Archivo Digital UPM: No conditions. Results ordered -Date Deposited. 2021-12-03T18:47:12ZEPrintshttps://oa.upm.es/style/images/logo-archivo-digital.pnghttps://oa.upm.es/2013-10-02T12:49:40Z2016-04-21T15:23:31Zhttps://oa.upm.es/id/eprint/15329This item is in the repository with the URL: https://oa.upm.es/id/eprint/153292013-10-02T12:49:40ZCuestionarios online como herramienta de aprendizaje: comparación de resultados con diversos modelos de cuestionarios.En este trabajo se comparan dos estrategias de evaluación a través de cuestionarios online: i) realizar un cuestionario de evaluación sumativa al final de cada bloque temático del curso o ii) permitir que, para cada bloque temático, el alumno pueda repetir varias veces un test (con realimentación de las respuestas correctas) y realizar un test similar después de esta fase de aprendizaje. Los resultados muestran varios beneficios de la segunda estrategia: una mayor participación de los alumnos hasta el final del semestre, una menor tasa de abandono de la asignatura y una mejor nota en los test. Sin embargo, para que el aprendizaje sea efectivo en el caso de repetir varias veces un test, se necesita un banco de preguntas suficientemente amplio para que exista una probabilidad baja de que las preguntas no se repitan. Este problema se ha resuelto utilizando la posibilidad que presentan algunas plataformas de gestión del aprendizaje de generar automáticamente variantes numéricas de una misma pregunta. Por último se observa un grado de acuerdo muy elevado por parte de los alumnos con el peso que los resultados de los tests tienen en la nota final del curso, en el caso bajo estudio.
Abstract: This work compares two different assessment strategies through online tests: i) summative assessment tests at the end of each part of the course or ii) the students can do selfassessment tests (with feedback) prior to a summative assessment test at the end of each part of the course. Results show that the second strategy yields several benefits: an increment in student participation till the end of the semester, a reduction in the drop-out rate and an increase in the mean test marks. However, a large item bank is necessary to effectively implement this second strategy. This problem has been addressed by using the automatic item generation facility offered by several learning management systems. Finally, a high proportion of the students agree with the weight proposed for these tests in the final course mark.Juan Carlos González de SandeJuan Ignacio Godino LlorenteLuis Arriero EncinasJuana María Gutiérrez ArriolaDavid Osés del CampoVictor Osma RuizRubén Fraile Muñoz2011-11-02T10:00:25Z2016-04-20T17:56:44Zhttps://oa.upm.es/id/eprint/9580This item is in the repository with the URL: https://oa.upm.es/id/eprint/95802011-11-02T10:00:25ZCheating and learning through web based testsThe use of web-based tests delivered through learning management systems has grown at university level in the last years. One of their key advantages is the possibility of creating tests with some degree of randomness that are automatically assessed in real time. Although the access to the learning management system resources is controlled for each student by means of personal username and password, the cheating among students when doing the tests cannot be avoided. However, if the students finally learn, in spite of cheating, the process could still be considered to be successful. In this work, the date, the required time to solve the test and the grades of quizzes undertaken by students through a web based learning management system are analyzed and they are compared to the grades obtained by the same students in a written test solved in an examination classroom under the supervision of the teacher. The course in which this study has been developed (Signals and Systems for Electrical and Electronics Engineering undergraduate students) is organized in 5 subjects and the students make a quiz on the web for each subject. At the end of the course the students make a final written exam that includes a true/false test. Around 50 questions for each subject of the course have been created. The questions are organized in 5 to 8 categories for each subject. The learning management system generates quizzes by arbitrarily selecting 1 or 2 items from the 5 to 8 categories in a given subject to complete a 10-item quiz. Due to the reduced number of items for each category and the large number of students that attend the course, several questions are repeated in quizzes generated for different students. The authors have noticed that some students work in groups to solve the quizzes. Some of them answer all the questions in a quiz in few minutes (less than 20 % of the time used by the most of their mates) and obtain high scores. When the scores of the same students in the final exam are analyzed, it is found that they also obtain good results. Then, it could be concluded that although they have found a way of cheating to solve the web quizzes, this is still pedagogically valid because they have learnt about the subject (they also obtain good results in the written test).Juan Carlos González de SandeRubén Fraile MuñozLuis Arriero EncinasVíctor Osma RuizDavid Osés del CampoJuan Ignacio Godino Llorente2011-07-27T09:39:58Z2016-04-20T16:51:34Zhttps://oa.upm.es/id/eprint/7909This item is in the repository with the URL: https://oa.upm.es/id/eprint/79092011-07-27T09:39:58ZAn error based mathematical module to enhance learning in signals and systemsDuring the last years, the lecturers at the Circuits and Systems Engineering Department at the E.U.I.T. de Telecomunicación at the Universidad Politécnica de Madrid are observing more and more serious mathematical errors in the different exams and exercises taken by the students. Although some of these mistakes can be considered unacceptable in engineering disciplines, it is possible for a student to pass the final exam regardless of these mistakes. In this scenario, and aware that results were getting worse and worse year after year, it was considered convenient, and almost indispensable, to develop math exercises that students must practice if they want to progress following a continuous and formative assessment method along their engineering studies. The first part of this work is to analyze basic mathematical errors in final exam exercises of the course “Signals and Systems”. We present and illustrate the most relevant errors detected during the last two years final exams of that course. The information obtained permits us to