STUDENTS ONLINE INTERACTION IN A BLENDED
LEARNING ENVIRONMENT – A CASE STUDY OF THE FIRST
EXPERIENCE IN USING AN LMS
Ivana Mijatovic, Jelena Jovanovic and Sandra Jednak
Faculty of Organizational Sciences, University of Belgrade, Jove Ilica 154, Belgrade, Serbia
Keywords: Interaction, Learning Outcomes, Blended Learning, Case Study.
Abstract: The main objectives of the research presented in this paper are to explore online interactions and
engagement of students who are using a Learning Management System (LMS) for the first time in their
studies, and the impact of different types of students’ online interactions on their learning outcomes. To
answer our research questions, we have conducted a semester-long study with 88 undergraduate students
enrolled in the Quality Engineering course taught in the blended learning mode. Our findings show that the
students perceived interaction as a dominant aspect of the online part of the course (done using the Moodle
LMS). Our findings also provide evidence that different types of interactions can influences different levels
of learning outcomes. If the acquisition of factual knowledge is desired, then interaction with learning
content is the most influential. However, if higher levels of learning outcomes are to be achieved, then more
interactive online communication is needed. The need for interaction is rising with increasing levels of
learning objectives (outcomes). Our findings also show that students’ involvement in more challenging
activities, in order to fulfil more demanding learning objectives (like application of knowledge or analysis,
synthesis and evaluation) increase their need for student-teacher and student-student interaction.
1 INTRODUCTION
Many universities have integrated e-learning into
their courses aiming to widen the opportunities for
learning as well as to leverage the potential that
novel technologies offer for advancing the learning
process. Many authors reported successful usage of
communication and collaboration tools offered by
today’s Learning Management Systems (LMSs)
when applied in a blended learning environment
(Kember, McNaught, Chong, Lam, & Cheng, 2010;
Barnard, Lan, To, Paton, & Lai,. 2009; Moon, 2007;
Ruiz, Mintzer, & Leipzig, 2006; Wiecha, Gramling,
Joachim, & Vanderschmidt, 2003; Moran &
Myringer, 1999). However, one should notice that
the reported research studies come primarily from
developed world countries, whereas few research
efforts have been devoted to the online teaching and
learning practices in those countries which are only
now starting to adopt e-learning practices. Being
teachers in such a country, our experience as well as
experience of our colleagues show that transferring a
part or all of teaching and learning into an e-
environment tends to suffer from the same problems
as face-to-face ‘traditional’ teaching – to be based
on one-way information transmission (i.e., ex-
cathedra lecturing).
In Eastern European countries teaching at higher
education institutions was and still is predominantly
based on traditional ex-cathedra lecturing and
transmission-based teaching, rather than
constructivist approaches that allow for the
development of critical thinking skills. Girgin and
Stevens (2005) defined transmission-based teaching
as a method where the instructor acts as an
authoritative source of expert knowledge and passes
on a fixed body of information to be practiced alone
and reproduced by students on demand. For many
university teachers, particularly in developing
countries, teaching methods incorporating students’
active participation (e.g., working on case studies,
class discussions, etc) represent new and innovative
ways of teaching. Bollag’s (1996) claimed that at
Eastern European universities, “students have little
control over their studies and few chances for
classroom discussion”. Mertova and Webseter’s
research (2009) revealed a need for a more
systematic move towards a student-centered
445
Mijatovic I., Jovanovic J. and Jednak S..
STUDENTS ONLINE INTERACTION IN A BLENDED LEARNING ENVIRONMENT – A CASE STUDY OF THE FIRST EXPERIENCE IN USING AN
LMS.
DOI: 10.5220/0003963804450454
In Proceedings of the 4th International Conference on Computer Supported Education (ESEeL-2012), pages 445-454
ISBN: 978-989-8565-07-5
Copyright
c
2012 SCITEPRESS (Science and Technology Publications, Lda.)
approach, as well as a need for contemporary
approaches to teaching, learning, and assessment in
higher education in the Central and Eastern
European countries.
The result of our previous research (Mijatovic &
Jednak, 2011), which included 433 undergraduate
students in Serbia, showed that only 31.9% of
examinees had experience with active participation
in classes in secondary school. This finding suggests
that passive learning is still dominant in Serbian
secondary schools. The same research showed that
students, who had more experience with active
participation in secondary school, prefer active
teaching methods, actively participate in a class, and
are more likely to be in the group of higher
achievers.
In such circumstances, fostering students’ active
participation and interaction can be a difficult task,
and every mean that can help in advancing the
learning process is more than welcomed. Even
though the usage of LMSs has become a well
established practice in developed countries, in
developing countries, it represents a new and
innovative practice, and reports of its educational
effects are still scarce. Aiming to contribute to this
research area, we have conducted a research (case)
study where we explored online interactions and
engagement of students who are using an LMS for
the first time in their studies. Specifically, we were
interested in examining the impact of different types
of students’ online interactions on their learning
outcomes.
