Impact of Teaching a Neuro-Computerized Course Through VLE to Develop Computational Thinking Among Mathematics Pre-service Teachers

Abstract

The current study aims to determine the effect of teaching a mechanic neuro-computerized course through virtual learning environments (VLE) to develop computational thinking among mathematics pre-service teachers. The neuro-computerized virtual learning environments (NCVLE) model was designed to be used to teach the mechanics course to third-year students of the mathematics department. To achieve the targeted learning outcomes, the study recruited (102) third-year students of the Faculty of Education and classified them into a control group of (50) students and an experimental group of (52) students. The experiment lasted for 14 weeks during one semester of the 2021-2022 academic year. The results agreed with most of what has been found from relevant literature and studies. Also, the results indicated that the NCVLE model played a vital role in the purposeful teaching, learning, and assessment processes and enhanced the learning of computational thinking.

Keywords

neuro-computerized virtual learning environment computational thinking digital learning strategies

Introduction

In the last two decades, virtual learning environments (VLEs) in the field of education have attracted scholarly interest due to the utilization of new methods of technology in the learning process (Sanfilippo et al., 2022). Using virtual reality (VR) as a tool for learning requires several pedagogical principles to orient the implementation of activities. Moreover, the method provides learners with attractive activities that promote the independent acquisition of knowledge and reflection on previous experiences (Albus et al., 2021). One of the main advantages of VR is that it can easily illustrate concepts more than a conventional classroom, such as those in mechanics sciences, does due to its dynamic presentation, such as simulation, which enhances understanding (Boulton et al., 2018; Calvert & Abadia, 2020). In this example, learning occurs through simulation of the natural environment, which develops observation and problem-solving skills (Fussell & Truong, 2022).

Currently, VR is undergoing rapid development and increased engagement from various individuals because it provides an interactive environment through computer programs and possibilities of making changes to such an environment (Hamilton et al., 2021). This level of flexibility enables movement around and within virtual spaces (Boos et al., 2016) and natural interaction with the environment through life-like scenarios, situations, and environments. Previous studies define VR as the assimilation of reality, training, and learning (Peteranetz et al., 2020; February; Ng & Chu, 2021; Christofi et al., 2018; Pappa et al., 2018).

Moreover, VR provides users with highly interactive environments and three-dimensional (3D) scenes, such that it has gained attention as a tool for learning because it can motivate and facilitate spatial cognition in the performance of tasks (Kuznetcova et al., 2019) and generate live and affective experiences (Gomez, 2020). VR is an excellent method for experimental learning and can replace conventional contexts due to place and time constraints or safety concerns for students (Cheng & Tsai, 2019).

Virtual learning environments (VLEs) represent a category of information systems that electronically support higher education (Radianti et al., 2020). In addition, they embody the features of varied Virtual Reality Environments (VREs) and their associated information (Leite et al., 2022). Based on these attributes, VREs are of importance to teachers because they can be used to identify fields of interest and corresponding VLEs to reinforce learning (Ryan & Poole, 2019).

Recently, technological advancements in neuro-computerized are rapid and widespread. Neuro-computerized pertains to the system that establishes brain-computer connections. Thus, the opportunities for collecting information from the human brain for sharing and manipulating information using computerized methods are unprecedented (Trenado et al., 2021). As a result, neuro-computerized advocates in “a new era” of reinforcement, improvement, or advancement gained strong and are considered “better than good” or “better than humans” (Bustamante & Navarro, 2022).

Purpose of the study

The study aimed to provide a procedural guide consisting of a collection of applied stages for creating neuro-computerized virtual learning environments (NCVLE) model and assess the effectiveness of it in developing computational thinking (CT) among pre-service teachers of the mathematics department. so, it intends to answer the following questions:

1. What are the procedural steps used to create NCVLE to develop CT for pre-service mathematics teachers?

2. What is the extent of support that can be provided by NCVLE to develop pre-service mathematics teachers’ computational thinking?

Development in technology has led to the extension of its functions to daily life. For example, scholars on mechanics in the field of mathematics provide evidence that the course requires the development of computerized thinking (Jocius et al., 2022; Saritepeci, 2020). Neuro-computerized plays a vital role in elucidating the relationship between computational thinking (CT) and problem-solving skills of students, especially those undertaking the course. In general, however, students should be trained in CT regardless of academic discipline. Toward this end, a virtual environment based on neuro-computerized can be used as the basic principle of the curriculum. Designing such a neuro-computerized environment can enable students to focus on enhancing competencies and employing current best practices. An important step in this process is matching the learning environment with the major components of mathematical thinking. Moreover, the cognitive processes of students during CT can give insight into the professional development necessary to incorporate CT in the classroom. Khlaisang and Songkram (2019) and Lin et al. (2020) identified several benefits of an effective VLE system, such as (1) better learning and (2) high levels of self-competency are achieved in VLEs compared with the traditional learning environment.

Another advantage of VLEs is that they can support a learning environment with multiple users in a virtual platform, which encourages interaction and CT (Moon et al., 2020). Thus, the current study aims to design a VLE model that develops CT among Pre-service teachers of the mathematics department at the Faculty of Education. Toward this end, the study defines VLEs as a type of environment that combines the advantages of a learning management system (LMS) and a 3D platform (Open Simulator) to resolve issues in mechanics. LMS offers varied contents, such as e-books, videos, and animations, with unlimited access using all forms of media (i.e., smartphones and computers). A virtual environment is provided to promote cooperation and communication among students for group projects that aim to reach solutions to problems related to mechanics. At the same time, VLE enables other groups to present constructive comments from other sites (Ceallaigh, 2022; Alamri et al., 2020; Raj & VG, 2022; Luo et al., 2022; Xue et al., 2022).

Conceptual Framework

Virtual Learning Environments

VLE is a basic learning tool that facilitates learning through cooperation regardless of time and place constraints (Luo et al., 2022; Cheng et al., 2020). A virtual environment is an electronic space with an interactive application using computers to induce a digital artificial environment (Muñoz-Cristóbal et al., 2017). Thus, VLE can simulate real environments, procedures, objects, and processes and supports learning activities inside classrooms and independent learning outside classrooms. Undoubtedly, the future will require extensive use of this method (Alamri et al., 2020; Raj & VG, 2022; Xue et al., 2022).

VLEs can include electronic web management systems, virtual reality experiences, augmented reality, or a combination of these (Chao, 2016). The virtual world is a mathematical metaphysics and is a destination for virtual world users as it supports imagination. Importantly, virtual reality (VR) is increasingly used for teaching and learning. Inoue (2012, 3407) defines VR as follows: “an interactive computer-based application that provides a synthetic digital environment – and thus, virtual reality provides a way to simulate environments, objects, actions, and processes.” In education, it is a unique computerized technology that allows learners to interact within a simulated or imaginative environment (Muñoz-Cristóbal et al., 2017). This type of learning represents a change from static to dynamic, from isolated to interactive, from particular to general, from invisible to visible, and from exclusive to inclusive (Alamri et al., 2020; Hamutoglu et al., 2020).

Virtual reality (VR) is typically referred to as a “learning environment” in the academic literature concerning its implementation in the classroom. As a result, the question of whether these terms can convey the uniqueness of VR as a new learning environment arises, given that for many years, “virtual learning,” “virtual education,” “virtual learning environment,” or “learning virtual environment” have been associated with online learning without students' physical presence (Mukasheva et al., 2023). However, VLEs go beyond VR in that they create a realistic learning environment that mimics the actual world, demonstrating genuine innovation. This is because they do not rely on the availability of VR headsets, simulators, or 3D bases.

Uses of Virtual Learning Environments

VLE is a new approach that supports creativity and provides learning opportunities among students. VLE is being used here for computational and computerized learning. Drawing similarities to human thinking and computers was done as far back as the 1950s with information processing theory (Miller, 1956). Cognitive development studies use information processing theory to explain memory encoding. It assumes humans do more than react to environmental stimuli. Humans analyze information. Experts believe the brain’s principles and functions are simple, but neural networks and their actions are quite strong (Wang et al., 2003). VLEs present many advantages in education, such as simulating real-world actions and scenarios (Ito et al., 2020; Peteranetz et al., 2020, February; Ng & Chu, 2021). Many educational institutions that aim to expand their activities are mainly interested in using virtual classroom systems because they reduce the costs of conventional classrooms, travel, and time (Wang et al., 2019; Liu et al., 2021).

