A Theoretical Framework for Analyzing Student Achievement in Software Education

Abstract

Software education and its value are strongly emphasized among basic university courses in the era of the fourth industrial revolution. Numerous university students use the internet and various software in their daily lives. However, there is a lack of awareness on the necessity and value of software education. Therefore, a systematic software education methodology for university students is required. Moreover, an educational strategy that meets the needs of students is required to provide students with more efficient software education. This study aims to analyze the intention of using software among students and build a classification scheme for educational intentions to achieve educational objectives by establishing a strategy for software education. Therefore, this study presents a strategic framework for which a 2 × 2 matrix is proposed based on hedonic motivation and effort expectancy. We examine various aspects of its practical application, and derive improvements through focus group interviews. An empirical experiment was carried out in a basic Python programming class in order to find variables in the framework. The framework was verified in terms of three aspects: utility, convenience, and elasticity. The software education strategy framework developed in this study will help achieve educational goals and resolve efficiency issues in the software education field in the future.

1. Introduction

The fourth industrial revolution, currently underway, is characterized by ubiquity, availability, and interaction at a global scale. The impact of technology on the economic and societal aspects is much more complex, changing the labor market, and affecting employment [1]. Platform businesses, such as Amazon, are destroying jobs by providing more at a lower cost with their technical prowess in diverse areas such as distribution and logistics [2]. It is also affecting the teaching methods in traditional higher education, shifting the focus from “how to teach” to “how to learn” [3]. Take Thailand as an example; there is a decreasing number of college students in traditional classrooms, and many are seeking ways to apply digital transformation throughout schools as technology is influencing teaching [4]. In this regard, Klaus Schwab emphasized the need to prepare a workforce to cooperate with intelligent machines and develop related educational models [5].

In this context, attempts are being made worldwide to keep up with social and environmental trends and strengthen software (SW) and artificial intelligence (AI) education. South Korea has amended the Software Promotion Act to provide SW education in elementary and middle schools [6]. Through SW-oriented university programs, advanced training is provided to computer major students, whereas SW education is offered to non-major students for the purpose of spreading the values of SW [7]. Chung, You, and Mun (2022) stressed the importance of AI education by showing that the frequency of the keyword “education” was high in the results of text mining analysis on AI research papers published in South Korea. They also mentioned the need for detailed educational plans, citing the weaknesses of the current educational methods [8]. Recently, a bill for the “AI Education Promotion Act” has been proposed, and is under discussion to implement AI education at all stages, from early childhood education to higher education [9].

The starting level of SW and AI education is programming education for communicating with computers. However, programming is difficult even for computer major students, and there are many difficulties in teaching programming languages to non-major students. In SW education, the weightage for practical exercise is high, unlike in classes that provide theory-oriented lectures. If students do not study each chapter in the proper order according to the curriculum, they will have difficulty in comprehending the next course, resulting in a large percentage of students quitting in the middle of the curriculum. Based on this, Terroso and Pinto (2022) used P5.js in classes for non-major students to learn programming easily [10]. Chun, Jo, and Lee (2021) conducted educational sessions using scratch and physical blocks, which helped to develop interest in students [11].

Despite the growing importance of SW education, there is still a lack of research on effectively educating non-major students. An appropriate strategy is required to develop a new form of education beyond the current form of teacher-oriented education, i.e., student-oriented customized education. Therefore, we propose an SW education strategy framework that is based on previous studies on learners’ characteristics. We also discuss the proposed model based on observations from the focus group interviews (FGI).

The purpose of this strategy Is not to determine the current position of the organization surrounded by its external environment, but to predict its future; an excellent strategy can be derived through in-depth examination of its meaning and purpose [12]. The purpose of this study is to improve programming education and provide proposals for nurturing SW human resources by providing guidelines of creating strategies through an in-depth investigation to respond to the fourth industrial revolution.

This study is structured as follows: Section 2 describes the theoretical background and experimental results of this research’s model. Section 3 defines the SW education strategy framework that consists of the factors selected. Section 4 debates the conceptual model with experts and derives its meaning. Lastly, Section 5 presents the discussion on the analysis, the debate results, and future research.

