and openness to experience. Table 1
summarises the traits associated with each personality
type.
Big Five Model Measurements. The most com-
mon measurements are the NEO Five-Factor Inven-
tory (the NEO-FFI) and the Big Five Inventory (BFI).
Many versions of the NEO-FFI have been developed.
Some of these versions have 240 questions, mak-
ing them time-consuming to complete (Laidra et al.,
2007). Thus, new short versions have been proposed.
However, these shorter versions suffer from reliability
problems. As a result, many researchers use the BFI,
which is considered more reliable than the NEO-FFI.
In addition, the BFI is free to use, and it exists in sev-
eral languages. Some versions are designed for chil-
dren and others for parents. In this research, we used
the BFI (46 questions) which was developed for chil-
dren (McInnerney and Roberts, 2004).
Table 1: A Summary of the Big Five Personality Traits
(adapted from (Wiggins, 1996)).
Personality Characteristics
conscientiousness
(a strong sense of purpose
and a high aspiration level)
Leadership skills, capability to make long-term
plans and often has an organised support network
Extroverted
(a preference for companionship
and social simulation)
Has good social skills and numerous friendships,
often participating in team sports and having
club memberships
Agreeable
(a willingness to defer to
others during interpersonal
conflicts)
Forgiving attitude and a belief in cooperation
Neurotic
( sadness, hopelessness
and guilt)
Low self-esteem and irrational
and perfectionistic beliefs
Openness to experiences
(a need for variety, novelty
and change)
Interested in different hobbies
and knowledgeable about foreign cuisine
3 METHOD
This study aims to investigate how different person-
alities interact with existing social components (chat)
in online learning system. In addition, we aim to
examine the effect of social interactions on learners
with varying personalities regarding their knowledge
gain and satisfaction.
Setup. We built an online learning system to teach
Microsoft Excel. The course consisted of 15 lessons
designed by the researchers, starting with simple top-
ics, such as drawing tables and visualising graphs.
From there, the course progressed to high-level top-
ics, such as mathematical and logical functions. We
built the website in two identical versions. One ver-
sion included a social component (chat); the other
lacked this component. In the version that included
chat, learners could use a button labelled ’Talk to a
friend’. When the learner clicked on this button, the
chat form appeared (Figure 1 ). In this version, learn-
ers believed that they were talking to another learner,
whilst, in fact, they were talking to a researcher who
was following a predefined script. We used this tech-
nique in order to control the conversation and to en-
sure that each learner experienced the same condi-
tions.
At the start, learners were required to set up a user-
name and a password. Subsequently, learners were
asked to supply demographic information, including
their age and gender. Learners were then asked to fill
in a BFI personality test. Finally, learners were asked
to fill in a pre-test related to the course itself.
Figure 1: A Screenshot of the Website Showing the Chat.
Participants. Before running the experiment, we
were granted ethical approval from four schools in
Saudi Arabia. Then, we sent a consent form to learn-
ers’ parents in which the school explained the pur-
pose of the experiment and informed them that all the
collected data would be anonymous and secure. The
learners and their parents were made aware that the
learners were free to dropout at any time.
After obtaining the consent forms, 194 learn-
CSEDU 2019 - 11th International Conference on Computer Supported Education
418
ers (91 boys, 103 girls) participated in the experiment.
The Classification of Personalities. In this study,
we were concerned with understanding how differ-
ent personalities would deal with existing social com-
ponents. Accordingly, after obtaining the personality
score from the BFI personality test, we classified each
personality type into high, medium and low. Figure 2
shows an example of the classification of the extraver-
sion personality. The cut-off points were arbitrarily
chosen. To perform the classification, we drew a his-
togram that shows the values of the personality di-
mension on the x-axis and the frequency of the learn-
ers with that personality type on the y-axis. Then, we
classified learners lower than µ σ as low. Learners
were assigned values for a specific personality trait
above µ + σ.
In this study, we believe there will be a strong ef-
fect on learners with extreme personalities. Conse-
quently, we only focus on the learners who are high
and low on each extreme of the personality type.