2 BACKGROUND
In the context of education, interactivity implies
'doing' as opposed to 'being' (present) (Downes,
2007); it assumes students’ active participation in
the learning process, rather than their passive
consumption of the content provided by teachers
(O’Connell, 2007). In other words, when considered
from the educational perspective, interactivity can be
equated with social and creative engagement, where
engagement is defined as “student-faculty
interaction, peer-to-peer collaboration and active
learning...” (Chen et al, 2008). The notion of
engagement defined in this way is fully consistent
with the Moore’s typology of interaction in distance
education (Moore, 1989). Specifically, by focusing
on learning events, Moore defined three types of
interaction: a) student–content interaction, b)
student–teacher interaction, and c) student–student
interaction. Anderson & Garrison (1998) extended
Moore’s framework with a few additional types of
interaction thus creating a model, named
Interactivity Triangle, which became a widely
accepted model of interactivity in learning settings
(Garrison & Arbaugh, 2007). Specifically,
Interactivity triangle has students, teachers and
content at its vertices. Each vertex is related with the
other two and with itself, so that, for example,
teachers are in interaction with students and content,
but they also interact among themselves.
Modern learning theories stress the importance
of interactivity in learning and call for social and
creative engagement of students. For example, the
Conversation theory argues that learning is a
continual conversation; with the external world and
its artefacts; with oneself; and also with other
learners and teachers (Pask, 1976). Likewise,
according to the Siemens’ Connectivism, the digital
age relies on connected learning which occurs
through interaction with various sources of
knowledge and participation in communities of
common interest, social networks, and group tasks
(Siemens, 2005). Connectivism also stresses the
importance of technology in the learning process and
the connection of individuals with technology as
well as with other individuals through technology.
Numerous studies have demonstrated the
benefits of interaction in the learning process,
especially, student-student and student-teacher
interaction. For example, Ellis (2001) reported the
following positive aspects of online interaction: 1)
access to peer knowledge, 2) availability of other
students to provide feedback, and 3) an opportunity
to reflect on the exchanged messages. Likewise,
Johnson & Johnson (1989) reported that learning
tends to be the most effective when students are in
the position to work collaboratively, express their
thoughts, discuss and challenge the ideas of others,
and work together towards a group solution to the
given problem. It has also been found that
interactions among students within a study group
facilitate the development of critical thinking skills,
skills of self-reflection and co-construction of
knowledge and meaning (Brindley et al, 2009).
On the other side, researchers and practitioners
alike have found that interaction is not something
that can be easily established in a learning
environment. This is primarily due to the
inappropriate course design (Brindley et al, 2009)
and/or the students’ lack of collaboration skills, such
as decision-making, consensus building, and dealing
with conflict (Finegold & Cooke, 2006). For
example, Mercer and Fisher (1997) wrote that
among different kinds of group discussions
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
446
(disputational, commutative, and exploratory),
exploratory discussions have the highest educational
value. However, a group of students can fail in
developing and sustaining an exploratory discussion
(e.g., when continuous disputes lead to frequent
breakdowns of communication). Though this is not
typical for graduate students, it tends to be frequent
in K-12 or undergraduate classrooms where students
are more susceptible to competition or immature
group behaviours (Yardi, 2006). Siemens (2002)
reported on similar findings. In particular, he defined
a continuum of four levels of student-student
interaction: 1) communication (‘talking’,
discussing); 2) collaboration (sharing ideas and
working together in a loose environment); 3)
cooperation (doing things together, but each
individual with his/her own purpose); 4) community
(striving for a common purpose). He found that in
online learning settings, interaction often does not
go beyond communication/collaboration and that
community level could be possible only in graduate
level programs with high learner-learner contact.
3 METHODS
3.1 Research Questions
The research work presented in the paper was driven
by the following three main research questions
(RQs):
RQ1: How do students perceive different forms of
online interactions in a blended learning
environment?
In the scope of this RQ, we were interested in the
students’ perceptions of different kinds of
interactions they have experienced within the LMS.
In particular, we aimed to explore the students’
experience with the following kinds of online
interactions:
1) Interaction with colleagues – did students
recognize their online interactions with other
fellow students as important and valuable?
2) Interaction with teachers – did students perceive
their interactions with the teacher as an
important and valuable experience?
3) Teacher’s feedback – was the teacher’s
feedback perceived as important and valuable?
4) Interaction with learning content – did the
students use LMS only as “course material
storage”? In other words, did they primarily
favour this kind of interaction?
RQ2: Whether and to what extent different types of
interactions influence different levels of learning
outcomes?
Relations between students’ perceptions of
different types of online interactions and different
levels of students’ learning achievements (i.e.,
learning outcomes) are definitely complex. In the
context of this study we were primarily interested in
exploring the association of four above stated types
of interactions (interaction with colleagues,
interaction with the teacher, teacher’s feedback and
interaction with content) and different types of
learning outcomes, namely acquisition of factual
knowledge, application of the acquired knowledge,
and Analysis, Synthesis and Evaluation (ASE) of
knowledge. These types of learning outcomes are
based on the Bloom's Taxonomy (Bloom, 1994)
which states that skills in the cognitive domain
revolve around knowledge, comprehension, and
critical thinking of a particular topic. For each type
of learning outcomes, we considered three possible
levels of students’ achievement: high, moderate and
low achievement.