2. Literature Review

Computer major students start to learn various programming languages and internalize relevant concepts in their first college year. However, the importance of programming education is relatively low for non-major students; these students find programming classes difficult [13,14,15]. Moreover, by measuring the relationship between students’ mental health factors and learning motivations, it has been inferred that a clear direction, rewards and recognition, and feelings of anxiety and depression increase the general fatigue of students [16]. The act of programming refers to coding the instructions to be given to the computer, which requires a programming language. Programming languages, such as Python and C, require constant use. If not, one’s ability to use them will decline. Therefore, the intention to use them continuously is important. The lecture used in the empirical study of this prior research deals with Python programming and basic computing skills. The Python programming chapter covers basic grammar, development environments, and simple projects. Computing skills deal with the overall and basic knowledge of SW technology.

Cheon, et al. (2022) conducted a Python programming class for non-major students and analyzed the intentions of 129 students regarding the continual use of the programming language to determine whether there were differences between the academic majors [17]. Specifically, they applied the technology acceptance process from the perspective that programming is a new technology for non-major students. As a result of performing multiple regression analysis using the unified theory of acceptance and use of technology by Venkatesh et al. [18], they found that performance expectancy (t = 6.624, p < 0.001), effort expectancy (t = 3.232, p < 0.01), and social influence (t = 5.664, p < 0.001) had a statistically significant effect on Python programming intention. However, the standardization coefficient of effort expectancy was found to be relatively low, unlike in the other two factors.

In addition, the use intention of students varied according to their majors despite providing the same course name, instructor, assignments, and test question type.

Based on an aforementioned study [17], which had confirmed the applicability of Python programming intention, Cheon et al., (2022) conducted a follow-up study to determine the reasons for low effort expectancy [19]. Using the extended unified theory of acceptance and use of technology by Venkatesh et al. [20], they analyzed 257 non-major students in required and elective liberal arts courses via a survey and structural equation modeling, and then evaluated the influence of behavioral intentions. The analysis showed that hedonic motivation (t = 4.42, p < 0.001) had the greatest influence on the dependent variable, followed by performance expectancy (t = 5.496, p < 0.001), social influence (t = 4.517, p < 0.001), and facilitating conditions (t = 2.982, p < 0.01). The path coefficient of effort expectancy (t = 0.774, p > 0.05) had no statistically significant influence.

Furthermore, as a result of examining the matrix using the influence on the dependent variable and index of each factor, it was found that hedonic motivation and performance expectancy were distributed in the first quadrant at the upper right part, whereas social influence and facilitating conditions were in the third quadrant at the lower left part, as shown in Figure 1.

To increase the continuous use intention of students, it was necessary to strategically give priority to hedonic motivation and performance expectancy in the first quadrant, and establish a plan for using them. In particular, there have been several attempts to induce hedonic motivation among students in various studies. For example, Boytchev and Boytcheva (2020) improved students’ scores and boosted their motivation by applying gamified evaluation to their classes designed to improve computer graphics skills [21]. Additionally, Manzano-León, et al., (2021) utilized online escape rooms as part of their educational strategy to promote motivation among students in the Department of Education and Social Psychology [22]. It turned out that using gamification elements was helpful in encouraging students to debate and participate. For these reasons, hedonic motivation will be used as the first factor in this study.

Furthermore, an independent sample t-test was performed to determine whether there was a difference in the behavioral intention between the required liberal arts courses and elective courses; it was found that the difference was not statistically significant.

Based on the aforementioned studies [17,19], Yoo, et al., (2020) conducted a study on the factors affecting the academic performance of students, with the goal of improving educational outcomes [23]. They collected the characteristics of 463 students and analyzed the predictive classification between the student characteristics and academic performance. The student characteristics consisted of Python programming intention-related independent variables, confidence, attendance, and the frequency of chatting in online chat rooms. The academic performance consisted of a score and grade. A data mining method was applied using multiple and logistic regression analyses, and finally, the importance of the variables was examined using a decision tree. According to the analysis results, the factors extracted through the feature selection consisted of attendance, effort expectancy, hedonic motivation, confidence, online chat frequency, and Python programming intention, as shown in Figure 2.