The value of the
personality
Histogram of Exp2TableTime2$op
Openness Personality
Frequency
0 1 2 3 4 5
0 5 10 20 30
!=2.6
!+"=3.9!-
"=1.31
Figure 2: The Classification of the Extraversion Personality.
Procedure. After obtaining the consent form, we
asked learners to fill in a demographic questionnaire,
the BFI personality test and the pre-test. Then, learn-
ers were divided equally into two groups: one group
used the website including the chat, and the other
group used the version without the chat. The two
groups were balanced in their age, gender, knowledge
level and personality score. Next, we asked learners
to use the website any time they liked; they were free
to dropout at any time.
After six months, most of the learners had either
dropped out or completed the course. Thus, after two
months, we asked learners to fill in a post-test that
related to the course and with the same number of
questions as the pre-test. Then, we calculated their
knowledge gain as:
Knowledge Gain = learners’ post-test results - learn-
ers’ pre-test results
We also asked learners to fill in a satisfaction
questionnaire. To measure their satisfaction, we used
the e-learner satisfaction tool (ELS) developed by
Wang (Wang, 2003). This tool consists of several
components, including a system interface, learning
content and system personalisation. The ELS tool
comprises 13 questions with a seven-point Likert
scale ranging from strongly disagree’ to strongly
agree’.
Hypotheses. In this study, we assume that there will
be variation in the response of the learners towards
the existing social elements. (Laidra et al., 2007) de-
scribe highly conscientious learners as always achiev-
ing well, and always making good progress. We hy-
pothesise that learners with this personality will use
the chat properly and that the conversation with the
learners will be on topics related to the course, most of
the time. These learners will always have high knowl-
edge gain and high satisfaction, regardless of whether
the system has social elements or not.
In contrast, highly extroverted learners, as de-
scribed by (Laidra et al., 2007), are likely to enjoy
making social relationships and interacting with oth-
ers. Thus, these learners will use the chat frequently,
and most of the messages will be off-topic. This may
enhance their knowledge gain and satisfaction.
Highly agreeable learners are usually described as
kind, and they like to collaborate with others. As a
result, we hypothesise these learners will spend their
time talking about the course and trying to help the
ones with whom they interact. This may enhance their
knowledge gain and their overall satisfaction.
4 RESULTS
In the study, we aimed to examine the influence of
the presence chat on learners with different person-
alities. For that, we looked to the number of mes-
sages sent from each personality (table 2). There was
a high number of messages sent from some personali-
ties compared to other personalities. Thus, we believe
it is interesting to find the kinds of messages received
from learners and to find if the kind of messages has
any effect on learners’ knowledge gain and their satis-
faction. Thus, we went to analyse the topic discussed
in the messages.
4.1 Content Analysis
In this study, we received almost 7,000 messages in
50 days from different learners with different person-
alities. Hence, as a first step, we cleaned our data
by removing all images and emojis. In the next step,
each message was independently coded by 2 annota-
tors, based on the message content. (We asked two na-
tive Arabic speakers to annotate each messages based
Investigating How Social Elements Affect Learners with Different Personalities
419
Table 2: The Number of Messages and the Number of Chats
Opened for Learners with Different Personalities.
Personality
Number
of learners
Number
of messages
Number
of chats opened
Average
(messages
per chat)
High
conscientious
22 1,238 53 23.3
Low
conscientious
18 1330 58 22.9
High
extraversion
22 3,317 109 30.4
Low
extraversion
25 343 26 13.19
High
agreeableness
16 1,076 44 24.4
Low
agreeableness
14 1305 36 36.25
High
neuroticism
18 978 46 21.2
Low
neuroticism
19 2486 94 26.44
High
openness
21 1,749 70 24.9
Low
openness
19 1654 55 30.7
Total 194 15476 591 26.1
on the codes presented in our codebook.) Table 3
shows the coding scheme for the messages. After that,
we measured the inter-rater reliability of our coding.