To answer this research question we needed to
explore individual influence of each type of
interaction as well as their overall influence.
Accordingly, we further decomposed RQ2 into the
following two research questions:
RQ 2.1: What associations (if any) exist between
different types of interactions and different types of
learning outcomes?
RQ 2.2: What is the overall effect (if any) of
different types of interactions on learning outcomes?
Figures 1 and 2 (http://jelenajovanovic.net/
ESEeL2012/supplementary-material.html) present
the variables relevant for addressing RQ2.1 and
RQ2.2, respectively, as well as the considered
relations between these variables.
RQ 3: What effects do students’ interactions have on
learning outcomes when observed with other
influential factors in a blended learning
environment?
To address this research question, besides
different kinds of online interaction, we also
considered the following factors: perceived
difficulty of the course content, perceived course
relevance, opinion about the course literature,
students’ acceptance of (i.e., confidence in) LMS
and perceived usefulness of LMS. We wanted to
explore if there is a correlation between the students’
perception of different kinds of interactions on one
hand, and students’ learning outcomes on the other,
when these additional factors are considered as well.
In addition, we wanted to identify the factors that act
STUDENTSONLINEINTERACTIONINABLENDEDLEARNINGENVIRONMENT-ACASESTUDYOFTHE
FIRSTEXPERIENCEINUSINGANLMS
447
as the strongest differentiators among the groups of
achievements (high, moderate and low) in the three
considered types of learning outcomes. Figure 3
(http://jelenajovanovic.net/ESEeL2012/supplementa
ry-material.html) shows the variables relevant for
addressing RQ3 and the considered relations
between these variables.
3.2. Study Design
To answer our research questions, we organized a
semester-long study in the scope of the Quality
Engineering course taught at the largest state
university in Serbia. Data were collected from
students’ assignments and the questionnaire they
filled in at the end of the semester. By assessing the
students’ assignments, we were able to collect the
data about the students’ learning achievements. The
students’ responses to the questionnaire provided us
with information about their perceptions of different
factors influencing their experience of the learning
process
.
3.3 Study Participants
The study participants were undergraduate students
of the 4
th
year who were enrolled in a 14-week
Quality Engineering course during the Winter
semester 2011. There were 120 enrolled students in
total. Aside from face to face classes, students had a
chance to use Moodle LMS, on a voluntary basis, to
download teaching material, work on different
interactive tasks and task oriented quizzes, and
participate in on-line discussions. After the semester
and the final exams were finished, a questionnaire
was offered to all 120 students, of which 98
(81.60%) responded. Ten questionnaires were
considered invalid due to being incomplete. Thus, 88
(73.3%) of the questionnaires were taken for
analysis.
3.4 Data Collection and Analysis
In order to collect the data required for addressing
RQ1, i.e., data about the students’ perceptions of
different kinds of online interactions they
experienced during the course, we applied Critical
Incident (CI) technique. In particular, we followed
the instructions given by Johnson and Gusatafsson
(2000) for conducting this technique. CI technique
typically involves an interview or a questionnaire in
which participants are asked to provide list of things
they like and dislike about the object of research. In
our research, we asked students to provide a list of
five things they liked and disliked about their
experience with the LMS they used in the course.
This method was chosen because we did not want to
have any kind of influence on students’ perception
of online learning they experienced. By recalling
and mentioning their experience, an examinee
provides evidence of specific experience which is
valuable and important for her or him. A case in
which some examinee has had specific experience
but she or he did not mention it, is possible.
However, in our case that kind of experience was
not important, valuable or worthy of note. After
collecting and coding critical incidents, we analyzed
occurrences of same or similar critical incidents. We
were specially focused on experiences in which any
type of interactions was mentioned.
To collect the data required for exploring RQ3,
i.e., data about factors other than online interactions
that influence students’ learning achievements, we
prepared a questionnaire and at the end of the
semester, asked the students to fill it in. The
questionnaire allowed us to gather the data about
students’ a) perception of the difficulty of the course
content; b) perception of the course relevance; c)
opinion about the course literature; d) acceptance of
(confidence in) the LMS; and e) perception of
usefulness of the used LMS. The entire
questionnaire is given in the Appendix (http://jelenaj
ovanovic.net/ESEeL2012/supplementary-material.ht
ml).
Information on students’ achievements was
based on their scores on the assignments which
included home works and mid/term and final exams.
The achievements are analyzed in three areas in
accordance with the types of learning outcomes
defined in the course curriculum:
1. Acquisition of factual knowledge: scores
obtained for answering multiply choice
questions or providing short answers based on
previously learned materials (factual
knowledge).
2. Application of the acquired knowledge: scores
obtained for applying accurate method in a new
situation and solving a quantitative task
(calculus).
3. Analysis, synthesis and evaluation (ASE):
scores obtained for the ability to: i) provide
adequate discussion of the results using proper
argumentation, ii) identify motives and causes,
iii) give argumentation in order to evaluate or
propose solutions, iv) discuss implications and
limitations.
Based on their scores, for each type of learning
outcome, the students were then divided into three
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
448
groups: higher achievers (scores above 80%),
moderate achievers (scores between 60% and 80%)
and low achievers (scores below 60%).
