Over the years I have come to realize, with some frustration, that you can expect your good students to learn data, definitions, problem types... but that it is unreasonable to expect them to connect information between different parts of your subject (or, God forbid, of different subjects), to learn to solve problems that are completely different of anything they have seen so far, to deviate from what you have taught them. I characterize this situation saying that our students have passive knowledge, instead of the more desirable active one.
With passive learning students retain the knowledge they are given and go no further. A pure passive student would be little more than a recording machine. Students with active knowledge recreate what they've been taught. The knowledge becomes theirs, they know the facts and the reasons. They can adapt their knowledge as facts change and they can expand their knowledge with new reasoning. Note that in this context active is not the same as industrious, passive does not mean lazy. An active student might not be energetic, many passive students are hardworking and responsible.
Passive knowledge is not inherently bad. It is the first step to learning. I would consider inappropriate to try to hammer in active knowledge to incoming freshmen on their first quarter. You must begin passive, activity comes later. But I've been teaching higher level courses during the last years and I haven't seen much activity among my students. I continue to see them very interested in knowing exactly what is it they are responsible for in the exam, weary of situations they haven't seen (and haven't seen solved) before. They handle ever bigger projects, true. But they simply have learned to put together an increasing number of passive skills. From time to time you do see some stroke of activity, but it is rare, too rare.
It is not surprising that students prefer passive learning, as it is easier to master passive knowledge than active one. And it is easier for us to teach for passive learning: we know exactly what to tell, we know exactly what to evaluate, we know exactly what to expect. Everything is under control. Because it is easier for me, and it is easier for them, teaching and learning are becoming passive.
But for society a passive-knowledge individual is less useful than an active-knowledge one. An active person will be more innovative, will adapt better to changes, is better suited to understand and gather insights from his environment. Of course, necessity teaches people to be active, but I would rather teach my students how to swim, than just build their muscles and throw them into the water.
An essentially active enterprise is writing. I have used the process of writing as the basis of a method to promote active learning, and I have used this method to teach a computer architecture course. The students spend most of their time, both in class and out, writing about the subject. By choosing topics, finding main and secondary ideas, selecting what is relevant and what is not, ordering the ideas into a solid line of reasoning, they actively learn about the subject.
The method presented here is essentially different from anything I know of. It is not teaching them to write in a writing across curriculum style [4], nor is it the use of writing to enhance students' communication skills [2, 6, 8]. Writing is not something added to the course. Writing is the course. In the following sections I will present the fundamentals of the method, how I implemented it, and the results achieved.
Writing is by no means the only way to move toward active knowledge. In an interesting and thorough paper Fekete et al. [3] show several techniques they have used to encourage their students to reflect on their learning process and state of knowledge. Another approach is the use of web-based on-line journals [7]. So given all these new techniques and technology, why use old-fashioned writing?
I chose writing to enhance active knowledge for two reasons. One is that knowing how to write is a valuable skill in itself. Nowadays employers demand good communication skills: they know our students have more than enough CS knowledge, so they want other `generic skills' [2] among which communication is one of the most needed. And in (my country) writing is a skill of which you hear no more once you've finished high school. The second and most important reason is that I personally find that writing demands active knowledge. To write a decent document it is imperative to have good lines of reasoning, illustrative examples that show exactly what you want, sharp definitions. You must not write too little, making your writing hard to follow, and what is more difficult, you must not write too much, muddling your explanations. You must make yours whatever you want to write about so that you can explain it your way. Writing is essentially active.
So, why not teach using the `traditional' lecturing style, and then make them write papers? Because this does not really enhance active knowledge. By definition, active learning is something they have to do themselves. If you link concepts from different subjects on the blackboard, you are active, but they are passively learning the links you show them. They have to establish their own links and that is something only they can do. So given that establishing these links, threading lines of reasoning, creating new and appropriate examples is what active learning is about, this is what we are going to do. All the time. In and out of the classroom. Through writing.
(The course I tought) is an advanced computer architecture course that is taught during the 3rd year of our computer science degree. It is a year-long course with two two-hours meetings a week. The main textbook is Hennessy and Patterson's classic text [5].
No writing course is offered at our school, so students come with little or no writing experience at the college level. Therefore during the course I must teach them writing fundamentals along with the computer architecture material. There was no required textbook for the writing part, but I heavily used two excellent books, one by J. Williams [10] on writing style, and another by Booth, Colomb and Williams [1] on the structure of a report. Two other books I used more sparingly were Writing for Computer Science by J. Zobel [11] and A Handbook for Scholars by M.C. van Leunen [9].