The framework for NCVLE design considers neuro-computerized in a virtual environment that promotes interaction among students to motivate learning experiences. It relies on cognitive and individualized learning concepts. In other words, learning occurs at the unique paces of learners (Ceallaigh, 2022).

NCVLE has many forms (i.e., VLE; Wang, 2020). However, it includes a definite number of stages, such as building the interface, using models for input specifications (i.e., targeted knowledge and skills), and establishing a model for learning processes (assessment) and learners. Figure 1 depicts the model.

Figure 1.

Model of the neuro-computerized virtual learning environment.

Within a NCVLE, students participate in a variety of activities. These activities are made up of a series of carefully planned problems that are connected together to demonstrate an integrated neural network. All of these tasks adhere to the computational thinking method in terms of formulating the problem in such a way that the neural computerized learning environment can assist in solving it, while providing a process for organizing and analyzing the data, and while creating a process for data representation through abstractions and simulations, such as the temperature of the soil when carrying out a measurement process for it. The process of offering solutions through algorithmic thinking and profiting from them later to solve similar problems connected to soil temperature or the proportion of salts in it, for example, begins with the provision of the relationship between the components of soil and temperature. Hence, this method of logical organization in offering solutions and solving problems increases the accuracy of the student’s learning and so develops his computational thinking skills and follows the same procedures in finding solutions to similar challenges in the future. The NCVLE model is responsible for all procedures that enhance learner performance, such as identifying common factors among soil components. Two main aspects require decisions, namely, what if intervention should take place. In many cases, intervention may not be required or decided by the system or appealed (i.e., learners asking for help when needed). Moreover, it could be achieved through accurate methods and by modifying the learning environment from an ideal viewpoint (Wu et al., 2021).

The model provides inputs to decide on educational alternatives and decode the most urgent intervention. Considering the range of decisions is important because it consists of two basic stages (Targeted knowledge and skills). Such an environment enables learners to add, change, or use opportunities according to their needs (Panjaburee et al., 2022).

At the lowest level, the model considers decisions based on a problem-solving context. For example, if a student is lagging in any of the steps or has trouble performing the last task, then monitoring of the educational environment and consultation with the model is required. Intelligent tutoring systems identify the type of interventions to be offered by giving hints or posing basic questions (Máñez et al., 2019). At the highest level, external decisions lend support by evaluating issues and curriculum groups. The role of artificial intelligence technology attracted scholarly attention and was used to develop models for decision-making (Nikiforos et al., 2020). The current study stresses one component of such efforts to investigate roles and improve estimations of learner status and decision-making on support.

However, ensuring that knowledge transfer occurs from VREs to the real world and vice versa is extremely important (Guerrero Alonso et al., 2022). In addition, virtual reality should be managed in a manner that enables VR to overcome problems in interactions with the natural environment and develop observation and problem-solving skills among students (Araiza-Alba et al., 2021). Thus, experimenting within a virtual communication platform is necessary to justify the use of VREs in supporting students by employing various methods, such as photographs of vast worlds.

It has become increasingly popular to use virtual reality (VR) in the classroom, as evidenced by recent systematic studies of VR applications in K-12 and higher education (Di Natale et al., 2020; Radianti et al., 2020). Radović et al., (2022) presented a process of designing an experiential learning environment (mARC) through an iterative design-based research course in a Master’s program for Educational Sciences. This Master is offered as distance learning and is mostly for teachers who want to get an academic degree and work while they study. To aid in the development of students' abilities and academic knowledge, their mARC model offers guidance and insights into the process of planning and constructing more immersive learning environments. According to Feyzi Behnagh and Yasrebi (2020), if used constructively, VR has the ability to promote constructivist learning; otherwise, it becomes nothing more than amusement and distractions in the learning environment. In addition, such advantages indicate that VLE is appropriate for higher education but requires caution in use.

Computational Thinking

This construct denotes the process of identifying and solving problems practically through an information agency (Shin et al., 2022). Specifically, Nordby et al., (2022). Define CT as the ability to:

a. Analyze a problem or process in parts;

b. Establish steps to produce a solution that fits the problem;

c. Simplify the problem without overlooking its complexity;

d. Apply previous knowledge and experience to new experiences in other fields.

CT in education promotes problem-solving skills applicable to all areas of specialization (Lockwood & Mooney, 2018). It helps students develop and understand the prerequisites for identifying and understanding real-world issues and forming links between prior knowledge and identifying problems (Strawhacker et al., 2017).

VLE promotes the statement, representation, and sharing of ideas as well as opportunities for experimentation, involvement in cognitive and rational processes, and presentation of formative comments at appropriate times (Martin et al., 2020). The basic hypothesis underlying the development of cognitive skills is to ensure knowledge transfer to other contexts, topics, and perhaps fields (Pinargote-Ortega et al., 2021). In other words, learners apply acquired knowledge and skills independently to new situations (Attias et al., 2022). Assessing CT acquired by students is a basic issue that requires standardized tools for accuracy (Varela et al., 2019; Palts & Pedaste, 2020). The general idea about the dimensions and suggested models of CT can be presented in three stages, namely, (1) identifying and (2) analyzing the problem, and (3) providing a solution. The three stages consider 10 CT, namely, (a) stating, (b), stripping, (c) restating, (d) analyzing a problem, (e) formulating, (f) aligning, and (g) setting the frequency of a computational design, and (h) automating, (i) generalizing, and (j) evaluating outcomes.

Saritepeci (2020) provided evidence of positive development in CT after experimentation. Durak and Saritepeci (2018). Proposed an extensive revision of the CT field and elucidate the lack of consensus on the components of CT. The authors concluded that CT includes eight components, namely, abstractions, manipulations, symbol system, concepts, analysis, frequent thinking, conditional logic, efficiency constraint, and correction of errors.

Methodology

Participants and Design

Quasi-experimental methods were used to identify the effect of teaching a mechanic neuro-computerized course through virtual learning environments (VLE) to develop computational thinking among mathematics pre-service teachers by comparing treated units. For this purpose, two groups were randomly selected, the first group being the experimental group, which consisted of 52 (male = 28, female = 24) third-year students from the Mathematics Department faculty of education (33% of a total of 150 students).

The second group was the control group which consisted of 50 (male = 26, female = 24) third-year students from the same department who undertook the same course and time frame. However, practical application was conducted using computers in the conventional mode (i.e., PowerPoint presentations and oral discussions).

The course is mandatory and uses programs, such as MATLAB version R2018b, LMS, and modern methods pedagogies suitable for VLE and the Internet. The course was implemented during the second semester of the academic year 2021–2022. Classes were held once a week for 4 hours (2 hours for lectures and 2 hours for practical application in a neuro-computerized virtual environment).

Designing the Neuro-Computerized Virtual Learning Environment

The NCVLE system was designed and tested from October 2021 to January 1, 2022. The experimental study was bead on the NCVLE to develop computational among students from February–April 2022. The researchers divided the experiment into the following stages (Liang et al., 2019):

First Stage

At this stage, input specifications are provided for the NCVLE using exploratory factor analysis (EFA). The study reviewed related literature to analyze and utilize related theories. The search consisted of the keywords CT, NCVLE system basis for determining input, process, output, and feedback, and principles, strategies, and technologies for teaching mechanics. Collected information was used to develop the framework for NCVLE and its beta systems. Afterward, tests were developed. Trainers of executors of NCVLE analyzed the data obtained from tests using EFA. Then, an experimental version of NCVLE was developed in terms of their suggestions.

Second Stage

The stage aims to develop the learning process model and assess the effectiveness of NCVLE in reinforcing and developing computational. The study investigated the use of the model through modern pedagogical and technological techniques; the pedagogical techniques that influence the environment’s relevance to the aims and expected learning results, the substance of instructional materials, and the teaching methods used; the technological context that involves the availability of technical equipment and adequate software for the environment’s operation (Mukasheva et al., 2023). A group of pre-service teachers with knowledge of information and communication technology was recruited. The experimental group used the developed NCVLE model for one semester (3 months), whereas exercises were assessed regularly after the end of teaching every week to provide appropriate feedback for the development of computational thinking.

Third Stage

The goal of the stage is to improve the acquisition of knowledge and skills by giving users the chance to use neural networks, which are made up of numerous connected neurons that can receive input from a variety of sources and then produce a distinctive output that can be transmitted to other neurons. This is how learning occurs, and input and output functions can be implemented using various learning techniques to create neural networks at the user’s own pace (Sun & Wang, 2019).

