**4. Research methodology and experiment design**

Considering the research objectives and the defined research hypotheses, we have designed an experimental setup as the test environment for the study. Since this research investigates about the student engagement in 3D MULE and the two variables which directly relate with the need of 3D MULE management, we decided to conduct a two part study as the experiment. First, students were allowed to experience the 3D MUVE supported learning environment as a part of their studies. We used Open Simulator for creating the learning island; contrast to Second Life, OpenSim gives us the complete autonomy on deciding the server environment and required configurations and scalability. We also realised that, it would provide sufficiently accurate data for the analysis if we use a student credit bearing course activity, which is conducted using 3D MUVE. It is important that students participate in a real learning session to provide accurate data for our study than following a hypothetical set of instruction, which does not provide them a true motivation as they have with their course learning activities. Moreover, such arrangement helps students to comfortably associate the experience they had when they answer the questionnaire.

With that in mind we considered to associate the study with two course modules that use the same learning aid developed in 3D MUVE. We also decided to increase the accuracy of the experiment data using a broad sample of participants by selecting the two course modules one from the undergraduate curriculum (CS3102 – Data Communications and

**H2** – Appropriate system environment management is a major factor of the success of 3D

For the second objective of this study, we investigated the impact of the above said two variables on the student engagement with the 3D MULE. Importantly, the engagement with the 3D MULE may not necessarily represent the student engagement with the learning, although there can be a positive correlation if the learning tasks are constructively aligned (Biggs, 1996). However, the opposite of the above condition is trivial; if the student engagement with the 3D MULE is low, then their engagement with the learning tasks that depend on 3D MULE, tend to be low as well, for obvious reasons. Nevertheless, since many researchers on 3D MUVE supported learning have indicated, as mentioned above, there are unique advantages of using 3D MUVE for teaching and learner support, which we may not be able to obtain through the other methods. If that is true, then deductively, we can presume that if students do not highly engage with 3D MUVE that host learning task, there

is a high tendency on them having less engagement with the learning tasks as well.

Furthermore, as we are expected to formulate policy considerations for 3D MULE management, those policy considerations should not negatively affect the student engagement with the environment. Therefore, to examine the influence on student selfregulation and system environment management on student engagement with 3D MULE,

**H3** – Students' self-regulatory behaviour has a positive and significant effect on student

**H4** – System environment management has a positive and significant effect on student

Considering the research objectives and the defined research hypotheses, we have designed an experimental setup as the test environment for the study. Since this research investigates about the student engagement in 3D MULE and the two variables which directly relate with the need of 3D MULE management, we decided to conduct a two part study as the experiment. First, students were allowed to experience the 3D MUVE supported learning environment as a part of their studies. We used Open Simulator for creating the learning island; contrast to Second Life, OpenSim gives us the complete autonomy on deciding the server environment and required configurations and scalability. We also realised that, it would provide sufficiently accurate data for the analysis if we use a student credit bearing course activity, which is conducted using 3D MUVE. It is important that students participate in a real learning session to provide accurate data for our study than following a hypothetical set of instruction, which does not provide them a true motivation as they have with their course learning activities. Moreover, such arrangement helps students to

comfortably associate the experience they had when they answer the questionnaire.

With that in mind we considered to associate the study with two course modules that use the same learning aid developed in 3D MUVE. We also decided to increase the accuracy of the experiment data using a broad sample of participants by selecting the two course modules one from the undergraduate curriculum (CS3102 – Data Communications and

Multi User Learning Environment management

we defined the following two hypotheses for the analysis.

engagement with 3D Multi User Learning Environments

engagement with 3D Multi User Learning Environments

**4. Research methodology and experiment design** 

Networks) and the other from the postgraduate (taught MSc program) curriculum (CS5021 – Advanced Networks and Distributed Systems). Information about the experiment samples is given in the Table 1.