In terms of the importance of the variables, it was found that hedonic motivation was the most important variable for grade “A” students, whereas attendance had the greatest impact on the academic performance of the grade “B” students. Although a previous study [19] showed that effort expectancy was not statistically significant, it was found to be an important factor—important enough to be chosen in feature selection for both score and grade. Furthermore, it is in the root node in the decision tree analysis of grade “B” students. Students with the lowest effort expectation were classified as those with the lowest scores. This means that grade “B” students find programming more difficult than grade “A” students; therefore, it is necessary to provide a method that simplifies students’ approach and makes learning easy. Therefore, effort expectancy will be used as the second factor in this study.

Lastly, it is confidence that motivates learners and suggests strategies that help to implement their plan [24]. Confidence, one of the most important motivational factors among learners, is known to be the major factor that determines math competency [25] and computer skills [26]. Additionally, Yoo, et al., (2022)’s multi-regression analysis results proved that confidence has a positive impact on grades [23]. Although not derived from the decision tree analysis, this study will consider confidence the third factor.

3. Research Method

To propose a research model based on the previous related studies, this study will proceed in the following way: First, check the descriptive statistics of the factors derived in Section 2, which are effort expectancy (EE3), hedonic motivation (HM2), and confidence (C3). Second, set the educational objective that meets the needs of the new environment in the fourth industrial revolution. Third, suggest an SW education strategy framework based on the educational objective.

The first is the descriptive statistics based on the description of the variables used in Yoo, et al., (2020) [23] and collected data. The measurements of hedonic motivation, effort expectancy, and confidence are visualized so that it is possible to check the chances of applying them to the SW education strategy framework. The questions for the variables are as follows, and respondents were asked to give their answers on a scale of 1 to 7.

• HM2: I find Python programming interesting.
• HM1: I find Python programming to be enjoyable.
• EE3: Python programming is easy for me to use.
• C3: I feel that if I work hard in this class, I can succeed.

Among them, HM1 and HM2 are related to hedonic motivation, so it is necessary to select variable that reflects it. HM2, EE3, and C3 are corresponding to score, so HM2 is used. This study uses grades separated by 40:60, as categories need to be used in the box–whisker plot generated in SPSS v. 27 (IBM Korea, Seoul and Korea).

The analysis results show that the average of C3 (M = 5.594, S.D = 1.40628) is the highest, followed by HM2 (M = 4.58, S.D = 1.649) and EE3 (M = 4.05, S.D = 1.713). The box–whisker plot of two variables is described in Figure 3. For HM2, the quartile ranges of Grade “A” and Grade “B” were found to be the same. For EE3, the quartile range of Grade “A” is 4–6, less scattered than that of Grade “B”. For C3, it is found that the quartile range of Grade “A” is significantly different from, and does not overlap with, that of Grade “B”. However, outliers were found in Grade “A”.

The second is setting an educational objective. Objectives accurately indicate the direction in which a manager believes an organization should move forward, and strategies can be set according to these objectives [27]. Our objective is to “train students in a way that can help them become knowledge workers needed in the era of the fourth industrial revolution”. Of course, IT subjects such as programming do not take priority over other subjects such as arts in that multidisciplinary education must be carried out for the educational purpose of nurturing the whole person [28]. However, more and more non-computing major students are becoming interested in and demanding computing courses, as computing is becoming more important throughout all fields [29]. With this backdrop, Leidig, Ferguson, and Reynolds (2015) suggest that computing education can be connected with “problem-solving skills and critical thinking”, one of the objectives of liberal arts education, “ethics”, and “information and digital literacy” [30]. It is more important for knowledge workers to perceive the outside world in that computers typically display situations that actually exist and ignore events in the perceptual domain, which are not clear facts [31]. This is one of the key competencies required of knowledge workers as they must face and analyze challenges faced by the organization and make decisions with a tremendous amount of data in front of them in the era of the fourth industrial revolution.