Then, we calculated the inter-rater reliability, which
is used to measure the level of agreement between
raters (Berelson, 1952) (Miles et al., 1994). The re-
sults show an inter-rater agreement of 97%.
Table 4 shows the number of messages in each cat-
egory sent from each personality.
Later, we looked in our data and tried to find
a common pattern in terms of learners’ personali-
ties. We found that highly conscientious learners of-
ten included a greeting in their messages and that
their messages generally remained on topic. Even
when conscientious learners discussed something off
topic, they were usually discussing the experiment
and what they needed to do to complete the experi-
ment. In contrast, highly extroverted learners often
sent multiple messages, began their messages with
greetings, and had off-topic discussions. For example,
they talked about football games and tourist places
to visit. Highly agreeable learners often used mes-
sages to introduce themselves and build relationships.
For example, some of these learners were asking if
they could have a real-time meeting. Highly open
learners preferred to talk about travel and fashion,
and they usually talked about their favourite places
to visit. These learners usually send links about the
course. Highly neurotic learners rarely opened the
chat dialogue. Most of their discussions were off
topic. They were complaining about their school and
homework. For example, one asked, ”Do you have
the same amount of homework as we have? I am tired
from all of it”. Figure 3 shows an example of the pat-
tern of messages from the different personalities.
Table 3: The Codebook Used to Define the Messages.
Code Definition Example
T
(Topic)
Messages relevant to the
topic (Microsoft Excel)
Which operation has priority in the
following formula in Excel:
3 2 + 1/4
O
(Off-topic)
Messages related to any
topic other than Microsoft
Excel (e.g, fashion, travel,
weather)
Who was the winner of the last football
game between Al-Hilal and Al-Naser?
G
(Greeting)
Messages representing all
greetings, such as ’hello’,
or welcoming messages,
such as’ thank you’.
How are you? Where are you from?
GM
(Gamifica-
tion)
messages related to
gamification, such as
points and badges.
How many points and badges
did you collect?
Table 4: The Total Number of Messages (On-topic, Off-
topic, Greeting and Gamification).
Personality In-topic Off-topic Greeting
Game-
fiction
Total
High conscientious
N 300 574 321 43 1238
% 24.2 46.3 25.9 3.6 100
Low conscientious
N 148 507 639 36 1330
% 11.1 38.1 48.1 2.7 100
High extraversion
N 571 1612 864 270 3317
% 17.2 48.5 26.02 8.1 100
Low extraversion
N 54 152 123 14 343
% 15.7 44.32 35.8 4.08 100
High agreeableness
N 186 419 398 73 1076
% 17.5 38.9 36.9 6.7 100
Low agreeableness
N 164 599 443 99 1305
% 12.5 45.9 33.9 7.7 100
High neuroticism
N 98 414 433 33 978
% 10 42.3 44.3 3.3 100
Low neuroticism
N 454 1118 743 171 2486
% 18.38 44.9 29.8 6.8 100
High openness
N 265 759 638 87 1749
% 15.5 43.3 36.4 4.8 100
Low openness
N 258 765 480 151 1654
% 15.5 46.2 29.2 9.1 100
Total number of messages= 289
Total number of messages= 39
Total number of messages= 98
a)
b)
c)
Greeting
messages
In
-topic
messages
Off
-topic
messages
Gamification
elements
Figure 3: A Sample from the Pattern of Learners’ Conver-
sation for: a) Highly Extrovert, b) Highly Conscientious, c)
Highly Agreeable learners.
Additionally, to understand the effect of the num-
ber and the type of messages on learners with differ-
ent personalities, we looked to the knowledge gain
and the satisfaction of the learners. Table 5 sum-
marises the results of the learners’ knowledge gain
in both versions, while Table 6 summarises the re-
sults of satisfaction. Both tables show variations be-
tween personalities in response to the chat. For exam-
ple, highly neurotic learners were not satisfied with
the chat. However, highly extroverted learners are
shown to have the most learners affected by the chat.