To perform data analysis required for answering
RQ2.1 we used descriptive statistics and cross-
tabulation with a Pearson’s Chi-Square statistic test
for each item. For measuring the relative strength of
an association between two variables we used
Pearson's Contingency Coefficient (C). First, we
tested whether there is any statistically significant
association and if so, we proceeded with measuring
the strength of the association. According to
Nargundkar (2004) and Crewson (2008), if the
values of the Pearson's Contingency Coefficient (C)
range between 0.5 and 1, there exists a strong or
high association; values between 0.3 and 0.5 indicate
a moderate association, and those between 0.1 and
0.3 show a low association.
In order to find out how different aspects of
blended learning influence students’ achievements
and how strongly each one of them differentiates
among the levels of achievements (questions RQ2.2
and RQ3), we conducted Discriminant Function
Analysis (DFA) (Harlow, 2005) based on the Wilks’
lambda statistic. The rationale for choosing this
method is its adequacy for answering our research
questions (RQ2.2 and RQ3) as well as its suitability
for the type of data we collected.
For the purpose of applying DFA, we identified
the following independent variables: X1 –
Interaction with colleagues, X2 - Interaction with
teachers, X3 – Teacher’s feedback, X4 – Interaction
with learning content, X5 - Perceived difficulty of
the course content, X6 – Perceived course relevance,
X7 – Opinion about course literature, X8 - Students’
acceptance of (confidence in) LMS and X9 -
Perceived usefulness of LMS. The first four
variables were defined using the CI technique. After
collecting and coding critical incidents, we analyzed
occurrences of the same or similar experiences by
following the propositions of the CI technique. For
each study participant we observed if he/she has
mentioned a particular experience (CI) and if so, was
it defined as positive or negative. Based on that, the
considered experience was characterized as
negative, not mentioned, or positive. Accordingly,
variables X1-X4 were assigned one of the following
values: 1 – negative, 2 – no mentioning, or 3 –
positive. Variables X5 – X9 are formed trough a
summated scale defined in the questionnaire used in
the study (see Appendix). Theory background for
variables X8 - Students’ acceptance of (i.e.,
confidence in) the used LMS and X9 - Perceived
usefulness of the used LMS are taken from
Technology acceptance model originally proposed
by Davis (1989).
DFA is conducted separately for each of the
three types of learning outcomes which are
considered as dependant variables with three
possible values (low achievers, moderate achievers
and higher achievers). Interpretation of DFA is
based on discriminant loading. According to Hair et
al. (2009), discriminant loadings are less affected by
multi co-linearity and thus more useful for an
interpretative process. According to the same
authors, discriminate loadings above ± 0.40 should
be used to identify substantive discriminant
(independent) variables. The canonical correlation
coefficient (CC) is used to reflect the percentage of
variance in the dependent variable explained by the
mutual influence of independent variables.
According to Harlow (2005), the substantial value of
canonical correlation is 0.30 or higher, where the
value of 0.30 corresponds to about 10% of the
variance explained.
4 RESULTS AND DISCUSSION
In this section we present and discuss the study
results from the perspective of our research
questions (see Section 3.1).
4.1 RQ1: How do Students Perceive
Different Forms of Online
Interactions in a Blended Learning
Environment?
Our results show that the main perceived positive
aspects, among all the identified aspects of using an
LMS in a blended learning environment are related
to the students’ engagement and interaction. The
majority of the students (90.9%) mentioned at least
one type of online interactions as either a positive or
negative experience. More details are given in Table
1 where we presents the results we got by analyzing
the students’ “critical incidents”.
As Table 1 indicates, majority of the students
(84.1%) mentioned human interaction (e.g.
interaction with colleagues and teacher, interaction
with colleagues and interaction with the teacher) as a
positive experience. In other words, human
interaction is seen as the dominant positive aspect of
students’ experience of the online part of the course.
Some students mentioned more than one experience
(CI) related to interaction (e.g., interaction with
colleagues and teacher, interaction with colleagues
and interaction with teacher). However, we found
STUDENTSONLINEINTERACTIONINABLENDEDLEARNINGENVIRONMENT-ACASESTUDYOFTHE
FIRSTEXPERIENCEINUSINGANLMS
449
Table 1: Students’ experiences related to different types of interactions.
Type of interaction
Critical incidents
Positive No mentioning Negative
number % number % number %
Interaction 74
84.1
14 15.9 0 0
Interaction with colleagues and teacher 24 27.3 64 72.7 0 0
Interaction with colleagues 36 40.9 52 59.1 0 0
Interaction with teacher 38 43.2 50 56.8 0 0
Teacher’s feedback 40
45.5
39 44.3 9
10.2
Feedback time 29 33.0 50 56.8 9 10.2
Feedback related to achievements 23 26.1 65 73.9 0 0
Feedback accuracy 3 3.4 85 96.6 0 0
Interaction with learning content 52
59.1
36 40.9 0 0
detached groups of cases in which interaction with
colleagues and interaction with the teacher are
mentioned separately (48% and 52%, respectively,
of all who mentioned human interaction). Even
though, in 27.3% of these cases we observed critical
incidents related to “interaction with colleagues and
the teacher” present together with separate
interactions (i.e., only with colleagues, or only with
the teacher), we will use separate interactions
(interaction with colleagues and interaction with the
teacher) as variables in further analysis.
