Effectiveness of Problem-Based Learning in the Unplugged Computational Thinking of University Students

Abstract

Computational thinking is recognized as a critical competency in contemporary education, preparing individuals to tackle complex challenges in a digitally pervasive world. In this quasi-experimental design study with pretest and post-test measures, the possibility of developing computational thinking from the field of didactics of mathematics in higher education students was investigated. This was performed via a problem-based learning (PBL) methodology using problem solving in the experimental group or, alternatively, focused on the analysis of solved problems in the control group. After the intervention, the control group experienced a statistically significant improvement in the scores obtained in the post-test measure. Thus, PBL and problem solving did not lead to an improvement in the students’ computational thinking, whereas the analysis of solved problems approach did. Therefore, the results suggested the potential benefits of this latter methodology for teaching computational thinking.

1. Introduction

Proficiency in computer and programming skills has become a key factor for success and adaptability in an increasingly digitalized world. The ability to understand and apply computer science and programming concepts is a valuable resource in the current era, promoting innovation and problem-solving across a wide range of fields [1].

For these reasons, the development of computational thinking has become of paramount importance in today’s society as it provides a structured and logical approach to addressing complex problems and making informed decisions. This type of thinking promotes higher cognitive processes, goes beyond programming, and focuses on problem solving and the ability to approach challenges in a systematic manner [2]. Starting this learning process from an early age allows students to acquire solid skills and a mindset to meet future technological challenges [3].

In addition to being relevant for students, it is also essential to focus on the development of computational thinking in future teachers. These professionals will not only transmit their knowledge but will also play a fundamental role in the formation of relevant 21st-century skills and competencies. Therefore, training teachers in computational thinking not only enhances their learning but also promotes quality education for the modern world, where technology and information play an increasingly important role in our daily lives [4].

With the purpose of fostering the development of computational thinking in teachers in training, a contextually integrated educational intervention was designed in the area of mathematics didactics. This educational intervention was supported by the recent inclusion of computational thinking in the primary education curriculum in Spain, specifically in the area of mathematics. Additionally, the practical and manipulative approach that characterizes mathematics teaching at this educational stage was leveraged, recognizing its importance in improving skills related to problem-solving [5].

An approach was chosen that would allow students to explore computer-related concepts and problem-solving without relying on digital technology. In other words, it was decided to develop computational thinking in an unplugged way [6]. The choice to develop “unplugged” computational thinking was due to the participants’ lack of prior programming or coding experience. The unplugged approach was adopted as a pedagogical strategy to introduce computational thinking to avoid affecting the research by the participants’ level of familiarity with the technological tools most commonly used in the computer field. This approach is especially beneficial in educational environments where technological resources may be limited or in situations where there is a desire to gradually introduce students to computational thinking before engaging more advanced digital tools [7].

Consequently, a quasi-experimental research design was carried out with a control group and an experimental group. In both groups, the promotion of computational thinking was pursued using a problem-based learning (PBL) methodology. In the experimental group, this was focused on problem-solving using Spanish playing cards, while the control group used a classical approach with solved problems as a resource to understand and evaluate solutions to the posed questions.

To assess the effectiveness of the methods used, the following research questions were formulated:

• RQ1: How do PBL methodologies influence the development of computational thinking?
• RQ2: Is problem-solving with Spanish playing cards more effective than the classical approach of analyzing solved problems in developing computational thinking?

Therefore, the main hypothesis of this research posited that both methods were effective in developing computational thinking; in addition, the secondary hypothesis suggesting that problem solving might be a more effective methodology in this context was also explored.

State of the Art

Computational thinking is a concept that has evolved in recent decades in the fields of education and computer science. Despite its growing importance in modern society, there is no single, universally accepted definition that fully and definitively captures what computational thinking entails. This phenomenon reflects the multifaceted and constantly developing nature of this field, as well as the diverse perspectives from which it can be approached [8].

Nevertheless, although there is no universally accepted definition of computational thinking, the literature on this concept shows notable consistency in the dimensions, components, skills, or concepts it encompasses [9]. Computational thinking is operationalized in the literature as a collection of concepts or components such as abstraction, decomposition, debugging, iteration, algorithms, evaluation, or generalization. These dimensions are recognized as essential to computational thinking, commonly researched, and closely linked to computing practices [10].

