Mobile-based Assessment: An exploration of Motivation and Acceptance factors.

Stavros Nikou (2017)

Ph.D. Dissertation, University of Macedonia, Department of Economics, Thessaloniki, Greece.

Mobile-based Assessment (MBA) is an emerging field in the wider context of mobile learning research that has the potential to provide new opportunities for assessment.

However, successful implementation of MBA depends on user acceptance. While there is a considerable number of studies investigating mobile learning acceptance, no research exists that investigates the driving factors that influence intention to use mobile technologies for assessment purposes. Moreover, the motivational aspects of Mobile-based Assessment have not been fully explored. A comprehensive theoretical MBA framework, based on a solid theoretical background is needed to addresses the motivational dimensions of MBA.

This thesis is aiming at introducing analytical models about the acceptance of Mobile-Based Assessment from students‟ and teachers‟ perspectives as well. The thesis also suggests a comprehensive framework regarding motivation issues in Mobile-Based Assessment.Digital Library

Integrating personalized emotional and physiological measurements into the technology acceptance model: a study of a computer based assessment system.

Vasileios Terzis (2012)

Ph.D. Dissertation, University of Macedonia, Department of Economics, Thessaloniki, Greece.

This thesis develops and evaluates a framework of methodologies, based on theories about the role of different factors of learning acceptance in order to achieve reliable recognition of user intentions to use an information system system. The main contributions of this thesis are summarized in the following points:

1. Development of the basic holistic model of acceptance of a computer based assessment system.
2. For the first time I developed a model which explains the long-term behavior of the user using the user's expectations regarding the system before they used the system with their perceptions after the use of the system.
3. For the first time, the effect of gender as a determinant of user's behavioural intentions to use the system.
4. For the first time introduced the personality characteristics in an acceptance model.
5. Incurred significant research findings, which could be valuable in the design of such future systems tailored to the personality characteristics of each user.
6. For the first time incorporated the cultural dimensions of user in an acceptance model regarding computer based assessment (CBA) systems. The results are important for further customization of CBA system to the characteristics of each user.
7. For the first time learning styles included in an acceptance model. These data are important because they provide the opportunity for further individualized a CBA system.
8. For the first time investigated the facial expressions of the user as a source of emotion recognition, within a CBA system, and identified characteristics that may mislead the recognition in this process. Moreover, for the first time provided data on how the student’s instant emotions experienced during the use of the CBA system affect his/her behavioural intention to use the system.
9. The effect of emotional feedback as a separate variable in the CBAAM. The findings are important, as it indicates that the emotional feedback is one factor that greatly influences user’s behavioural intention to use the system.
10. For the first time, I used electroencephalograph (EEG) as a suitable and reliable tool for the measurement of user’s perceptions regarding usefulness, ease of use and playfulness of the system. Research findings have confirmed that the EEG may actually be an effective tool in order to predict and determine user’s behavioural intentions.

Digital Library

Affective artificial intelligence in education: affect recognition and feedback in the context of a self-assessment test system

Christos N. Moridis (2011)

Ph.D. Dissertation, University of Macedonia, Department of Economics, Thessaloniki, Greece.

Computerized self-assessment test systems can be an integral part of any e-learning system. Moreover, preparation tests through computerized self-assessment test systems, to help students before final exams, can be essential to any educational or learning process. Although lack of emotional recognition and emotional feedback capabilities of traditional tools of e-learning has been acknowledged as one major limitation in the recent decade, there have been no previous efforts to incorporate affective handling capabilities into self-assessment test systems.