Table 1. Details about the experiment samples and associated course modules

It is important to mention that these two course modules have different learning objectives and tasks while representing different levels in Scottish Credits and Qualification Framework (SCQF) (SCQF, 2007). Therefore, the subject content in each level had dissimilar objectives, and the assessment tasks were slightly diverse, although students used the same learning tool. However, for this study, we are focused on student engagement with the environment and the impact on that from the two aspects, their view of self-regulation and the system management. In that regard, we can conclude that, although students had a different level of the same learning task on the same learning aid, both samples had the same characteristics with respect to our measure. Therefore, the experiment has not been affected by the learning tasks given in the modules; hence can be considered as a single sample consisting of 59 students for the data analysis.

The learning environment, Wireless Island (Sturgeon *et al.*, 2009), is a dedicated region for facilitating learning and teaching wireless communication. It was developed as a research project which provides interactive simulations with various configuration settings for students to explore and learn. It also includes supplementary learning content such as lecture notes, lecture media stream and a museum to depict the history of wireless communication development. Figure 2 shows the island layout (left-side image) with different interactive content and places for student learning. The right-side image of the Figure 2 shows the interactive simulation on teaching *Exposed Node Problem* in wireless communication.

To facilitate small group learning with a less competitive environment interaction, we decided to have 6 students per region. Therefore, five regions were created in the OpenSim environment and loaded the Wireless Island on each. This resulted in an

3D Multi User Learning Environment Management –

image - the enlarged map view of the island aelous3

**5.1 Observations of student behaviours** 

**5. Results and analysis** 

hypotheses testing.

considered.

An Exploratory Study on Student Engagement with the Learning Environment 115

Fig. 3. Map view of the experiment setup (a small square represents an island), right side

Based on the experiment design described above, our data gathering was conducted during the laboratory sessions by observing the avatar interactions and then at the end of the session through a questionnaire. Let us first discuss the observed student engagements during the laboratory session and then analyse the collected questionnaire data for the

For the preliminary analysis on observed student activities following scenarios are

Avatar appearance change or outfit creation is an entertaining activity for any 3D MUVE user. However, the ease of creating attractive appearances within a limited time can be one of the critical determinants for student motivation on appearance change. Notably, the OpenSim with Hippo viewer gave an additional step on changing user appearance compared to Second Life. Users first have to create body-part, edit it and then wear to change the shape of the default avatar. Without this step, students could not change the gender of their avatars (default avatar has the view of a female user shown in the top left picture of the Figure 4). However, few students spent more time during their lab session and created more sophisticated shapes, clothes and appearance. Postgraduate students showed relatively less enthusiasm on changing their avatar shape, whereas many undergraduates went to a further step by comparing the avatar appearances with their friends'. In overall, students showed different preferences on spending time for appearance change and their commitments towards complex shape creation. However, we believe that the student commitment for making their avatars look good should not be underestimated as it could take substantial portions of their learning time. Constructive arrangements, such as, pre-sessions for familiarity and entertainment before the actual learning engagement, are therefore, highly encouraged.

identical learning set up for each student, and students were given their assigned region as their home place to start the learning task. Region information is given in the Table 2 and the corresponding 3D MUVE map, and an island map is shown in the Figure 2. The root island (Learn) remained as an empty island with everyone to access as a sandbox so that students can try their desired content creation and other activities without affecting the learning environment.

Fig. 2. Wireless Island overview with layout – left side; right side – an interactive learning aid for simulating Exposed Node Problem in Wireless communication


Table 2. Experiment region information and their relative positions on the Grid

In the Figure 3, a block of square represents an island (256 m x 256 m virtual area), and the islands are distributed as shown in the map, to minimise adjacency problems and to simulate the isolated island look and feel. Tiny green dots indicate the student avatar positions when the image taken.

The second part of the data gathering based on a questionnaire with 15 questions divided into two sections: Avatar Behaviour – 7 questions, and 3D MULE management – 8 questions. Additionally, five open-ended questions were included to help students to express their opinions openly. The 8 questions in the 3D MULE management section had some relevance on the two factors that we are investigating, self-regulation and environment management; however, the questions did not directly represent the variables. We decided to confirm through the statistical analysis, therefore, treated as 8 related questions.

Fig. 3. Map view of the experiment setup (a small square represents an island), right side image - the enlarged map view of the island aelous3