Computational thinking, as a formal concept, emerged with the advent of computers. However, although the term and the central idea gained relevance in the digital era, the components or dimensions that constitute computational thinking preexist modern computers [7]. Before the proliferation of digital technology, there were already methods and techniques for problem solving that, although not called “computational thinking”, shared similarities with its principles. With the arrival of computers, the concept of computational thinking emerged, giving a new nuance and relevance to problem-solving, focusing on the use of connected tools and specifically programming [11].

For this reason, the development of computational thinking through unplugged problem-based learning is presented as a coherent and highly relevant alternative in current education [12]. This approach is based on the recognition of computational thinking as a set of essential cognitive skills that enable individuals to approach complex problems logically and efficiently and not simply as a technological skill [6]. By placing a central emphasis on problem solving or analysis as the core of computational thinking education, the development of algorithmic, critical, or analytical thinking is promoted, among other higher-order cognitive processes that are essential not only in the field of computing but in virtually all disciplines and everyday life contexts [13].

Given that computational thinking has an inherent connection with computer science and computing, an approach to problem-solving for beginners in this field can be proposed from the perspective of computational logic and basic data structure [14]. In this direction, the balance between accessibility and complexity of the search and sorting algorithms positions them as an ideal context for challenging students of various educational stages. Its understanding and elaboration provide a solid foundation for learning computer science and enhancing problem-solving skills [15].

In turn, in the stages of early childhood and primary education, there is a convergence between solving problems related to computing, coding, or programming and the use of fundamental mathematical skills [16]. This similarity manifests itself when applying mathematical strategies to approach and solve tasks of this nature. For example, logic, sequential reasoning, and the understanding of patterns and relationships, all essential elements in mathematics, become crucial tools in the process of internalizing computational concepts [17]. For these reasons, the area of mathematics is ideal for the development of computational thinking via the resolution of unplugged algorithmic problems in primary education [18,19].

There are numerous unplugged materials that have been used to develop computational thinking skills; among the most popular are robotic kits, paper tasks, or board and card games, the latter being the most common [12]. Board and card games encourage interaction and high-level thinking, aligning with programming and coding principles [20]. These resources have a long history of use in education and have demonstrated significant teaching and motivational results in developing computational thinking [21,22,23,24].

On the other hand, several studies have shown that the computational thinking of pre-service teachers is influenced by multiple attitudinal, emotional, and behavioral factors, such as self-efficacy, confidence, interest, and willingness to learn. Additionally, the applicability of the acquired knowledge and the possibility of using it in the future is crucial for improving their skills related to this type of thinking [4]. Given the influence of such factors on the computational thinking of teachers in training, opting for the implementation of active and participatory methodologies in the learning process of computational thinking skills can be a strategy to motivate and reinforce these aspects in the educational process [25,26,27].

Based on these considerations, it was decided to carry out quantitative research with a quasi-experimental design involving a control group and an experimental group composed of undergraduate degree students in primary education. The aim of the research was to evaluate the impact of a PBL methodology applied using two different approaches. On one hand, problem-solving using Spanish playing cards as unplugged material was employed in the experimental group, and on the other hand, a classical approach to analyzing solved problems of the same type was used in the control group.

The intervention tasks for both the experimental and control groups focused on simulating computational problems. The tasks were designed to introduce students to situations that required creating, analyzing, optimizing, or debugging search and sorting algorithms. The aim was to develop an intervention useful for their future careers that would motivate and interest the students, promote attitudinal factors, and be based on the application of prior knowledge to solve challenges.

To assess the impact of the implementation of such tasks, the following hypotheses were formulated:

H1. 

Students who face problems related to data sorting and searching through the use of cards achieve statistically significant improvements in their computational thinking skills.

H2. 

Students who analyze solved problems related to data sorting and searching with cards achieve statistically significant improvements in their computational thinking skills.

H3. 

Students who face problems related to data sorting and searching through the use of cards obtain, in statistically significant terms, higher development of computational thinking with respect to students who analyze the same type of solved problems with a classical approach.

2. Materials and Methods

The participants were students from the Primary Education Degree at the Faculty of Education, Economics and Technology of Ceuta. The study sample consisted of two pre-established groups from the second and third years of the degree; therefore, the sampling used in the research was non-probabilistic by pre-established groups.

