Relación entre el enfoque de aprendizaje en el rendimiento académico universitario. Un estudio de caso

- 7 - ISSN: 1133-8482 Píxel-Bit. Revista de Medios y Educación RELACIÓN ENTRE EL ENFOQUE DE APRENDIZAJE EN EL RENDIMIENTO ACADÉMICO UNIVERSITARIO. UN ESTUDIO DE CASO. RELATIONSHIP BETWEEN LEARNING APPROACH IN THE UNIVERSITY ACADEMIC PERFORMANCE. A CASE STUDY. Dra. Ana Isabel Vázquez-Martínez [email protected] Universidad de Sevilla. Escuela Técnica Superior de Ingeniería de Edificación. Departamento Construcciones Arquitectónicas-II. Avda. Reina Mercedes, nº 4, 41012, Sevilla (España) En este artículo se analiza si existe relación entre el enfoque de aprendizaje y el rendimiento académico de los alumnos, mediando como variable independiente el empleo de WEbCT. Se realiza en la Escuela Técnica Superior de Ingeniería de Edificación (ETSIE), de la Universidad de Sevilla, sobre una muestra de 176 alumnos. En función de las necesidades del análisis estadístico se recurrirá a la t de student, t de Welch, U de Mann-Whitney, ANOVA (F de Snedecor o F de Welch) o la prueba de Kruskal-Wallis. Los resultados obtenidos indican que no existe correlación entre el enfoque y el rendimiento académico, con docencia mediada por plataforma educativa. Palabras clave: rendimiento académico, WebCT, enfoque de aprendizaje. This article analyzes the correlation between the learning approach and the academic performance of students, the use of WebCT is an independent variable. The study is done in the Building Engineering School (ETSIE) at the University of Seville, the sample size is 176 students. Depending on the statistical analysis the t test, Welch's t, Mann-Whitney, ANOVA (F of Snedecor or Welch F) or Kruskal-Wallis test are used. The results do not indicate a correlation between the learning approach and the academic performance, the teaching methodology includes a virtual learning platform. Keywords: academic performance, WebCT, learning approach. Nº 42 Enero 2013 - pp. 7-21 ©2013 Píxel-Bit. - 8 - Vázquez-Martínez, A. I. Píxel-Bit. Revista de Medios y Educación 1. Introduction. Coll (2007, p. 179) indicates that learning is a process in which «students learn, the contents that are object teaching and learning, and teacher who helps students to build meaning and attribute meaning to they do and learn». We wanted to start this chapter with this significant paragraph we understand is vigorously related to the implementation of the European Higher Education Area. Biggs (1999, p. 62) suggests that learning is «a way of interacting with the world. As we know, our conceptions of things and phenomena change and see the world differently. The acquisition of the information, in itself, does not provide such changes but so does the way we structure this information. Thus, education is concerning that conceptual change, not the acquisition of information». The Lisbon Declaration (2007) suggests that universities should move towards a teaching model, in which the student is the center of the learning process. This is enhanced when it asserts that those processes must adapt to highly variable needs that students present, and that must also ensure full accessibility to the means available. The study of learning approaches has been used, in its different versions, to learn, diagnose and solve conflicts of learning as either single instrument or in combination with others. Hernández Pina (2005, p. 29) defines learning approaches as «intentional phenomenon of the individual that is directed towards the world around him. It is not something that is within the student, but rather is set to how the student learning experience». To Corominas, Tesouro and Teixidó (2006, p. 446) the «learning approaches arise from consideration synergistic content area of study, the demands of context and awareness that students are learning». And also indicate that «learning approaches include the individual (genetic, cognitive style, prior experience) and the situation in which the behavior occurs. A person in a context is not simply the sum of the person and the context» (p. 446). Marton and Svensson (1979) argue that the approaches focus on the characterization of the conceptions about the student interprets the world around him, especially the content and the learning context, and the act of leaning in itself (p. 471). Therefore Entwistle and Tait (1996) understand that this new