Cross-Disciplinary Analysis of Cooperative Learning Dimensions Based on Higher Education Students’ Perceptions

This study analyzes the perception, mainly on motivation, interpersonal relationships, and learning outcomes, of higher education students from seven university disciplines derived from Cooperative Learning (CL) activities undertaken in the course of their degree studies. The cross-disciplinary dimensional analysis on cooperative learning included students’ motivation, academic performance, students’ relationships within cooperative groups, group organization, and teacher involvement. The study was carried out using a questionnaire validated by a number of CL experts. The subsequent analysis of a sample of 162 student’s perceptions on the CL dimensions provided first, positive students’ perceptions regarding satisfaction, motivation, learning outcomes, and interpersonal relationships, and second, that differences between university degrees on CL were significant, suggesting a strong dependence of cooperative dimensions on the implemented approach.


Introduction
Cooperation is seen as an educational methodology that is based on providing students with the skills for sustainable development, and relies mainly on integrating group strategies into personal development. Education for sustainable development, when developed in tertiary systems, enhances students' ability to understand and interact with the environment and with society in order to best identify individual and collective impacts [1]. Therefore, university teachers should be aware of transferring sustainability skills and contents to enhance sustainable competences to the teaching profession. It entails promoting student relationships within groups and fostering relationships in socialization and learning with the environment and with the society [2][3][4]. Cooperative learning is directed towards promoting group development during strategic action, interpersonal relationships, students' affective, cognitive, and social development [3,14,36,[45][46][47][48] and is considered a methodological tool capable of responding to the individual needs of 21st century students [10,11,18,46,49]. CL creates positive interpersonal relationships which give personal and academic support, improves mental health and well-being (including self-esteem and social skills), and promotes a positive attitude towards the university experience, as it involves student interaction and fosters relationships in socialization and learning [3,4,50].
CL is employed in various disciplines within the higher education system, although the majority of research on CL is carried out within faculties of education [50,[55][56][57]. In all fields, CL is present in the learning objectives and content of the degree studies. Some studies compare the effectiveness of CL with traditional university learning methods in different university programs; however, few transdisciplinary studies exist on this topic. Therefore, this study aims to contribute to the existing literature by providing empirical evidence on the effects of cooperative learning through a cross-sectional analysis of various university degree studies.
This study is mainly addressed to respond to how students from various university disciplines perceive Cooperative Learning Activities (CLA). Towards this purpose, first, the relationship between the grades obtained in CL activities and the final grade awarded for the subject was analyzed, then a survey was designed to determine the satisfaction students felt with the activities. The survey took into account student motivation, learning assessment, group relationships, CLA organization, and the involvement of the teaching staff. Therefore, the objective of this study is to gain information on the perceptions that university students from different degrees have after implementing in-class cooperative educational approaches, with special interest on students' perceptions on teacher involvement as well as on students' motivation, organization, and internal member relationships.

Context of the Study
This study was conducted within the framework of the Network of Educational Innovation in Cooperative Learning (XIDAC), at the University of Girona (UdG), during the academic year 2018-19. XIDAC members lecture on a variety of degrees which include Environmental Sciences, Chemistry, Mechanical Engineering, Business Administration and Management, and Primary School Education. Table 1 shows the university subjects encompassed in this study and describes the structure and main characteristics of the Cooperative Learning Activities (CLAs) covered in this project. All teaching staff involved in XIDAC ensured that all activities undertaken through Cooperative Learning complied with Johnson and Johnson´s five criteria [20,58]: (a) Positive interdependence, (b) face-to-face interaction, (c) individual responsibility, (d) small group interpersonal skills, and (e) group processing. In MP, groups of 3 students are asked to analyze the manufacturing process of a consumer good, present their work in a poster session, and then defend it in an oral presentation.

Participants
The total number of students involved was 299, and the percentage of response of the survey proposed was 54.18%. Therefore, the total number of answers to the survey was 162. Table 2 shows the number of students of each subject in the study, the number of answers given, and the percentage of each Cooperative Learning Activity (CLA) per subject.