These learners were satisfied with the presence of the
chat. However, this negatively affected their knowl-
edge gain, perhaps because learners were busy talking
CSEDU 2019 - 11th International Conference on Computer Supported Education
420
Table 5: The Summary of the Results of the Knowledge
Gain for the Personalities.
Personality
Knowledge gain
in version includes
chat
Knowledge gain
in version without
chat
The benefit
of the chat
N Mean Sd N Mean Sd
High
conscientious
22 2.12 1.5 12 2.58 1.7 -0.46
Low
conscientious
16 2.5 0.45 13 2.15 1.86 0.35
High
extraversion
23 1 2.1 24 2.0 1.45 -1.04
Low
extraversion
22 2.04 1.9 12 1.3 1.55 0.71
High
agreeableness
16 1.2 2.4 17 2.17 1.81 -0.97
Low
agreeableness
15 1.06 2.3 12 1.91 1.44 -0.85
High
neuroticism
18 0.87 2.2 15 1.2 2.27 -0.33
Low
neuroticism
26 1.61 1.9 14 2.5 1.28 -0.89
High
openness
21 1 2.4 16 2.37 1.5 -1.37
Low
openness
26 1.34 1.5 14 1.64 1.9 -0.3
Table 6: The Summary of the Results of the satisfaction for
the Personalities.
Personality
Satisfaction in
version includes
chat
Satisfaction in
version without
chat
The benefit
of the chat
N Mean Sd N Mean Sd
High
conscientious
22 6.67 0.6 12 6.1 0.76 0.57
Low
conscientious
16 6.4 0.78 13 6.07 0.73 0.33
High
extraversion
23 6.64 0.58 24 6.1 0.49 0.54
Low
extraversion
22 6.6 0.53 12 6.01 0.73 0.59
High
agreeableness
16 6.34 0.9 17 6.2 0.96 0.14
Low
agreeableness
15 6.4 0.78 12 6.3 0.77 0.1
High
neuroticism
18 5.1 0.7 15 6.3 0.87 -1.2
Low
neuroticism
26 6.3 0.78 14 6.3 0.75 0
High
openness
21 6.5 0.78 16 6.3 0.83 0.2
Low
openness
26 6.3 0.87 14 6.3 0.8 0
and chatting, not concentrating on the course itself.
5 DISCUSSION
This study tried to investigate the behaviour of the
presence of the chat in an online learning course on
learners with different personalities.
Highly conscientious learners are described by
(Laidra et al., 2007), as those learners who always do
their job without the need for external factors. These
learners spent their time in the chat talking about the
course itself or about off-topic subjects that were still
related to the experiment. For example, they asked
each other how much time they needed to finish the
course or if they needed to contact someone when fin-
ishing an experiment. This may explain why these
learners were more satisfied in the version which in-
cludes the chat. Further, highly conscientious learn-
ers always have their own trigger to motivate them.
For that, these learners have the same level of knowl-
edge gain in both versions (including and not includ-
ing chat).
Highly extroverted learners spent their time chat-
ting about topics not offered by the course. This chat-
ting enhanced these learners’ satisfaction. However,
the results from the knowledge gain differ from those
in related work that suggest that the existence of so-
cial components enhances these learners’ outcomes
(Swan and Shih, 2005). The knowledge gain of the
highly extrovert learners was worse in the version in-
cluding chat. This result may be explained by the
fact that highly extrovert learners are easily distracted.
However, this result cannot be guaranteed and we may
need to run another study to validate this.
Highly agreeable learners ranked second in send-
ing the highest number of messages. Furthermore,
most of the messages were classified as a ’greet-
ing’. These learners tried to build a relationship with
their conversational partner. For example, they asked,
’How are you?’, ’Where are you from?’ and ’Can
we meet somewhere?’. The messages sent from this
personality adversely affected their knowledge gain.
These learners’ knowledge gain was better in the ver-
sion that without the chat. Further, the satisfaction of
these learners was almost the same in the both ver-
sions.
