Personalized Flipped Classroom

Depósito de Investigación de la Universidad de Sevilla https://idus.us.es/ This version of the conference paper has been accepted for publication, after peer review and is subject to Springer Nature’s AM terms of use, but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: https://doi.org/10.1007/978-981-99-0942-1_109 Personalized Flipped Classroom María Luisa Sein-Echaluce Department of Applied Mathematics, School of Engineering and Architecture, University of Zaragoza, [email protected], ORCID: 0000-0002-6873-0996 Ángel Fidalgo-Blanco Department of Geological and Mining Engineering, School of Mining and Energy Engineering, Universidad Politécnica de Madrid, [email protected], ORCID: 0000- 0003-4034-7757 José Luis Martín-Núñez Instituto de Ciencias de la Educación, Universidad Politécnica de Madrid, joseluis.marti[email protected], ORCID:0000-0002-4102-6069 Amparo Verdú Vázquez Building Technology Department, Universidad Politécnica de Madrid, [email protected], ORCID: 0000-0003-2802-5114 Laura García Ruesgas Graphic Engineering Department, Higher Technical School of Engineering, University of Seville, [email protected], ORCID: 0000-0002-1856-6550 In this research work, a Personalized Flipped Classroom model has been developed and applied, consisting of carrying out a personalized lesson at home. The model allows students to learn the lessons and carry out micro-activities based on their level of knowledge and preparation. The model is designed by establishing groups of students according to their level of knowledge and, in this way, personalized activities are designed for each group, which are carried out cooperatively and collectively. The model was applied in an engineering subject and to carry out the study of the experience, a control group with 38 students and an experimental group with 35 students were established. The results of the experience show an improvement in the performance of the students as a result of the personalized activities designed. In addition, the designed model has encouraged students to self-organize themselves by levels of knowledge in other complementary activities to the Flipped Classroom (FC) method, such as tutorials. 2 Keywords: Flipped Classroom, Micro Flip Teaching, Personalized learning 1. Introduction The Flipped Classroom (FC) innovation method has been used successfully for more than 20 years. Initially, this method arose to activate learning in the classroom and more specifically in theoretical teaching contexts. The basic approach of this method is that the students receive the lesson outside the classroom, for example, by watching a video recorded by the professor at home, and later, during the face- to-face class, carry out a set of practical activities (homework) increasing the participation of the student. This method is popularly known as “the lesson at home and the homework in class” [1]. The completion of homework implies the use of high cognitive activities and if these are done in a professor-supervised and participatory way, it makes this method a very useful tool for active methodology. Thus, FC can be considered as an active methodology [2-3]. The first published applications appeared in 2000 [4-5] in a context where the use of Internet was not yet widely implemented. Therefore, these first models were based on carrying out activities prior to class (for example, watching the video of the theoretical lesson recorded by the professor) and then carrying out participatory activities in class, although the students could ask their questions after watching the video. This model, although it lasts today, has been modified after the incorporation of the Internet. A large volume of publications has been generated on the FC method, among which those that detect improvements in learning stand out, are applied to new contexts (not only to theoretical classes) and generate new functional models. Regarding the improvements in learning, it is highlighted that the students positively value the incorporation of the FC method [6–8], since it optimizes the time spent on learning [9], increases the motivation of the students [10] and gives rise to other improvements such as: reduction in the complexity of the subject, an increase in class attendance, improvements in performance when faced with complex exam questions, or increase in peer learning [11]. The FC method began to be used in theoretical classes since in these the students usually limit themselves to listening and taking notes. Therefore, it is applied in a context where it is assumed that students have a passive attitude. However, due to the success in improving learning, this method is beginning to be used in contexts where it is assumed that students are more active. FC has also been applied in practical contexts such as laboratories, projects and practical classes [12–14]. It has also been applied to promote the development of skills such as teamwork [15]. Regarding functional models, FC has undergone a great evolution. All models maintain the initial functionality "lesson at home and homework in class". However, the set of activities and processes that define the lesson at home and homework in class have changed considerably, mainly with the incorporation of the Internet. 