To learn a skill you have to do it repeatedly. I divided the course into five cycles. In each of them we would go through the process of writing, each time in more depth. The cycles lasted five or six weeks depending on the length and difficulty of the subject to be worked on. During the cycle we worked on a theme of computer architecture and emphasized an aspect of writing. Figure 1 shows the architecture and writing themes of each cycle. At the end of the cycle the students handed in a report on a topic chosen by them within the theme of the cycle. This report was evaluated according to two criteria: the content, paying special attention to its clarity and the quality of their reasoning, and the quality of the writing.
| Architecture theme | Writing theme |
| Arithmetic | Edition |
| Processors | Clarity of phrases |
| Pipelining | Structure of paragraphs |
| The memory unit | Structure of sections. |
| Introductions | |
| Input/Output | Tables and graphics |
The five reports represented two thirds of the final grade. An exam in which they had to solve problems from a given list made up the rest of the grade.
To create a report you first gather information, then choose a topic and an approach to the topic, write a rough draft, and finally elaborate this draft until you arrive to a finished report (or you run out of time). The cycle structure was designed to mimic this process.
Before the beginning of the cycle the students were given a carefully selected list of minimum reading they should do. The first meeting of the cycle was an overview of the architecture theme. After that the students were supposed to read the assigned material and select a topic for their report. To give them time to do so the second meeting of the first week covered the cycle's writing theme.
During the rest of the cycle the students wrote the report. To guide them through the process the students were given exercises at each meeting. These exercises helped them move through the writing process and were in accordance to the writing theme. At the end of each meeting they were assigned a task -create a list of main and secondary ideas for each section, write the introduction, prepare all figures and tables- that should be ready by the next meeting. To give them time to complete large tasks, I would insert a problem solving class. This allowed me to add problem solving skills to the course and to cover some material that had not been covered in the lecture.
This I have described is an ideal cycle. The structure of the actual cycles evolved quite a bit during the year as the students evolved. The more they mastered the process, the more active they became, the less they needed me. The last cycle consisted in the architecture and writing lectures, a few hints and let them work on their own. They knew what they had to do.
To interest the students in the experiment I created an atmosphere of adventure: I didn't know what was going to happen, neither did they so let's all explore together, work hard, have as much fun as possible, and if they put in an honest effort they would be accordingly compensated through their grade, independently of the outcome. In this manner the students became involved and added many suggestions during the course, especially at the method review sessions I set up at the end of each cycle. It was a kick for me to see how their suggestions were better expressed and more convincing at the end of the year than at the beginning.
The 26 students that enrolled the course were teamed into 13 pairs. During the first cycle I kept the students on a short leash. For instance I gave three different topics to the theme (arithmetic) and told them to stick to one of them. I also made sure there was at least one `solution' at the end of each writing exercise. The reports were surprisingly good.
Despite the good results the students were worried. One of their worries was the lack of uniformity of what they were learning: as each team explored different venues, each mastered different material. They wanted to know which was the fixed set of knowledge they had to master. It took some time for them to realize that there was no fixed set of knowledge. That it is neither important nor valuable to know three chapters of a book. What is important and very valuable is to be able to learn, understand, and explain those three chapters. Realizing this was one of the turning points of the course.
Another of their worries was their sense of being lost when they began writing. They didn't know how to decide on an approach to the topic or how to shape their ideas. They thought that it was a problem of lack of knowledge, so they suggested I spend two weeks lecturing instead of just one meeting. Although this was against the spirit of the method, I went along with the suggestion and spent the first two weeks of the second cycle lecturing. The results were terrible. The students found out that you always are at a loss when you begin, and that more lectures didn't help: you will not know how to approach your topic until you start working on it, and that you will not know how to shape your ideas until you start writing them down. They discovered that time dedicated to (active) writing was useful, time dedicated to (passive) lecturing, much less so. Fortunately, it had been their idea, so they learned a lot from this experience.
I considered it important to give ample feedback to the students on their work. Writing comments on their reports was one-way and insufficient, so I set appointments with each team to review their report in detail. In this way they could tell me why they had chosen a given way to explain a concept, we could search for alternatives, discuss the advantages and disadvantages of each one. I spent up to two hours per team per cycle, but it was time very well spent.
