Higher Education Students’ Assessments towards Gamiﬁcation and Sustainability: A Case Study

: Some academics are pushing for teaching sta ﬀ to implement didactical practices to obtain sustainable behavior. Moreover, a fundamental element of gamiﬁcation in all educational stages is the noteworthy e ﬀ ect it has on student motivation. In our study, we examined higher education students’ assessments in the University of Malaga on the design of educational gamiﬁcation and the teaching criteria that relate to gamiﬁcation, and its e ﬀ ect on student’s motivation, the applications of gamiﬁcation, and the attitude towards sustainability. A questionnaire was administered to students studying for degrees in education from the University of Malaga. A total of 187 valid surveys were obtained. Partial least squares structural equation modeling was computed for these data based on an exploratory–conﬁrmatory model and using SmartPLS version 3.3.2. A positive relationship between assessment of students’ motivation and sustainability attitude was found. We found that the opinions of the students about the role of their training and teaching design material in education would a ﬀ ect students’ motivation and the sustainability behavior, although the power of this last construct is moderate. M.J.S.-V. and J.J.L.O.; writing—original draft preparation, M.J.S.-V. and M.D.B.-M.; writing—review and editing, M.D.B.-M.; visualization, M.J.S.-V. and J.J.L.O.; supervision, M.J.S.-V.; project administration, M.J.S.-V. authors


Introduction
In recent years, there has been significant growth in the scientific literature on the implementation of non-traditional teaching methods in higher education to renew teacher training programs and improve teaching-learning processes [1][2][3]. Within a wide variety of constructive and innovative practical proposals, it is worth highlighting gamification as a pedagogical concept that is acquiring a wide epistemological and didactic development [4][5][6].
Following Deterding, Dixon, Khaled, and Nacke [7] gamification is defined as the use of game design elements in non-game related fields, and it means the implementation of this methodology in many areas that are different from where video games originated.
In the educational context, it is a procedure based on a system of scoring and recording rewards that provides students with tools for developing skills, solving problems related to social behavior, improving their performance, as well as greater involvement in the processes of teaching and Finally, it should be noted that the use of gamification within a classroom not only contributes to an improvement in the students' academic training, but also to a substantial improvement in emotional and social terms [37]; therefore, it not only promotes more intuitive, direct, and dynamic cognitive constructions but also focuses on shaping the students' integral development, improving and facilitating communicative processes, autonomous learning, and teamwork, which reconcile very well with the need to learn, enjoy, and maintain a high expectation towards university learning.