Method of Neural Instruction Networks for Learners

To accelerate the speed of learning about this network, learning occurred by providing a group of well-selected examples (training category). Such examples were divided into two types of training categories presented by the network as follows (Liu et al., 2017):

Supervised Learning of Artificial Neural Networks

This method is based on artificial neural networks (ANNs) (Ghanaati et al., 2020). A lab specialist presented training data for the mechanic course through the network in pairs, namely, the input and target forms.

Unsupervised Learning

This type of learning occurred by forming the training category with input and without automatic presentation of the objective through the network. Learning styles are built on the ability of the ANN to discover the distinguished features of shapes and layouts presented and the ability to develop an internal representation of such shapes without previous knowledge or examples (Zhang & You, 2017).

In this type of learning, the course on the environment was stored through the neural network. In other words, exercises are presented in the form of input and output beams. Moreover, this type of learning requires a standard for determining the similarities in beams and instructions (Kamilaris & Prenafeta-Boldú, 2018).

Developing the Network

The development of the network relied on a number of different dependencies. Each one was designed to function with particular neural networks that have distinguishing characteristics. The novel protocol was utilized in the process of developing the face network. The function new requires four inputs as determinates (Baalaaji & Bevi, 2021) as follows:

⁃ A matrix includes the minimum and maximum values for each input beam item, which can be replaced by min-max (p). This matrix determines the smallest and largest values in the input field.

⁃ A matrix includes the number of neurons in each layer of the network.

⁃ A matrix includes the names of activation dependencies for each layer.

⁃ Name of training dependency used.

Materials and Methods

The content of the units was designed by using the artificial neural network to develop computational thinking using a virtual environment for pre-service teachers' mathematics major faculty of education. Hence, the mechanic neuro-computerized course was designed according to the following steps:

Artificial Neural Network Equations

The following equations were used to build an artificial neural network model to predict soil penetration resistance as an example to teach students the application of mechanics to students of the Faculty of Education.

C I = {[3.62 C_{r} e^{- 0.1 θ / (1 + C_{r})}] + [0.0066 γ / (1 + 2 C_{r})]} e^{π / (1 + 2 C_{r})}

C_{r} - \frac{% C l a y}{(% S i l t + % S a n d)}

P D = {0.029 C I + [30.66 γ {V_{a}}^{2} (1 - \cos β) / g]} a w N_{b},

Where:

CI = cone penetration, MPa;

C_r = ratio of clay to silt and sand;

θ = soil moisture content, % w/w; e = exponent.

PD = plow draught, kN;

γ = soil specific weight, kN/m³;

V_a = actual forward speed, km/h;

β = moldboard tail angle, deg.;

g = gravitational constant, m/s²;

a = cut depth, m;

w = furrow cut width, m;

N_b = number of plow bottoms.

ANN Models

Neural networks are extremely appropriate for function fit problems. A neural network with sufficient features (i.e., neurons) can fit any data with arbitrary accuracy. They are largely well-matched for non-linear questions. Given the non-linear nature of real-world events, neural networks are effective runners for resolving problems. The current study is written for the developers of the MATLAB programming language (Abidoye et al., 2020). A set of steps were followed to do so as follows:

Designing ANN Models

Designing ANN models involves several systemic procedures. In general, it requires five basic steps, namely, (1) collecting and (2) preprocessing data, (3) building and (4) training the network, and (5) testing the performance of the model (Figure 2).

Figure 2.

Basic flow of designing the artificial neural network model.

Data Collection

Collecting and preparing sample data is the first step in designing ANN models. Measurement data of virtual density, porosity, and hydraulic connection for 3 years period from 2015 to 2017 (Canziani et al., 2017 May).

Data Preprocessing

After data collection, three data preprocessing procedures are necessary to train the ANN, namely, (1) solving the problem of missing data, (2) normalizing, and (3) randomizing data.

Building the Network

After constructing the data sets, an ANN architecture was created. As previously discussed, Feed-forward neural networks have become one of the popular paradigms in the rapidly expanding field of neural models, a sufficiently complex feedforward network can carry out complex pattern recognition tasks using only very elementary processing units, and there are programming algorithms' that resemble closely a learning process. The ANN consisted of three inputs (i.e., virtual density, porosity, and hydraulic connection; one for each variable of the installation structure), 14 hidden neurons, and one output as a target (i.e., soil penetration) for the desired uniformity (Figure 3). The model is tasked to develop the pre-service teachers CT because of the input data (Baker et al., 2016).

Figure 3.

Architecture of the artificial neural network.

Training the Network

During training, the weights were adjusted to render actual outputs (predicted) close to the target (measured) outputs of the network (Negrete, 2018).

The MATLAB utility “nn train tool” was used to train the ANN (Qadir et al., 2020). This tool opened a data manager window, which enables importing, creating, using, and exporting the experiment data of neural networks. Figure 4 displays the screen captions of the ANN training windows obtained using the “nn train tool” toolbox in MATLAB.

Figure 4.

Architecture of the artificial neural network (ANN).

Testing the Network

To test the performance of the model, the model was exposed to unseen data. To evaluate quantitatively assess the performance of the developed ANN models and verify whether an underlying trend in their performance was observed, statistical analyses of the coefficient of determination (R²), root mean square error and mean bias error were conducted (Espejo-García et al., 2020).

Evaluating the Neural Networks

Figure 5 shows the diagram of the predicted values of the model compared with observed values for training, testing, and validating the data. The solid line represents the best fit of the linear regression line between outputs and targets. The R-value is an indication of the relationship between outputs and targets. If R = 1, then an exact linear relationship between exists outputs and targets, whereas a value close to zero indicates no linear relationship between outputs and targets. For this example, the training data indicate a good fit. Moreover, the validation and test results exhibit R values greater than 0.8. The scatter plot represents data points with poor fit. The overall determination coefficient is 84%, which indicates that the ANN predicts 84% of the actual soil penetration value. As such, the study infers that the model is an accurate predictor of the abovementioned values and recommends it for use based on management inputs of soil-related factors.

Figure 5.

The plot of training, testing, and validating data of accounting performance.

Research Tool

Computational Reasoning Thinking Test

The validity of the test was verified and modified according to student and course characteristics based on the comments of all referees specialized in the field of computer education and information technology.

Ease and Difficulty Factors (Wilson, 2005)

The two researchers calculated the variability factors and difficulty of the test questions through a computational reasoning test. The best difficulty factor ranged from 40% to 75%, whereas that for the test items ranged from 0.40 to 0.73. The levels of discrimination ranged from 0.07 to 0.1, which indicated that the factors of difficulty and discrimination were suitable for the objective of the study.

Reliability

The test-retest method was used to determine the reliability of the test. Table 1 presents the correlation between the test and retest conducted on the experimental samples.

Table 1.

Pearson’s Correlation Coefficient Between the Test and Retest of the Pilot Study.

(CT) test questions	Test		Retest		R	Sig
(CT) test questions	M	SD	M	SD	R	Sig
Introduction to computational thinking	4.56	1.35	4.56	1.22	0.81	0.01^a
Decomposition	3.03	1.54	3.96	1.67	0.83	0.01^a
Pattern recognition	4.36	1.08	4.03	1.11	0.93	0.01^a
Abstraction	4.50	1.90	4.11	1.10	0.95	0.01^a
Algorithm	5.20	1.23	5.29	1.65	0.78	0.01^a
Evaluating solutions	4.93	1.59	4.60	1.45	0.71	0.01^a
Total	25.20	8.49	24.56	8.89	0.81	0.01^a

^aNote. R = 0, 36; where n = 50 (52–2).

Table 1 indicates that the total correlation score (0.81) is high and appropriate for conducting the study using the computational reasoning test.

Group Equation

An equation was made between the experimental and control groups by administering the computational reasoning/thinking test (test items = 54) to the members of the two groups (control − experimental) before experimenting to ensure that the results are consistent between the groups.

Table 2 indicates the t-values for the total scores of the computational reasoning test and compares the total results of the pre-test for computational reasoning/thinking between the experimental and control groups (introduction to CT = 10; decomposition = 7; pattern recognition = 8; abstraction = 9; algorithm = 10; evaluating solutions = 10)

Table 2.

The Equation Between Experimental and Control Groups in the Computational Reasoning Test.

