ADAPTIVE LEARNING METHODOLOGY FOR UNDERGRADUATE LEARNING OF STATISTICS AND PHYSICS
Elena Evtimova1, Temenoujka Fuller2
1Chair of Natural Science, Physics Section,
Department of Language Teaching International Students – IFS
at Sofia University „St. Kliment Ohridski“
27, Kosta Loulchev St., 1111 Sofia, Е-mail: aloniar@abv.bg
2 Cooperative Learning Center,
Central Arizona College, USA, Е-mail: temenoujka.fuller@centralaz.edu
Abstract: Contemporary computer technology integration within the process and structure of education changes the methods from presenting or lecturing to designing learning environments with a self-directed learning management as a strategy to direct students' learning independence by using multi-dimensional ongoing self-assessment. Educators use guided questions to direct students' attention. In this study, a set of learning quizzes for physics and statistics are presented and briefly discussed. Different forms of objective tests, such as multiple choice, matching response, true/false quizzes, short essay, or filling-in-the-blank, are used to map an infrastructure to guide and facilitate students’ independent learning on the key details of the chapters in physics and statistics. Students in statistics worked individually in a time that was convenient for them. An immediate feedback supported students' independent study and minimized the misconceptions. Physics quizzes utilized Internet links and Java applets to support the learning process. Final evaluation shows that the average achievement is improved with more than one standard deviation compared to the reading pretests conducted in the self-learning phase of the project. Also, according to a qualitative analysis conducted with experimental samples, students' commitment, personal involvement and active self-directed learning from complex technical texts improved considerably.
INTRODUCTION
Adaptive Learning Systems and computer based education. For the fiscal year 2005, the US industry of instructional technologies expected to realize between $42 and $142 billion from educational products [1]. Unfortunately, a Meta-analysis developed in [2] to unify the results of educational research on the impact of instructional technologies on learning, shows only a modest positive effect which is not cost effective. Thus, the National Institute of Standards and Technology of the USA [3] calls for a specific new approach that goes by the name of Adaptive Learning Systems (ALS). ALS should be precisely tailored to the needs of learners and educators. The program seeks comprehensive solutions which recognize that, in order to properly address user requirements, instructional technologies must be more flexible and scalable with respect to all fundamental aspects of net-centric, Web-based instruction – content, delivery, search, and quality of service [4].
This paper presents research-based development of asynchronous learning instrument developed to fit the learning demand of a diverse students' population. The learning demand for particular learning services as it is defined in [5] and [6] is measured by the time invested by students for out-of-the-classroom learning with technologies or individualized instruction. Design, development and implementation of ALS, as well as any other type of educational products, will require experimental and theoretical support by Type 1 and Type 2 developmental research [7]. Type 1 developmental research is characterized by a research based development of prototypical products which implies inclusion of empirical evidence of their quality. The Type 2 research is for generating methodological directions for the design and evaluation of such products and hence the investigation is oriented toward a general analysis of design, development, and/or evaluation processes. This paper presents Type 2 developmental research enriched with traditional studies conducted in the Department of Language Teaching International Students (DLITS), Sofia University, Bulgaria and The Learning Center of Central Arizona College, the USA. In 2001, this approach was developed and applied since then for chemistry classes and later in 2005 for statistics classes with objective tests available online. During 2005 and 2006 guided quizzes were successfully used with students enrolled in physics classes in DLTIS. A blended technology for development of learning supplements to guide students' learning was used successfully for undergraduate statistics class.
EXPOSITION
1. Transition from Customized Learning to Adaptive Learning Systems. It is unlikely that the producers of software products for education will discover adaptive learning systems without research. Educators need to address the demand for adaptive instruments for simulations research of students' learning needs as a real distribution in the time and to approach the development of learning by customization, observation and new customization. The trend of new educational system is to be oriented toward the individual learning demands and so the teacher needs to readjust to the idea about adaptive learning. One way to approach this problem is by paying close attention of students as individuals instead of stereotyping them with standardized tests after standardized instructions. Individualized learning needs are critical conditions for control and development of supportive environments for students' achievement. Bloom [6] found that "students who are individually tutored learn twice as much as students instructed in the conventional manner." This study is an indication that research conducted in tutoring centers or learning departments will provide qualitative and quantitative information for design and development of Adaptive Learning Systems (ALS).
The time invested in learning with instrument developed for discipline specific learning skills reflects students’ desire to utilize learning support, such as one-on-one tutoring, computer-assisted self-tutorials, or working in cooperative learning settings. Learning time invested by the students in a learning center or online/computer based tutorials is observable and measurable variable used in a longitudinal study on learning demand [8]. This study provides a baseline for personalized learning systems for self-directed in or out of the classroom study with or without academic support. In the old traditional educational research, the individual learning needs were averaged and the average value was stated as a descriptive measure of students needs. Learning needs were presented by a "subject-matter expert approach" or "performance technology approach” [9]. For the teacher-centered traditional classroom, the main building block was any systematic instructional design tailored to the average measure of the learners needs. Students in traditional classroom were considered responsible to “get” whatever the designers build into the curriculum. The goal of modern education is to change teacher-centered approach with adaptive learning systems.