The first two cycles were tough, but enriching. Afterwards everything went smoother. By the fourth cycle the students understood that writing difficulties were generally due to knowledge deficiencies. Overcoming these difficult passages was not a matter of looking up words in a dictionary, but of thinking and understanding better what was it you wanted to describe or explain. By the last cycle they didn't need me much; they were perfectly capable of learning and writing on their own.
Not everything was perfect. Problems were a problem; also students became very proficient in circumventing tough spots. The method is geared toward conceptual understanding, and this helps acquire problem solving skills, but only indirectly. I haven't found a convenient way of adding problem solving into the method. Solving problems on the blackboard was unsatisfactory. The students found this boring and against the spirit of the method. I also tried to get the students to write their own problems. But soon decided that this was beyond their skills and quickly, and perhaps prematurely, discontinued the practice.
Non-central concepts that were difficult to understand or explain rarely appeared in the reports. Students used their freedom to choose topics to circumvent tough spots. This is more a slight disappointment than a serious drawback: the students worked hard, and they shouldn't be blamed for taking the easy way out sometimes. I couldn't find an easy way of conserving their freedom in choosing a topic within the theme and an approach to this topic, but make them write about some specific sub-topic. It is a point to work on further.
To assess the worthiness of the method I had long and frequent talks with the students, handed out an anonymous survey at the end of the course and observed. I would qualify the experiment as a success. The goals I had in mind at the beginning were accomplished, other positive results I hadn't even thought of were achieved.
As I had hoped, the students were active throughout the year. To be able to write reports they considered of adequate quality they created their own knowledge out of the lectures, the required readings provided, and other material they sought on their own. They also became proficient writers; their reports ranged from acceptable to excellent. They were clear, easy to read, focused.
Furthermore, the students became independent. At the end of the year they were capable of reading, learning, understanding and writing a report with very little assistance. They also were clear on how well they knew the subject. They had a self-confidence on their knowledge, and themselves, that I had missed on most of my students in previous years.
The students generally enjoyed the experience, although some of them didn't care too much for all the writing they had to do. They valued the workload as high, but not excessive, and worth it, as they considered they had learnt much more than in other courses. They acknowledged that their learning capacity had increased, and that they had a different approach to learning. Most of them would recommend the course to their friends.
Despite considering the experience a success, I am in no way advocating the use of this method as the way of teaching everything. It is a method geared toward conceptual material, so it would not be well suited for courses such as CS/1 in which you want the students to learn skills. Each team requires a fairly large amount of attention by the instructor, so it is not well suited for large classrooms. The changes sought are profound, and require time, so the advantages would be less if used in a course that lasts only a quarter or even a semester. Large part of the success was due to the cooperation of the students, I doubt they would be so enthusiastic if every subject was taught with this method.
On the up side, profound changes are also long lasting. The activity and the improved learning capacity of the students should spill over to any other course they take (it is my intention to follow-up on the students that took this course to see if this is so). Using this method in just a few courses throughout the curriculum should suffice to promote active knowledge, and learning, in all courses.
When I began I had a clear idea of what I wanted to achieve, but a foggier idea of how to do it. Although is does not show in this paper, the method has evolved during the year as I was ready to do all kind of changes as they were needed. Here are some tips if you want to try this method. Many of them are useful when trying any teaching innovation.
| Write a ten line abstract of your future report. Pass this abstract to three other groups. These groups will write two questions each they would like to see answered in your report. Given the questions, rethink your report and rewrite your abstract. |
| Take a section of your rough draft. Write the introductory part of each paragraph. Rewrite these sentences as one paragraph. Is it a good summary of the section? Are there any sentences that should be added or deleted? If you have added a sentence, add a corresponding paragraph; If you have deleted a sentence delete (or rewrite) the corresponding paragraph. Has the section improved? |
| Given a topic-based outline, change it into a point-based outline. For instance, change ``Caches'' into ``The cache is the heart of a memory hierarchy: You need a good cache design if you want good memory performance'' |
| Create a new diagram simplifying (removing something) from each diagram you've made. Decide if the added detail of the old diagrams were worth the added complexity. |
To promote active knowledge I have designed a teaching method that uses the process of writing as the basis of learning. It has been implemented in an advanced computer architecture course. The course was divided into five cycles, each one with its theme and the students learned the subject by learning how to write a good report on each theme. The experience was successful. The students not only learned actively and wrote proficiently, they became independent, with raised self-confidence. This method probably can only be used under adequate circumstances, but its benefits should be profound and long-lasting.