Axis (CT) test questions	Experimental (n = 52)		Control (n = 50)		DF	t	Sig
Axis (CT) test questions	M	SD	M	SD	DF	t	Sig
Introduction to computational thinking	4.80	1.67	4.06	1.53	51	1.64	0.11
Decomposition	3.65	0.66	3.10	1.54	51	1.71	0.08
Pattern recognition	4.65	1.66	4.14	1.78	51	1.40	0.06
Abstraction	4.56	1.40	4.95	1.33	51	1.20	0.30
Algorithm	5.12	1.03	5.06	1.33	51	1.80	0.40
Evaluating solutions	4.17	1.97	4.91	1.08	51	1.08	0.51
Total (pre-test)	23.01	8.11	24.31	9.04	51	1.83	0.69

The study found that the total mean scores of the control and experimental groups in CT were 23.01 and 24.31, which confirms the absence of statistically significant differences and the usability of the equation for CT before experimenting.

Data Collection and Analysis

The two researchers verified the CT test, which aims to determine the students’ level of CT in a virtual environment based on a neuro computerized. The test consists of 54 items (i.e., introduction to CT = 10; decomposition = 7; pattern recognition = 8; abstraction = 9; algorithm = 10; evaluating solutions = 10)

The test is in a multiple-choice format. The NCVLE model was first administered to the experimental group. To determine whether statistically significant differences between the experimental and control groups in terms of CT, the two researchers conducted tests on CT before and after teaching.

Results

The CT test includes six segments to achieve the research objectives and is applied through the NCVLE. Results were analyzed using SPSS version 21 to determine whether statistically significant differences exist between pre- and post-administration between the groups in terms of CT/reasoning.

The following sections present the results.

Averages, standard deviations, and standard error in the axes of testing CT

Table 3 indicates shows that the mean scores of the pre-and post-tests for introduction to CT reached 2.98 and 8.10, respectively, with a standard deviation of 1.053, which is normal and reflects convergence in responses. In other words, the NCVLE platform enables students to develop problem-solving skills. The mean scores for the pre-and post-tests for decomposition were 2.69 and 5.65, respectively, with a standard deviation of 0.968, which reflects a significant result in terms of the desire of students to develop their skills. Moreover, the mean scores of the pre-and post-tests for pattern recognition reached 2.98 and 6.37, respectively, with a standard deviation of 0.908. This finding indicates that the NCVLE platform provides students with opportunities to review problems related to mechanics. The mean scores of the pre-and post-tests for abstraction reached 3.15 and 7.50, respectively, with a standard deviation of 0.897. Moreover, the platform reflects a consensus in the mean scores for the pre-and post-tests of algorithms at 3.44 and 7.94, respectively, with a standard deviation of 1.074, which reflects the participants’ positive attitude toward NCVLE. Lastly, the mean scores of the pre-and post-tests for evaluating solutions were 3.48 and 7.98, respectively, with a standard deviation value of 1.075, which is attributed to the LMS platforms that support learning management systems in universities. All findings indicate improvement in CT.

Table 3.

Pre-and Post-Test Level of Computational Thinking in the Experimental Group.

Paired-sample statistics
Segment		Test	M	SD	Std. error Mean	Correlation	Sig
1	Introduction to computational thinking	Post	8.10	1.053	.146	.141	.317
1	Introduction to computational thinking	Pre	2.98	.671	.093	.141	.317
2	Decomposition	Post	5.65	.968	.134	−.044-	.754
2	Decomposition	Pre	2.69	.701	.097	−.044-	.754
3	Pattern recognition	Post	6.37	.908	.126	.154	.277
3	Pattern recognition	Pre	2.98	.754	.105	.154	.277
4	Abstraction	Post	7.50	.897	.124	−.113-	.427
4	Abstraction	Pre	3.15	.777	.108	−.113-	.427
5	Algorithm	Post	7.94	1.074	.149	.079	.579
5	Algorithm	Pre	3.44	.539	.075	.079	.579
6	Evaluating solutions	Post	7.98	1.075	.149	.173	.220
6	Evaluating solutions	Pre	3.48	.57702	.0800	.173	.220

^*Note. N = 52.

Differences Between the Pre-and Post-Tests of the Axes of Computing Thinking for the Experimental Group

All finding exhibit α < 0.05. Therefore, statistically, significant differences exist between the pre-and post-tests in favor of the post-administration of the test segments. Introduction to Computational Thinking had the greatest mean score 5.115 and standard deviation 1.166, while Algorithms and Evaluating Solutions had the lowest average degree 4.500 and standard deviation 1.163 and 1.128 respectively Table 4.

Table 4.

Differences Between the Pre-and Post-Tests (Experimental Group).

Paired-sample test
Test segments	Paired Differences					t	df
	M	SD	Std.error Mean	95% confidence interval
	M	SD	Std.error Mean	Lower	Upper
Introduction to computational thinking	5.115	1.166	.162	4.791	5.440	31.643	51
Decomposition	2.962	1.220	.169	2.622	3.301	17.503	51
Pattern recognition	3.385	1.087	.151	3.082	3.687	22.444	51
Abstraction	4.346	1.251	.173	3.998	4.694	25.060	51
Algorithms	4.500	1.163	.161	4.176	4.824	27.898	51
Evaluating solutions	4.500	1.128	.1565	4.185	4.814	28.744	51

^*Note. α < 0.05 for all segments.

Comparison of Experimental and Control Groups (Post-Test for Computational Reasoning)

Table 5 indicates that the mean difference between the two groups is in favor of the post-administration with higher standard deviations in most experimental test segments compared with the control group.

Table 5.

Mean scores, standard deviations, and standard error of the post-test (experimental and control groups).

Group statistics
Test segment (post-test)	Group	M	SD	Std. Error mean
Introduction to computational thinking	Experimental	8.10	1.360	.189
Introduction to computational thinking	Control	4.62	1.053	.146
Decomposition	Experimental	5.6538	.96781	.13421
Decomposition	Control	3.0962	.86907	.12052
Pattern recognition	Experimental	6.3654	.90811	.12593
Pattern recognition	Control	3.4423	.66902	.09278
Abstraction	Experimental	7.5000	.89662	.12434
Abstraction	Control	3.6346	.59504	.08252
Algorithm	Experimental	7.9423	1.07400	.14894
Algorithm	Control	4.2885	.72319	.10029
Evaluating solutions	Experimental	7.9423	.99830	.13844
Evaluating solutions	Control	4.2500	.73764	.10229

^*N = 50.

The study infers the following results based on the results in Table 6.

⁃ Introduction to CT = 1.328. This finding indicates homogeneity in variation holds (p > .05) with a t-value of 14.596 and a significance level of .000 during analysis. That is, statistically, significant differences exist at a level of 0.001 between the mean scores of the two groups in favor of the experimental group.

⁃ Pattern recognition = 2.957. This result also indicates homogeneity in variance (p = .198) with a t-value of 18.688 and a significance level of .000 during analysis. In other words, statistically, significant differences exist at a level of 0.001) between the mean scores in favor of the experimental group.

⁃ Abstraction = 2.889. This value indicates homogeneity in variance (p = .215) with a t-value of 25.902 and a significance level of .000 during analysis. This finding indicates statistically significant differences at a level of 0.001 between the mean scores in favor of the experimental group.

⁃ Algorithm = 2.087. However, this result points to homogeneity in variance (p = .252) with a t-value of 20.350 and a significance level of .000 during analysis. This finding represents statistically significant differences at a level of 0.001 between the mean scores in favor of the experimental group.

⁃ Evaluating solutions = 1.324. This finding indicates homogeneity in variance(p = .253) with a t-value of 21.450 and a significance level of .000 during analysis. Thus, statistically, significant differences are observed at a level of 0.001 between the mean scores in favor of the experimental group.

Table 6.

t-Values, mean differences, and standard deviations of differences between the two groups for the computational thinking/reasoning test.