Multiple studies on learners' needs have shown the importance of cognitive learning styles, affective and behavior orientations, and emotional differences. A meta-cognitive cooperative learning [10] and a scaffolding learning approach applied to the zone of proximal development [11], [12] are only two examples for cognitive approach to classroom practices. The Vygotskian notion [11] of the zone of proximal development often abbreviated ZPD, is the gap between a learner's current or actual developmental level determined by independent problem-solving ability and the learner's emerging or potential level to solve problems with appropriate help. The notion of ZPD implies that a student's development is determined by the social interaction and collaborative problem-solving opportunities. Learning is inherently a social, dialogical process. That is, given a problem or task, people naturally tend to find more perspectives by discussing the problem with others. Technologies can support this conversational process by connecting learners. In this paper, the technology-based design of online or computer-based supplements is used for students to try problems outside of the classroom. The conversation in the classroom is after students had sufficient time to try the problems. Students work with their personal speed of solving problem, and later on in the classroom, the instructor can engage learners in mindful processing of information without the intervention of formal instruction, more like a facilitator of the process.
The affective and behavioral components of students' personality had been excluded from the cognitive studies for years. One of the first researches on the emotional and volitional components of learning, presented in [13] and [14], is calling for a multidimensional assessment, and this would be a step towards desirable Adaptive Learning Systems (ALS). The purpose of ALS is to guide students to becoming highly self-motivated, self-directed, intentional and independent learners [14].
In this paper, the individual assessment is presided by individualized instruction and self-assessment in new learning facilities called learning centers or learning departments. Our goal is to conduct face-to-face action research on student learning development embedded into the process of education as needed and not only in the terminal points (at the beginning and the end of the semester). At the moment, the development of adaptive learning system requires intensive research in specialized centers. In this paper, the development of adaptive instruments for simultaneous research and individualized learning environment are presented and discussed.
2. Multidimensional Assessment - Mapping Meaningful Ideas. The paper is based on the long-running study on multidimensional assessment and evaluation.
Mapping Meaningful Ideas is a learning method for directing students to analyze, visualize, relate and make sense of information. In a meaningful learning, supporting instructor creates opportunities for all students for acquiring, interpreting, analyzing, evaluating and manipulating information. Instructors help in the Zone of Proximal Development and allow students to do the rest of the learning individually (Figure 1).
To lead each student, the instructor needs information and feedback. Online supplements provide opportunities to work independently with each student by flexible design of activities: Socratic discussion boards, mind mapping, quizzes, and projects. The exact set of online supplements depends on the distribution of students' personalities [14], and the distribution of students learning skills. In some cases the instructor may facilitate an easy example to help some students in the zones of proximal development; however, the computer-generated statistics will prevent instructor from over-patronizing students and doing the work for them. There is time for instructor to present and to lecture; the difference is in proportion of traditional teacher-centered classroom compare to learning classroom in which students discuss, discover, build new knowledge, creatively work problems together. The level of judgment is minimized to provide opportunities for self-fulfillment and learning independence.
The individualized learning curve presented on Fig. 1 is selected arbitrarily. The distribution of learning curves may be very complex and dynamical. As far as instructor keep in mind the differences and provide opportunities and guidance, students will develop skills and knowledge body. Support and understanding of students' differences are powerful tools for adaptive learning systems. The adaptive learning instruments serve as measurement of students' needs and feedback. They save time for instructor to focus on students' personality profile, the management of learning, and the best sequencing of material.
Meaningful learning means that for each task, instructors of the future will consider students' learning preferences, represented as Transforming/Conforming domain, and will select learning strategy to guide students towards the real problems. Students have different learning curl for different problems. The goal of the education is to move from artificial tasks to real life task no matter how complex they are.
On the picture above, every student has a predisposition to be educated in a certain way – some would like mentoring and deductive education, some would prefer step by step and concrete presentation. Everybody is comfortable with certain teaching method. Students perform well in their personal (natural) mode. On the other side of the learning equation is the learning task. Each learning task requires certain skills and knowledge body.