Gamification and Sustainability in the Higher Education
Education for sustainable development concentrates its efforts on solving the planetary crisis that surrounds us, recognizing global problems such as the loss of biodiversity and climate change [38]. It relies on the involvement of various social institutions, including the university, as a place in which to acquire conceptual and procedural bases and values with which to exercise the profession from the perspective of ethics and care for the environment [39]. In this context, education is a key element and a priority to involve new generations to carry out personal actions that contribute towards respecting the environment and increasing social awareness [40][41][42][43].
In 2015, within the framework of the United Nations, world leaders unanimously adopted the document "Transforming our world: The Agenda 2030 for Sustainable Development", constituting the most ambitious global agreement in our recent history. This Agenda came into force on 1st January 2016, with the aim that by 2030, the world and all societies would be able to move towards a better, more sustainable, and inclusive future [44]. The Agenda defines the 17 Sustainable Development Goals, alluding to the need to promote education for sustainability, reorienting training plans at all levels of education, including university [45][46][47]. In the report derived from the previous document, the need to create learning spaces focused on students is highlighted, promoting a playful, innovative, and participatory pedagogy, capable of enhancing motivation and academic performance through attractive teaching methods such as gambling. In this way, it is possible to promote self-regulated learning based on an interdisciplinary approach to environmental conflicts. Given this panorama, the United Nations Educational, Scientific and Cultural Organization (UNESCO) and the United Nations (UN) are in favor of the use of video games in the educational space to provide training in topics related to social conflicts and environmental protection [48]. In the university context, in particular, there are more and more teachers committed to the use of gamification as an avant-garde methodology that favors learning through the implementation of virtual teaching experiences. According to Gatti, Ulrich, and Seele [49], the elements of game positively influence both the motivation and the experience of the student in terms of their pro-environmental attitudes, reaching higher levels of conceptual knowledge, awareness, and sensitivity towards ecological conflicts. In the same line, other authors suggest that serious environmental video games or game-playing platforms will improve the cognitive, emotional, and behavioral potential of students in problems related to climate change. When these factors converge with social interaction, the possibilities of implementing actions that will result in an improvement in environmental reality are increased [48,50,51].
Despite all the declarations of intent made by international organizations and the importance they give to the learning/teaching attitudes or patterns of behavior contributing to increased sustainability, there is a great deal of research that focuses on analyzing the process of curriculum greening at the university, which reveals the necessary involvement of teaching staff to generate changes in teaching following the sustainability paradigm [39,[52][53][54][55][56][57][58][59] The implementation of education for sustainable development in education and higher education requires a commitment to reorienting learning experiences to make students aware of their environmental responsibilities, capabilities, and motivations [60].
Following Langendahl et al. [61], based on the gamification defined by Deterding et al. [7], gamification can be used as a means to engage students in management education for sustainability. Research in support of gamification, as an active and innovative methodology, for education for sustainable behavior is scarce [49,61]. The purpose of this study is, on the one hand, to identify the Sustainability 2020, 12, 8513 4 of 20 assessment of higher education university students about the convenience of the use of gamification as an active methodology in the teaching-learning process, as well as the implementation of education for sustainable development in higher education. On the other hand, to test these new ideas established by Gatti [49] and Langendahl [61]. Concretely, we analyze the effects of implementing the gamification on motivation, on different applications of the gamification, and on the development of attitude towards sustainable behavior-all refer to the perceptions of the university students interviewed-based on a model of structural equations. These students of education from university will be teachers in the future.
The literature described above allows us to consider a conceptual exploratory-confirmatory model. Several authors point out that the training and design of gamification affects motivation; the following hypotheses are confirmatory: H1 and H2 are established based on [28][29][30][31][32] and [48].

Hypothesis 1 (H1).
Training and teaching criteria about gamification are directly positively related to motivation.

Hypothesis 2 (H2).
Training and teaching criteria about gamification are directly and positively related to applications of gamification.

Hypothesis 3 (H3).
Training and teaching criteria about gamification are directly and positively related to attitude towards sustainability.

Hypothesis 4 (H4).
Applications of gamification are directly related to attitude towards sustainability. Now we test if there was a relation between the rest of the construct. The following hypothesis are descriptive-exploratory with no established sign: Hypothesis 5 (H5). Global motivation directly affects applications of gamification.
Hypothesis 6 (H6). Global motivation directly affects attitude towards sustainability. The final theoretical model includes four constructs with 12 items (from a total of 26 initial considered items): applications of gamification with two items; training and teaching criteria (in the design of gamification and the minimum criteria of the activities of subjects to learn) named with two items; global motivation with three indicators; attitude towards sustainability with two items.

Data and Measures
The data collection method used for this analysis was a questionnaire developed specifically for gamification and sustainability based on an established scale used in the studies by Aznar-Díaz, Raso-Sánchez, and Hinojo-Lucena [62] and the study by Domínguez Valerio, Moral Cuadra, Medina Viruel, and Orgaz-Agüera [63]. Most of the students had previous experience with gamification and sustainable behavior at university. The questionnaire includes items of gamification and sustainability. It was distributed online among the students at the University of Malaga's "virtual campus", based on the Moodle platform. Item descriptions are included in Table 1.

Profile of Respondents
This study involved a sample of 187 students of education at Malaga University. It was a voluntary questionnaire and anonymous. Data collection was performed by lectures between March and April of the year 2020. A total of 138 were female (73.8%) and 49 males (26.2%). The average age was 22.97 years old and age ranged between 18 and 43 years old. The profile of respondents is listed in Table 2. Table 1. Initial items of the questionnaire.

Gender
Age Degree Name of the Degree Course

M1_motivation
Students' motivation can be higher with these resources.

M2_learningquality
If the work in the classroom takes place in a fun context, then the quality of learning improves.

M3_rewards
The challenge of obtaining rewards, points or medals in a video game can be a stimulus for students when facing an educational activity.

M4_lessmotivothers
The gamification in some educational activities or disciplines can imply the lack of motivation in other educational activities in which more traditional methodologies are followed.