Independent-sample test
Equal variances assumed (post-test)	Levene’s test for equality of variances		t-test for equality of means
	F	Sig	t	df	Sig. value (2-tailed)	Mean difference	Std. error (difference)	95% confidence interval (difference)
	F	Sig	t	df	Sig. value (2-tailed)	Mean difference	Std. error (difference)	Lower	Upper
Introduction to computational thinking	1.328	.168	14.596	102	.000	3.481	.238	3.008	3.954
Decomposition	2.799	.197	14.179	102	.000	2.5576	.1803	2.1999	2.91548
Pattern recognition	2.957	.198	18.688	102	.000	2.9230	.1564	2.6128	3.23333
Abstraction	2.889	.215	25.902	102	.000	3.8653	.1492	3.5693	4.16138
Algorithms	2.087	.252	20.350	102	.000	3.6538	.1795	3.2977	4.01999
Evaluating solutions	1.324	.253	21.450	102	.000	3.6923	.1721	3.3508	4.03373

The Support of NCVLE Provides to Develop CT

Keeping track of what happened during the experiment’s practical application allows us to draw conclusions about the support that NCVLE can give in developing computational thinking along the following lines:

1. Pedagogical and Technological Tactics: Learners are asked to decompose the target problem into multiple smaller parts of the algorithm programming tasks if the problem is presented in a model learning task within soil mechanics exercises. This is done if the problem is presented in a model learning task. Learners are tasked with identifying the fundamental aspects of an issue and then speculating on how the performance of individual problem-solving strategies relates to the overall functionality of computing algorithms.

2. Assessment of Action: refers to the application of the original algorithms. It seeks to investigate unanticipated errors during the development of soil mechanics algorithms. On the basis of test findings, corrections are made to the algorithm programming applications until data processing is successful. The focus of assessments is on students' work, error correction, and efforts to alter their programming structures in order to progress.

3. Feedback Generator: Generation of feedback has been regarded as a crucial intervention approach for strengthening computational thinking skills. Using a data-driven automated grader capable of providing immediate and precise feedback for algorithmic programming tasks, this method addresses these obstacles. The feedback generator facilitates the editing and development of computer algorithms for students.

4. Development of knowledge and skills: After the problem has been defined, the learners are given the task of creating symbolic representations of the soil mechanics algorithm symbols in order to simulate the neural computing algorithms. This helps the learners develop their knowledge and skills. A model that divides the development of skills in computational thinking into three stages is proposed here: (1) determining the problem, (2) finding a solution to the problem, and (3) analyzing the results of the solution.

Discussion and Implications

The study aimed to verify the development of CT in a virtual environment designed using the neural computing method. The use of VR was limited to teaching soil mechanics. Moreover, the study employed the CT test (Román-González et al., 2017).

The results of using the NCVLE in terms of test segments (Palts & Pedaste, 2020), which represented six topics are like those of other studies (Trenado et al., 2021). Thus, the conclusions can be evaluated regarding the test topics. One of the main reasons underlying the results is the environmentally compatible learning content of the soil mechanics course, learning experience, and teaching material.

Students are restricted to materials published in the neuro-computerized environment produced by VR designers (neuro-typed et al., 2019). Thus, researchers aim to design a model that meets the individual needs and learning outcomes of students. However, the skillset necessary to produce fully neuro-typed VLEs remains very demanding (Chang et al., 2020). In this regard, experience in teaching mechanic courses (or any course) that are fully based on appropriate NCVLE is regarded as technical competence. It is one of the potential solutions for developing CT/reasoning in contrast to other computer-generated environments.

Creating content for VLEs is relatively easy and customizable to suit the needs of students (Boulton, et al., 2018). However, diverse fields of learning will continue to experience constraints in providing materials related to NCVLE. In this manner, the learning environment should be designed to be convenient and easy to learn.

Implications for Outcome Measures and Assessment Instrumentation

Relying on CT tests to assess the results of learning soil mechanics was one of the most common assessment tools used in many studies on CT. Despite the existence of extensive discussions on the advantages and disadvantages of using the CT test, many scholars consider it the best tool for verifying progress in CT. This notion is in line with those of other studies that report the ability of assessment and evaluation tools to encourage comprehensive learning of all thinking skills. The CT test is typically administered after an experiment. The researchers noted that none of the studies went beyond learning CT only. Most of them recommended further studies to investigate the role of virtual environments in learning. However, the nature of the assessment tool is suitable for learning outcomes and relies on the use of brief answers, which leads to an appropriate measure that is more appropriate for examining the depth of learning achieved (Palts & Pedaste, 2020). In this manner, students gain more opportunities to demonstrate the acquisition of computing skills for any subject. Moreover, the research on NCVLE could benefit by expanding the definition of the elements that constitute the outcome of neuro-modulatory learning. This aspect can be achieved by comparing pre-and post-test scores to experimental groups or by comparing post-test scores between experimental and control groups to demonstrate how using the NCVLE environment can induce deeper understanding through experiential learning.

Implication of Intervention Characteristics for Learning Outcomes

The study examined the effectiveness of using the NCVLE model in experimental and applied settings and the implications of assessing its suitability for teaching (i.e., soil mechanics in the current study). Most of the results of studies that employ the NCVLE model indicate that the novelty of the model may hinder the learning process to a certain extent at the beginning of the experiment. This case is especially true for students without prior experience with these technologies. This notion agrees with that of Saritepeci (2020), who cited those results after exposure to an experience indicate a substantial positive development in CT levels.

In the same manner, Ceallaigh (2022) found that the first session of learning VR led to low performance, which is like studies on traditional methods. However, participants are adaptable and learn about the environment and master the functions of neuro-computerized environment programs. By navigating sessions in succession during a course that employs the NCVLE technology, studies can demonstrate its advantage over traditional methods.

Nevertheless, addressing the potential negative impact of NCVLE as an educational tool at the beginning of an experiment is extremely important, especially when results are directly compared with another method. Luo et al., (2022) highlighted the comparison between experimental and control groups. Notably, the authors stressed that the novelty of NCVLE may impede learning outcomes and its application in classrooms. Therefore, students' familiarity with VR technology should be considered, especially when comparing results between groups. In other words, participants should be trained in several sessions as a means of mitigating issues in developing their computer skills. In addition to the intervention, students' exposure to NCVLE tends to be short due to the relatively limited number of studies that rely on web-based conventional books (Xue et al., 2022). Argued that using NCVLE as a form of blended or multimodal learning is a better option (Alamri et al., 2020).

Learning Outcomes in NCVLE

The current study examined learning outcomes across areas of procedural knowledge. To a large extent, the most common field was the teaching of procedural skills related to neuro-computing for soil mechanics. The results of the experimental group exceeded the scores of the control group in the post-test for CT for the six segments, namely, introduction to CT (t-value = 14.596; average difference = 3.481), decomposition (t-value = 14.179; average difference = 2.55,769), pattern recognition (t-value = 18.688; average difference = 2.92,308, abstraction (t-value = 25.902; average difference = 3.86,538), algorithms (t-value = 20.350; average difference = 3.65,385), and evaluating solutions (t-value = 21.450; average difference = 3.69,231). These findings are in line with those of other studies (Ceallaigh, 2022).

Moreover, the results imply that NCVLE technology enables users to gain new insights and ideas, which may be difficult to gain using traditional methods. The study identified scientific topics, in relation to the application of NCVLE in education. However, other scientific disciplines that require an abstract or conceptual understanding (i.e., mechanics) can benefit from the perception provided by NCVLE. The results of the current study demonstrated the advantage of using NCVLE in education and pointed to positive results in developing computer skills. The neuro-learning environment was used to successfully simulate the field of mechanics given the nature of mathematics courses (i.e., soil weight, water content, soil classification, soil compressibility, soil mixture, groundwater, soil penetration resistance, soil particles, soil saturation, soil porosity, and soil mechanics).

Future research should examine the application of neuro-computerized environments in an expanded educational context and knowledge transfer to real-world scenarios. For example, users can interact with virtual images and symbols as NCVLE experiments aim to understand how virtual learning can be applied to real-life situations (Ryan & Poole, 2019).

Implications for Future Practice

The current study developed a set of computer skills among students using NCVLE. Previous studies observed that NCVLE can be used to develop computer skills and cognitive and procedural knowledge (Yen et al., 2018), by expanding learning outcomes to include potential benefits, such as increased understanding of tools in a computerized environment and ability to identify complex topics. Technical practice can benefit from the relationship between learners and the NCVLE. Thus, the model should be comprehensively analyzed to examine learning outcomes that exceed test scores for computational reasoning. Furthermore, the study identified future areas of improvement, which address the variables and expand the scope of the research. As Peteranetz et al., (2020, February) suggested, the novelty of NCVLE can impede learning outcomes due to the lack of familiarity with technology. Therefore, considering the limitation of the training period is necessary for free movement between the elements of the NCVLE, which can relieve anxiety about the novelty of the technology (Vertesi et al., 2020).