identify the main lacks, difficulties and defaults of the students. The second part of this work is to develop a training module in order to the students can practice as many times as they want with simple exercises dealing with the topics where frequent errors are detected. After practicing they must pass an initial test to make sure that students have acquired the adequate basic mathematical background and skills to progress successfully in the mentioned course. The questions and exercises have been written using different formats, most of them to be compatible with Moodle platform requirements.Luis Arriero EncinasDavid Osés del CampoVíctor Osma RuizJuana María Gutiérrez ArriolaJuan Carlos González de SandeRubén Fraile MuñozCésar Benavente Peces2011-07-27T08:03:03Z2016-04-20T16:51:51Zhttps://oa.upm.es/id/eprint/7916This item is in the repository with the URL: https://oa.upm.es/id/eprint/79162011-07-27T08:03:03ZA proposal for the evaluation of Final Year Projects in a competence-based learning frameworkThe authors propose a system for the assessment of Final Year Projects (FYPs) whose educational outputs have been defined previously in terms of competences. For building the proposal, eleven pre-defined competences were ranked and a different weight was assigned to each one. The ranking was made individually by all the authors following a blind two-step process. The first step consisted in ordering the competences by relevance and the second step in grading that relevance for each pair of competences having consecutive positions in the list. As a result, an overall weight was computed for each competence and the final proposal was produced by averaging the individual proposals. In addition, three moments are defined for the assessment of FYPs: the FYP process itself, the written report and the oral presentation. Bearing in mind this, the competences that can be evaluated in each moment are identified and a specific assessment form for each moment is also proposed.Rubén Fraile MuñozIrina Argüelles AlvarezJuan Carlos González de SandeJuana María Gutiérrez ArriolaCésar Benavente PecesLuis Arriero EncinasDavid Osés del Campo2011-03-28T08:28:31Z2016-04-20T13:03:28Zhttps://oa.upm.es/id/eprint/3499This item is in the repository with the URL: https://oa.upm.es/id/eprint/34992011-03-28T08:28:31ZA case study: Final exam versus continuous assessment marks for electrical and electronic engineering studentsIn this work we analyze the final marks obtained by the students of the course “Signals and Systems” of the Electrical and Electronic Engineering degree at the E.U.I.T. de Telecomunicación in the Universidad Politécnica de Madrid. Students within this course are assessed by two different methods: the first one consists of only a final exam while the second one implies students following a continuous and formative assessment method in which they are required to solve some open exercises almost every week and to fulfil a self-evaluation test every two weeks all along the whole semester. During the academic year 07/08, all the students of the course were assessed by both methods. Herein, we analyze the final marks (in a 0-10 scale) obtained by the students and compare the pass and fail rates obtained with both assessment methods when the minimum mark for passing is set to 5. This work was done with a sample of 210 students that were divided in 7 different groups with different lecturers. We have found that more than 70 % of the students where classified in the same group, either pass or fail, with both assessment methods. As for the rest, 15 % of the students failed with the continuous assessment method but succeeded in the final exam and the remaining 15% passed in the continuous assessment and failed in the finals. For these two groups, observations indicate that the first one corresponds to students that either deliberately chose to be evaluated only with a final exam, thus not presenting the required coursework, or presented medium quality works (marks around 4) and did a greater effort for preparing the exam. As for the last group, it should be highlighted that their effort during the course allowed the majority of them to obtain marks above 3.5 points in the final exam. Based on these results, we made simulations so as to have an insight in what would be the student’s final marks when both assessment methods were combined with different weightings. Nowadays, this kind of combination is a common practice in many Universities, but usually the final exam has a higher weight (60 to 80 %) in the final student’s marks. We found that even if the final exam’s weight was only 25% and the continuous assessment’s weight was 75%, only 3% of the students that had final exam marks lower than 3.5 would obtain a weighted mark over 5.0. Consequently, we conclude that continuous assessment gives practically the same pass/fail rates as the final exam.Juan Carlos González de SandeLuis Arriero EncinasCésar Benavente PecesRubén Fraile MuñozJuan Ignacio Godino LlorenteJ. GutierrezDavid Osés del CampoVíctor Osma Ruiz2010-06-28T10:17:12Z2016-04-20T12:55:00Zhttps://oa.upm.es/id/eprint/3374This item is in the repository with the URL: https://oa.upm.es/id/eprint/33742010-06-28T10:17:12ZMimo Systems Low complexity SVD Implementation AnalysisThis paper analyses the implementation of the singular value decomposition (SVD) using approximation to the exact computation for MIMO systems in the case of modulation-mode and power assignment set-up. The study developed in the paper focuses on the use of low complexity algorithm with low computational load oriented to the use of devices with limited resources as FPGA, highlighting some of the advantages and drawbacks against more sophisticated devices. The implementation of the SVD is analyzed through the algorithms that efficiently perform the required computations, seeking for computationally efficient solutions that provide parallelism and low complexity. The CORDIC algorithm seems to be a good candidate for this task since it can efficiently compute the singular value decomposition. It is shown that this algorithm provides an efficient tool for SVD computation with appropriate accuracy and the computational complexity obtained and the required resources make it feasible to be implemented on an FPGA device. System performance degradation is analyzed compared with conventional and exact method for SVD obtaining some key conclusions.César Benavente PecesLuis Arriero EncinasDavid Osés del CampoFrancisco Javier Ortega GonzálezJosé Manuel Pardo Martin