M5_teachermotivation
The implementation of gamification in the classroom can be more motivating for teachers when working.

V1_noproblemclass
The integration of video games in the classroom is not at all problematic.

V2_allcurriculum
Gamification is effective in the educational environment if it includes the whole curriculum.

V3_performance
Game-playing is compatible with the needs of performance and achievement assessment, as established by current educational regulations.

V4_gametraining
Teachers must follow training courses on the application of gamification techniques in the educational environment in order to guarantee its effectiveness.

V5_Mincriteria
The implementation of gambling in the classroom requires certain minimum criteria for the selection of games by teachers.

A1_linguisticomm
Playful work in the classroom can contribute to the acquisition of linguistic communication skills by students.

A2_mathescience
Playful work in the classroom can contribute to the acquisition by students of mathematical and basic skills in science and technology.

A3_digital
Playful work in the classroom can contribute to the acquisition of digital skills by students.

A4_learningtolearn
Playful work in the classroom can contribute to the acquisition by students of skills for learning to learn.
A5_spirit entrepreneurial Playful work in the classroom can contribute to the students' acquisition of initiative and entrepreneurial.

A6_cultural
Playful work in the classroom can contribute to the students' acquisition of cultural awareness and expression.

A7_socialcivic
Playful work in the classroom can contribute to the students' acquisition of social and civic skills.

A8_family
It is important to involve the family in order to make an adequate pedagogical use of video games.

A9_gametraditional
Game-playing is a more effective pedagogical strategy than traditional teaching-learning methodologies.

A10_extraeffort
The planning and design of learning situations in which the use of video games is integrated involves an extra effort on the part of teachers.

A11_distractgoal
The implementation of game-playing in the classroom can distract students from their pedagogical goal to focus more on the playful aspect.

S1_sustainliving
Everyone should receive education that teaches the knowledge, perspectives, values, issues and skills for sustainable living in a community.

S2_nextgeneration
The current generation must ensure that the next generation inherits a community at least as healthy, diverse and productive as the one today.

S3_sustainatschool
Teaching the principles of sustainability must be integrated into all subjects and at all levels of schooling.

S4_environinclusion
Do you think teachers have been interested in including environmental content in their subjects.

Scale of Items
The Likert scale used is from 1 to 4. Although the scale was previously established and validated [62,63], we ran a reliability test (coefficient of Cronbach's alpha) using statistical software SPSS v25 for the measurement scale on the construct and, further, in a stepwise process, some items were removed from each factor to increase the alpha coefficient. Assessment of the measurement scale's internal consistency revealed a coefficient alpha (CA) with a range of values from 0.64 to 0.91 for the respective items of each construct, which suggests an internal consistency for these scales (for the CA, see Table 3). We obtained a descriptive analysis of the questionnaire's items and then we applied a quantitative approach of a structural equation modeling based on variance estimated by partial least squares-structural equation modeling (PLS-SEM) to analyze the relationships between different dimensions of gamification in education and the future teachers' attitudes towards sustainability. Table 3. Cronbach's alphas.

Constructs and Ítems
Cronbach's Alphas Considering that a 3 on the Likert scale represents an agreement, we focused on those items with a mean over this value. Based on the original dimensions of the survey's, the base of a gamification study [62]: the averages above 3 in the items of student motivation, quality teaching, obtaining rewards, and teacher motivation stand out in "global motivation"; in "viability of gamification" the relevant averages are those of training in gamification techniques and of the minimum criteria for selection of games by teachers; in "applications", the higher averages are presented in linguistics, mathematics, and science, digital competencies, learning to learn, cultural awareness, and expressions, social and civic competencies, as well as the role of the family in the appropriate pedagogical use of video games; finally, in "sustainability behaviors", the averages of greatest magnitude are those related to sustainability in living, at school, and next-generation (Table 4).