In addition, a qualitative follow-up analysis is beneficial for the first-hand experience of using NCVLE and for highlighting concerns regarding the lack of familiarity with the NCVLE technique. The greatest concern in this regard is the environmental use of tools for NCVLE. However, more subtle forms of learning beyond mere recall of information require elucidation. Multiple-choice tests can be used to assess students' in-depth understanding and applied knowledge of CT. Future research should consider the nature of these interventions in a sound theoretical framework (Adams et al., 2021). In this manner, sound learning objectives and specific assessment methods can be set. The study demonstrated that NCVLE is an effective tool for developing CT and for problem-solving qualifications (Lockwood & Mooney, 2018). Therefore, future studies research should focus on the potential of NCVLE as a tool for acquiring CT. A strong body of evidence indicates that NCVLE experiments can elicit high levels of response, which should be explored further in the educational context (Pappa et al., 2018). Moreover, whether this ability can adopt a perspective that can lead to higher areas of learning, such as assessing problem-solving skills for computational reasoning or creating new solutions as a direct result of insights perceived from NCVLE should be investigated to enable researchers and mathematics pre-service teachers to critically verify VLE tools for solving issues related to mechanics.

Limitations

The experiment of the present study faced many challenges during its implementation in the context of the Mechanics course for developing computational thinking, as students immersed themselves in the experiment using a virtual learning environment based on neuro computerized. It took them time to integrate and accept the new environment. Students training in the new environment needs further clarification, as the current study was an attempt to test the concept of the neuro-computerized environment. Future research must have a similar distinction to avoid the difficulties that students faced with the experiment in the present study. However, future research should also investigate the effect of designing a neuroscientific virtual environment on the learner and the motivational results of acquiring computing skills. Students’ level of interest does not necessarily diminish during the experiment because the neuro-computerized environment requires students to take responsibility for their learning and observe their academic and technological content by asking questions, conducting evaluation standards, and following feedback processes during the learning process. To achieve this, the teacher can use some strategies such as encouraging students to be involved in the new environment and integrating technological knowledge and academic content to weave and support learning.

Among the limitations, it was noted that current results can be generalized in specific fields such as: (chemistry - physics - biology) for pre-service teachers. As the sample was small and restricted to scientific disciplines, future research should include a larger number of participants for a wide range of multiple contexts.

Studying the effects of a neuro-computerized approach among higher education students may be of particular interest to scientific disciplines. Generalization is also hampered by the relatively short duration of the experiment and its limitations to one semester. Future research should span a longer period across diverse sets of courses for multiple disciplines. Moreover, it should be noted that the current data reflected the participants' perceptions of their level of participation in the experiment and the teacher’s perceptions of their skills performance. Ideally, future research should aim to measure both aspects in an objective way to reveal the benefits of this learning approach in a neuro-computerized environment.

Results contribute to revealing the importance of digital transformation of learning environments for higher education majors under emergency conditions and the irregularity of students attending conventional classrooms and giving them the ability to rely on themselves completely. Although we all would like to measure data as realistically as possible, the conclusions of the analysis are limited because they do not adequately reflect everything that is happening in the education setting. Therefore, more research is needed in non-linear modeling and non-linear testing of neuro-computerized.

Finally, neuro-computed virtual reality has potentialities as new educational technology to increase objective learning outcomes and face community emergencies such as a Covid 19 pandemic when this environment is used as a principle for learning design. However, further research is still needed to determine how it can be implemented in a variety of disciplines.

Conclusions

The study contributes to the literature by providing evidence that a neuro-computerized VRE can motivate mathematics pre-service teachers (i.e., added value) to build an NCVLE specific to the course. Another added value is increased motivation to identify mechanics problems. Moreover, the model promotes the acquisition of CT in an interactive and pleasurable manner, especially for students who lack CT. Moreover, the study finds that the students' computational thinking skills enable them to solve complex problems more easily, either by dividing the problem into manageable and easier-to-understand parts, or by comparing the solution to previous problems and searching for alternative or common solutions, allowing them to focus on the important information only, and pulling out specific differences to make one solution work for multiple problems. - abstraction; creating a step-by-step solution to a problem - algorithms. This plan is applicable to everyone, regardless of expertise, task, or age.

The participants assumed that they learned nothing at the initial stage of the applications. However, the opposite is true as evidenced by the post-test results. Hence, the outcome of using NCVLE in higher education is dependent on conditions provided by schools, which indicates numerous challenges. Thus, further research is required before introducing the model on a large scale. Importantly, research that includes more participants should be conducted to ensure whether using NCVLE as a support tool in university education is truly possible and to identify the possibilities and challenges that accompany the use of this technology.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors extend their appreciation to the Deputyship for Research and Innovation, Ministry of Education in Saudi Arabia for funding this research work through the project no. (IFKSURG-2-1678).

Data Availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

ORCID iDs

Yousri Attia Mohamed Abouelenein

Mohamed Ali Nagy Elmaadaway

Author Biographies

Yousri Attia Mohamed Abouelenein is an Associate Professor of Educational Technology - Damietta University - Egypt. An academic program development expert at Damietta University and manager of Electronic Services. His research interests focus on designing digital learning environments and digital teaching. and the integration of ICT in education - and learning analytics.

Mohamed Ali Nagy Elmaadaway is an associate professor with a focus on the intersection of education and technology; research interests include e-learning, computer-based instruction, distance education, educational software, LMSs, AI, and expert systems. I am a consultant for numerous e-learning development initiatives in a variety of nations and have a long track record of international research and publishing in the fields of e-learning and computer science in education.

References

Abidoye

L. K.

Bani-Hani

El Haj Assad

AlShabi

Soudan

Oriaje

A. T.

(2020). Effects of environmental and turbine parameters on energy gains from wind farm system: Artificial neural network simulations. Wind Engineering, 44(2), 181–195. https://doi.org/10.1177/0309524X19849834

Adams

Feng

Liu

J. C.

Stauffer

(2021). Potentials of teaching, learning, and design with virtual reality: An interdisciplinary thematic analysis. In: Intersections across disciplines (pp. 173–186). Springer. https://doi.org/10.1007/978-3-030-53875-0_14

Alamri

Lowell

Watson

S. L.

(2020). Using personalized learning as an instructional approach to motivate learners in online higher education: Learner self-determination and intrinsic motivation. Journal of Research on Technology in Education, 52(3), 322–352. https://doi.org/10.1080/15391523.2020.1728449

Albus

Vogt

Seufert

(2021). Signaling in virtual reality influences learning outcome and cognitive load. Computers and Education, 166, 104154. https://doi.org/10.1016/j.compedu.2021.104154

Araiza-Alba

Keane

Chen

W. S.

Kaufman

(2021). Immersive virtual reality as a tool to learn problem-solving skills. Computers and Education. p. 104121. https://doi.org/10.1016/j.compedu.2020.104121

Attias

Vigneri

Dimitrov

(2022). Rethinking modular multi-exponentiation in real-world applications. Journal of Cryptographic Engineering, 1–14. https://doi.org/10.1007/s13389-022-00287-w

Baalaaji

Bevi

(2021). Design of a novel chaotic neural network based encryption system for security applications. Journal of the Chinese Institute of Engineers, 44(5), 424–439. https://doi.org/10.1080/02533839.2021.1919558

Baker

Gupta

Naik

Raskar

(2016). Designing neural network architectures using reinforcement learning. https://arXiv_preprint_arXiv:1611.02167

Boos

Chu

Cuervo

(2016, June). Flashback: Immersive virtual reality on mobile devices via rendering memoization. In: Proceedings of the 14th annual international conference on mobile systems, applications, and services (p. 291–304). https://doi.10.1145/2906388.2906418

10.

Boulton

C. A.

Kent

Williams

H. T.

(2018). Virtual learning environment engagement and learning outcomes at a ‘bricks-and-mortar’university. Computers and Education, 126, 129–142. https://doi.org/10.1016/j.compedu.2018.06.031

11.

Bustamante

J. C.

Navarro

J. J.

(2022). Technological tools for the intervention and computerized dynamic assessment of executive functions handbook of research on neurocognitive development of executive functions and implications for intervention. IGI Global, p. 310–339. https://doi.10.4018/978-1-7998-9075-1.ch014

12.

Calvert

Abadia

(2020). Impact of immersing university and high school students in educational linear narratives using virtual reality technology. Computers and Education, 159, 104005. https://doi.org/10.1016/j.compedu.2020.104005

13.

Canziani

Culurciello

Paszke

(2017, May). Evaluation of neural network architectures for embedded systems. In: 2017 IEEE international symposium on Circuits and systems (ISCAS) (pp. 1–4). IEEE. https://doi.org/10.1109/ISCAS.2017.8050276

14.