approach provides a valid conceptual framework for understanding the differences in student learning. And to Duff (2004), they try to understand the processes of learning that students continue to evaluate the learning experiences of these and the way they face, give meaning to them. To Ramsden (1992) the approaches are associated with how and what students learn rather than how much they learn. In this respect we consider the contribution of Valle Arias, Cabanach Gonzalez, Núñez Pienda and Gonzalez-Perez (1998), who after a detailed study suggest that: The perceived capability influences both internal and causal attributions in academic self-concept, while an incremental conception of intelligence should influence the use of a deep learning approach. At the same time, prior attainment on academic self-concept influences on causal attributions and current academic outcomes. Furthermore, it is argued that the perception of the evaluation criteria, the type of subject, teaching style, and task characteristics affect learning approaches (p. 397). - 9 - ISSN: 1133-8482 Píxel-Bit. Revista de Medios y Educación The first that employed the construct approach were Marton and Säljö in 1976 with university students in Sweden, employing the terms of deep approach and shallow approach to learning to refer to the way in which students faced reading research articles from the focus of qualitative approach developed from a phenomenographic orientation. Hernandez Pina, Garcia Sanz Martinez Clares, Hervas Avilés and Maquilon Sánchez (2002, p. 490) submits that «a Shallow approach to learning is clearly at odds with the objectives and principles of what should be a university education». Selmes (1988, in Pozo, 1996, p. 207) establishes the characteristics for the shallow approach that set out in Table 1. The features that Selmes attributed to the deep approach are gathered in the Table 2. However, Kember (in García Berbén, de la Fuente, Justicia & Pichardo, 2005, p. 259) states that the focus of the students belong to a continuum, in which the shallow and the deep focus occupy the ends. · Shallow: extrinsic motivation. Memoristic strategy. Mechanical learning. Quantitative conception of learning. · Intermediate 1: motivation fundamentally extrinsic. Memoristic Strategy, but uses the understanding to facilitate it. Quantitative conception of learning, but considers that is necessary not just memorization but understanding must be involved in a lesser extent. · Understanding and memorization: intrinsic and extrinsic motivation. Initially seeks to achieve understanding, but with the aim of memorization. Qualitative conception of learning, is in a state of equilibrium between memorization and meaning construction. · Intermediate 2: intrinsic motivation. Uses memory strategies, but after having understood the material. Qualitative conception of learning. · Depth: student motivation is intrinsic, strategies are aimed to understanding the task. The student understands that learning is reached with the construction and revision of the material to be learned. Table 1. Features of shallow approach according Selmes. Source: Well, 1996, p. 207 Isolation It focuses on the procedural elements of the task Tendency to treat the material as if it's isolated from other materials Considers that the task consists of discrete part It focuses on tasks elements Memorizing Consider that the context of the task requires memorization of material The student defines the task as a memory task The student states his intention to memorize the material Passivity The task is defined by another person Indicates a thoughtless or passive approach of the task Indicates teacher dependence Try the material externally - 10 - Vázquez-Martínez, A. I. Píxel-Bit. Revista de Medios y Educación 2. Methodology. 