Survey Design
To construct the survey, first, an extensive bibliographic search on cooperative learning was carried out that served to define the elements, indicators, parts, and questions of the survey. Second, the survey was submitted to a validation process that used the judgments of external experts, who confirmed the usefulness of this instrument. Third, we carried out a pilot test with a group of students from different university degrees to evaluate if they could easily understand the questions of the survey. From this process, we concluded that the survey met the requirements needed to achieve the aims of our research.
Students involved in this study were asked to answer online the designed survey (Google Forms). The survey was available for 1 week after completion of the subject. All answers are individual and anonymous.
The survey consisted of 20 statements divided in 5 dimensional blocks and evaluated the following: (1) Level of motivation; (2) learning outcomes and time spent on the CLA; (3) relationship with the rest of the members of the group; (4) organization of the CLA; and (5) involvement of the teaching staff (Table 3). Table 3. Survey statements by thematic blocks.

Dimension Block
Code Statements Figure   B11 The CLA increased my motivation for the subject B12 The CLA made me more participative My relationship with the teaching staff improved with this activity B53 The teaching staff organized this activity well All responses were ranked in 6 categories: (0) No response/do not know (NR/DK); (1) strongly disagree; (2) somewhat disagree; (3) neither agree nor disagree; (4) somewhat agree; and (5) strongly agree (Table 3). In order to statistically analyze the results, the variation in satisfaction between categories was considered constant and "NR/DK" responses were not taken into account for the statistical treatment of the results. In addition, mean values for each block were also considered a metric of satisfaction for each block of statements.

Statistical Analysis
A two-fold analysis was carried out ( Figure 1). Firstly, the initial analysis focused on the relationship between CLA grades and the final grade for the subject based on differences between CLAs grades and subject grades. This was conducted using the Kruskal-Wallis test, as most of the parameters analyzed were not distributed normally. Secondly, student satisfaction was determined by analyzing the responses to the survey described in Table 3 above. SPSS software (version 23, 2015, SPSS Inc. Chicago, IL, USA) was used to estimate Pearson's correlation coefficients of the mean values of each block of responses (with a bilateral significance of 0.01), and in order to analyze possible relationships between types of responses and blocks of statements. Differences between degree studies were analyzed using the Kruskal-Wallis test for non-parametric data, as none of the responses were normally distributed.
In addition, a Principal Component Analysis (PCA) was conducted to analyze the association between the statements and type of response. This multivariate analysis approach is used to analyze interrelationships among a large number of variables and to explain these variables in terms of their common underlying dimensions (Principal Components (PCs) [59,60]). Its goal is to reduce the information contained in a large number of original variables into a smaller set of PC variables with a minimum loss of information. In this study, we used PCA to group statements with similar responses, and to analyze students' perceptions of the various aspects of CLAs. After applying PCA, a varimax rotation was conducted. This enabled a "clean up" of the PCs by increasing the participation of variables with a higher contribution, while simultaneously reducing that of variables with a lower contribution. In this procedure, the number of original variables contributing to each PC (called henceforth varifactors (VF)) is reduced at the cost of a loss of orthogonality [59,60]. The dataset used for the PCA included all the responses obtained from the surveys ranked from 1 to 5, thus excluding responses "NR/DK".

Relationship between Grades Obtained in CLAs and Final Grades for Subjects
The CLAs analyzed in the study were carried out in various degree subjects. The main characteristics of the activities are described in Tables 1 and 2, and the relationship between the grades obtained for CLAs and the final grade for the subject for each student are shown in Figure 2. These parameters are important, as they may condition how CLAs are evaluated in the survey.  Figure 2 shows that in the majority of subjects, the highest CLA grades relate to the highest final grades in the subject, showing a significant correlation between these two parameters (R = 0.584). However, there were some differences between subjects and the way in which the CLA grades were evaluated. For example, in Design of Organic Molecules (DOM) and Physical Education Didactic Resources (PEDR), all students obtained a CLA grade ≥ 8. In the case of Hydrogeology (HG), World and Spanish Economy (WSE), Manufacturing Processes (MP) and Physical Education Didactics (PED), grades varied from 2 to 9.5; and finally, in Physical Education (PE), the CLA grades ranged from 5.6 to 9.3. Regarding the final grades of the subject, in DOM, values varied from 2 to 9; in HG, WSE, MP, and PED, students obtained final grades ranging from 4 to 9. The highest grades were observed in PEDR that varied from 6 to 10.
Consequently, when the CLA grades for the various subjects were compared using the Kruskal-Wallis test, PEDR, and DOM, CLA grades could be considered similar (p-value = 1.000); however, significant differences were detected between these two subjects and the rest of the subjects (p-values < 0.001-0.049). Significant differences were also detected between WSE and PE CLA grades (p-value = 0.042).
When the final grades of the subjects were analyzed, differences between WSE and the rest of the subjects were detected (p-values ranging < 0.001 and 0.010), as were differences between PEDR and subjects such as HG, PE, and PED (with p-values between < 0.001 and 0.002). In this case, students obtained the lowest grades in WSE (6.00 ± 1.06), and the highest grades in PEDR (7.94 ± 0.99).
Finally, it was observed that for 76.7% of the students, the cooperative learning activities contributed to obtaining a better final grade, with no significant differences detected between subjects (p-value = 0.273).