3 This incorporation allows new activities to be included in the lesson at home, such as, answering students' doubts, incorporating debates and even videoconference sessions with students. New processes were also added to serve as a link between the "lesson at home" and "homework in class". One of these methods is the so-called MicroFlipTeaching (MFT) where the performance of micro-activities is incorporated during the "lesson at home" to later work with the results of the same during the "homework in class" [6, 7, 14]. Likewise, one of the aspects that has contributed to revolutionizing the FC method derives from incorporating a new functionality into the model: “obtaining information on the state of student learning during the lesson at home to personalize homework in class”. With this idea, previous work has been done to promote collective intelligence [16-17] and new models of Education 4.0 [18]. This research work is part of the aforementioned idea: to obtain information on student learning during the lesson at home. Specifically, a new model is proposed that incorporates personalized learning during the lesson at home. For this, the MFT model is modified incorporating said functionality. There are already previous works where it is shown that CF favors the adaptation of the course to the students' own learning pace [19], as well as the increase in the responsibility of the students over their own learning [12]. It has also been successfully applied in contexts similar to home lessons, such as MOOCs [20-21], and its impact on students has been studied [22-23], even generating a new MOOC model [ 24]. The objectives of this research work are: • To define, develop and apply a new FC model that incorporates personalized learning in the "lesson at home". • To obtain groupings by levels of knowledge of students. • To analyze the impact of learning from the application of personalized activities at the level of knowledge of each group of students. The following sections present the proposed model, the context of application of the model, its results and the conclusions. 2. Functional Model It is proposed to create a new functionality that allows the "lesson at home" to be adapted to the level of knowledge of the students and that this functionality can be applied to any FC model. Figure 1 describes the functional application of the activities of the "lesson at home" that has the objective that the students acquire specific knowledge. It is applied to the MFT model [6, 7, 17] whose main feature within the "lesson at home" phase is that it incorporates a micro-activity associated with the acquisition of the concept. The process is described in Figure 1 with the following sequence: 4 • First, a diagnostic assessment is carried out to find out the level of prior knowledge that each student has regarding a specific topic. The activity can be carried out through one or several questionnaires and based on the answers or the score obtained, students will be grouped according to their level of knowledge (Figure 1-a). • Secondly, contents are presented (expository part of the concept) and personalized micro-activities based on the result of the diagnostic assessment: o Figure 1-b represents the expository part of the concept that you want the students to acquire. The concept is personalized to the level of knowledge of the students, for example, assigning one or another video depending on the level. o Figure 1-c represents the micro-activity adapted to the level of knowledge of the students, allowing the activity to be more or less complex. Likewise, it is possible to determine groupings according to the learning style of the students and personalize the contents offered simply by alternating the sequence between the activities that Figures 1-b and 1-c represent, without doing all the different contents. For example, if a student's learning style is pragmatic, he or she will want to experiment first before turning to theory. So, the student may be offered activity 1-c and the student will use activity 1-b to help to carry out the micro-ac- tivity. If another student has the theoretical style, first will carry out the activity that represents Figure 1-b and later the one that represents Figure 1-c. Figure 1. Personalized Lesson In this case study, the personalized lesson under the MFT model has been applied, as shown in Figure 2, with the following processes for the lesson at home and homework in class. Home lesson: • Figure 2-a. The students carry out the personalized lesson following the activities explained in Figure 1. • Figure 2-b. The asynchronous communication system, for example one or several forums, is used for students to formulate their doubts both about the concept and about the personalized micro-activity, and both professors and students can answer. It can also be used to reflect or create and participate in debates, or as a dynamic system, for example by sending a message to all students. 5 • Figure 2-c. The synchronous communication system, for example a videoconferencing system, can be used to do a collective tutorial, an important clarification or for dynamic systems, for example conducting a webinar to highlight something important that the teaching staff considers. • Figure 2-d. It represents the acquisition of information about the student's learning status. The sources of information are all the activities that have been described previously (the personalized lesson, the synchronous and the asynchronous communication). Likewise, there is another source of information that is provided by the records of the online platform used, for example dates of entry, length of stay, resources consulted, etc. Homework in class: • Figure 2-e. Activities to be carried out in face-to-face class, or synchronous online, based on the results of the micro-activities. • Figure 2-f. Masterful micro-lessons are prepared by the teaching staff based on the information obtained from the different activities that the students have carried out during the "lesson at home". • Figure 2-g. Practical activities that can be traditional, for example doing a problem, or activities organized by the groups associated with the level of knowledge, such as collective tutorials associated with the level of knowledge. In the latter case, mixed groups can be made (formed by students with different levels of knowledge) or as the professor staff considers appropriate. Figure 2. MFT Model with personalized lesson 6 3. Context The application of the model was carried out in the subject "Computer Science and Programming" corresponding to the first year of the Degree in Energy at the Universidad Politécnica de Madrid. It was applied in two programming laboratory groups, a control group with 38 students and another experimental group with 35 students. Laboratory classes are made up of 12 sessions of 2 hours each. In both groups, 5 sessions were applied under the MFT model. The personalized lesson was applied to the experimental group while it was not