2.1. Objectives and hypotheses. The main objective of this research is to Test the relationship between dominant learning approach of students and their academic performance. As have employed two different teaching methods, you can set two subgoals: · Check the relationship between dominant learning approach of students and their academic performance during the first semester of the course. · Check the relationship between dominant learning approach of students and their academic performance during the second semester of the course. Hypothesis Hypothesis 1: There are differences in the performance of students in the subject Materials 1 in function of the dominant learning approach. Since different methodologies used in the two semesters we formulate the following sub-hypotheses. Sub-hypothesis 1.1: About the performance in the first quarter: H0: No significant differences in student achievement in the subject Materials 1 during the first quarter with teaching methodology based on the use of WebCT learning platform, depending on the dominant learning approach. H1: Significant differences in student achievement in the subject Materials 1 during Table 2. Features of deep approach according Selmes. Source: Well, 1996, p. 207 Personal integration • Intention to create a personal interpretation of the material • Emphasizes the importance of personal interpretation comparing with those of another person • Indicates the desire to relate the task to the personal situation outside the immediate context • Intent to link ideas and personal experiences with the topic of the task • Indicates the desire to link the task / concept with everyday situations • Consider the task as a part of personal development Interrelations • Intent to connect the parts of the task each other • Intent to relate the task with other relevant knowledge • Relate what you know about another problem with a new problem • Match the previously studied materials with new materials or new materials with future materials • Intent to relate material from different sources • Think proactively in the relations between the parts of the material • Try to relate the aspects of a problem Transcendence • Intention to focus on the meaning of the content • Intent to think about the underlying structure of the task • Try to use some of the material to represent all, or a text to represent a kind of text - 11 - ISSN: 1133-8482 Píxel-Bit. Revista de Medios y Educación the first quarter with teaching methodology based on the use of WebCT learning platform, depending on the dominant learning approach. Sub-hypothesis 1.1: About the performance in the second quarter: H0: No significant differences in student achievement in the subject Materials 1 in the second quarter with traditional teaching methodology without using the WebCT learning platform, depending on the dominant learning approach. H1: Significant differences in student achievement in the subject Materials 1 in the second quarter with traditional teaching methodology without using the WebCT learning platform, depending on the dominant learning approach. 2.2. Tool. It has been used Questionnaire Revised Study Process (R-CPE-2F), adapted by Recio Saucedo (2004, 2007) of inventory R-SPQ-2F (Bigs, Kember & Leung, 2001). The goal of this questionnaire is to identify the predominant learning approaches in students from the exhibit, namely: shallow and deep. Each of the approaches, while two subscales comprises as follows: · Deep Approach: DA = DM (deep motivation) + DS (deep strategy) · Shallow Approach: SA = SM (Shallow motivation) + SS (surface strategy) The questionnaire is resolved from a Likert type scale of five possible answers, and each is associated with a score of 1 to 5: never or rarely (1) Sometimes (2), half of the time (3), often (4) always or almost always (5). To determine the reliability of the questionnaire of learning approaches R-SPQ2F was calculated Cronbach’s alpha coefficient for each of the two sets of 10 items that are diagnosed with each of the two learning approaches of the main scale, shallow approach and deep approach, and each of the four sets of five items that diagnose the four learning approaches subscale, namely, shallow-motive, shallowstrategy, deep-motive, deep-strategy. The coefficients obtained for each approach are set out in Table 3. 2.3. Population and sample. The population consists of 315 students enrolled in four of the ten groups in the first year of the Higher Technical School of Engineering Building at the University of Seville. The sample consisted of 176 students (55.87% of the population), who are the students who completed the questionnaire of learning approaches R-SPQ-2F of Biggs. The sample was chosen in an incidental way, that is not randomly, since it aims to achieve contextual information so that the results found in the same reverse (Gil Flores, Rodriguez Gomez & García Jiménez., 1995, p. 224), or opinion sampling method (Sabariego, 2009, p. 148), or intentional (Cohen & Manion, 1990, p. 139). But also understand that it can be considered a convenience sampling DEEP MOTIVE DEEP STRATEGY SHALLOW MOTIVE STRATEGY SHALLOW DEEP APPROACH SHALLOW APPROACH .478 .568 .609 .575 .68 .739 Table 3. Cronbach’s alpha coefficient by approach - 12 - Vázquez-Martínez, A. I. Píxel-Bit. Revista de Medios y Educación (Cohen & Manion, 1990, p. 138), or causal or accessibility (Sabariego, 2009, p. 148) simultaneously because informants are individuals closest to the formative action, in order to report on it, and know their organismic characteristics. 