The study was conducted during class hours and was taught in the subjects of Teaching and Learning Mathematics for Primary Education (TLMPE) in the second year and Design and Development of the Mathematics Curriculum for Primary Education (DDMCPE) in the third year. Participation in the study was optional, and the research was approved by the Ethics Committee of the University of Granada (registration number: 3500/CEIH/2023). The sample consisted of 31 students (5 males and 26 females), where the experimental group (hereafter group A) was composed of 2 males and 11 females, and the control group (hereafter group C) included 3 males and 15 females.

After applying the pretest and analyzing the homoscedasticity of the groups, no significant differences were observed between the groups in terms of variances in the results. Therefore, considering that the demographic characteristics and the size of the groups were similar, the third-grade group was selected as the control group and the second-grade group as the experimental group.

The Computational Thinking Scale (CTS) [28] was used as a pretest and post-test data collection instrument. This five-point Likert-type scale consists of 29 items that can be grouped into five factors: creativity (8 items), algorithmic thinking (6 items), cooperativity (4 items), critical thinking (5 items), and problem-solving (6 items). The CTS measures students’ computational thinking skill levels and has been validated in higher education students. Additionally, reliability values for the CTS above 0.7 were obtained, calculated from Cronbach’s Alpha coefficient, in the pretest and post-test measurements of the global scale (obtained from the 5 factors) and the three dimensions analyzed.

2.1. Data Analysis

The Shapiro–Wilk normality test was applied to the data sets selected due to the sample size not exceeding 50 individuals. In turn, Levene’s tests were conducted to examine the homoscedasticity of the variables under consideration. Finally, hypothesis testing was performed using both parametric analyses with independent and related samples T-tests, as well as non-parametric analyses with Mann–Whitney U tests. The entire data analysis process was carried out using SPSS statistical software, version 25.

2.2. Procedure and Materials

The research was developed in five sessions per group. The pretest application in group A took place on 4 May 2023, while the post-test was applied on 16 May 2023. Group C received pretest administration on 8 May 2023 and post-test administration on 23 May 2023. Each group participated in an intervention consisting of three sessions, with each session lasting two hours.

The teacher responsible for both subjects, whose class time was used to conduct the research, developed the activities for both groups. The subjects of TLMPE and DDMCPE were selected because they belong to the field of mathematics didactics and due to the recent introduction of computational thinking into the primary mathematics curriculum.

In addition, it was decided to work with a PBL methodology, following the methodological line of the subjects to be taught in the groups. Consequently, an intervention program was implemented to improve computational thinking skills in the experimental group, based on problem-solving using manipulative materials and selecting cards for their versatility to simulate computational elements. Meanwhile, in the control group, the methodology focused on the analysis of solved problems without manipulative material.

When planning the intervention for the experimental group, the model for developing computational thinking skills by Palts and Pedaste [29] was used as a framework to investigate cognitive development through card games. Based on the three-stage sequential structure proposed by this model, a learning scenario was created in which the dynamics of card games were adapted specifically to address the phases of defining the problem, solving it, and analyzing the solutions in a playful and algorithmic context.

The intervention tasks were designed to simulate computational situations of data searching or sorting in which higher-level concepts were worked on in a context of basic unplugged coding or programming. More specifically, the purpose of using cards was to generate a learning situation in which each card emulated a variable or data point, and grouping cards simulated a vector or matrix. For example, when a card is face down on the table, and its value is unknown, it represents a “variable”, whereas if it is face up, it is an “integer data type”, making a similarity with computational elements along with mathematical concepts [19].

The different problems posed to the students of the experimental group were designed to generate the need to create search or sorting algorithms, varying the difficulty of the problems as the different objectives set for each session were achieved. The groupings for the tasks changed in each session, working both individually, in pairs, and in large groups.

To simulate the behavior of a computer, different rules were proposed depending on the problem presented. For example, the condition “you can only have one card face up on the table at any time” simulated the use of an auxiliary variable as a resource when searching for an element in a vector. Similarly, “you can only move a card from its position if you exchange it with another card” represented the swapping of positions of two data points in a vector when performing a sorting algorithm.

Additionally, depending on the problem posed, students were asked to count how many cards they turned over during the process of solving the task in order to compare resolutions in the large group and see who could solve the problem with the fewest number of flips. This approach incorporated the concepts of iteration and debugging.

Furthermore, in some of the assigned tasks, students were instructed to clearly document on paper the steps they followed in solving each problem so that if someone else read their “report”, they could reproduce those steps perfectly. This was based on the computational thinking proposal through algorithmic explanations by Peel, Sadler, and Friedrichsen [18], which focuses on creating handwritten algorithms that detail a sequence of steps to explain a process.

