Chapter 8 Pilot Study
8.2 Methodology
8.2.2 Relaxation training session
Based on preliminary studies (see Chapter 5), relaxation exercises and positive affirmation were identified as the components of the relaxation session of the affective tutoring system (ATS) framework. The user-centred trials (see Chapter 6) indicated that many participants were unable to follow the relaxation regime without prior training. So a relaxation training was incorporated at an early stage of the session to help them to follow the regime correctly.
Researchers across a broad array of disciplines have identified the fact that a comfortable environment is one of the key factors for the success of relaxation exercises (e.g. Mary, 2006; Tacken, 1989; Davis, 1992; Tusek & Cwynar, 2000). Two important factors were identified for making experimental room comfortable; the seating arrangement of the participants and the lighting. Each of the participants was provided with adequate space to avoid distraction from other participants and the lighting was subdued in the experimental room, see Figure 8-1.
Figure 8-1: The pictures of the relaxation training session. 8.2.3 Procedure
This study used a between-subject methodology based on an experimental and a comparison group. The experimental group is referred as the DD + DI group and the comparison group is referred as the DD group. While the experimental group consisted of participants who were asked to use both the domain-dependent (DD) and domain-independent (DI) strategies to help regulate their state of well-being, the comparison group consisted of participants who were presented only with domain- dependent strategies to help regulate their well-being.
There were 14 stages in this pilot experimental study. These stages were further grouped into five sub-phases. The complete flow chart of the experimental design is presented in Figure 8-2.
Figure 8-2: The complete experimental design flow chart
Learning activities
Appraising student’s initial state of well-being
Learning activities
o Reporting self-belief level of a lesson (self-report)
o Answering the questions
1 2 4 5 3 Primary appraisal phase Constructing lesson plan Relaxation exercises Primary reaction phase (DI strategies)
Appraising student’s state of well-being
Secondary appraisal phase
Reacting to student’s state of well-being (Using DD strategies)
Determining the student’s final affective state Secondary reaction phase (DI strategies)
Pre-test session (students are asked to select two lessons)
Appraising student’s final state of well-being
Repeat 4X
Primary reaction phase (DD strategies) Secondary reaction phase (DD strategies) Pre-test Relaxation training
Constructing lesson plan
Post-test session (identical with the pre-test material) Post-test Remarks: DD refers to domain-dependent DI refers to domain-independent No action DD + DI group DD group
Relaxation training for the DD and DD+DI groups
Self-reporting the change in student’s state of well-being
Presenting student with example of code
6
Relaxation exercises No action
DD + DI group DD group
Self-report the change in student’s state of well-being
7 8 7 9 8 10 11 12 13 14 Self reporting Self reporting Sub-phase 1 Sub-phase 2 Sub-phase 3 Sub-phase 4 Sub-phase 5
Participants were asked first to take the data-structures pre-test session (Stage 1). Using paper-based materials, participants were required to select and complete two of the four available Data Structures topics (i.e. recursive, linked list, stack and tree – see Appendix J for details). They were given about 20 minutes to complete both of their selected topics. No supporting materials such as notes or books were allowed or provided. Identical pre-test materials were used by both the experimental group and the comparison group (see Section 8.2.1 for details).
After the pre-test, all participants were presented with relaxation training (Stage 2). During this stage, they were asked to follow a narrated relaxation script which taught them the correct method of doing the relaxation exercise and the positive affirmation activity. Three identical relaxation training cycles were conducted to ensure the participants had mastered the relaxation skills. In order to create an identical environment for both groups, all participants were required to undertake the same relaxation training even though the DD group did not actually do the relaxation exercises later during learning.
Following relaxation training, participants started to work with the system itself. All participants were asked to self-report their initial affective states via an online version of the PANAS questionnaire (Stage 3). Using the Well-being formula (Eq. 4.1- see Chapter 4 for details), the participant’s state of well-being was calculated. This was followed by the primary reaction phase (Stage 4). The primary reaction phase varied according to which group the participant was in. In the DD version, which was also the comparison group for the experiment, no activities were deployed as a reaction to the participants’ affective state. In contrast, in the DD+DI version, the participants were asked to do the domain-independent strategies (i.e. breathing exercises and positive affirmation activity). At the end of the relaxation exercises, the DD+DI group were asked to report any change in their state of well-being (Stage 5).
Having completed the primary reaction sub task, all participants were then asked to construct a session plan; there were four topics (i.e. recursive, linked list, stack and tree) to choose from (Stage 6). Of the four available topics, they were required to select the same two topics they had chosen for the paper-based pre-test. Subsequent to
the construction of a session plan, participants were provided with an example of an upcoming lesson and were also provided with step-by-step instruction for it (Stage 7).
The participants were then presented with the first of their four lesson4 interfaces and their associated activities (Stage 8); i.e. the notes, examples and the problem itself which the participants were required to complete. However, before they could start to answer the questions, they were required to self-report on their level of self- confidence, to indicate how confident they were about solving the problem presented, using a scale of 1 to 3 (1 = not confident and 3 = very confident).
For each of the four lessons, participants were given 15 minutes and allowed six attempts to complete that lesson. They were allowed to use all the helpful features incorporated into the software environment. These included examples and notes and they were allowed to navigate freely within the environment (see Chapter 6 for details).
A secondary appraisal of the state of well-being of all of the participants’ was carried out after they had submitted their answers (Stage 9). Each participant’s state of well- being was computed based on three variables: their self-belief level, their independence level and the number of attempts they had made at solving the problem. They were inferred to be in a state of positive well-being if they had produced the correct answers for the problem in less than 15 minutes and with not more than 6 attempts. Conversely, participants who failed to complete the lesson within 15 minutes or had made more than 6 attempts in the problem were inferred by the system to be in a state of negative well-being. See Chapter 4 for details of the secondary appraisal mechanism.
Next, the secondary reaction phase of the ATS framework began (Stages 10, 11). The DD+DI group were again presented with the domain-independent strategies; i.e. relaxation exercise and positive affirmation before reporting any change in the state of their well-being. In contrast, the DD group’s members were asked to self report any
change in their state of well-being without doing the domain-independent (relaxation) strategies.
Next, both groups of participants were presented with similar domain-dependent strategies (Stage 12). For instance, if a participant was inferred to be in a high state of positive well-being, the system suggested that the participants move two steps ahead in his session plan. Likewise, the system advised a participant who was inferred to be in a state of negative well-being either to do an easier lesson for his next task or to redo the current lesson. The complete mechanism of the domain-dependent strategies is presented in Table 6-3 (for participants who were inferred to be in a state of positive well-being) and Table 6-4 (for participants who were inferred to be in a state of negative well-being) of Chapter 6.
The learning session cycle of the primary and secondary appraisal and reaction phases together with problem solving continued until the entire session plan was completed (each participant had four cycles to complete).
Upon the completion of all the learning sessions, participants were asked to again self-report their affective state using the online PANAS questionnaire. Finally before they left the lab, they were asked to take the paper-based post-test on Data structures. Identical materials to the pre-test were used for the post-test, and again with students being tested on the same topics they had been pre-tested on and had worked on with the system.
As for the qualitative data collection, six randomly selected participants from each of the DD and the DD+DI groups were interviewed at the end of the experiment (the interviews were carried out in the Malay language). These interviews were conducted using a semi-structured approach. The interview transcripts are presented in Appendix M.