Analysis of Student Satisfaction with the CLAs
Students' satisfaction with CLAs was analyzed using the survey described in Table 3. The results are summarized in Figure 3. For motivation statements (B1), 52.2% of the students felt that CLAs had increased their motivation for the subject; 71.4% thought that CLAs made them more participative, and 61.7% liked this type of activity because it made them collaborate more with fellow students. Differences between the subjects were observed. In particular, the highest percentages for B1 statements were obtained for HG, DOM, PE, PED, and PEDR that ranged from 60 to 95%. In contrast, MP and WSE showed the lowest values (30 to 60%).
Regarding learning assessment and time spent (B2 statements; Figure 3), 55.1% of the students thought that CLAs encouraged them to be more involved in their individual work; 56.9% felt that these activities helped them learn how to learn; 57.4% felt more responsible for their individual learning; 56.8% thought that this kind of activity made them feel more responsible for fellow student's learning; 61.8% agreed that CLAs improved group learning; 57.2% of the students thought they had understood the subject better as a result of CLAs; and finally, 53.6% thought they had learned more in the time spent on this activity than working by themselves. When comparing the different subjects, in general, B2 statements followed a similar trend than B1 ones. Except for B21 and B24, for which MP and WSE showed similar percentages than DOM and HG (30 to 50%), for the other B2 statements, MP and WSE were the subjects with the lowest values, 30 to 55%. In contrast, the other subjects showed percentages ranging from 60 to 100%.
When the students were asked about their relationship with the other members of the group (Figure 3, B31, B32, and B33), 84.2% of them felt that their fellow students listened to their opinions during the activity, and 89.3% felt they listened to their fellow students' opinions during the CLAs. In addition, 69.4% agreed that CLA stimulated critical thinking. In these cases, MP and WSE also showed lower percentages (40 to 60%) than the other subjects (75 to 100%).  Table 2 Regarding the organization of CLAs (B4 statements, Figure 3, B41, B42, B43, and B44), 70.2% of students thought that their group organization was productive and satisfactory; 64.2% liked CLA dynamics; 58.8% thought that the classroom, furniture, and distribution of the furniture were suitable for carrying out the activity; and 65.9% agreed that the supporting material supplied by the teaching staff was appropriate for carrying out this type of activity. MP and WSE were also the subjects with the lowest percentages for these B4 statements (35 to 60% vs. 65 to 100%).
Finally, when the students were questioned about their relationship with the teaching staff (teaching staff's involvement in B5), 70.7% felt the lecturer/professor was more motivated than in other subjects; 56.6% thought that their relationship with the lecturer/professor had improved as a result of CLA; and 65.9% thought that the teaching staff had organized the activity well. Similar to the other statements, the lowest percentages were obtained for MP and WSE, which ranged from 30 to 55%. The other subjects showed percentages that varied from 70 to 100%.
To analyze possible differences between subjects/degrees, a non-parametric test (Kruskal-Wallis test) was carried out, as none of the responses to the statements were normally distributed. The results showed ( Figure 3):

1.
In Block 1-Motivation, the main differences were detected between MP and WSE and the remaining subjects (with p-values between 0.44 to <0.001 in most of them), showing lower valuations for this block of statements in these two first subjects.

2.
In Block 2-Learning assessment, differences in responses were mainly detected due to the lower values obtained in WSE and MP than some of the other subjects (e.g., PE, HG, and PEDR), or the higher valuations obtained of PE than other subjects (e.g., HG, MP, and WSE).

3.
In Block 3-Group relationship, the differences between subjects were minimal and only detected between PE and WSE in statements B31 and B33 (p-value of 0.022 and 0.001), and between PED and MP in B33 (p-value of 0.006). Responses from PE and PED showed the highest values.