For simpler problems, these written explanations did not pose much difficulty. However, as the complexity or laboriousness of the tasks increased, students, when expressing their solutions, grouped steps, created functions, and represented “conditionals” or “loops” using everyday language.

An example of a problem posed is as follows: “Find 9 cards of the same suit. Shuffle them and place them face down, making a 3 × 3 square. You can only have two cards face up on the table at any time, and you can only move a card from its position if you swap it with another card. Following these rules, you must arrange the cards in ascending order (from left to right and from top to bottom). Write step by step how you solve it” (Figure 1).

On the other hand, in the control group, group dynamics were implemented to introduce computational thinking concepts in a more theoretical way than in the experimental group. In the first session, each student was assigned one or two components of computational thinking, and they were provided with an explanatory document detailing these components (algorithmic thinking, abstraction, conditional logic, iterative thinking, recursive thinking, and debugging), as well as the general concept of computational thinking. During this initial session, each student worked individually to create a summary, diagram, or concept map highlighting the most relevant information about computational thinking and its assigned components.

In the second session, students were grouped according to the components of computational thinking that had been assigned to them in the previous session, and each group was asked to share and discuss the information summarized in the previous session. Additionally, they were provided with three solved problems similar to those worked on in the experimental groups, but with detailed solutions. The students were asked to identify parts of these solutions where the assigned computational thinking component was manifested and to record them for use in the next session.

Finally, in the third session, the students were regrouped so that each group consisted of members who had worked on different components of computational thinking. Each student explained their assigned component to the rest of the group and how it was reflected in the solved tasks they had previously analyzed. To conclude the session, the teacher posed six questions to encourage group discussion and reach a consensus response. An example question was, “What is the difference between iterative thinking and conditional logic?”.

Regarding the role of the teacher in the interventions, they were responsible for explaining each activity, observing the students’ solutions, providing help, and offering tasks of varying complexity to students as needed to ensure gradual learning.

3. Results

Means and standard deviations for each variable were calculated, distinguishing between the results obtained in the pretest and post-test, as well as differentiating between groups. The results are presented in

Table 1. Means and standard deviations for CTS and subdimensions in pretest and post-test measures.

Group	Test	Global Scale		Algorithmic Thinking		Critical Thinking		Problem Solving
		M	D	M	D	M	D	M	D
A	Pretest	101.000	7.906	19.385	3.548	17.539	4.034	13.000	3.416
	Post-test	101.923	8.190	19.539	3.843	18.308	3.591	13.615	3.754
C	Pretest	95.611	8.493	18.833	3.400	17.000	4.201	13.111	3.341
	Post-test	98.944	9.831	19.056	4.893	17.722	3.611	13.889	3.270

.

It is observed that the means of the variables generally increase in both the experimental group and the control group. As for the standard deviations, a different pattern is observed, as they only decrease in three of the eight variables analyzed.

Regarding the Shapiro–Wilk normality tests for the results of the scale and its subdimensions, differentiating between groups, both at the time of pre-intervention data collection and afterward, the results of the statistic along with the degrees of freedom and p-values are shown in

Table 2. Shapiro–Wilk normality tests.

Group	Test	Global Scale			Algorithmic Thinking			Critical Thinking			Problem Solving
		S-W	df	p	S-W	df	p	S-W	df	p	S-W	df	p
A	Pretest	0.916	13	0.219	0.925	13	0.292	0.943	13	0.496	0.953	13	0.643
	Post-test	0.930	13	0.340	0.952	13	0.630	0.952	13	0.626	0.814	13	0.010
C	Pretest	0.951	18	0.446	0.953	18	0.475	0.927	18	0.173	0.926	18	0.167
	Post-test	0.989	18	0.998	0.967	18	0.745	0.935	18	0.239	0.952	18	0.462

.

In the analysis, a significance level of 0.05 was applied, so the null hypothesis was accepted when p-values were greater than 0.05. Consequently, it was concluded that the distribution of the data for the “problem-solving” variable did not follow a normal distribution in the post-test of group A, while the other variables to be analyzed did. Subsequently, the analysis of homogeneity of variance was then performed, applying a significance level of 0.05. The results (shown in

Table 3. Levene’s test for homogeneity of variance based on the mean.