Thirdly, based on the objective and related studies in Section 2, this study suggests an SW education strategy framework for non-majors with a focus on the 2 × 2 matrix presented in Figure 4 below.

First, we will explain each axis. The effort expectancy on the vertical axis refers to the ease of using programming language. Students need sufficient practice and familiarity to increase the effort expectancy, which requires a step-by-step educational approach. The hedonic motivation on the horizontal axis refers to the degree of pleasure or satisfaction felt while creating programs. Students may feel a sense of achievement when they encounter various errors in the code but can solve the problems by themselves.

The objective of education in the SW education strategy framework is to develop Beginners (the students in the third quadrant in the “as-is” state) into Skilled individuals (students in the first quadrant in the “to-be” state). The Beginners in the third quadrant account for the majority of students; they have low effort expectancy and hedonic motivation. Through appropriate and high-quality theoretical and practical education, the Skilled students in the first quadrant are highly confident because they find programming easy. Furthermore, their sense of satisfaction is high as they watch their own programs operating effectively and generating values.

There are two methods of developing students from Beginners to Skilled individuals. First, students can develop from Challengers in the fourth quadrant into Skilled individuals. Although Challengers have not yet received SW education like Beginners, they have a high interest in SW, and tend to solve problems without giving up. Such students can be developed into Skilled individuals by increasing their effort expectancy through step-by-step education. Second, Beginners can be developed first into Theorists through repeated learning while constantly arousing their interest in SW and setting their learning difficulty lower than that of Challengers. If excessive information is provided to Beginners at once, their interest may wane. Therefore, a suitable amount of study and practice, combined with feedback, should be provided; particularly, the students should be encouraged [32]. Through this process, Theorists become familiar with basic syntax and SW concepts. Moreover, new challenges and accomplishments can be expected when students are provided with customized solutions according to their academic majors and interests. Based on this, students can be ultimately developed into Skilled individuals. At this stage, it is important for the students to consciously practice programming as a habit [32], and assignments with sufficient practice time and difficulty level should be provided during practice sessions.

Finally, it is confidence that drives students to execute the plan based on strategies. Kim and Mauborgne (2017) state that a humanistic process is needed as one of the core elements in the blue ocean shift strategy, which is about boosting people’s confidence and making them implement their plans on their own [33]. With this in mind, instructors should continue to inspire confidence in students so that they can overcome anxiety and obstacles on their own while moving towards a to-be state. As a result, confidence has the effect of boosting students’ learning speed in the matrix.

4. Result of Expert Interview with the Model

We conducted focus group interviews (FGI) with experts in the fields of education and information systems to systematically verify the effectiveness of the analysis framework [34,35]. Nine interviewees participated (three professors, three SW engineer experts, and three researchers) in the interviews over the course of 2 days on the 12–13 October 2022. When designing the interview questions, we excluded general-level questions as much as possible, and chose questions that could elicit expert-level opinions.

The FGI on the framework was conducted with the in-depth interview method. There are two steps in this process: FGI, which is a preliminary step, and content analysis, which is a result analysis step. A group of users (professors, researchers, SW engineers) were invited in order for us to check the user experience of, and improvements to, the framework. The respondents in the interview were classified into different groups depending on their characteristics. The analysis procedure was detailed according to the content analysis definition of the prior research for framework analysis. The interview answers were dissected by sentence, and pre-processing was conducted. Then, the data were divided by unit, and an evaluation index was extracted based on core consistencies [36].

The analysis model for this content analysis is described in Figure 5.

Step 1: Sentences are classified as “advantage” units and “improvement” units.

Step 2: Perform a pre-test to find representative words. These words are integrated and classified depending on their similarity.

Step 3: Check the hierarchy of words. Evaluate and categorize the similarities of sub-concepts in the structure.

Step 4: Perform a statistical analysis through SPSS and then identify relations between objects.

There were six interview questions: three common questions and three free-answer questions. The common questions were about the framework’s usefulness, application flexibility, and ease of use. The free-answer questions were about the characteristics and expert evaluations of the framework [37,38].