4.
In Block 4-Organization and Block 5-Professor involvement, differences were mainly detected between WSE and MP, and the remaining subjects. However, these differences were not constant for all the statements. For example, in statement B41, the differences were only observed between WSE and PE (with a p-value of 0.009); in B44, differences were detected between WSE and HG, PE, PED, and PEDR, and between MP and PEDR (with p-values ranging < 0.001-0.042); and in B51, MP showed lower valuations than HG, PE, and PEDR (p-values of < 0.001-0.015), and WSE, than PE (p-value of 0.001).
A Principal Component Analysis was used here as a way to synthetize student's responses to the satisfaction survey, and to determine the various trends and relationships between statements and answers. The final solution included 3 components that explained 69.18% of the total variance. The value of the Bartlett chi-square statistic, by applying the sphericity test, was 2215.9 (for 190 degrees of freedom and a minimum significance level of <0.001), confirming that the survey variables were correlated. The measure of sampling adequacy (MSA) obtained by the Kaiser-Meyer-Olkin (KMO) measure was 0.918.
VF1, with 29.97% variance, was mainly composed of statements related to B5 (teaching staff's involvement), B4 (CLA organization), and B1 (student motivation), as well as statements B2, B25 (related to the improvement in group learning), and B22 (related to the process of learning to learn). In this regard, Figure 4d shows that the students who gave B1 the highest valuations were also those with the highest scores for VF1. 2.
VF2 (22.50% of the variance) mainly comprises B2 statements (related to assessment of learning outcomes and time spent); B11 and B12 are related to motivation and involvement in the activity. Therefore, the students who gave the highest scores for VF2 were mainly those who gave the highest scores for Block 2 (Figure 4e).

3.
VF3 represents 16.71% of the total variance. This comprised B3 statements (relationship block), and one of the Block 4 statements (B41) related to group organization. As with previous VFs, a high correlation is observed between VF3 and B3 mean values (R 2 = 0.867; Figure 4f). Table 4. Loadings of the 20 statements analyzed on the 3 significant varifactors (VF). Statements are detailed in Table 3. Underlined values represent relevant loadings. According to the results of the study, most students who felt motivated by CLAs (the highest score for B1 statements), also thought that these activities were well organized (the highest scores for B4 answers) and that the lecturer/professor was heavily involved in the activity (the highest scores for B5; Figure 4). This perception was independent of the time spent on these activities, how the concepts related to the subject were learned (represented by VF2), or the relationship with the group and its dynamics (represented by VF3, and statement B41).

