Empathic Pedagogical Conversational Agent for Development of Computer and Research Competencies: A Research Plan

Abstract

Empathic Pedagogical Conversational Agents (PCAs) are learning tools that can favor adaptive learning in acquiring skills. Although the authors studied the efect of the learning articulated by the empathic PCAs in computer competencies, it is only focused on student perceptions. In contrast, many studies have considered the learning performance in the research competencies, but an assessment of the student perceptions and the qualitative approach is necessary. Thus, the main goal is to understand the adaptive learning articulated by an empathic PCA and its impact on the development of competencies in distributed systems and educational research in higher education. To this end, a framework is constructed to evaluate the learning outcomes of this type of learning tool (learning performance and student perceptions). The research will be a mixed method quasi-experimental. It will collect and analyze quantitative and qualitative data and integrate the information into two quasi-experiments. The main expected results are to improve the integration of empathic PCAs in the education ifeld through the construction of evidence-based algorithms. Moreover, through the integration of Artificial Intelligence and Natural Language Processing to favor advancement in the construction of better PCAs incorporating emotional features. These results will be disseminated through a compendium of publications.

Keywords

• Research
• Competencies
• A

1. Introduction

In the Organisation for Economic Co-operation and

Development countries member, the average studentteacher ratio in higher education is fiteen to one in public it is dificult for educational institutions to respond in a personalized way to the development competencies of each student. In particular, the development of computer cation [2, 3, 4, 5]. Information and Communication Technology (ICT) mediated learning is a mode of education that supports the solution [6]. Specifically, Pedagogical

Conversational Agents (PCAs) are learning tools that can

favor adaptive learning in acquiring skills [7, 8, 9]. These agents, also known as educational chatbots, can function as independent tools or be integrated into Intelligent

In the set of PCAs, there are agents with empathic

abilities; that is, they evoke an empathic reaction in the learner, which is their main diference from the others.

Emotions are integral to the educational experience, influ

Proceedings of the Doctoral Consortium of the 18th European ConferPortugal. 0000-0003-4634-6288 (E. Ortega-Ochoa) search Competencies is the doctoral research plan for the general question: How does the adaptive learning articulated by an empathic PCA afect the development of comin higher education? Based on the problem, the main goal is to understand the adaptive learning articulated by an empathic PCA and its impact on developing computer and research competencies. To this end, a framework is constructed to evaluate the learning outcomes of this type of learning tool, specifically learning performance and student perceptions. 1.1. Justification petencies in students through an online environment has promoted researchers to propose and assess learning tools such as empathic PCAs [10, 11, 12], as well as has favored scientists to work on establishing a reliable way to assess these competencies [5, 22].

2.1. Empathic Pedagogical

Conversational Agents

The results are of interest in the literature and prac- This section reports on the implementation and evaluatice. On the one hand, when it comes to the PCAs in tion stage of 13 empathic PCAs, constituting the existing e-learning, the current trends require more research on literature and relevant topics for this research plan. For the assessment of learning performance and perceptions more information on the Systematic Literature Review about learning promoted by empathic PCAs [13, 19]. protocol, the systematic review registration number is Scientific research suggests the assessment of learning osf.io/jnf3x. performance and student perceptions in a general and To begin, previous studies have predominantly favored particular way on applying PCAs in diferent domains the experiment as the research design for implementing [13, 14, 15, 16], whose design and development are fo- empathic PCA, with a quantitative approach being the cused on achieving student-teacher interaction, that is, norm. However, four of these studies opted for a mixed incorporating suficient empathic abilities. On the other methods design to comprehensively assess the efectivehand, the assessment is also of interest in the practice ness of this learning tool [13, 14, 15, 16]. This approach given that it will allow responding to the need to pro- allows for a more comprehensive evaluation, particularly mote adaptive learning. There is a need to strengthen the in the later stages of the intervention, as all reports inonline teaching and learning process in higher education clude a posttest. It is worth noting that only seven of institutions’ courses and projects [20]. Specifically, there these studies incorporated a pretest, rendering comparis a need to reinforce computer and research skills in isons with students’ initial states impossible in the other higher education [2, 3, 4, 5]. For that, the study considers studies. Instead of a pretest, four studies utilized a control competencies in distributed systems of the Computer group as an alternative resource.