2.3.1. Sociodemographic profile of the initial sample. The sample for the study of learning approaches was constituted by 176 students who filled in the questionnaire of learning approaches R-CPE-2F, of which 60 (34.09%) were women and 116 (65.91%) men. In graph 1 shows the distribution of students by group and gender. 2.3.2. Identification of the approaches. Recio Saucedo (2004, p. 99) proposes determining the intensity of the approach on the basis of the difference between the scores that a student obtains between deep and shallow approach. The idea is that as the minimum score that can be obtained in each approach (deep or shallow) of the main scale is 10 (10 items that may have a minimum score of 1) and the maximum score that can be obtained in each approach is 50 (10 items that can have a maximum score of 5), that the smallest difference that can exist between scores on each approach is 1 and the largest possible difference is 40. In this way if the difference between the scores of approach is between 1 and 13 points is considered low intensity if between 14 to 26 is considered medium intensity, and if it is between 27 and 40 is considered high intensity. With these precisions, the assignment to approach learning in the main scale for the students reflects the following distribution (Table 4) based on the dominant approach (approach with the highest score): Predominantly we find deep approach students (139) versus shallow approach (30), and there are 7 students with the same score Graph 1. Sample distribution by gender and group  Grupo 1 Grupo 2 Grupo 3 Grupo 4 38 31 35 12 16 19 15 10 0 5 10 15 20 25 30 35 40 número alu mno s mujeres hombres - 13 - ISSN: 1133-8482 Píxel-Bit. Revista de Medios y Educación in both approaches so do not ascribe to any of them. As the intensity is observed that in both approaches highlights the low intensity level. 2.4. Data analysis. When you employ as grouping variable for contrast the learning approach dominating the main scale, which has two categories, we will use the Student’s T, Welch’s T, or MannWhitney’s U depending on whether or not comply the cases of normality and homoscedasticity. And when we employ as grouping variable to carry out the contrast the dominant learning approach subscale, which has four categories that will be used ANOVA (Snedecor’s F or Welch’s F, depending on the course of homoscedasticity) or the Kruskal-Wallis depending on supposition of normality. In all cases it is intended to determine whether there are significant differences in the means of scores on the subject of Materials 1, depending on the dominant approaches aforementioned for a significance level of 95% (á = .05). 3. Results 3.1. Influence in the performance with the use of WebCT platform It first analyses the potential influence of the categories of the main scale learning Table 4. Distribution of the dominant learning approachess. LEARNING APPROACHES N. STUDENTS % APPROACH INTENSITY High Med. Low Deep 139 78.98 2 42 95 Shallow 30 17.04 0 3 27 Undefined 7 3.98 Table 5. Normality test for the qualifications of 1st quarter (1C) by main scale approach. 1st QUARTER QUALIFICATIONS APPROACH KOLMOGOROV-SMIRNOV Stadistic gl significance Did. Unit 1 Shallow .108 25 .200* Deep .101 124 .003 Did. Unit 2 Shallow .189 22 .039 Deep .207 117 .000 Did. Unit 3 Shallow .129 20 .200* Deep .045 111 .200* mid-term-1 Shallow .132 26 .200* Deep .116 130 .000 *. This is a lower limit of the true significance. - 14 - Vázquez-Martínez, A. I. Píxel-Bit. Revista de Medios y Educación approaches. The normality test of Kolmogorov-Smirnov (Table 5) indicates that the distribution is not normal have any categories of the grouping variable «Focus» with p <.05 in the qualifications of the didactic unit 1, didactic unit 2 and mid-term 1. While grades in the didactic unit 3 