Prueba	Global Scale		Algorithmic Thinking		Critical Thinking		Problem Solving
	F	p	F	p	F	p	F	p
Pretest	1.099	0.303	0.072	0.791	0.018	0.893	0.164	0.689
Post-test	1.069	0.354	1.387	0.248	0.246	0.624	0.009	0.924

) indicated that the variances of the scale, as well as the variances of the sub-variables, are equivalent between the groups. For all homogeneity tests, the degrees of freedom df1 was 1, and df2 was 29.

Based on the obtained results, comparisons were made between the pretest and post-test results in both groups A and C for the variables “global scale”, “algorithmic thinking”, and “critical thinking” using the intragroup paired samples T-test. Additionally, the “problem-solving” variable was compared using the parametric intragroup paired samples T-test in group C and the non-parametric Mann–Whitney U test in group A. The results of the analysis are shown in

Table 4. Paired sample T-tests and Mann–Whitney U tests.

	Global Scale			Algorithmic Thinking			Critical Thinking			Problem Solving
	t	df	p	t	df	p	t	df	p	t	df		p
GC-GC Post-pre	2.236	17	0.039	0.356	17	0.726	1.725	17	0.103	1.268	17		0.222
	t	df	p	t	df	p	t	df	p	Z		p
GA-GA Post-pre	0.817	12	0.430	0.262	12	0.798	1.382	12	0.192	−0.908		0.364

.

Significant differences were observed (using a significance level of 0.05) when comparing the pretest and post-test means of the control group in the “global scale” variable, while no such differences were observed in the other variables of the same group or in those of the experimental group.

Since different interventions were carried out in both groups and in order to analyze whether the effect of the interventions between the groups is statistically different, the differences in scores between the pretest and post-test of the control and experimental groups were compared using an independent samples T-test, as the variables met the assumptions of normality and equality of variances (F = 1.901; p = 0.179). The test statistic was t = −1.203 with 29 degrees of freedom associated and a significance value p = 0.239.

Since the p-value (0.239) was greater than the significance level used of 0.05, the null hypothesis was not rejected. In other terms, this meant that there was not enough evidence in the data to conclude that there was a statistically significant difference between the means of the differences in the pretest and post-test measures of the two groups.

4. Discussion

This research delves into a topic of great relevance in contemporary education: the development of computational thinking in teachers in training [1]. In line with the observations of Caeli and Yadav [7], who warn about the risk of the widespread use of digital devices and intuitive interfaces in our daily lives diminishing appreciation for the importance of understanding the internal workings of computational tools, it is essential to promote the development of computational thinking as a means to understand how these technologies work. The centrality of this topic in the digital era makes research in this field not only relevant but necessary.

The main finding of this research determined that PBL methodology with a classic, analytical, and unplugged approach had a positive impact on improving computational thinking among teachers in training. The approach applied, based on the analysis of solved problems, not only responds to a curricular mandate [5] but also aligns with the methodologies and content proposed for the subjects in the area of mathematics education for undergraduate degree students in primary education.

The intervention carried out in the control group was based on understanding the concept of computational thinking and the skills that compose it and then analyzing how these are integrated into problem-solving. As noted in review [4], this way of introducing computational thinking to teachers in training, focused on their professional development, has been the most common and has shown the most favorable results in the literature.

In the studies [4,19], it is observed that integrating computational thinking with the professional subjects of teachers is a favorable strategy for cultivating this type of thinking. The results of the present work are in line with these findings, as strategies inherent to mathematics teaching have been employed, integrating degree contents and fundamental concepts of computational thinking. This teaching practice has not only promoted computational thinking but has also acted as an essential pedagogical tool for teachers, offering resources applicable in their future careers.

While fostering the development of computational thinking through an analytical PBL methodology, an unplugged approach has also been utilized for this purpose. In the literature, unplugged computational thinking has not only shown significant learning outcomes in early childhood and primary education stages but has also been successfully applied in secondary and higher education, although to a lesser extent [2,6,12,18]. In harmony with the results presented in the literature, those shown in this research suggest that it is possible to develop unplugged computational thinking in higher education.

However, despite the evidenced potential of this unplugged approach to developing computational thinking, the results of this study also infer that not all unplugged methodologies significantly foster the development of computational thinking. The PBL methodology applied for solving unplugged sorting and searching problems using cards did not result in a statistically significant improvement in the computational thinking of the students. According to the results from [6], the effectiveness of this methodology could be influenced by the educational stage of the students, as the efficacy of unplugged approaches is greater in primary education students than in secondary or university students.