Ceallaigh

T. Ó.

(2022). Designing, navigating and nurturing virtual learning spaces: Teacher educators’ professional development priorities and potential pathways. Teaching and Teacher Education, 115, 103697. https://doi.org/10.1016/j.tate.2022.103697

15.

Chang

S. C.

Hsu

T. C.

Jong

M. S. Y.

(2020). Integration of the peer assessment approach with a virtual reality design system for learning earth science. Computers and Education, 146, 103758. https://doi.org/10.1016/j.compedu.2019.103758

16.

Chao

P. Y.

(2016). Exploring students’ computational practice, design and performance of problem-solving through a visual programming environment. Computers and Education, 95, 202–215. https://doi.org/10.1016/j.compedu.2016.01.010

17.

Cheng

Kolbe

Grant

Eller

Hales

Symon

Griswold

Eppich

(2020). A practical guide to virtual debriefings: Communities of inquiry perspective. Advances in Simulation, 5(1), 18–19. https://doi.org/10.1186/s41077-020-00141-1

18.

Cheng

K. H.

Tsai

C. C.

(2019). A case study of immersive virtual field trips in an elementary classroom: Students’ learning experience and teacher-student interaction behaviors. Computers and Education, 140, 103600. https://doi.org/10.1016/j.compedu.2019.103600

19.

Christofi

Kyrlitsias

Michael-Grigoriou

Anastasiadou

Michaelidou

Papamichael

Pieri

(2018). A tour in the archaeological site of choirokoitia using virtual reality: A learning performance and interest generation assessment. In: Advances in digital cultural heritage (pp. 208–217). Springer. https://doi.org/10.1007/978-3-319-75789-6_15

20.

Di Natale

A. F.

Repetto

Riva

Villani

(2020). Immersive virtual reality in K-12 and higher education: A 10-year systematic review of empirical research. British Journal of Educational Technology, 51(6), 2006–2033. https://doi.org/10.1111/bjet.13030

21.

Durak

H. Y.

Saritepeci

(2018). Analysis of the relation between computational thinking skills and various variables with the structural equation model. Computers and Education, 116, 191–202. https://doi.org/10.1016/j.compedu.2017.09.004

22.

Espejo-Garcia

Mylonas

Athanasakos

Fountas

(2020). Improving weeds identification with a repository of agricultural pre-trained deep neural networks.Computers and Electronics in Agriculture, 175, 105593. https://doi.org/10.1016/j.compag.2020.105593

23.

Feyzi Behnagh

Yasrebi

(2020). An examination of constructivist educational technologies: Key affor-dances and conditions. British Journal of Educational Technology, 51(6), 1907–1919. https://doi.org/10.1111/bjet.13036

24.

Fussell

S. G.

Truong

(2022). Using virtual reality for dynamic learning: An extended technology acceptance model. Virtual Reality, 26(1), 249–267. https://doi.org/10.1007/s10055-021-00554-x

25.

Ghanaati

Sjöblom

Faghani

(2020). Design of a virtual test cell based on GMDH-type neural network for a heavy-duty diesel engine. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 235(2–3), 436–445. https://doi.org/10.1177/0954407020960614

26.

Gomez

L. I.

(2020). Immersive virtual reality for learning experiences In: Radical solutions and eLearning (pp. 183–198). Springer. https://doi.org/10.1007/978-981-15-4952-6_12

27.

Guerrero Alonso

P. D.

Merino Ceballos

Moreno Llorca

Ros Candeira

Zamora

(2022). Data model, E-infrastructure services, and the virtual research environment (VRE) In: The landscape of the sierra Nevada (pp. 369–384). Springer. https://doi.org/10.1007/978-3-030-94219-9_22

28.

Hamilton

McKechnie

Edgerton

Wilson

(2021). Immersive virtual reality as a pedagogical tool in education: A systematic literature review of quantitative learning outcomes and experimental design. Journal of Computers in Education, 8(1), 1–32. https://doi.org/10.1007/s40692-020-00169-2

29.

Hamutoglu

N. B.

Gemikonakli

Duman

Kirksekiz

Kiyici

(2020). Evaluating students experiences using a virtual learning environment: Satisfaction and preferences. Educational Technology Research and Development, 68(1), 437–462. https://doi.org/10.1007/s11423-019-09705-z

30.

Inoue

(2012). Virtual reality learning environments. In Seel

N. M.

(Ed), Encyclopedia of the sciences of learning. Springer. https://doi.org/10.1007/978-1-4419-1428-6_651

31.

Ito

Hearne

Mill

Cocuzza

Cole

M. W.

(2020). Discovering the computational relevance of brain network organization. Trends in Cognitive Sciences, 24(1), 25–38. https://doi.org/10.1016/j.tics.2019.10.005

32.

Jocius

O’Byrne

W. I.

Albert

Joshi

Blanton

Robinson

Andrews

Barnes

Catete

(2022). Building a virtual community of practice: Teacher learning for computational thinking infusion. TechTrends: For Leaders in Education and Training, 66(3), 547–559. https://doi.org/10.1007/s11528-022-00729-6

33.

Kamilaris

Prenafeta-Boldú

F. X.

(2018). A review of the use of convolutional neural networks in agriculture. The Journal of Agricultural Science, 156(3), 312–322. https://doi.org/10.1017/S0021859618000436

34.

Khlaisang

Songkram

(2019). Designing a virtual learning environment system for teaching twenty-first century skills to higher education students in ASEAN. Technology, Knowledge and Learning, 24(1), 41–63. https://doi.org/10.1007/s10758-017-9310-7

35.

Kuznetcova

Glassman

Lin

T. J.

(2019). Multi-user virtual environments as a pathway to distributed social networks in the classroom. Computers and Education, 130, 26–39. https://doi.org/10.1016/j.compedu.2018.11.004

36.

Leite

W. L.

Xing

Fish

(2022). Teacher strategies to use virtual learning environments to facilitate algebra learning during school closures. Journal of Research on Technology in Education, 1–15. https://doi.org/10.1080/15391523.2022.2110335

37.

Liang

Hong

Wang

Zeng

(2019). Neural network modeling and single-neuron proportional–integral–derivative control for hysteresis in piezoelectric actuators. Measurement and Control, 52(9–10), 1362–1370. https://doi.org/10.1177/0020294019866846

38.

Lin

K. Y.

K. C.

Hsiao

H. S.

Chang

Y. S.

Chien

Y. H.

(2020). Effects of web-based versus classroom-based STEM learning environments on the development of collaborative problem-solving skills in junior high school students. International Journal of Technology and Design Education, 30(1), 21–34. https://doi.org/10.1007/s10798-018-9488-6

39.

Liu

T. C.

Lin

Y. C.

Wang

T. N.

Yeh

S. C.

Kalyuga

(2021). Studying the effect of redundancy in a virtual reality classroom. Educational Technology Research and Development, 69(2), 1183–1200. https://doi.org/10.1007/s11423-021-09991-6

40.

Liu

Wang

Liu

Zeng

Liu

Alsaadi

F. E.

(2017). A survey of deep neural network architectures and their applications. Neurocomputing, 234, 11–26. https://doi.org/10.1016/j.neucom.2016.12.038

41.

Lockwood

Mooney

(2018). Computational thinking in secondary education: Where does it fit? A systematic literary review. International Journal of Computer Science Education in Schools, 2(1), 41–60. https://doi.org/10.21585/ijcses.v2i1.26

42.

Luo

Han

Zhang

(2022). Prediction of learning outcomes with a machine learning algorithm based on online learning behavior data in blended courses. Asia Pacific Education Review, p. 1–19. https://doi.org/10.1007/s12564-022-09749-6

43.

Máñez

Vidal-Abarca

Kendeou

Martínez

(2019). How do students process complex formative feedback in question-answering tasks? A think-aloud study. Metacognition and Learning, 14(1), 65–87. https://doi.org/10.1007/s11409-019-09192-w

44.

Martin

Dennen

V. P.

Bonk

C. J.

(2020). A synthesis of systematic review research on emerging learning environments and technologies. Educational Technology Research and Development: ETR and D, 68(4), 1613–1633. https://doi.org/10.1007/s11423-020-09812-2

45.

Miller

G. A.

(1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63(2), 81–97. https://doi.org/10.1037/h0043158

46.