Exploratory-Confirmatory Model and Reliability Test
As was aforementioned, the theoretical model includes 4 constructs with 12 items. Concretely, the construct, which is related reflectively with the items: A04, the contributions to learning to learn; A05, the acquisition of initiative and entrepreneurial; A06, the acquisition cultural awareness and expression; A07, the acquisition of social and civic skills, represents the applications of the gamification and the latent variable was named so. The construct related with V4, teachers must follow training courses on the application of gamification techniques in the educational framework in order to guarantee its effectiveness; V5, the implementation of gamification in the classroom requires certain minimum criteria for the selection of games by teachers, is named training and teaching criteria. Items M1, student's motivation; M3, student's rewards; M5, teachers' motivation, represent the motivation either by the students or by the teacher and was called global motivation. Finally, there is a construct related with S1: everyone should receive education that teaches the knowledge, perspectives, values, issues and skills for sustainable living in a community, and S2: the current generation must ensure that the next generation inherits a community at least as healthy, diverse and productive as the one today, we understand this construct as the attitude towards sustainability. SEM is a statistical approach that allows us to evaluate a theoretical model and allows us to test all the relationships in a model at the same time among constructs and items, and among the latent variables. Data from a sample size of 187 completed questionnaires with no missing values are adequate for the PLS-SEM analysis method as it meets the minimum threshold of 100 samples established by earlier studies [64]. Furthermore, Chin [65] and Hair et al. [66] recommended that the minimum sample size should be 10 times the largest number of paths appointed to a particular construct. Later on, as shown in Figure 1, the maximum number of paths appointed to a latent variable was three in this study; a sample of 187 valid questionnaires was acceptable. PLS-SEM was performed on the collected data using SmartPLS 3.3.2 software [67].
The assessment of the PLS-SEM was carried out in two steps: Firstly, the outer model, and, secondly, the inner model. In the measurement (outer) model, each construct was related with their items reflectively (mode A); as a result, the evaluation of the measurement model included: Item reliability, Cronbach's alpha (CA), composite reliability (CR), average variance extracted (AVE), and discriminant validity. Concerning the structural model, an assessment of the structural model was conducted computing R2 and the coefficient Q2 by Stone-Geiser [68,69], and the constructs' collinearity. The significance of the path coefficients and the magnitude of the standardized beta-path coefficients were the second to last steps of the inner model's assessment [70] taking into account that the higher the magnitude of the path coefficient is, the stronger is the effect, in other words, the higher the path's relevance. The bootstrapping of 5000 re-samples of a sample size of 187 generated confidence intervals and assessed the significance of the path. The model is a mix of confirmatory-exploratory model, as a result, a two-tailed test was considered as only option that can be selected from the two available options: one or two tails. Lastly, mediation analysis was performed.
adequate for the PLS-SEM analysis method as it meets the minimum threshold of 100 samples established by earlier studies [64]. Furthermore, Chin [65] and Hair et al. [66] recommended that the minimum sample size should be 10 times the largest number of paths appointed to a particular construct. Later on, as shown in Figure 1, the maximum number of paths appointed to a latent variable was three in this study; a sample of 187 valid questionnaires was acceptable. PLS-SEM was performed on the collected data using SmartPLS 3.3.2 software [67]. The assessment of the PLS-SEM was carried out in two steps: Firstly, the outer model, and, secondly, the inner model. In the measurement (outer) model, each construct was related with their items reflectively (mode A); as a result, the evaluation of the measurement model included: Item reliability, Cronbach's alpha (CA), composite reliability (CR), average variance extracted (AVE), and discriminant validity. Concerning the structural model, an assessment of the structural model was conducted computing R2 and the coefficient Q2 by Stone-Geiser [68,69], and the constructs' collinearity. The significance of the path coefficients and the magnitude of the standardized beta-path coefficients were the second to last steps of the inner model's assessment [70] taking into account that the higher the magnitude of the path coefficient is, the stronger is the effect, in other words, the higher the path's relevance. The bootstrapping of 5000 re-samples of a sample size of 187 generated confidence intervals and assessed the significance of the path. The model is a mix of confirmatoryexploratory model, as a result, a two-tailed test was considered as only option that can be selected from the two available options: one or two tails. Lastly, mediation analysis was performed.

Results of the PLS-SEM
We followed a systematic process of examining the loadings following recommendations [71,72].

Measurement Model
The measurement model included constructs related to their respective items reflectively (mode A in SmartPLS); the evaluation of the measurement model includes: Item reliability, Cronbach's alpha (AC), composite reliability (CR), and average variance extracted (AVE) as it was mentioned in Section 3. Convergent validity and reliability are shown in Table 5. Scale items loaded on their respective

Results of the PLS-SEM
We followed a systematic process of examining the loadings following recommendations [71,72].