Engineering Degree at UOC and educational research of Secondly, the authors consider two variables when the Basic Education Degree at UNAE. assessing the efectiveness of empathic PCA in learning.

The following sections will present the literature re- The first variable, learning performance ( 1 ), encompasses view, research questions and goal, method, and potential content, procedures, or attitudes [13, 12, 23, 19]. The ethical issues. Finally, the research contribution to the preferred data collection method for this variable is the problem solution in the Technology Enhanced Learning test, with the specific content dependent on the domain (TEL) domain will also be described. and objectives related to empathic PCA. A strictly quantitative approach has been consistently employed for 2. Literature Review evaluation. The second variable, student perceptions ( 2 ), is multidimensional, with most reports evaluating Empathic PCAs are educational chatbots that can facil- the afective bond dimension. Other dimensions, such itate the development of skills. An empathic agent is as interaction enjoyment (e.g., [15, 16]) and confidence “a synthetic character that evokes an empathic reaction perception (e.g., [24, 14]), are also considered. Questionin the user” (p. 310) [21]. For instance, the [10] and naires, surveys, and interviews serve as the primary in[11] agents focused on computer competencies, and [12] struments, encompassing quantitative and open-ended agents focused on research competencies. Recent studies questions. These findings align with prior research, such have evidenced the need to configure chatbots that in- as the work by [25] on Intelligent Virtual Agents, where corporate empathic abilities to mitigate frustrations and instrument reuse is rare, and exploring new dimensions conversation breaks [10]. Emotions play a crucial role remains a prevailing trend in empathic PCA evaluation. in education by afecting students’ motivation, attention, Thirdly, the types of feedback employed play a sigmemory, communication, problem-solving, and overall nificant role in achieving positive outcomes in learning well-being. Furthermore, research has suggested quan- performance and student perceptions [15, 16] and aptitative and mixed assessments of their results. On the pear to exhibit a positive correlation [13, 19]. For enother hand, computer and research skills are a complex hanced learning performance, cognitive and empathic and broad set of competencies highly valued in higher feedback, hints, and bimodal feedback, are deemed esseneducation [2, 3, 4, 5]. The goal of developing these com- tial [12, 23, 19]. Additionally, analyzing and commending student progress has shown a positive impact [23]. On the other hand, to foster positive student perceptions, cognitive and afective feedback, scafold design, chatbot book talk and social afective cues, coherent facial expressions, specific characteristics of Embodied Conversational Agents, support for students with significant levels of anxiety, or popular culture topics empathic are required [10, 11, 24, 14, 19, 26, 27, 28]. Notably, studentspecific factors, such as gender [ 26], can influence these outcomes. In summary, most feedback types share common attributes, with variations in their impact on the two evaluated variables. Figure 1 illustrates the resulting framework for assessing the learning outcomes of empathic PCA. + Learning Performance +

+

Student Perceptions

Learning Outcomes 2.2. Competencies Development necessary. Likewise, because of the need to have a solid framework to develop and evaluate the set of competencies, this study is based on the contributions of [3], who conceptualize the competencies in eight scientific activities: problem identification, questioning, hypothesis generation, instrument construction and redesign, evidence generation, evidence evaluation, drawing conclusions, and communicating and examining.

This section focuses on reports related to computer and

research competencies.

Although the authors studied the efect of the learning articulated by the empathic PCA in computer competencies, it is only focused on student perceptions. First, in the quasi-experiment of [10], they concluded that, in the web design domain, the use of specific types of afective and cognitive feedback has a positive efect on the afective state. However, studies are needed to validate the 3. Research Questions and Goal efectiveness of such cognitive and afective PCA abilities.

Second, [11] found that afective dialogue, based on en- The research question is: How does the adaptive learning couragement phrases, positively impacts the motivation articulated by an empathic PCA afect the development of students, female students, and engineering students. of competencies in distributed systems and educational The authors concluded that afective feedback signifi- research in higher education? Sub-questions (SQ) [29]: cantly impacts motivation, particularly in these cases.