Even though interaction with colleagues in an
LMS was recognized as a positive, important and
valuable experience by 40.9% of all the participants,
none of them mentioned only this (among all
observed) type of interaction. Explaining the
positive aspects of interaction with colleagues,
students wrote: “I like the chance to work with
colleagues I even do not know well” or “It is
interesting to discus with colleagues about task
problems related to course”.
While 43.2% of the students perceived their
online interactions with the teacher as an important
and valuable experience, only in three cases (3.4%)
this was the only type of observed interaction.
Explaining the positive aspects of this type of
interaction some students wrote: “It’s great to have
a chance to argue with teacher” or “discussion with
teacher is available whenever I need her”.
Teacher’s feedback was perceived as important,
valuable and mostly positive experience in 55.7% of
cases, but only 3 students (3.4%) mentioned only
this (among all observed) type of interaction. Closer
look on “critical incidents” in this area suggests that
the students saw the teacher as a dominant source of
desired feedback. We did not find any report on the
experience with colleagues’ feedback. “Fast
reaction and responding on inquiry”, “fast
clarification of confusions”, “late answering” or “I
had a chance to get right answers” are critical
incidents of ours examinees related to the teacher’s
feedback. All critical incidents observed in this area
can be sorted in three groups: timely feedback,
feedback on student’s achievements and feedback
accuracy (right answers).
However, one-way information transmission
proved as still very important. The results indicate
that students’ experience with the other course’s web
site, where (two-way) interaction was not possible,
shaped students’ experience with new learning
environment. Interaction with learning content is
mentioned in 59.1% of cases, but 11 (12.5%)
students mentioned only this (among all observed)
type of interaction. Critical incidents in this category
were: “all course materials are on one place, I can
reach it any time I want”, “all information could be
found on one place” or “I can download what I
need”.
4.2 RQ 2.1: What Associations (if any)
exist between Different Types of
Interactions and Levels of Learning
Outcome?
Table 2 presents the results we obtained by
analyzing associations between different types of
interactions and different types of learning outcome.
A statistically significant, strong and positive
association is found between interaction with
learning content and the acquisition of factual
knowledge. Furthermore, a statistically significant,
positive but moderate association is found between:
interaction with colleagues and the acquisition of
factual knowledge; teacher’s feedback and Analysis,
synthesis and evaluation (ASE) of knowledge.
Finally, a statistically significant, positive but low
association is found between: interaction with
colleagues and ASE; teacher’s feedback and
application of knowledge. We did not find any
evidence that the experience of interaction with the
teacher had statistically significant influence on the
students’ achievements.
These results suggest that if acquiring factual
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
450
Table 2: Crosstabs' statistics and measures of association between different types of interactions and types of learning
outcomes (N=88).
Types of interactions
Factual knowledge
df = 2
Application
df = 2
Analysis, Synthesis and
Evaluation (ASE)
df = 2
Interaction with colleagues
Chi - Square 12.633
**
5.423 6.222
*
C 0.379** 0.266*
Interaction with teacher
Chi - Square 0.955 5.903 3.953
C
Teacher’s feedback
Chi - Square 1.956 6.588* 11.255**
C 0.274 0.358**
Interaction with learning content
Chi - Square 26.488** 1.031 5.455
C 0.548**
* significant for p < 0.05; ** significant for p < 0.01;
C - Pearson's Contingency Coefficient (0.5 - 1 strong or high; 0.3 - 0.5 moderate and 0.1-0.3 low association)
knowledge is desired, then students’ focus is
expected to be on interaction with the learning
content. On the other side, if higher levels of
learning outcomes are wanted, then two-way
interaction, i.e., communication and collaboration
are needed. Finally, the results related to interaction
with teacher and teacher’s feedback suggest that
teacher is seen as an authoritative source of expert
knowledge; students are not used to or not
encouraged enough to discuss with teacher as equal.
4.3 RQ 2.2: What is the overall Effect
(if any) of Different Types of
Interactions on Learning
Outcomes?
Even though the majority of our examinees reported
reliance on multiple types of interaction (Section
4.1), we did not find strong and statistically
significant association between different types of
interactions. Multi co-linearity in DFA is identified
by examining tolerance values. The tolerance values
for all of the independent variables were larger than
0.10, so the level of multi-collinearity was
acceptable in this analysis (Hair, 2009).
With respect to the acquisition of factual
knowledge (the first two rows of Table 3), values of
squared canonical correlations suggested that mutual
influence of the observed types of interactions is
positive, and corresponded to 36 % of the variation
between the group of higher achiever and the group
comprising moderate and lower achiever. The
second discriminant function suggests that mutual
influence of all types of interactions could explain
about 13% of difference between moderate and
lower achiever. The discriminant loadings of
independent variables that highly contribute, among
other variables, to the assumption that the students
will be in the group of higher or moderate and lower
achievers are bolded in Table 3. For the acquisition
of factual knowledge, interactions with learning
content (X4) have dominant effects on variations
between higher and moderate and lower achievers,
whereas interaction with colleagues (X1) has
dominant effect on variation between moderate and
lower achievers.