• Usefulness of framework—whether the potentials of students for development in the future could be predicted through the analysis framework [39].
• Application flexibility—whether the framework can be applied to general students without being significantly affected by the analysis target students’ majors and the current SW industry’s characteristics.
• Ease of use—whether the analysis methodology of this framework can be applied easily in various fields [40,41].

These three questions are as follows.

Question 1a.

Is this framework appropriate as a methodology for showing the trajectory of academic achievement? Can the trajectory be properly used in education?

Question 1b.

This framework intends to show how students’ academic performance can change. What role does this framework serve to achieve this? What aspects need to be improved?

Question 2a.

Can this framework be applied to various learning processes in the SW industry?

Question 2b.

The SW industry has a range of the latest technologies and fields of work. What aspects of this framework need to be improved to be used in various SW fields?

Question 3a.

Is it easy for professors and researchers to use this framework?

Question 3b.

Does this framework have any characteristics that make it easy to apply to learning strategies? What improvements are needed for professors to use this framework simple and easy?

These three questions ultimately aimed to determine whether the framework had a logical structure and could be reliably applied in other studies.

In the case where the interviewees provided ambiguous answers, we used a supplementary method to measure the framework’s effectiveness more clearly. The answers to the questions were given on a five-point Likert scale from “Strongly agree” to “Strongly disagree”. The questions were designed to approach the opinions of the interviewees more objectively through the answers.

In the results of the answers to the questionnaire questions, high scores of greater than 4 points were received on an average for usefulness of the framework and application flexibility. However, a score of 3.2 points was received for the ease of use, which was lower than the scores of the other factors.

The analysis showed that this score was received because the answers were given considering the resources required for the categorization of students in the process of applying the framework to actual student data. It was inferred that the categorization process should be more objective and quantifiable.

The three free-answer questions and their main answers are as follows.

(1)

How can the Beginner and Challenger be classified?

The biggest difference between the Beginner and Challenger lies in whether the student can determine whether the encountered problem can be solved using SW. Beginners have an interest in learning SW, and tend to use it to solve various problems. Challengers have various knowledge levels for SW, and tend to attempt challenges to learn new technology.

From this perspective, Challengers, unlike Beginners, attempt to learn technologies that are different from those of the past. Furthermore, Challengers explore more advanced SW knowledge and actively participate in innovative domains created in the fields of self-driving vehicles, Internet of Things (IoT), 3D printing, AI, virtual reality, etc. Meanwhile, Beginners continue learning while maintaining their interest in the current SW industry.

(2)

What kind of education is required for Beginners and Challengers? How can the level of difficulty be adjusted?

An effective education strategy for Beginners has aspects similar to those observed in continuing education. For the Beginner group, it is most appropriate to provide SW education programs that can be easily understood, even by novices to programming education. For example, subjects that appear often in the media and that students are interested in, such as courses on app programming or smart IoT programming, are appropriate.

Therefore, Republic of Korea SW-oriented university education is appropriate for Beginners. This is because the main goal of the SW-oriented universities in the Republic of Korea is to train high-quality SW specialists and nurture future skilled people who can apply SW to their respective majors and fields of specialty. In education for Beginners, it is important to provide an appropriate level of difficulty so that Beginners can maintain a continued interest in SW education. If the difficulty level of SW education is too low or too high, the interest of Beginners in education may decline considerably. Furthermore, diverse interests need to be aroused by applying SW education to real-world problems, as is achieved via problem-based learning.

Challengers need to cultivate the ability to recognize problems around them, view them from a logical perspective, and use SW as a tool to solve them on their own. From this perspective, Challengers need education that involves planning projects and analyzing/designing their major elements on their own. Therefore, it is essential that instructors with practical experience in the information technology field provide one-on-one guidance for a certain period. For example, SW-oriented universities in South Korea have been gradually increasing class hours on practical projects. The difficulty level of practical education is high for students owing to its nature. Therefore, it is necessary to ensure that students are rewarded with increased competence according to the level of difficulty. Information should be provided to students to let them track their progress, which will help complete their courses.