The difference between students learning equilibrium and the task demand is what should be learned, or so-called learning curl. Students perform well when the task is in the area of their personal preferences, but they grow when they challenge themselves to work with uncomfortable, unnatural methods. The challenge is something that makes us fly in the learning space and feel the flow. "Teachers usually oversimplify most ideas in order to make them more easily transmittable to learners. In addition to stripping ideas out of their normal contexts, the ideas are distilled to their simplest form so that students will more readily learn them. In fact, problems are multiple components and multiple perspectives and cannot be solved in predictable ways like the canned problems at the end of textbook chapters. Actually the need is to engage students not only in solving simple problems but rather complex and ill-structured problems as well." [16] To avoid the illusion that problems are simple, well design learning will use difficult problems with sufficient time and help for students outside of the classroom. The classroom would be reserved for the final discussion. In flexible learning model, the instructor is not just an expert who knows everything; the instructor will provide learning opportunities and experiences for all students. The decentralization of the classroom is the first and most important condition for the adaptive learning systems.
The latter underlined quality of learning process is a very important part of the whole environment especially when the students study a special discipline (physics in our case) in a language that is not native for them. Here is one of the main points to stress over by the presence of the teacher and individual tutoring.
The Physics quizzes are designed to measure the knowledge and achievements of high school students attending the preparatory class in physics at the Department of Language Teaching International Students. They are comparable to SAT II Physics Tests in level and difficulty. At the same time the quizzes are prepared thoroughly enough in order to give a backing in students’ creative learning manner. Although the SAT tests were taken as a model, there are at least two essential differences: the time for doing any of the quizzes is not fixed – it depends on the students’ knowledge, skills and attitude; each of the quizzes contains links to Java simulations of the processes and phenomena over which they are built and guided solutions. So, when students meet with an unfamiliar term or concept, they have the opportunity to learn these concepts and to try the problem again. The embedded help allows students to gain self-reliance and fluency in the subject exploration. The quizzes are composed in a manner that is congruent with all the categories in the Cognitive Domain of the Bloom’s Taxonomy: 1. Knowledge of terminology; facts, principles and generalizations, theories and structures. 2. Comprehension: Realizing the meaning of information. 3. Application: The use of learned information in new situations. 4. Analysis: Examining and understanding the organization structure of information to develop divergent conclusions by identifying motives or causes, making inferences, and/or finding evidence for generalizations. 5. Synthesis: Creatively applying prior knowledge and skills to produce a new whole. 6. Evaluation: Judging the value of material based on personal values/ opinions, resulting in an end product, with a given purpose. These physics quizzes are organized onto the material studied in college courses such as in the books: [16 – 18]. They comprise the chapters on: Mechanics, Heat, Kinetic Theory, Thermodynamics, Fluids, Waves, Electricity and Magnetism, Direct and Alternating Current Circuits. An example of a simple quiz used during the process of experiment is given in appendix 3.
The paper presents some research on students' learning orientations and demands and individualized instruction supporting a modern vision of the education as a set of strategic objectives for self-directed learning. The implementation of ALS requires different perspectives to be united in a multipurpose set of learning opportunities adaptive to students' personal differences. Learning centers and departments qualified to provide help to student demands are the key to success of the Adaptive Learning System. Here the educators have to create courses and tests that are loosely structured environments that promote challenging discovery. Self-directed goals depending on individual preferences are encouraged; however, the instructor needs to keep the appropriate level of challenges for each project. The entire course and consequently the corresponding tests should be built around learning freedom for each student to view and share content from his/her perspectives. All computer-based supplements should be short, concise and with sufficient support as needed.
4. Method. The modern education is impossible without computers. In this study, technology is used to design learning environments, adaptive to students individual and group learning needs, in which relevant learning services are accessible for a diverse population. This vision concerning education poses questions about observable and measurable variables of students learning in- or out-of-the-classroom. What part of students' learning educators could measure and control? The answer to this question touches upon the method of education. A new method for research of students' learning curves requires observation directly into the dynamic of the process. One way to observe individual learning is by designing learning environments to measure the utilization of learning services and projects. Depending on the way students utilize particular learning environment, the educator enhances that environment to develop desire for learning, commitment, engagement, motivation and involvement of each and every individual without limitations and learning discriminations. Therefore, the learning needs depend on the learning environment designed by the educators, and students' needs of support reflect teachers' accomplishment to design appropriate learning environments. Hopefully, due to technology integration, instructors can assess students' needs with statistical study embedded into online self-directed learning instruments. The aim of online research, integrated into learning supplements is to allow personal learning freedom and pre-emptive self-study for all students. In the case of using self-directed study with standardized tests online, there is no need of a proctor; the stress is minimized by allowing non-synchronized testing at time and place comfortable for the students. Students learn while the computers record statistical data such as the number of attempts to solve test’s problems, key strokes, total usage time, time distribution during the day, and scores (Fig. 2 and Fig. 3). This means that students construct meaning about the new material, depending on their personal knowledge and ability to utilize different resources such as books, online applets, online sites, help, online track tutorials, or, even, online correspondence.