Measurement Model
The measurement model included constructs related to their respective items reflectively (mode A in SmartPLS); the evaluation of the measurement model includes: Item reliability, Cronbach's alpha (AC), composite reliability (CR), and average variance extracted (AVE) as it was mentioned in Section 3. Convergent validity and reliability are shown in Table 5. Scale items loaded on their respective factors with loadings ranging from 0.73 to 0.91; the coefficient alpha ranged from 0.64 to 0.87; the average variance extracted ranged from 0.81 to 0.91, confirming convergent validity; composite reliabilities ranged from 0.58 to 0.77 supporting reliability for all constructs.  Table 6 displays the results for the Fornell-Larcker procedure [73] to assess discriminant validity, which is satisfactory for all constructs, meaning that construct measures different concepts. The same conclusion was obtained with the results of the Heterotrait-Monotrait (HTMT) ratio criteria, as the values are below 0.85 [74].

Structural Model
The predictive power was measured by the coefficient of determination R2 and the Q2 of Stone-Geiser. Values of 0.67, 0.33, and 0.19 for R2 should be considered substantial, moderate, and weak, respectively [65]. The R2 values of each endogenous latent variable in this study were 0.128, 0.157 (all aforementioned lower than weak), and 0.234 (lower than moderate) for applications of gamification, global motivation, and attitude towards sustainability, respectively. The coefficient's Q2 were all greater than one. The measure named effect size (f2) is a complementary measure and it evaluates the increment in R2 by the omission of a construct.
We employed a variance inflation factor (VIF) to check the existence of harmful collinearity among predictors if any VIF is over 3.3 following the advice of Hair-Jr, Hult, Ringle, and Sarstedt [75]. In our study, the VIF obtained indicated that there is no problematic collinearity ( Table 7). All the above-mentioned values are displayed in Table 7.  Figure 2 offers a comprehensive visualization of the path model for the estimated model. The results of our estimated path coefficient (direct effects) are included in Table 8.   The results of our estimated path coefficient (direct effects) are included in Table 8. The hypotheses H1, H3, and H4 are supported. The results reveal the positive and significant direct effect of training and teaching criteria on the global motivation and attitude towards sustainability. Lastly, global motivation affects applications of gamification with sign-positive. Figure  3 summarizes the mediations under consideration.

Mediation Analysis
The mediating effect, indirect effect, or mediation involves a third variable intermediating the relationship between an exogenous variable and an endogenous variable, in this case, all variables are latent variables. That is, the effect of the independent construct on the dependent construct is mediated by the third construct named mediating latent variable or mediator [76].
Considering that the direct effects of the applications of gamification and global motivation are not significant to the attitude towards sustainability, as well as the training and teaching criteria to the applications of gamification. It is interesting to study the mediating role of said constructs. So, we went further and propose a more general hypothesis: instead of direct relations, we changed H1 to H6 to positive relations between each pair of constructs, and we added four relations of mediations, three sub-models with a two-path mediated effect, and one with three-path mediated effect. We have already tested the hypotheses: H1 to H6. Next, we test the mediation effects:

Mediation Analysis
The mediating effect, indirect effect, or mediation involves a third variable intermediating the relationship between an exogenous variable and an endogenous variable, in this case, all variables are latent variables. That is, the effect of the independent construct on the dependent construct is mediated by the third construct named mediating latent variable or mediator [76].
Considering that the direct effects of the applications of gamification and global motivation are not significant to the attitude towards sustainability, as well as the training and teaching criteria to the applications of gamification. It is interesting to study the mediating role of said constructs. So, we went further and propose a more general hypothesis: instead of direct relations, we changed H1 to H6 to positive relations between each pair of constructs, and we added four relations of mediations, three sub-models with a two-path mediated effect, and one with three-path mediated effect. We have already tested the hypotheses: H1 to H6. Next, we test the mediation effects: Hypothesis 7 (H7). Global motivation mediates between training and teaching criteria and attitude towards sustainability with a two-path mediated effect.

Hypothesis 8 (H8).
Applications of gamification mediates the relationship between the training and teaching criteria, and the attitude towards sustainability with a two-path mediated effect.