(3)

What kinds of customized solutions are there for developing Theorists into Skilled?

Typical solutions include basic/intermediate/advanced SW courses, SW hackathons, and SW competitions. Because of the nature of the Theorist, the capabilities of the Theorist group have various distributions. Therefore, students in the Theorist group need to be supported so that they can take part in various activities according to their capabilities. Therefore, it is essential to provide customized solutions continuously and consistently so that as many Theorists as possible can participate and maintain their interest in the courses to develop their capabilities.

Particularly, it can be very helpful for students with an intermediate or lower SW skill level in the Theorist group to experience SW education. From this point of view, it is necessary to build mentor–mentee relationships by providing assistant instructors to the entire Theorist group in the SW education programs, and let the students participate in the customized solutions.

At the beginning of the interview, we explained the direction of FGI and content analysis to professors and researchers from the user group, and the SW curriculum used in modeling the methodology to SW engineers. During the FGI process, we prepared for the interviewer to collect content analysis data after the interview was over.

By analyzing all the phrases in the interview, we extracted 94 characteristics, 38 advantages, and 23 improvements. A pre-test was conducted on the data and exceptions were deleted. Then, the data were divided by concept with duplicate and similar items being deleted. As a result, the data were organized into 15 “advantage” units and 9 “improvement” units. These units were then divided into a total of seven categories. We analyzed units in each category and conducted SPSS’ multiple response analysis for analytical factor values, deriving four “advantage” items and three “improvement” items in the strategic framework. These seven categories received a high rating of 10% or higher. Exceptional cases of the methodology were identified by performing feedback on the three “improvement” units [42].

5. Conclusions and Discussion

In this study, we proposed a framework by strategically analyzing student characteristics. The goal of this study was to support professors in the analysis of the distribution of students, and provide more systematic guidance to provide customized education focusing on non-major students.

For this framework, we proposed a 2 × 2 matrix based on the two characteristics analyzed in previous studies. The framework showed the development process of students, from the Beginner group that were new to SW into the Skilled group.

For the diversity of the analysis results, we classified two paths—the Theorist and Challenger groups—in the development stages from Beginner to Skilled, based on previous studies. Challengers are a group that has basic knowledge of SW and are enthusiastic about learning new technologies.

In this regard, Challengers are different from Beginners; they attempt challenges to learn technologies that are different from those of the past. Therefore, they have relatively high hedonic motivation compared to Beginners, and seek advanced SW knowledge, participating actively in innovative domains created in fields such as AI and virtual reality.

Beginners are a group that continues learning while maintaining their interest in the current SW industry. They can eventually progress into the Skilled group by continuously cultivating their hedonic motivation and effort expectancy. Each stage in this framework was derived rationally and logically through FGI, and we confirm that the stages are suitable for use when educating non-major students.

There are two limitations to the methodology proposed in this framework.

The first one is the scope of the students not majoring in SW. Because some of those students are in fields that are deeply related to the SW field, more detailed sub-categorization should be considered in follow-up research. In this study, we defined non-SW major students as students whose curricula did not directly include SW education.

The second limitation is that this study was based on a school curriculum, and focused on a semester-based analysis. Therefore, there is some difficulty in applying the proposed framework to short-term SW courses, which are often offered in current online education programs; it is necessary to consider a more sophisticated mechanism.

In follow-up studies, we will need to refine the framework by introducing short-term SW education programs, such as nanodegrees.

Owing to the nature of the SW field, non-major students often take interest in SW and begin studying. Moreover, in many cases, experts in other fields use computing technology for their work [43,44]. Therefore, this framework can be used to examine the tendencies of people entering the SW field and the relevance of SW education to their work.

The prior research of this framework analyzed the learning intentions of Python programming class students. The relationship between learning factors was identified through regression analysis of data and logistic regression analysis. Then, decision tree analysis was applied to the results to look into the importance of the variables. Based on these data, we designed a matrix using hedonic motivation and effort expectancy to raise students’ intentions for continuous learning. Additionally, the methodology was constructed by applying the confidence value identified in the multiple regression analysis as the third variable.