When the application of knowledge is
considered, mutual influence of the observed types
of interactions is positive and corresponds to 24 %
of the variation between the group of higher
achievers and the group of moderate and lower
achievers (Table 3, rows 3 and 4). The second
discriminant function failed statistical significance.
Dominant discrimination ability in this area has
teacher’s feedback (X3).
When higher ASE skills are considered, the
observed types of interactions can explain 12% of
the variation between the group of higher achievers
and the group consisting of moderate and lower
achievers, and 9% of the variation between groups
of moderate and lower achievers (Table 3, last two
rows). In this area, interactions with colleagues (X1)
and the teacher (X2) have dominant effects on
variations between higher and moderate and lower
achievers, but teacher’s feedback (X3) has dominant
effect on the variation between moderate and lower
achievers.
4.4 RQ 3: What Effects do Students’
interactions have on Learning
outcomes When Observed with
other Influential Factors in a
Blended Learning Environment?
The tolerance values for all the independent
variables were larger than 0.10, thus indicating that
the level of multicollinearity was acceptable in this
analysis (Hair, 2009). In area of acquiring factual
knowledge, all observed independent variables
STUDENTSONLINEINTERACTIONINABLENDEDLEARNINGENVIRONMENT-ACASESTUDYOFTHE
FIRSTEXPERIENCEINUSINGANLMS
451
Table 3: Summary statistics for discriminant function analysis.
Discriminant Loadings
(Structure correlations)
Canonical
correlation
Dependent variable - learning outcomes X1 X2 X3 X4
Factual knowledge; Λ = 0.553 (p = 0.000) (separate higher achiever
from moderate and lower)
0.410 0.137 0.037
0.846
0.604
Factual knowledge; Λ = 0.870 (p = 0.009) (separate moderate from
lower)
0.689
-0.038 0.384 -.356 0.360
Application; Λ =0.738 (p = 0.001) (separate higher achiever from
moderate and lower)
-0.458 -0.384
0.669
0.193 0.488
Application; Λ =0.969 (p = 0.454) (separate moderate from lower) -0.062 0.900 0.450 0.075 0.176
Analysis, Synthesis and Evaluation (ASE); Λ = 0.793 (p = 0.012)
(separate higher achiever from moderate and lower)
0.656 0.504
0.210 0.078 0.357
Analysis, Synthesis and Evaluation (ASE); Λ = 0.906 (p = 0.042)
(separate moderate from lower)
-0.357 0.309
0.849
-0.161 0.306
X1 - Interaction with colleagues
X2 - Interaction with teacher
X3 - Teacher’s interaction
X4 - Interaction with learning content
Table 4: Summary statistics for discriminant function analysis.
Discriminant Loadings (Structure correlations)
Canonical
correlation
Dependent variable X1 X2 X3 X4 X5 X6 X7 X8 X9
Factual knowledge; Λ = 0.356 (p =
0.000) (separate higher achiever
from moderate and lower)
.295 .092 .037
.551
.345 .100 -.272 -.269 -.033 .750
Factual knowledge; Λ = 0.814 (p =
0.034) (separate moderate from
lower)
.492
.009 .303
.417
-.157 .297 -.093 .220 .142 .431
Application; Λ =0.415 (p = 0.001)
(separate higher achiever from
moderate and lower)
.154
-
.063
.297 -.075
.690
-.042 -.018 -.317 -.098 .623
Application; Λ =0.679 (p = 0.000)
(separate moderate from lower)
.327 .383
.438
.133 .157 -.072
.682
-.022 .255 .567
Analysis, Synthesis and
Evaluation;
Λ = 0.388 (p = 0.000) (separate
higher achiever from moderate
and lower)
.536
.248 .281 .227
.580
.234 -.047 .288 -.027 .648
Analysis, Synthesis and
Evaluation;
Λ = 0.669 (p = 0.000)
(separate moderate from lower)
.356 .072 .220 .334 .232 .368 -.182
.634
.063 .575
X1 - Interaction with colleagues
X2 - Interaction with teacher
X3 - Teacher’s feedback
X4 - Interaction with learning content
X5 - Perceived course content difficulty
X6 - Perceived course relevance
X7 - Attitude toward course literature
X8 - User acceptance (confidence) of LMS
X9 - Usefulness of LMS
DV - Dependent variable, Λ - Wilks' Lambda
contribute 56% of the variations between higher
and moderate and lower achievers, and 19% of
variations between moderate and lower achievers
(Table 4, the first two rows). For this type of
learning outcome, interaction with learning
content (X4) has dominant effect on variations
between higher and moderate and lower
achievers. The variation between moderate and
lower achievers is primarily determined by the
interaction with colleagues (X1) and interaction
with learning content (X4).
In area of application of knowledge, all
observed independent variables contribute to 40%
of the variations between higher and moderate
and lower achievers, and 32% of variations
between moderate and lower achievers (Table 4,
rows 3 and 4). For this type of learning outcome,
the perceived course content difficulty (X5) has
dominant effect on the variation between higher
and moderate and lower achievers, whereas
teacher’s feedback (X3) has dominant effect on
the variation between moderate and lower
achievers.