applied to the control group. In this work, the diagnostic assessment activity was not applied, this work was left to the choice of the students themselves so that they could choose their own level of knowledge. Both the concepts and the micro-activities were organized on three levels: • Level 1. Students with a low level of knowledge regarding the subject which they were going to work with. • Level 2. Intermediate level, students with sufficient knowledge to carry out the micro-activity and understand the concept. This level is the one assigned to the students in the control group. • Level 3. Advanced level, for students who already had prior knowledge of the subject matter to be studied. It was presented as a challenge. No dynamization activities were carried out, nor strategies to enhance student participation, since the aim was to test whether the personalized lesson learned could serve as stimulation and motivation by itself. Both in the experimental group and in the control group, during the face-to-face session, the results of the micro-activities were discussed (errors, well-resolved examples, complementary ways of doing them, etc.). 4. Results The micro-activities results were very low, only 18.5% of the control group and 22.8% of the experimental group carried out the assignments. No significant differences were found in terms of the level of participation. Therefore, it was initially considered that the experience had not been successful, since in both groups, the experimental and control, MFT was used in the same way. It was not possible to obtain information to group the students into different levels and subsequently carry out activities personalized to each level. However, in session number 4, the representant of the students in this course and at the request of the students of the experimental group, without the professors being aware of it, delivered a list of students grouped into three levels of knowledge in 7 order to carry out personalized and collective tutorials at each level. Table 1 indicates the number of students grouped in each level. Table 1. Number of students by level of knowledge Level Nº Students Level 1 4 Level 2 10 Level 3 3 The total number of students grouped by level of knowledge was 17, which represents 48.6% of the total number of students in the experimental group. For a week, collective tutorial sessions were held for each level. At that time for the control group, no tutorials were given because no one requested them. After two weeks, part of the students of the control group, informed by the experimental group, requested a collective tutoring, but they did not want to be grouped by levels of knowledge as in the experimental group. To find out if the students had been grouped correctly in the groups, the average grade of each student obtained in the first call for the laboratory exam was observed (see Table 2). Table 2. Average grade of the students by group (18 points maximum) Level Nº Students Level 1 5.75 Level 2 6.21 Level 3 13.76 In order to know the level of effectiveness of the actions of personalized tutorial action by knowledge groups, the contrast of qualifications between the experimental group and the control group is first exposed (Table 3). In the second row, the minimum qualification necessary to make average with the theoretical exams of the subject is highlighted and, therefore, it would allow him to pass it. If the mark does not reach the minimum, the students will fail regardless of the mark they obtain in theory. Table 3. Percentage of students with scores in each range and for each group Group Grupo Experimental Control Passed: 5 ≤ x ≤ 10 31.03% 31.50% Minimum: 3.5 ≤ x <5 31.03% 26.31% Failures: x <3.5 37.93% 42.10% Subsequently, the qualifications of the students who participated in the personalized actions grouped by levels (only from the experimental group) are analyzed. In this case, the percentage is reflected in Table 4. 8 Table 4. Percentage of students with scores in each range who participated in personalized tutorial Group Personalized tutorial Passed: 5 ≤ x ≤ 10 46.66% Minimum: 3.5 ≤ x <5 20% Failures: x <3.5 33.33% 5. Conclusions The objectives regarding the definition of the model and its application in a university context were met. The objective of making groups based on the level of knowledge was also met, although it has been carried out in a different way from what was previously established. From experience, it can be deduced that the inclusion of a lesson learned adapted to the level of knowledge, where the students choose their level of knowledge voluntarily, does not provide more motivation than a traditional FC method or, in this case, the MFT method. The free choice of the students to locate themselves in a specific level of knowledge does not seem to be a good option, since all the participating students both in the control group and in the experimental group have chosen the same level of difficulty, the intermediate one. Initially, they do not like to opt for extreme levels and prefer to be more conservative and stay in the middle. But the idea of organizing learning by levels of knowledge was reconsidered by the students themselves, who were grouped by level of knowledge based on the activities carried out in class with the experimental group (since the professor showed the exercises for the different levels only for the experimental group). In this sense, the impact has been greater than that of carrying out the personalized lesson since it was carried out by 20% of the enrolled students, while in the groups carried out by the students themselves, participation was 48.6 %. The grouping of the students by levels of knowledge is valid, as reflected in the average grade obtained in Table 4. The personalized activities, as initially proposed, contributed to improve the academic results of the participants. As future work, the personalized lesson will be applied with the diagnostic system so that the students do not have to choose their level of knowledge and dynamic techniques will also be applied to improve the participation of the students in the personalized FT method.