In the FGI results, “promptness” accounted for the highest proportion of all categories. The education strategy framework has the characteristic of being a tool that can be directly applied to the educational field. In other words, the first goal is for professors to quickly determine students’ learning performance. The second largest proportion following that of promptness is “elasticity”. This indicates the features that characterize the SW field. While the SW industry has a variety of structures, some similarities are maintained through all development processes. Plus, there are similarities in human resources, equipment, tools, languages, etc. This forms the SW industry with high mobility and similarity compared to other industry fields. In this respect, elasticity is one of the major traits that uphold this framework. In terms of the intuitive nature of the framework and the empirical aspect of education, the results of units of reliability (trajectory reliability, universality, storage of results, reflection of information) require rapid application. To meet this condition, it is necessary to continue to supplement the framework in a long-term and detailed manner. Applying exceptional variables to the items marked as negative in

Table 1. The conceptual content survey by analyzing unit concepts.

Category	Positive/ Negative	Concept	Number of Concept	Response	Ratio (%)
Convenience	positive	Convenience, easy application, robustness, stability, accessibility, familiar structure, easiness	6	11	11.82796
Elasticity	positive	Elasticity of the field, flexibility, easy use, simple programming, simplicity, assignment of values, equations, procedural changes	7	13	13.97849
Promptness	positive	Intuitiveness, 2 × 2 matrix structure, formalization, single structure	4	16	17.2043
Reliability	positive	Reliability of trajectories, universality, storage of results, reflection of information	5	10	10.75269
Elasticity	negative	Exceptions, technological changes, framework updates, students’ tendencies	5	12	12.90323
Convenience	negative	Statistical data needed, “semester” unit, a long time needed, framework changes	4	11	11.82796
Reliability	negative	New directions, exceptions, difficulty in identifying changes	3	12	12.90323

can be considered in future for the framework proposed in this study. For example, we may be able to judge future achievement by reflecting certain values’ biases in the learning performance trajectory. By applying weighting elements to the framework this way, it will be possible to update the framework in a more sophisticated manner.

The proposed framework does not deeply incorporate exceptional or environmental factors, such as an innate talent for SW knowledge. In the case of students who have a talent for SW and algorithms, we assumed that their hedonic motivation and effort expectancy were already positioned outside the framework, or they would move directly into the Skilled group, bypassing certain paths [45,46]. In addition, other exceptional factors should be considered in follow-up studies to perform advanced analysis. The UTAUT model used in this study was generated based on eight related theories, including the technology acceptance model (TAM). Effort expectancy is related to perceived ease of use in TAM [47]. According to research by Arpaci (2017), self-efficacy, which refers to a judgment of a student’s ability to use a system, statistically affects perceived ease of use [48]. In other words, it is expected that raising self-efficacy by differentiating the difficulty between Beginners and Challengers will increase effort expectancy, and that hedonic motivation is related to perceived enjoyment [49]. Arpaci and Durdu (2019) and Arpaci (2021) applied the Scratch language, which was taught to computer science students, to TAM, and found that perceived enjoyment had a statistical impact on attitude, and the parameter does have an impact on continuous use intention [50,51]. In particular, perceived enjoyment has the greatest impact on attitude, which makes it deeply related to existing research [19]. Therefore, it is necessary to look into the applicability of TAM in future research. Furthermore, each student’s academic performance can be associated with his/her hedonic motivation and effort expectancy to design a method for modeling the competency indices of students.

In the era of digital transformation, the potential that digital technology holds in the labor market is huge. With the spread of digital technology, the degree of automation is constantly increasing in major fields in the SW industry. Work productivity is increasing, and job quality is improving. From the beginning of 1980 to the second half of the 2010s, wages grew rapidly along with technological maturity in the information industry. The convenience provided by digital technology plays a large role in this phenomenon. This process is directly related to changes in digital and production processes, and is evolving into the form of a smart factory. In this regard, SW education provides students with job stability in the labor market. This framework will help students locate where they are at in the SW education stage and serve as a guide in an educational curriculum to become an advanced SW engineer.