Nevertheless, like the previous study, there are no results • What is the acquisition level of competencies in on the impact of such empathic learning tools on learning distributed systems and educational research of performance. Thus, considering the framework (Figure adaptive learning articulated by empathic PCA? 1), the assessment of the cognitive and afective feedback (SQ1) on the learning performance of computer competencies • What are the student perceptions of adaptive is necessary. Because of the need for a solid framework learning articulated by empathic PCA in develto evaluate the set of competencies, this study is based on oping competencies in distributed systems and the contributions of [5], who conceptualize user compe- educational research? (SQ2) tence in three factors: conceptualization of competence, • Is there a relationship between empathic PCA measurement methods, and knowledge domains. feedback with acquired competencies in dis

At the same time, ICT-mediated strategies favor the tributed systems and/or student perceptions dedevelopment of research competencies. In this sense, veloping educational research competencies? defining their evaluation is essential, for both learning (SQ3) tools and domains. [12] discussed the impact of Multi- • Are there significant diferences in acquired comagent Intelligent Tutoring System feedback on student puter and research competencies between the learning performance in the research methods domain. pretest and posttest, and between the control and The authors found that the agents’ cognitive support pos- experimental groups? (SQ4) itively impacts students with low rejecty sensibility in confusion regulation. On the other hand, agents’ em- The main goal is to understand the adaptive learning pathic support positively impacts students with high re- articulated by an empathic PCA and its impact on develjecty sensibility in confusion regulation. In contrast to oping computer and research competencies. Professors the studies on computer competencies, this study con- will know whether there is a relationship between the siders the learning performance, but an assessment of three variables (empathic PCA feedback, learning perthe student perceptions and the qualitative approach is formance, and student perceptions) and whether there are diferences in learning performance. These issues are relevant to understanding how PCA works in e-learning.

4. Method

The research will be a mixed method quasi-experimental (convergent core design) because it will use predefined groups. It will collect and analyze quantitative and qualitative data and integrate the information into two quasiexperiments [30]. The research will collect a qualitative component during the Randomized Controlled Trial ([31]). The purpose is to understand and depict processes experienced by the experimental groups. Specifically, the parallel database variant will be used. That is, two parallel data strands will be collected and analyzed independently and only joined in interpretation (see Figure 2). The data source for the quantitative approach will be the level of computer and research proficiency (learning performance), for which the test will be used, and student perceptions of adaptive learning articulated by PCA in a generalized manner, for which the survey will be used. The survey will allow data collection on perceptions, which will be generalized to the population, but individual experiences will not be analyzed in depth [32]. The data source for the qualitative approach will also be the perceptions. In this case, these will be in a particular way for which reflective practice will be used [ 33]. Reflective practice will be a safe space for participants to express their feelings, perspectives, and biases regarding the experience; notwithstanding, its purpose may be limited in the hypothetical case that students’ external circumstances reduce the depth of their responses [34].

4.1. Data Collection Techniques and

Instruments

The instruments’ constructs are identified in Figure 1. In

the quantitative approach, the tests are according to the competencies stipulated in the syllabus and the question- 4.2. Procedure naire is organized into sections (constructs) [32]. The questionnaire consists of two parts, both with 7-item Likert-type closed-ended response options. The first part is on student perceptions, which has twenty questions: interaction enjoyment (5), confidence perception ( 9 ), and affective bond (6). The second part is on learning outcomes and consists of one question. Regarding the reflective questionnaire, the free-association reflective questionnaire has three reflection themes corresponding to the constructs mentioned above. The content of the questionnaires is appropriate because it encompasses the three perception constructs that a user may have with TEL [35, 36]. First, student perceptions regarding interaction enjoyment encompass ease of access and use. Second, confidence perception of the content encompasses the

The two approaches (quantitative and qualitative) will fol

low the same guidelines. A plan will be designed considering the course syllabus for the implementation period. Communication with the participants will be through the noticeboard and email, which will be sent by the researcher. The first message will be the doctoral research summary. The following messages will be the pedagogical guidelines of the implementation plan. The instruments’ administration will be electronic. The test will be applied at the start and end, and the questionnaires will be applied at the end of the implementation (see Figure 2). The instruments will be virtualized on the Qualtrics platform. The test and questionnaires will be pilot-tested before final administration. The researcher will oversee data collection at the institutions, UOC and Randomized Controlled Trial

Probability Sampling

Pre-Implementation Test

Implementation Period by Course Experimental empathic PCA + learning by doing

Control Learning by doing Post-Implementation Test

Questionnaires ( 2 ) Quantitative Analysis

Descriptive and inferential statistics Interpretation Reflexive Questionnaire

Qualitative Analysis

Content analysis Integration and Interpretation (SQ2) • Learning performance (SQ1) and student perceptions (SQ2) • Relation between constructs/variables (SQ3) • Significant diferences in learning performance (SQ4) • Cross-tabulation of student perceptions • Interpretation of how the qualitative findings enhance the quasi-experiment veracity of the information in the empathic PCA interventions. Third, student perceptions of afective bonding encompass the ability of the agent to establish an empathic connection.