Discussion
This research analyzed university students' perceptions of cooperative learning activities carried out within the framework of subjects taught in various university degrees. The study looked at whether the CL activities helped improve students' final grades in each subject. In general, it was observed that all CL activities contributed to students obtaining a better overall grade for the subject. From these results, it can be concluded that these activities have improved students' learning experience and the fact that they obtain better grades may also result in their positive assessment of these types of activities. In this regard, the results coincide with the works which state that CLA favors accomplishing academic goals [13,21,[42][43][44][45][46]61].
After analyzing the responses to blocks of statements on perspectives, results show that responses from students in the MP and WSE on the Mechanical Engineering and Business Administration and Management degrees (respectively), were generally the opposite to responses of students who study subjects PE, PED, and PEDR in the Primary School Education degree. In contrast, the remaining two subjects, DOM and HG on the Chemistry and Environmental Sciences degrees (respectively) occupy a more intermediate position. From a general point of view, a probable explanation could be related to possible entry barriers for students of these degrees, who may not be used to these types of learning activities. Traditionally, teaching staff in these degrees use CLA less than in those in education degrees, where this type of training is introduced in both theory (the students will be future teachers), and practice (in the classroom, students carry out collaborative learning activities). Moreover, the education degree at the University of Girona traditionally employs these types of collaborative activities. They are gradually introduced in the second year in PE, increasing to a number of CL activities in several final year subjects, e.g., PED and PEDR, a fact that may explain why the results obtained are higher than those of the initial year subjects [50,[55][56][57][58].
The above analysis shows that the final and overall assessment of CLA is always positive in all subjects and degrees. This should encourage teaching staff in any degree subject to consider carrying out this type of activity. Student satisfaction with CLA is related to increased motivation resulting from the course contents, academic performance in relation to the effort put in, relationships with members of the working group, how the activity is organized, and the extent of the lecturer/professor's motivation. The questionnaire was thus designed to address these aspects.
By analyzing the interrelationship between CLA components, results show that the variables in which CLA increased student motivation (B11) and how this activity improved group learning (B25) positively correlate with all aspects related to how the activity was organized: Organization of the activity itself, material, space, and distribution of students. Furthermore, these variables also correlate with the teaching staff's motivation as perceived by the student, the dynamics of the activity, and how this activity encourages collaboration between peers. Regarding the teaching staff's motivation, active intervention and constant feedback from the lecturer/professor is considered necessary in order to carry out an autonomous, shared, and critical learning process [8][9][10]. Similarly, Figure 4 shows that the subjects in which the best CLA grades are obtained coincide with those in which students claim to be more motivated, DOM, PE, and PEDR. As noted by other authors [62][63][64], students retain more knowledge in cooperative learning situations, as they have to explain it to others. In addition, the fact that they have to learn something efficiently in order to explain it and teach it to others enables them to learn in a more reflexive and critical way [40,51,65].
In addition, the study also detected that students perceived an improved understanding of the subject and were more motivated as a result of carrying out CLA. Along these lines, Tran [33] and Laguador [28] point out that when students perceive an improvement in their learning within a cooperative context, their attitude towards learning also improves. Likewise, in the second varifactor obtained through the PCA, the fact that the CLA makes the student more participatory (B12) correlated with his or her individual work and learning (B21, B23), the learning process itself (B22), feeling responsible for fellow students' learning (B24), and with a better use of time during learning (B27).
As expected, a correlation was observed between variables related to the functioning of the group such as feeling heard and listening to peers, stimulation of critical thinking, and the organization of the group itself. This suggests that the student feels that he or she is part of the group and that he or she can also take an active role in it. In contrast, Figure 4f shows that students in subjects where the group is formed by the teacher (MP) and who have worked in specific cooperative groups (MP and WSE) gave the items in block B3 (Team relationship) of the survey the lowest scores. These results highlight, among other studies, that cooperative learning creates positive interpersonal relationships which are characterized by personal and academic support [3,4,14]. In this regard, promoting interaction within the group of students, fosters relationships through socialization and learning [66][67][68].

Conclusions
In this interdisciplinary study, we have analyzed how students from various university degrees have assessed their participation in cooperative learning activities. Overall, the results of this research show that these activities have enabled students to obtain better grades in their subjects. A questionnaire on aspects of learning such as motivation, degree of learning achieved, and interpersonal relationships was analyzed, and findings show that most students evaluate cooperative learning activities positively. This is because they feel more involved in the learning process, and thus a greater sense of responsibility towards their learning. Findings also show that higher student motivation is related to the activity being well organized as well as the degree of motivation students perceive from the teaching staff, and that this higher motivation in turn leads to a better rating of the activity. In addition, students consider that these activities encourage them to adopt a more active role in the classroom and that this in turn improves their learning process, increases the responsibility they assume for their fellow students' learning, and helps them use study time more efficiently. Finally, we also observed that these activities promoted positive interpersonal relationships between group members as well as with the teaching staff.
Moreover, results affirm that although there are slight differences in student satisfaction with cooperative activities, depending on the university degree where it is carried out, the overall results are positive. Thus, no limitations should be put on using this type of activity in any subject in a university degree course, as it provides a space where learning is promoted by fostering better social skills in students. It also teaches them the skills that will make them more competent in the workplace and gives them the opportunity to access better job opportunities in the 21st century work environment. Finally, results also pointed out that the degree of student satisfaction depended on factors such as the role taken by the teaching staff, how the activity is organized in class, the relationship between classmates, and the subject studied.
In summary, the main contribution of this study is that it has been shown that cooperative learning can be applicable to any university degree. The lowest percentages of student satisfaction have been obtained for those subjects in which, traditionally, this type of learning is less employed, reinforcing the idea that cooperative learning cannot be done on an on-off basis, but a process is needed for the student to learn to cooperate. Based on all the above, universities should strongly consider adopting it as a teaching methodology because it leads to more creative, flexible, critical, and innovative people, essential values for citizens of the 21st century. We would like to mention that a limitation of the current study is the size of the sample, which will be enlarged in future studies.

Conflicts of Interest:
The authors declare no conflict of interest.