Hypothesis 9 (H9).
Global motivation mediates the relation between training and teaching criteria, and applications of gamification with a two-path mediated effect.

Hypothesis 10 (H10).
The relationship between applications of gamification and attitude towards sustainability is sequentially mediated with a three-path mediated effect. Hayes and Scharkow [77] point out that the best approach for detecting a mediation effect is the bias-corrected bootstrap interval, although percentile intervals provide an adequate compromise concerning Type-I errors.
In the case of mediation reported in H8, global motivation does not mediate between training and teaching criteria and attitude towards sustainability, and neither does between application and attitude towards sustainability (H9). Concerning applications of gamification, it does not mediate between the training and teaching criteria and attitude towards sustainability, and neither does it between global motivation and attitude towards sustainability. In the case of the training and teaching criteria and applications of gamification (H7), global motivation mediates between them. It is worth noting that global motivation does not directly influence the applications of gamification. T-statistics and p-values of direct, indirect, and total effects are displayed in Table 9.
Finally, concerning the H10 mediation, the relationship was rejected between training and teacher criteria and attitude towards sustainability is sequentially and positively mediated by global motivation and applications of gamification. On the other hand, applications of gamification do not impact attitude towards sustainability (H5), and the rest of the hypotheses (H1, H2, H3, H4, and H6) are supported, not rejected, at a level of significance of 5%. Figure 5 shows the total effects. Our results show the relevance of global motivation with a positive impact on applications of gamification, as well as on attitude towards sustainability. The role of training and teaching criteria positively affects applications of gamification, global motivation, and attitude towards sustainability. Our objective was reached, we conclude that our study contributes to the empirical analysis of the influence of gamification on motivation and sustainability.  In the case of mediation reported in H8, global motivation does not mediate between training and teaching criteria and attitude towards sustainability, and neither does between application and attitude towards sustainability (H9). Concerning applications of gamification, it does not mediate between the training and teaching criteria and attitude towards sustainability, and neither does it between global motivation and attitude towards sustainability. In the case of the training and teaching criteria and applications of gamification (H7), global motivation mediates between them. It is worth noting that global motivation does not directly influence the applications of gamification. T-statistics and p-values of direct, indirect, and total effects are displayed in Table 9. Table 9. Direct, indirect, and total effect.

Paths
Letter in Figure 3 or Our results show the relevance of global motivation with a positive impact on applications of gamification, as well as on attitude towards sustainability. The role of training and teaching criteria positively affects applications of gamification, global motivation, and attitude towards sustainability. Our objective was reached, we conclude that our study contributes to the empirical analysis of the influence of gamification on motivation and sustainability.