For the highly neurotic learners, we hypothesised
that these learners will be demotivated because of the
chat. These learners did not use the chat, and there
were very few messages sent from this personality.
Further, most of the messages sent from these highly
neurotic learners were from those students with very
high values in other dimensions, such as agreeable-
ness. This may be explained by the nature of the
study, as the chat was optional, and the learner had
to take the initiative.
Highly open learners were shown to have the same
level of satisfaction in both versions (including and
not including chat). However, the knowledge gain
was negatively affected by the chat. This might have
happened because of these learners’ lack of interest in
the chat. These learners are described by (Costa and
McCrae, 2008) as learners who are more likely imag-
inative and enjoy unusual events. Thus, the chat may
not interest these learners.
In our study, we noticed that some learners opened
the chat dialogue without starting a conversation.
Moreover, when learners started a conversation, they
Investigating How Social Elements Affect Learners with Different Personalities
421
were unsure with whom they were speaking. This
may have prevented learners from participating ac-
tively in conversations. (Cobb, 2009) pointed out that
learners speaking to anonymous entities may feel un-
comfortable participating in such conversations. In
addition, delays in sending and receiving messages,
for example, due to a poor Internet connection, could
have affected the learners’ behaviour.
Thus, because of the previous shortcomings, we
may need to conduct a further experiment with more
realistic chat. We can, for example, make the learn-
ers contact each other and record the conversation be-
tween them. In addition, we can repeat the same ex-
periment, but take the initiative to discern how differ-
ent personalities will respond.
6 CONCLUSIONS
Because of the lack of physical interaction be-
tween learners in online courses, many studies have
suggested incorporating social elements into those
courses to prevent learners from feeling isolated. In-
teractions with teachers and other learners can be
achieved using chats, discussion boards and emails
(DeSchryver et al., 2009). However, some research
has claimed that some learners do not prefer to talk
to others, while others may get distracted because of
these interactions (Laidra et al., 2007). Some learners
may also feel insecure in online interactions (Tu and
McIsaac, 2002). Because of these varied responses to
social components, we designed this study to inves-
tigate how different personalities respond to existing
social components, such as chat.
The results from our study confirm the variation of
the effect of chat on the different personalities. Some
learners enjoyed using the chat. Some personalities
spent their time talking about off-topic subjects rather
than the course, while others preferred to build rela-
tionships and introduce themselves. This variation in
the response to the chat affected learners’ knowledge
gain and satisfaction. For example, Some personali-
ties are not expected to show any difference in their
knowledge gain and satisfaction either with or with-
out the social element, such as highly conscientious
learners. However, other learners are more satisfied
where there is a social element. Meanwhile, some
learners have less knowledge gain where there is chat,
such as the highly extrovert learners. These learners
have the highest number of messages. Most of their
messages are about topics other than the course itself,
for example: fashion, travel and sport. To enhance the
knowledge gain of these learners, we may need to ob-
serve the behaviour of these learners, and then direct
the conversation back to the topic itself. For example,
if the learners start to talk about an off-topic subject,
one might say, ’This is okay, but let’s talk about the
course’.
This study provides insight into the type and num-
ber of messages sent by different personalities. How-
ever, in this experiment, only a few learners with
extreme personalities were included. In addition,
there was a positive correlation between personalities,
which may have resulted in bias. In this experiment,
the learners had to take the initiative and start a con-
versation; as such, some learners chose not to talk.
Thus, we could not examine the effect of the chat on
them. Furthermore, the researchers responded to the
learners, which may have resulted in delayed or unin-
teresting responses. Thus, further studies need to be
conducted to have a better understanding of the effect
of social components on learners. In this study, we
used personality as a stable characteristic related to
individuals’ behaviour. However, there are other char-
acteristics that may be considered (e.g. learners’ ef-
fective state, mood and learning style). After building
a good understanding about what is needed and liked
by each learner, we can build a system that provides
a dynamic place for learners’ conversations. This can
be done by controlling and directing the conversation
for some learners or by encouraging some learners to
interact more with others.