The planning of the intervention carried out in this study with the experimental group was based on well-founded theoretical models [16,29], justified instructional approaches [18], and coherent contexts [11,13]. It involved methodologies and materials that have been shown to be effective at different educational stages [21,22,23,24,25,26,27], considering the students’ prior knowledge in computer science, programming, and coding [13] and taking into account the attitudinal, emotional, and behavioral factors of teachers in training [4].

Despite this, in dissonance with the results obtained in [20,21,22,23,24], the use of card games as unplugged material has not significantly favored the development of computational thinking in the context of this research. Furthermore, the results obtained in this intervention differ from those observed in [25,26,27], where the use of a gamified methodology favored the development of computational thinking. Additionally, the use of theoretical models by Palts and Pedaste [29], Popat and Starkey [16], and Peel, Sadler, and Friedrichsen [18] also have not favored significant development of computational thinking.

The absence of significant results using an unplugged PBL methodology could be determined by several factors, among which attitudinal or emotional elements should be considered. These elements can act as barriers or facilitators in the learning process, and therefore, it is essential to delve deeper into the analysis of these components to understand their impact on the research [4]. It is important to note that this research was conducted within the framework of mathematics, a discipline that sometimes may not generate the expected level of motivation in students. Therefore, it is necessary to consider how perception and interest in the subject may be influencing the assimilation of computational thinking.

The results obtained in the experimental group may also suggest that the implemented design could benefit from methodological improvements. The unplugged activities carried out were designed to be solved using knowledge familiar to the students, but also introducing challenging tasks. However, the assimilation of computational concepts requires a learning process that involves higher-order thinking or high-level cognition, and the tasks performed might not be aligned with the current level of the students [13,14].

5. Conclusions

After exhaustively analyzing the results of this research, it can be affirmed that the proposed objectives have been met. The impact of the implementation of two educational interventions based on the PBL methodology has been evaluated, focused on one hand on solving problems related to sorting and searching data using cards as unplugged material and, on the other hand, focused on the analysis of solved problems of the same type. The hypotheses proposed for this study have been tested, and only the one stating that students who analyze solved problems related to sorting and searching data through cards achieve statistically significant improvements in their computational thinking skills can be accepted.

This research faced several limitations that must be considered when interpreting its results. Firstly, it is important to note that the sample size used was relatively small, and the sampling method was not random. Consequently, the results of the study are applicable only to the chosen research respondents and cannot be generalized.

Additionally, the limitation posed by the use of a single scale as the primary method of measurement must be considered. Using only one instrument to collect information can be insufficient to evaluate concepts that are, by nature, more multifaceted [8,9,10]. This could lead to an incomplete or partial understanding of the effectiveness of the interventions carried out.

These limitations highlight the need for a cautious interpretation of the findings and suggest the importance of considering future research that could address these restrictions to provide a more comprehensive and generalizable view.

In order to deepen the results obtained, it was decided to carry out additional qualitative approach research, with the main objective of describing and analyzing the perceptions of trainee teachers in relation to their initial practices focused on the development of computational thinking.

In [30], the qualitative findings revealed a widespread predilection for interactive and reflective learning. However, this approach also exposed new limitations, such as variability in the reception of challenges according to the different skill levels of the students and the need for more flexible teaching strategies that would accommodate diverse learning styles. Additionally, the choice of unplugged materials, while offering an innovative approach, does not include a comparison with methodologies that integrate digital devices, which could provide a more balanced understanding of computational thinking in current educational environments.

As future lines of research, it is proposed to first study the competence level of the students involved and their attitudes towards mathematics present in the group to adapt the proposed activities, as well as to graduate and space out the increase in difficulty of the activities.

It is also proposed to design and create specific tools and methods for data collection. These instruments, developed ad hoc, would be specially aimed at evaluating and measuring computational thinking in the specific context of the educational interventions carried out in the classroom, which could provide more valuable and reliable results regarding the theme under study.

In conclusion, while certain approaches of the PBL methodology appear to offer benefits in the development of computational thinking, there are still areas that require additional research and methodological refinement to fully understand their impact and potential in the educational field. These findings shed light on the complexity of fostering computational thinking in the higher education environment, the diversity of perspectives from which it can be approached, and the intricate nature of the concept.