This Ph.D. thesis is a first step towards this direction. In the context of this Ph.D. affect recognition and affective feedback methods were developed and evaluated for use during a computerized self-assessment test. The main contributions and conclusions of this Ph.D. are summarized in the following paragraphs:

1. Learners' mood models during an online self-assessment test were developed and evaluated. The proposed mood recognition models are easy to implement in a system. So far, there has been no applicable computational model for affect recognition during an online test. Moreover, the relevant results indicate that the assumptions underlying the mood recognition methods may prove useful for future research.
2. Additionally, affect was recognized through students' facial expressions during a self-assessment test. A self-assessment test may be a procedure involving particular facial expressions that could be misleading. For instance, the facial expression of a student who is trying to concentrate in order to answer a question could be misinterpreted as an angry emotion, while in fact the student is merely focusing on the test procedure. Cases like this were identified, so as to provide future systems with information in order to register only facial expressions triggered by relevant emotions during a computerized test. To the best of my knowledge this is the first attempt to evaluate facial expressions, as an emotional recognition source during a self-assessment test. Results indicated that facial expressions can be a reliable source of recognizing emotional states during a computerized test. Moreover, useful results concerning the emotional states experienced by students during a test were provided
3. In order to handle students' fear, happy, and sad emotions, empathetic Embodied Conversational Agents (ECAs), were designed and implemented. Few research studies have employed empathetic ECAs for the purpose of emotional regulation. Nevertheless, none of them tested the effect of different types of emotional facial expressions combined with different types of empathetic behaviour (parallel and reactive empathy). Results indicated that an ECA performing parallel empathy with a relevant to the student's emotion facial expression may cause this emotion to persist. Moreover, results showed that an ECA performing parallel and then reactive empathy (displaying a relevant to the student's emotion facial expression for parallel empathy and a different from the student's emotion facial expression for reactive empathy) appeared to be effective in altering an emotional state of fear to a neutral one.
4. The analysis of the affective feedback reward strategy developed and evaluated in this Ph.D., aimed at regulating students' state and trait anxiety, revealed that gender differences should be taken seriously into account when designing affective feedback strategies for a self-assessment test system.
5. Another attempt towards affective feedback was implemented by evaluating the impact of empathetic agents as feedback to human emotions for improving brainwave activity conducively to learning, as measured by the electroencephalograph (EEG). The analysis showed that empathetic ECAs can indeed have a significant influence on alpha and beta brainwave activity, and thus improve brainwave activity towards learning. To the best of my knowledge this is the first attempt to evaluate the impact of empathetic ECAs in terms of their capacity to modulate brain rhythms in order to be beneficial to learning activities.

Digital Library

Information and communication technology investments evaluation: a real options framework of analysis

Georgios N. Angelou (2010)

Ph.D. Dissertation, University of Macedonia, Department of Economics, Thessaloniki, Greece.

In this PhD thesis we develop a number of decision analysis models for evaluating Information and Communication Technology (ICT) investments in the joint presence of uncertainty and competition. The target is to analyze investments risks, goals and constrains, estimate the optimum deployment strategy and finally evaluate the overall ICT business. Viewing ICT investments as real options (ROs), we model flexibility of implementing ICT business and combine them with various decision analysis techniques, such as game theory (GT), goal programming (GP), fuzzy logic (FL), analytic hierarchy process (ΑΗΡ) and SWOT analysis, for modelling specific ICT business characteristics in a holistic decision analysis perspective.

It is the first time in the literature where the aforementioned techniques are integrated for modeling the specific ICT business characteristics in a common decision analysis framework. Analytically, the existing ROs models are strictly quantitative, while ICT investments experience tangible and intangible factors and the latter can be mainly treated by qualitative analysis. Moreover, ROs analysis in itself brings to the "surface" a number of factors that cannot be quantified, at least easily, by existing ROs models and methodologies. In this work, we enhance the quantitative analysis of the ROs introducing further qualitative option thinking. Our work suggests the management and business analysts, which adopt ROs, to recognize qualitatively the factors affecting the investment value and treat them in a ROs perspective. The results from our models may change the conclusions extracted by the typical ROs approach given by the literature. The ability to hold the option and delay the investment depends on the balance between a large number of criteria, which some of them can be treated qualitatively while some others quantitatively. The results of the thesis prove that the combination of quantitative and qualitative analysis, under a multicriteria perspective, could provide different conclusion for an ICT business concerning its optimum deployment strategy and its overall performance comparing to single quantitative analysis. Finally, we apply the proposed models and methodologies in real case studies from broadband networks and services, e-learning and firms' reorganization business fields, showing how they can be formulated and solved. The cases analyzed prove the usefulness and efficiency of the proposed models and methodologies. Digital Library