In area of Analysis, Synthesis and Evaluation
of knowledge, all observed independent variables
contribute to 42% of the variations between
higher and moderate and lower achievers, and
33% of variations between moderate and lower
achievers (Table 4, last two rows). In this area,
the perceived course content difficulty (X5) and
interaction with colleagues (X1) have dominant
effects on variations between higher and
moderate and lower achievers, whereas User
acceptance of LMS (X9) has dominant effect on
variation between moderate and lower achievers.
When observed with others influential factors,
interactions still play an important role in
achieving higher scores for all types of learning
outcomes. The same results can be seen in this
analysis as it was in previous: interaction with
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
452
learning content is more important in acquiring
factual knowledge, while teacher’s feedback and
interaction with colleagues are needed for
application and analysis, synthesis and evaluation
of knowledge.
5 CONCLUSIONS
The main objective of the study reported in this
paper was to explore online interactions and
engagement of students who are using an LMS
for the first time in their studies, as well as the
impact of different types of students’ online
interactions on their learning outcomes. Our
findings show that students perceive interaction
as the dominant feature of learning supported by
an LMS. Majority of the students (90.9%)
mentioned at least one type of online interaction
as a relevant experience when studying with the
support of an LMS.
Even though students clearly articulated the
difference between interaction with colleagues,
teachers and learning content (the types of
interaction defined in the works of Moore, 1989
and Anderson & Garrison, 1998) some specific
differences can be observed. While the keywords
related to interaction with colleagues are
“discussions” and “opinion and experience
exchange”, when reflecting on their
communication with the teacher, students more
often used words like “response” and “answering
questions”. These results suggest that teacher is
seen as an authoritative source of expert
knowledge. One of the consequences of ex-
cathedra lecturing and transmission-based
teaching is that students are not used to and often
not encouraged enough to discuss with teacher as
equal. This conclusion is further supported by the
fact that the students did not recognize
colleagues’ feedback as worthy of note. In a
broad sense our findings are compliant with those
of Siemens (2002) – we have provided an
evidence of the existence of only first two levels
of interaction with colleagues: communication
(talking or discussing) and collaboration (sharing
ideas and working together in a loose
environment).
Furthermore, our findings provide evidence
that
different types of interactions can influence
different levels of learning outcomes. If the
acquisition of factual knowledge is desired, then
interaction with learning content is the most
influential. On the other side, if higher levels of
learning outcomes are to be achieved, then
communication (i.e., two-way interaction) is
needed. In a nutshell, the need for interaction is
increasing with the increasing level of desired
learning objectives (outcomes). Additionally, our
findings show that students’ involvement in more
challenging activities, in order to reach more
demanding learning objectives (like application
of knowledge or analysis, synthesis and
evaluation) increase their need for student-teacher
and student-student interaction.
The presented research and generalization of
data display some limitations due to the small
sample and only one course considered. In
addition, the limitation is in the fact that we
observed the presence of different types of
interactions, but not their quality. Also, the
examinees in our research are taken as a relatively
homogenous group and the impact of the specific
needs and motives were not taken into account.
In order to further verify our findings and
make them more widely applicable, in our future
work we intend to organize another study with
more students enrolled in different courses. We
also plan to make use of the Practical Inquiry
framework (Garrison et al, 2001) to analyze the
messages that students exchange in online
communication channels in order to assess the
quality of their interaction. Additionally, we
intend to examine other potentially influential
factors on students’ achievements such as
different teaching strategies and students’
motivation.
ACKNOWLEDGEMENTS
This research has been funded by the Republic of
Serbia Ministry of Education and Science
(Program of Basic Research 2011-2014,
Infrastructure for Technology Enhanced Learning
in Serbia No.47003).
REFERENCES
Anderson, T., Rourke, L., Garrison, D. R., & Archer,
W. 2001. Assessing teaching presence in a
computer conferencing context. Journal of
Asynchronous Learning Net-works, 5(2), 1-17
Barnard, L., Lan, W.Y., To, Y.M., Paton, V.O., and Lai
S.-L. 2009. Measuring self-regulation in online and
blended learning environments. Internet and Higher
Education 12(1): 1–6
Bloom, B. S. 1994. Reflections on the development and
use of the taxonomy in Anderson, Lorin W. &
Lauren A. Sosniak, eds., Bloom's Taxonomy: A
Forty-Year Retrospective. Chicago National
Society for the Study of Education
Bollag, B. 1996. Reform efforts appear stalled at
colleges in Eastern Europe. The Chronicle of
Higher Education, 43(8), A-59.
Brindley, J. E., Walti, C., and Blaschke, L. M. 2009.
Creating effective collaborative learning groups in
an online environment. The International Review
of Research in Open and Distance Learning, 10(3),
Re-trieved from: http://www.irrodl.org/index.php/
irrodl/article/view/675/1271
STUDENTSONLINEINTERACTIONINABLENDEDLEARNINGENVIRONMENT-ACASESTUDYOFTHE
FIRSTEXPERIENCEINUSINGANLMS
453
Chen, P., Gonyea, R., & Kuh, G., 2008. Learning at a
distance: Engaged or not? Innovate, 4(3).