5. Data Collection and Visualization. Here we present two visualizations of data collected in asynchronous learning instruments provided with Black Board for a class of Introductory Statistics. The online supplement is used strictly to guide students learning outside of the learning center. Students enrolled in statistics classes, used the supplements only outside of the classroom. The initial reaction was negative; the supplements were difficult and demanding. Students' complains, such as: "is this an online class," were honest and hard for instructor to ignore. However, with the time, students not only transformed to independent learners; they became fluent in using online supplements (Fig. 2 and Fig. 3). At the end of the fall 2006 semester, students frequently asked if they may have more discussions online. The time of learning online was increased for each student, and the engagement in classroom changed the proportion of teacher/students time of speaking in favor of students. Students were more aware of their strengths and learning challenges, and initiated discussions about better learning plans that would work for them. Instead of artificial problems, typical for all statistics classes, students worked enthusiastically on their projects and used original data generated by their surveys to self-teach in the classroom the last three chapters.
Physics students, trained by these tests, besides science content were also English language learners; their skills in English language comprehension had significantly developed through the learning and testing process. The monitored physics groups had four students during the academic year 2004/2005 and three students during the next academic year 2005/2006. Each learner in these small groups was tutored almost individually as needed for the investigation. In figure 4, the formative results of all their attempts to solve the chapter quizzes are represented. Figure 5 displays the students mastering advance shown at the final exam.
![Text Box: Fig.5. Monitored results of the final exam are in agreement with the observations of Bloom’s Taxonomy [7]. The curve-fit of the trend-line is a polynomial of order 2.](ALS_tests_files/image013.gif){kind=small}
CONCLUSIONS
The advantages of asynchronous learning outside of the classroom can be summarized as follow:
1. The lecturer (instructor) acts as a designer and facilitator of learning environment (Appendix 1). The goal is for the student to gradually improve his/her confidence in data processing and interpretation;
2. The lecturer (instructor) facilitates discussion in classroom to provide opportunity for the student to build confidence as an autonomous learner;
3. The lecturer (instructor) posts test materials online (Appendix 2) to allow the student to work according to his learning curve, and to develop professional competence (Appendix 3).
In all cases presented in this paper, individualized learning services were available for all students. Constant monitoring of students’ learning progress provided valuable ongoing information for flexible decision-making, and increases relative learning independence. The educators had sufficient data to conduct research on students learning dynamics in order to create learning opportunities, adaptive to students' personal differences. For example, by careful integration of challenging scientific content into lessons on English for Special Purposes, the curriculum will provide opportunities for cross-discipline learning. The design of online or face-to-face classroom should be based on the real distribution of individuals' needs and demands. Standardized tests will inform individuals on a daily basis about their readiness to enter into a dialog online or in the classroom. The holistic learning approach of mapping meaningful ideas by each individual student was successfully utilized in this study by a combination of tutoring and ongoing testing. With the help of new technologies and educational techniques for accelerated learning, Bloom’s observations [7] were tested and proved to be valid.
Appendix 1
Report: Student X
|
Chapters |
No of Attempts |
Highest Score |
|
|
Lowest Score |
|||
|
1
|
11 of 4
|
100%
|
0%
|
|
2 |
4 of 4 |
84.20%
|
84.20%
|
|
3 |
5 of 4 |
100%
|
26.30%
|
|
4 |
1 of 4 |
80% |
80% |
|
5 |
7 of 4
|
90% |
0% |
|
6 |
1 of 4 |
57.90%
|
57.90%
|
|
7 |
2 of 4 |
65%
|
60%
|
|
8 |
1 of 4 |
90%
|
90%
|
|
Average: |
|
88% |
58%
|
|
Standard deviation |
|
14% |
30%
|
The learning outcome is 30% for this student. The standard deviation is smaller for highest grade. The number of attempts is reduced with the time from 11 to 1-2 attempts.
Appendix 2
Students have all supplements available online. One frequently used supplement is the discussion board. Although the statistics class is 100% face-to-face class, the discussion board was used to provide media for communications. In this class, the discussion board was linked to course documents. It is interesting that students worked on course documents almost as much as on assignments on the bases of which students are graded. Also, students frequently open the syllabus and use external links. The table below is only a small part of automatically generated data about students' usage of discussion board in week 3 of the semester.
|
Forum: Intro to Statistics |
|
|
||||||||||||||||||
|
Date: Sat Sep 02 2006 15:49 |
|
|||||||||||||||||||
|
Author: Fuller, Temenoujka <tfuller@centralaz.edu> |
|
|||||||||||||||||||
|
|
|
|||||||||||||||||||
|
|
|||||||||||||||||||
Appendix 3
A sample of a physics quiz on one of the experimental themes is given below:
Quiz about FLUIDS
Check
how much you've learned so far about the topics of Fluids with this
self-scoring quiz.