Discussion and Conclusions
This study had several purposes, first, to analyze the perceptions that higher education students from Malaga University have about the use of gamification as an innovative methodology in the teaching-learning process. Related to this objective, the results reveal that these higher education students perceived gamification as an active methodology that has positive effects on the motivation of both students and teachers. In this sense, the participants show that obtaining rewards, points, or medals are tools that stimulate students in the development of an educational activity. These outcomes are consistent with other studies that have analyzed how the use of innovative techniques, including gamification, promotes and improves the motivation, perception, commitment, and social awareness of students in learning [19][20][21]29]. Concerning the viability of gambling in the teaching-learning context, this study highlights that the perceptions that the implementation of gamification in the classroom requires teachers to have adequate training in this methodology and its practical application to the classroom context. Moreover, following Aróstegui-Barandica and Darretxe-Urrutxi [30], teachers must organize, plan, and select the appropriate digital games for the teaching topic. The objective is to encourage curiosity towards learning new content; to adapt to the learning pace of the students; to motivate and generate a good climate in the classroom. Concerning the educational applications of gamification, the participants perceived that these applications in the classroom can contribute to the acquisition of skills, among which is learning to learn, digital skills, and social and civic values. These contributions are in line with other research [33,34], which indicates that gamification and the appropriate use of serious video games are key tools for teaching content, skills, abilities, and improving creative capacity. The study by Villalustre and Del Moral [32], carried out with students of the Degree in Pedagogy, obtained similar results, showing that the use of game dynamics applied in the realization of a project increased motivation, the level of satisfaction with the task entrusted, knowledge of theoretical contents, and the development of key generic competences.
Secondly, students' assessments of how education for sustainable development is carried out in the university context were studied. The results show that the respondents' attitudes are in favor of receiving an education based on the knowledge of values and skills for a life marked by sustainable nuances, as well as the power to ensure that the next generation has an environment as healthy as possible. These results are also in line with other studies that focus on giving students tools to respond to conflicts related to global degradation [39,41,56,60]. Similarly, participants have expressed that teachers are not actively involved in making the principles of education for sustainable development a reality. The green curriculum does not only include the incorporation of environmental contents in the teaching programs, but it is also necessary to integrate global changes in the teaching processes that link theory with practice, action with reflection and, in short, a global conception of education that can guarantee the development of the individual, as is pointed out by Aleixo et al. [40] and Murga-Menoyo and Novo [43].
Thirdly, following Dichev and Dicheva's recommendations [36], we have contributed to an empirical work that investigates. Concretely, we have estimated a PLS-SEM in which the effects of the training and teaching criteria of gamification in the educational context, students' and teachers' motivations, applications of gamifications, and attitudes towards sustainability are related.
The results of the PLS-SEM model are limited as they refer only to the higher education students from the University of Malaga towards gamification and education for sustainable behavior; therefore, the sample is not representative of the population of all students at this university. However, the results obtained are in line with the findings of other studies that have been carried out.
Specifically, the introduction of gamification in the educational framework is perceived as a positive influence on the motivation of students. H1 teacher training and planning in gamification directly and positively affects global motivation, student motivation and teacher motivation, and is supported in line with findings of several authors [19][20][21][29][30][31][32]. Furthermore, the global motivation mediates totally between training and teaching criteria, and applications of gamification.
Additionally, the motivation has a direct and positive influence on the acquisition of pro-environmental attitudes (H6), together with greater awareness and sensitivity towards environmental conflicts, which are included in the construct named attitude towards sustainability, being a descriptive result. The most relevant finding are the total effects of this exogenous construct-training and teaching criteria-on attitude towards sustainability [49,61] and global motivation. Thus, as a result of these students' assessments, the key role of the teacher, as the main architect, is being responsible for motivating and encouraging students to develop sustainability skills, which come into play as he or she plans and designs educational activities with the environment as the central focus [48][49][50][51].
Among other conclusions, it can be pointed out that the use of active methodologies in higher education, such as gamification, requires that future education teachers assume a change of role to respond to the needs of today's society. To do so, they need to have acquired generic professional skills that are adjusted to the changes generated by the technological impact. Furthermore, their teaching methodology requires clear, rigorous planning and minimum criteria for the selection of teaching and learning materials to motivate both students and teachers; however, although the attitude towards gamification and pedagogical innovation is positive, maybe there are lecturers, who are reluctant to innovate methodology. This may be due to the demand for time and the imbalance it can create in terms of research when, up to now, research and academic papers are more relevant in the university's system of rewards compared with university teaching. It must be kept in mind that the various digital resources are continually becoming obsolete, which means that teaching staff must constantly update their knowledge.
On the other hand, we must stress the existence of technological limits to reach the maximum levels of interactivity and participation in the development of gamification, either because all students do not have technological resources, or because the participation at a collaborative level among students (through groups) is not sufficiently developed. However, the continuous digital advances will progressively overcome some of these limitations. Similarly, it should be measured whether the increase in motivation and participation of students is transferred in parallel to an increase in the acquisition of knowledge and skills. It is also necessary to reconcile the different student profiles with the teaching style used by the teaching staff, making use of gamification.
Taking into account the contributions derived from this study concerning sustainability, we highlight that although there is a favorable predisposition on the part of teachers and students concerning a green curriculum, the reality shows that the training received in the university context related to the development of sustainability is scarce, and teachers do not have the necessary innovative tools with which to promote sustainability skills. For all these reasons, we need to use the active methodologies, such as gamification, that guarantee the motivation of students not only in terms of cognitive learning results but also in their attitudes towards sustainable development. However, this last result related to the education for sustainable behavior is merely descriptive-exploratory based on education students' assessment from one grade and one university and needs to be compared with future researchers' results. Future lines of research should continue to investigate the potential of teaching tools as a learning resource, of an education in values, and civic and environmental attitudes in the university context, above all, with more empirical studies and with control groups.