Moon

Lee

Choi

G. W.

(2020). A conceptual framework for teaching computational thinking in personalized OERs. Smart Learning Environments, 7(1), 6–19. https://doi.org/10.1186/s40561-019-0108-z

47.

Mukasheva

Kornilov

Beisembayev

Soroko

Sarsimbayeva

Omirzakova

(2023). Contextual structure as an approach to the study of virtual reality learning environment. Cogent Education, 10, 1. https://10.1080/2331186X.2023.2165788

48.

Muñoz-Cristóbal

J. A.

Gallego-Lema

Arribas-Cubero

H. F.

Martínez-Monés

Asensio-Pérez

J. I.

(2017). Using virtual learning environments in bricolage mode for orchestrating learning situations across physical and virtual spaces. Computers and Education, 109, 233–252. https://doi.org/10.1016/j.compedu.2017.03.004

49.

Negrete

J. C.

(2018). Artificial neural networks in mexican agriculture, A overview. International Journal of Research in Agriculture and Forestry, 5(7), 1–4.

50.

D. T. K.

Chu

S. K. W.

(2021). Motivating students to learn STEM via engaging flight simulation activities. Journal of Science Education and Technology, 30(5), 608–629. https://doi.org/10.1007/s10956-021-09907-2

51.

Nikiforos

Tzanavaris

Kermanidis

K. L.

(2020). Virtual learning communities (VLCs) rethinking: Influence on behavior modification—bullying detection through machine learning and natural language processing. Journal of Computers in Education, 7(4), 531–551. https://doi.org/10.1007/s40692-020-00166-5

52.

Nordby

S. K.

Bjerke

A. H.

Mifsud

(2022). Computational thinking in the primary mathematics classroom: A systematic review. Digital Experiences in Mathematics Education, 8, 27–49. https://doi.org/10.1007/s40751-022-00102-5

53.

Palts

Pedaste

(2020). A model for developing computational thinking skills. Informatics in Education, 19(1), 113–128. https://doi.org/10.15388/infedu.2020.06

54.

Panjaburee

Komalawardhana

Ingkavara

(2022). Acceptance of personalized e-learning systems: A case study of concept-effect relationship approach on science, technology, and mathematics courses. Journal of Computers in Education, 9(4), 681–705. https://doi.org/10.1007/s40692-021-00216-6

55.

Pappa

Ioannou

Christofi

Lanitis

(2018). Preparing student mobility through a VR application for cultural education In: Advances in digital cultural heritage (pp. 218–227). Springer. https://doi.org/10.1007/978-3-319-75789-6_16

56.

Peteranetz

M. S.

Morrow

P. M.

Soh

L. K.

(2020, February). Development and validation of the computational thinking concepts and skills test In: Proceedings of the 51st ACM technical Symposium on computer science education (p. 926–932). https://doi.org/10.1145/3328778.3366813

57.

Pinargote-Ortega

Bowen-Mendoza

Meza

Ventura

(2021). Peer assessment using soft computing techniques. Journal of Computing in Higher Education, 33(3), 684–726. https://doi.org/10.1007/s12528-021-09296-w

58.

Qadir

Gazder

Choudhary

K. U. N.

(2020). Artificial neural network models for performance design of asphalt pavements reinforced with geosynthetics. Transportation Research Record: Journal of the Transportation Research Board, 2674(8), 319–326.

59.

Radianti

Majchrzak

T. A.

Fromm

Wohlgenannt

(2020). A systematic review of immersive virtual reality applications for higher education: Design elements, lessons learned, and research agenda. Computers and Education, 147, 103778. https://doi.org/10.1016/j.compedu.2019.103778

60.

Radović

Hummel

H. G. K.

Vermeulen

(2022). Design-based research with mARC ID model: Designing experiential learning environments. Learning Environments Research, 25(3), 803–822. https://doi.org/10.1007/s10984-021-09394-7

61.

Raj

N. S.

(2022). Early prediction of student engagement in virtual learning environments using machine learning techniques. E-Learning and Digital Media. p. 20427530221108027. https://doi.org/10.1177/20427530221108027

62.

Román-González

Pérez-González

J. C.

Jiménez-Fernández

(2017). Which cognitive abilities underlie computational thinking? Criterion validity of the computational thinking test. Computers in Human Behavior, 72, 678–691. https://doi.org/10.1016/j.chb.2016.08.047

63.

Ryan

Poole

(2019). Impact of virtual learning environment on students’ satisfaction, engagement, recall, and retention. Journal of Medical Imaging and Radiation Sciences, 50(3), 408–415. https://doi.org/10.1016/j.jmir.2019.04.005

64.

Sanfilippo

Blazauskas

Salvietti

Ramos

Vert

Radianti

Majchrzak

T. A.

Oliveira

(2022). A perspective review on integrating VR/AR with haptics into STEM education for multi-sensory learning. Robotics, 11(2), 41. https://doi.org/10.3390/robotics11020041

65.

Saritepeci

(2020). Developing computational thinking skills of high school students: Design-based learning activities and programming tasks. The Asia-Pacific Education Researcher, 29(1), 35–54. https://doi.org/10.1007/s40299-019-00480-2

66.

Shin

Bowers

Roderick

McIntyre

Stephens

A. L.

Eidin

Krajcik

Damelin

(2022). A framework for supporting systems thinking and computational thinking through constructing models. Instructional Science, 50(6), 933–960. https://doi.org/10.1007/s11251-022-09590-9

67.

Strawhacker

Lee

Bers

M. U.

(2017). Teaching tools, teachers’ rules: Exploring the impact of teaching styles on young children’s programming knowledge in Scratch Jr. International Journal of Technology and Design Education, 28(2), 347–376. https://doi.org/10.1007/s10798-017-9400-9

68.

Sun

Wang

(2019). A review of the artificial neural network surrogate modeling in aerodynamic design. Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, 233(16), 5863–5872. https://doi.org/10.1177/0954410019864485

69.

Trenado

Pedroarena-Leal

Ruge

(2021). Considering the role of neurodidactics in medical education as inspired by learning studies and music education. Medical Science Educator, 31(1), 267–272. https://doi.org/10.1007/s40670-020-01176-9

70.

Varela

Rebollar

García

Bravo

Bilbao

(2019). Skills in computational thinking of engineering students of the first school year. Heliyon, 5(11), Article e02820. https://doi.org/10.1016/j.heliyon.2019.e02820

71.

Vertesi

Dogan

Stefanidis

(2020). Usability evaluation of virtual learning environments: A university case study. In Online teaching and learning in higher education (pp. 161–183). Springer. https://doi.org/10.1007/978-3-030-48190-2_9

72.

Wang

(2020). Intelligent employment rate prediction model based on a neural computing framework and human–computer interaction platform. Neural Computing and Applications, 32(21), 16413–16426. https://doi.org/10.1007/s00521-019-04019-w

73.

Wang

Liu

Huang

(2019). Learning performance prediction via convolutional GRU and explainable neural networks in e-learning environments. Computing, 101(6), 587–604. https://doi.org/10.1007/s00607-018-00699-9

74.

Wang

Liu

Wang

(2003). Discovering the capacity of human memory. Brain and Mind, p. 89–198. https://doi.org/10.1023/A:1025405628479

75.

Wilson

(2005). Constructing measures: An item response modeling approach. Erlbaum Associates. https://doi.org/10.1016/j.evalprogplan.2005.07.008

76.

Peng

(2021). How to foster pre-service teachers’ stem learning design expertise through virtual internship: A design-based research. Educational Technology Research and Development, 69(6), 3307–3329. https://doi.org/10.1007/s11423-021-10063-y

77.

Xue

Huggins-Manley

A. C.

Leite

(2022). Semisupervised learning method to adjust biased item difficulty estimates caused by nonignorable missingness in a virtual learning environment. Educational and Psychological Measurement, 82(3), 539–567. https://doi.org/10.1177/00131644211020494

78.

Yen

M. H.

Wang

C. Y.

Chang

W. H.

Chen

Hsu

Y. S.

Liu

T. C.

(2018). Assessing metacognitive components in self-regulated reading of science texts in e-based environments. International Journal of Science and Mathematics Education, 16(5), 797–816. https://doi.org/10.1007/s10763-017-9818-2

79.

Zhang

T. H.

You

X. Y.

(2017). The use of genetic algorithm and self-updating artificial neural network for the inverse design of cabin environment. Indoor and Built Environment, 26(3), 347–354. https://doi.org/10.1177/1420326X15609772