REFERENCES
Ally, M. (2004). Foundations of educational theory for on-
line learning. Theory and practice of online learning,
2:15–44.
Anderson, T. (2008). The theory and practice of online
learning. Athabasca University Press.
Aragon, S. R. (2003). Creating social presence in online en-
vironments. New directions for adult and continuing
education, 2003(100):57–68.
Berelson, B. (1952). Content analysis in communication
research.
Claridge, G. (1977). Manual of the eysenck personality
questionnaire (junior and adult): Hj eysenck and sybil
eysenck hodder and stoughton (1975). 47 pp., together
with test blanks and scoring keys for junior and adult
versions. specimen set£ 1.80.
Cobb, S. C. (2009). Social presence and online learning: A
current view from a research perspective. Journal of
Interactive Online Learning, 8(3).
Costa, P. T. and McCrae, R. R. (2008). The revised neo
personality inventory (neo-pi-r). The SAGE handbook
of personality theory and assessment, 2(2):179–198.
DeSchryver, M., Mishra, P., Koehleer, M., and Francis,
A. (2009). Moodle vs. facebook: Does using face-
book for discussions in an online course enhance per-
ceived social presence and student interaction? In
CSEDU 2019 - 11th International Conference on Computer Supported Education
422
Society for Information Technology & Teacher Edu-
cation International Conference, pages 329–336. As-
sociation for the Advancement of Computing in Edu-
cation (AACE).
Garrison, D. R. (2007). Online community of inquiry re-
view: Social, cognitive, and teaching presence is-
sues. Journal of Asynchronous Learning Networks,
11(1):61–72.
Hofstee, W. K. (1994). Who should own the definition
of personality? European Journal of Personality,
8(3):149–162.
Hogan, J. and Hogan, R. (1989). How to measure employee
reliability. Journal of Applied psychology, 74(2):273.
Judge, T. A., Higgins, C. A., Thoresen, C. J., and Barrick,
M. R. (1999). The big five personality traits, general
mental ability, and career success across the life span.
Personnel psychology, 52(3):621–652.
Kehrwald, B. (2008). Understanding social presence in text-
based online learning environments. Distance Educa-
tion, 29(1):89–106.
Laidra, K., Pullmann, H., and Allik, J. (2007). Personal-
ity and intelligence as predictors of academic achieve-
ment: A cross-sectional study from elementary to sec-
ondary school. Personality and individual differences,
42(3):441–451.
Lowenthal, P. R. (2010). Social presence. In Social com-
puting: Concepts, methodologies, tools, and applica-
tions, pages 129–136. IGI Global.
McInnerney, J. M. and Roberts, T. S. (2004). Online learn-
ing: Social interaction and the creation of a sense
of community. Educational Technology & Society,
7(3):73–81.
Means, B., Toyama, Y., Murphy, R., Bakia, M., and Jones,
K. (2009). Evaluation of evidence-based practices in
online learning: A meta-analysis and review of online
learning studies.
Miles, M. B., Huberman, A. M., Huberman, M. A., and
Huberman, M. (1994). Qualitative data analysis: An
expanded sourcebook. sage.
Richardson, J. and Swan, K. (2003). Examing social pres-
ence in online courses in relation to students’ per-
ceived learning and satisfaction.
Swan, K. and Shih, L. F. (2005). On the nature and de-
velopment of social presence in online course discus-
sions. Journal of Asynchronous learning networks,
9(3):115–136.
Tu, C.-H. and McIsaac, M. (2002). The relationship of so-
cial presence and interaction in online classes. The
American journal of distance education, 16(3):131–
150.
Wang, Y.-S. (2003). Assessment of learner satisfaction with
asynchronous electronic learning systems. Informa-
tion & Management, 41(1):75–86.
Wiggins, J. S. (1996). The five-factor model of personality:
Theoretical perspectives. Guilford Press.
Investigating How Social Elements Affect Learners with Different Personalities
423