4.3. Data Analysis Techniques

Descriptive and inferential statistics will be the data analysis techniques for the quantitative approach and content analysis for the qualitative approach. On the one hand, descriptive statistics will allow the systematic presentation of the student’s data, and inferential statistics will allow an estimation of the population parameters and perform statistical analyses to answer the research questions [37]. Specifically, ( 1 ) to assess the correlation between the variables of learning performance, and student perceptions, Spearman’s rank correlation coeficient will be used, and ( 2 ) to compare the results of the pretest and posttest for both the control and experimental groups, a mixed-design analysis of variance will be used. The software to be used for these analyses will be SPSS [ 38], version 27.0. On the other hand, content analysis [39] will reduce the volume of words and phrases in a matrix format [40]. Based on the student perceptions variable of Figure 1, the codes will be constructed. That is, the codes will be elaborated following deductive coding. The units of analysis will be each student’s responses. In this sense, the analytic scheme will be developed before the analysis. The coding themes task will follow the strategy of [40]. The software that will facilitate this analysis will be NVivo [41], version 12. The integrated results will be performed on a joint display.

4.5. Mixed Methods Validity

The Mixed Method Research (MMR) validity is based on each approach and strategy specific to this method. The strategies for the quantitative approach are construct validity granted by the positive consequences of previous research, and reliability granted by piloting the initial version to construct definitive versions [ 42]. The strategies for the qualitative approach are communicating the results to the participants, reporting divergent results, and examining the results with professors [43, 44]. The strategies for MMR are addressing the internal and external threats identified in the literature review, providing a justification for qualitative data collection and its use, and considering unobtrusive data collection [30].

5. Potential Ethical Issues

The research has an ethical and moral commitment. The rules are in the research ethics protocol. Those responsible for the educational process and selected students will be informed of the objectives and phases and how to access their results. The evidence will be the communication emails and the informed consent; the latter will guarantee data confidentiality. Personal data will only be necessary for monitoring student participation during the quasi-experiment. Once data collection is complete, personal data will be separated from the data set to ensure privacy. Each student will be coded by cases during data analysis. The personal data will be 4.4. Sample stored on the researcher’s computer. The data from the coded cases will be stored in Mendeley Data, for open The population will be students of two courses: Dis- access. On the other hand, only the researcher can match tributed Systems, first semester 2023/2024, and Educa- the coded cases with the personal information, which tional Research: Theoretical and Epistemological Bases, will be private. Therefore, ethical criteria will always be second semester 2023/2024. In each course, the re- maintained to certify free and responsible collaboration, searcher will select an identical sample for the two ap- especially in data treatment. proaches using one-stage cluster probability sampling, given that the units of analysis are grouped into courses.

A general rule is that if the number of courses equals 3 6. Ph.D. Project’s Contribution or less, the sample will be equal to the population. Otherwise, the sample size will be calculated using Equation The novelty of the research is that it is framed in one 1, applying a confidence level of 95 % and an estimated of the key technologies and practices. Specifically, this error of 5 %. The Distributed Systems course is expected study is linked to the UOC’s Responsive Teaching and to have three courses as a population, approximately 253 Learning Processes and Outcomes in Online Education students. The Educational Research Project course is research line, in Education and ICT (e-learning). Accordexpected to have two courses as a population, approxi- ing to [6], technology, especially Artificial Intelligence mately 60 students. One time the size sample is obtained, (AI) and Natural Learning Processing (NLP), applied to the researcher will choose the number of courses that learning tools in practice, can enhance student learning are closest to the students’ size sample using Simple Ran- experiences. In this regard, this study provides a valudom Sampling; half of the courses will be assigned to the able contribution in two critical areas through assessing control and the other half to the experimental group. the adaptive learning articulated by the PCA. First, the positive or negative study results will improve the inte ℎ ∗ /22 ∗ (1 − ) gration of virtual agents in education by constructing ℎ = ( 1 ) evidence-based algorithms. Second, integrating AI and ( − 1) ∗ 2 + /22 ∗ (1 − )

Acknowledgments

With the support of a doctoral grant from the Universitat Oberta de Catalunya (UOC). Moreover, this research plan was created with the support of the Ph.D. supervisors, Thanasis Daradoumis and Marta Arguedas.