Retrieved from: http://innovateonline.info/pdf/
vol4_issue3/Learning_at_a_Distance-__Engaged_
or_Not_.pdf
Davis, F. D., 1989. Perceived usefulness, perceived
ease of use, and user acceptance of information
technology. MIS Quarterly, 13(3), 318-339.
Downes, S., 2007. Learning networks in practice.
Emerging technologies for learning, vol. 2.
Retrieved from: http://partners.becta.org.uk/page_
documents/research/emerging_technologies07_cha
pter2.pdf
Ellis, A. 2001. Student-centred collaborative learning
via face-to-face and asynchronous online
communication: What's the difference?
Proceedings of the 18th Annual Conference of the
Australian Society for Computers in Learning in
Tertiary Education, Melbourne, pp. 169−177.
Finegold, A., & Cooke, L. 2006. Exploring the
attitudes, experiences and dynamics of interaction
in on-line groups. Internet and Higher Education, 9,
201-215.
Garrison, D. R., Anderson, T., & Archer, W. 2001.
Critical thinking, cognitive presence, and computer
conferencing in distance education, American
Journal of Distance Education, 15(1), 7-23.
Garrison, D. R. & Arbaugh, J. B. 2007. Researching
the community of Inquiry Framework: Review,
Issues, and Future Directions. The Internet and
Higher Education, 10(3), 157-172.
Girgin, K. Z., Stevens, D. D. 2005. Bridging in-class
participation with innovative instruction: use and
implications in a Turkish university classroom.
Innovations in Education and Teaching
International, 42(1), 93-106.
Hair, J. F., Black, W. C., Babin, B. J. and Anderson, R.
E. 2009. Multivariate data analysis – a global
perspective. New Jersey: Pearson Education Inc.
Harlow, L. L. 2005.The Essence of Multivariate
Thinking, Lawrence Erlbaum Associates,
Publishers, London
Johnson M. D., & Gusatafsson A. (2000). Improving
Customer Satisfaction, Loyalty and Profit – An
Integrated Measurement and Management Systems,
The University of Michigan Business School
management series, John Wiley & Sons.
Johnson, D. W., & Johnson, R. T. (1989). Cooperation
and competition: theory and research. Edina, MN:
Interaction Book Company
Kember, D., McNaught C., Chong, C. Y. F, Lam, P.,
and Cheng K. F. 2010. Understanding the ways in
which design features of educational websites
impact upon student learning outcomes in blended
learning environments. Computers and Education
55(3): 1183-1192
Mercer, N., & Fisher, E., 1997) The importance of talk,
In P. Wegerif& P. Scrimshaw (Eds.), Computers
and talk in the primary classroom (pp. 13-21).
Clevedon, UK: Multilingual Matters.
Mertova, P., Webster, L., 2009. The academic voice in
English and Czech higher education quality.
Quality Assurance in Education, 17(2), 140-155
Mijatovic, I., Jednak, S., 2011. Attitudes toward Active
Participation as Predictors of Student Achievement
- Exploratory Research from Serbia. The New
Educational Review, Vol.24, No.2:258-270
Moon, Y.-L. 2007. Education Reform and
Competency-Based Education. Asia Pacific
Education Review 8(2): 337-341
Moore, M. G., 1989. Editorial: Three types of
interaction. American Journal of Distance
Education, 3(2), 1-7.
Moran, L., Myringer, B., 1999. Flexible learning and
university change. In: Harry, K. (ed.), Higher
education through open and distance learning,
London, UK: Routledge, pp. 57–71
O’Connell, J. (2007). Creative Web 2.0 learning. A talk
given at Christian Teacher Librarians Association
conference, Sydney, 23 May 2007. Retrieved from:
http://www.slideshare.net/heyjudeonline/creative-
web-20-learning
Pask, G., 1976. Conversation theory: Applications in
education and epistemology. Amsterdam: Elsevier
Ruiz, J. G., Mintzer, M. J., and Leipzig, R. M., 2009.
The Impact of E-Learning in Medical Education.
Academic Medicine 81(3): 207-212
Siemens, G., 2005. Connectivism: Learning theory for
the digital age. International Journal of
Instructional Technology and Distance Learning,
2(1), January 2005.
Siemens, G., 2002. Interaction. E-Learning Course.
October 8, 2002. Retrieved May 19, 2008, from ht
tp://www.elearnspace.org/Articles/Interaction.htm
Wiecha, J. M., Gramling, R., Joachim, P., &
Vanderschmidt, H., 2003. Collaborative e-learning
using streaming video and asynchronous
discussion boards to teach the cognitive foundation
of medical interviewing: a case study. Journal of
Medical Internet Research, 5(2), published online
doi: http://dx.crossref.org/10.2196%2Fjmir.5.2.e13
Accessed 26th April 2011
Yardi, S., 2006. The role of the backchannel in
collaborative learning environments. In
Proceedings of the 7th international conference on
Learning sciences (ICLS '06). International Society
of the Learning Sciences 852-85
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
454