you meet the assumption of normality from having both the category «shallow» and category «deep» p values p > .05. The fulfillment of the normality case on the ratings of glass allows use Student’s t-test (Table 6), while for the qualifications of the didactic unit 1, didactic unit 2 and mid-term-1 will have to use the non parametric MannWhitney U test (Table 7). The Levene test confirms that is fulfilled homoscedasticity case of to be its significance p = .215> .05 leading to choose the Student t test (equal variances were assumed) as a more appropriate alternative analysis to compare the average grade both groups. Consequently, there are significant differences in ratings of the didactic unit 3 between students shallow dominant approach (X = 4.1150, S = 2.19884) and the deep dominant focus (X = 4.9808, S = 1.65567), t (129) = 2041, p = .043, with a small effect size d = .3594. The non parametric Mann-Whitney U indicates no significant differences between shallow and deep approaches on ratings of the topics: Didactic unit 1, U = 1274, p = .161; Didactic unit 2, U = 1227.5, p = .731, mid-term1 U = 1363.5, p = .121. Analysed the influence of main scale categories of learning approaches, secondly is studied the possible influence of the categories of the subscale on the learning approaches. The normality test of Kolmogorov-Smirnov (Table 8) shows that the distribution is not normal to have some category of for the grouping variable «Approach» with p <.05 in the qualifications of the didactic unit 1, didactic unit 2 and midterm 1. While on the ratings of the didactic unit 3 If it is fulfilled the supposition of Table 6. Student’s T test for glass qualifications grouped by main scale approach. Table 7. Mann-Whitney’s U test 1C qualifications grouped by main scale approach. Levene test T-test for equality of averages F Sig t gl significance (bilateral) Did. Unit 3 Assumed equal variance 1.552 .215 -2.041 129 .043 Not assumed equal variances -1.677 23.039 .107 Test Statistics U de Mann-Whitney W de Wilcoxon Z significance Did. Unit 1 1274.000 1599.000 -1.402 .161 Did. Unit 2 1227.500 1480.500 -.343 .731 mid-term-1 1363.500 1714.500 -1.553 .121 - 15 - ISSN: 1133-8482 Píxel-Bit. Revista de Medios y Educación normality to take into the four categories of the subscale values for p> .05. The fulfillment of the normality assumption on the ratings of glass allows us to use the analysis of variance (ANOVA) (Table 9), while for the qualifications of the didactic unit 1, didactic unit 2 and mid-term one will have to use the non parametric Kruskal-Wallis test (Table 10). The test of homogeneity of variances (Levene test) ANOVA on the ratings of the didactic unit 3 gave a significance p = 118 which when greater than ,05 indicates that the supposed of homoscedasticity meets, so we use Snedecor’s F from the ANOVA. Consequently, no significant differences in ratings of the didactic unit 3 between students of the four dominant approaches the subscale F (4.130) =. 697, p = .595. The non-parametric Kruskal-Wallis test indicates no significant differences between the four approaches subscale scores on topics: Did. Unit 1, X2 (3, N = 137) = 2529, p = 0.470; Did. Unit 2, X2 (3, N = 126) = 1232, p = 0.745; Part-1 X2 (3, N = 143) = 1814, p = .612. The tests do not confirm our hypothesis in any of the grades, except in glasses by main scale approaches. Accordingly, the null 1st QUARTER QUAL APPROACH Kolmogorov-Smirnov Stadistic gl significance Did. Unit 1 Shallow / Motive .164 12 .200* Shallow / Strategy .185 14 .200* Deep / Motive .102 90 .023 Deep / Strategy .195 21 .036 Did. Unit 2 Shallow / Motive .174 10 .200* Shallow / Strategy .218 13 .092 Deep / Motive .209 84 .000 Deep / Strategy .204 19 .036 Did. Unit 3 Shallow / Motive .169 10 .200* Shallow / Strategy .137 12 .200* Deep / Motive .058 78 .200* Deep / Strategy .105 19 .200* mid-term-1 Shallow / Motive .225 12 .096 Shallow / Strategy .141 15 .200* Deep / Motive .115 93 .004 Deep / Strategy .212 23 .009 *. This is a lower limit of the true significance. Table 8. Normality test of 1st quarter qualifications by the subscale approach. Levene test ANOVA Snedecor Welch F Sig F gl-1 gl-2 Sig F gl-1 gl-2 Sig Did. Unit 3 1.881 .118 .697 4 130 .595 .457 4 29.494 .767 Table 9. ANOVA for qualifications of glasses grouped by subscale approaches. - 22 - Vázquez-Martínez, A. I. Píxel-Bit. Revista de Medios y Educación