NLP will allow advancement in the construction of better conversational agents, such as ChatGPT [45], but in this case, also incorporating education and emotional features. Therefore, the project contributes to scientific progress because it crosses several disciplines, resulting in the design of learning experiences. Educational and Psychological Testing, American Educational Research Association, 2014. [43] J. W. Creswell, C. N. Poth, Qualitative Inquiry and Research Design: Choosing Among Five Approaches, 4 ed., SAGE Publications, Inc., 2017. [44] J. W. Creswell, D. L. Miller, Determining Validity in Qualitative Inquiry, Theory Into Practice 39 (2000) 124–130. doi:1 0 . 1 2 0 7 / s 1 5 4 3 0 4 2 1 t i p 3 9 0 3 _ 2 . [45] OpenAI, ChatGPT: Optimizing Language Models for Dialogue, 2022, November 30. URL: https:// openai.com/blog/chatgpt.

References

2021 1 ( 2021 ) 18 - 33 . doi: 1 0 . 3 3 9 0 / d i g i t a l 1 0 1 0 0 0 2 . tecture and interaction protocol for pedagogical[19] A. Oker , F. Pecune , C. Declercq , Virtual tutor and empathic agents in 3D virtual learning environ-

pupil interaction: A study of empathic feedback ments , Multimedia Tools and Applications 77 ( 2018 )

as extrinsic motivation for learning , Education 27661-27684. doi:1 0 . 1 0 0 7 / s 1 1 0 4 2 - 0 1 8 - 5 9 4 2 - 4 .

and Information Technologies 25 ( 2020 ) 3643 - 3658 . [28] E. H.-K. Wu , C.-H. Lin , Y.-Y. Ou , C.-Z. Liu , W.-K.

doi:1 0 . 1 0 0 7 / s 1 0 6 3 9 - 0 2 0 - 1 0 1 2 3 - 5 . Wang, C.-Y. Chao, Advantages and Constraints of [20] E. G. Ortega-Ochoa , A Pedagogical Conversa- a Hybrid Model K-12 E -Learning Assistant Chat-

tional Agent for Tutoring in the Development bot , IEEE Access 8 ( 2020 ) 77788 - 77801 . doi:1 0 . 1 1 0 9 /

of Educational Research Projects, in: D. Gural- A C C E S S . 2 0 2 0 . 2 9 8 8 2 5 2 .

nick , M. E. Auer , A . Poce (Eds.), Innovative Ap- [29] V. L. Plano-Clark , M. Babbie , Research Questions in

ings of 'The Learning Ideas Conference' 2022 , New Social & Behavioral Research, SAGE Publications,

York , United States, 2023 , pp. 544 - 556 . doi:1 0 . 1 0 0 7 / Inc., 2010 , pp. 275 - 304 . doi:1 0 . 4 1 3 5 / 9 7 8 1 5 0 6 3 3 5 1 9 3 .

9 7 8 - 3 - 0 3 1 - 2 1 5 6 9 - 8 _ 5 1 . [30] J. W. Creswell , V. L. Plano-Clark , Designing and [21] L. Hall , S. Woods , The Importance of Similarity conducting mixed methods research , 3 ed., SAGE

in Empathic Interaction , in: C. Ghaoui (Ed.), En- Publications, Inc., 2018 .

cyclopedia of Human Computer Interaction , Idea [31] D. W. Grissmer , A Guide to Incorporating Multiple

Group Reference , 2006 , pp. 303 - 310 . doi:1 0 . 4 0 1 8 / Methods in Randomized Control Trials to Assess

9 7 8 - 1 - 5 9 1 4 0 - 5 6 2 - 7 . Intervention Efects, 2 ed., American Psychological [22] A. Opitz , M. Heene , F. Fischer , Measuring scientific Association , 2016 .

reasoning - a review of test instruments , Educa- [32] D. De Vaus , Surveys in Social Research, 6 ed., Rout-

tional Research and Evaluation 23 ( 2017 ) 78 - 101 . ledge, 2013 . doi:1 0 . 4 3 2 4 / 9 7 8 0 2 0 3 5 1 9 1 9 6 .

doi:1 0 . 1 0 8 0 / 1 3 8 0 3 6 1 1 . 2 0 1 7 . 1 3 3 8 5 8 6 . [33] M. Alvesson , K. Sköldberg , Reflexive Methodology: [23] A. Munshi , R. Rajendran , J. Ocumpaugh , G. Biswas, New Vistas for Qualitative Research, 3 ed., SAGE

R. S. Baker , L. Paquette , Modeling Learners' Cogni- Publications, Inc., 2017 .

tive and Afective States to Scafold SRL in Open- [34] F. D. Billups , Qualitative Data Collection Tools: De-

J. Zhang (Eds.), UMAP '18: Proceedings of the cations, Inc ., 2021 . doi:1 0 . 4 1 3 5 / 9 7 8 1 0 7 1 8 7 8 6 9 9 .

26th Conference on User Modeling , Adaptation [35] S. Lee , J. Choi , Enhancing user experience with

and Personalization , Singapore, Singapore, 2018 , conversational agent for movie recommendation:

pp. 131 - 138 . doi:1 0 . 1 1 4 5 / 3 2 0 9 2 1 9 . 3 2 0 9 2 4 1 . Efects of self-disclosure and reciprocity , Inter[24] E. Ayedoun , Y. Hayashi , K. Seta , Toward Personal- national Journal of Human-Computer Studies 103

ized Scafolding and Fading of Motivational Support ( 2017 ) 95 - 105 . doi:1 0 . 1 0 1 6 / j . i j h c s . 2 0 1 7 . 0 2 . 0 0 5 .

in L2 Learner-Dialogue Agent Interactions: An Ex- [36] S. Mendez , K. Johanson , V. M. Conley , K. Gosha,

Technologies 13 ( 2020 ) 604 - 616 . doi:1 0 . 1 1 0 9 / T L T . sis of children' interaction with an AI chatbot and

2 0 2 0 . 2 9 8 9 7 7 6 . its impact on their interest in reading , Interna[25] S. Fitrianie , M. Bruijnes , D. Richards , A . Abdul- tional Journal of Doctoral Studies 15 ( 2020 ) 373 - 392 .

rahman , W.-P. Brinkman, What are We Measur- doi:1 0 . 2 8 9 4 5 / 4 5 7 9 .

ing Anyway ? - A Literature Survey of Question- [37] R. Bruhl , Understanding Statistical Analysis and

naires Used in Studies Reported in the Intelligent Modeling, SAGE Publications, Inc ., 2018 . doi: 1 0 .

Virtual Agent Conferences , in: C. Pelachaud, 4 1 3 5 / 9 7 8 1 0 7 1 8 7 8 8 6 6 .

J.-C. Martin (Eds.), IVA '19: Proceedings of the [38] International Business Machines Corporation , IBM

19th ACM International Conference on Intelligent SPSS Statistics , 2020 . [Computer Software].

Virtual Agents , Paris, France, 2019 , pp. 159 - 161 . [39] K. Krippendorf , Content Analysis , 4 ed ., SAGE Pub-

doi:1 0 . 1 1 4 5 / 3 3 0 8 5 3 2 . 3 3 2 9 4 2 1 . lications, Inc., 2019 . doi:1 0 . 4 1 3 5 / 9 7 8 1 0 7 1 8 7 8 7 8 1 . [26] M. R. Scholten , S. M. Kelders , J. E. W. C. V. Gemert- [40] M. B. Miles , A. M. Huberman , J. Saldana , Qualitative

Pijnen , An Empirical Study of a Pedagogical Agent Data Analysis: A Methods Sourcebook , 3 ed., SAGE

as an Adjunct to an eHealth Self- Management In- Publications, Inc., 2014 .

tervention: What Modalities Does It Need to Suc- [41] QSR International, NVivo, 2018 . [Computer Soft-

Psychology 10 ( 2019 ) 1 - 14 . doi:1 0 . 3 3 8 9 / f p s y g . 2 0 1 9 . [42] American Educational Research Association, Amer-

0 1 0 6 3 . ican Psychological Association, National Council [27] T. Terzidou , T. Tsiatsos , H. Apostolidis , Archi- on Measurement in Education, The Standards for