Document View

Doing summaries reinforces connections among the new ideas students must learn and creates connections between new ideas and prior knowledge. A summary has four defining features: it is short, it tells what is most important to the author, it is written in one's own words and it states the information one needs to study.

Milli (pseudonym) proudly explained how she raised her score from 26% on the first psychology test to 79% on the second test. "It was that strategy you taught." The strategy, summarization, is very hard for students to pick up on their own, but it can be taught directly. Many students appreciate strategy instruction that makes summarization less mysterious. In this article I will describe a way to teach summarization that can be applied to all content areas, and I will explain the factors that make it effective.

A summary written for content area reading has four defining features: (a) It is short, (b) it tells what is most important to the author, (c) it is written "in your own words," and (d) it states the information "you need to study." We can make this powerful learning strategy available to high school and college students by teaching students to think about the passage and relate the ideas to one another to construct a summary rather than select sentences from the passage.

At the colleges and community colleges of our large U.S. urban university, diversity means that my classes have students of every race and religion along with immigrants from around the world. A class of 25 students may have people from 5 to 15 countries. Usually their ages range from 17 to mid SOs. Many of them need noncredit "developmental" reading or writing courses when they enter the university. Most of our students work and are responsible for children or older relatives while studying "full time." All my students know that the content area reading they do for school differs from reading for their own pleasure in two ways. One is that to study, they have to understand and remember new information. The other is that someone else has decided what should be learned.

Studying and remembering

One of the most common strategies college students use to study is to underline parts of a text and read the text over and over (Ruddell & Boyle, 1989). Many students view learning as something mystical that will happen to them if they reread faithfully. They have a vague model of memory as a collection of discrete facts, probably neatly listed on separate file cards in their brains. Information processing research provides a more sophisticated model of memory and more effective ways to learn from text.

The best model of semantic memory, memory for school information, is a network (maybe hypertext) of related ideas, grouped into interlinked concepts (Reynolds, Sinatra, & Jetton, 1996). Each time you use an idea, that idea and the ideas associated with it are alerted, or primed. The more an idea is primed, the stronger the memory. Thus, the more connections or links between ideas and the better organized the connections are, the more readily an idea can be recalled and applied. This is the reason rehearsal study methods, like rereading a chapter or reviewing file cards, are limited in their effectiveness (Ruddell & Boyle, 1989). They will strengthen individual ideas, but will not be very effective at creating connections among ideas. The paucity of connections limits cues for recall of the ideas and also limits the ways the learner can retrieve the information to apply it in new situations.

The strongest study strategies are those using elaboration. To elaborate, the student explicitly creates associations for the information to be learned. A good example of elaboration is Ciardiello's (1998) method in which students are taught to construct high-level divergent questions so that they are actively thinking about the material they read. A simpler form of elaboration is found in Cornell notetaking (Pauk, 1997) in which students write an example of each concept they study. The major drawback of elaboration is that it requires additional thinking, time, and energy, sometimes distracting students from the concepts they set out to study. Many students have had little training in elaboration and find it difficult. They prefer rehearsal, which makes minimal demands on cognitive capacity. Unfortunately, they do not realize that rehearsal often leads to rote learning and limits transfer resulting in minimal recall of what was studied.

The third type of study strategy, organization, reaches a balance between effort and effectiveness (McCormick & Pressley, 1997). In order to organize new information, students must draw on prior knowledge and pay attention to the nature of relationships among ideas. As they organize information by constructing a summary, an outline, or a graphic organizer, students generate links among the new ideas they are studying. They must use prior knowledge of the topic in order to organize the new information, so connections are made among related concepts. New information that is linked with the network of prior ideas will be more easily recalled, and can be used in more adaptive ways. In particular, writing a summary requires evaluating the information that is read to determine what is important enough to include. It also requires transforming long detailed passages into terse statements of the gist of the information. This "deep" processing primes concepts related to the new information and improves recall.

Ineffective student-Venerated strategies

Over the years students have told me that they have trouble writing summaries because it is hard to figure out what to put in and what to leave out. Like the students in earlier studies (Byrd, 1990; Winograd, 1984), my incoming students read the sentences in a passage and choose some sentences to copy or paraphrase while leaving out the other sentences. This "copy and delete" method is counterproductive. We discourage copying or close paraphrasing as plagiarism; it is against school rules and is illegal. In addition, while there are books and articles with paragraphs in which an explicit topic sentence states the main idea of the whole, many passages do not have explicit topic sentences. Most important, in order to enhance learning, summarization should be a process in which the ideas of a passage are related to one another, weighed, and condensed; a process of synthesis, not selection. What makes summarization a powerful study strategy are the connections formed as the reader groups ideas into associated concepts and the concepts into interrelated hierarchical networks or schemata. This grouping extends cognitive capacity.

A college student who has to read a 25- to 30-- page chapter each week for each subject faces thousands of sentences of information. Remembering thousands of separate facts is a very difficult task. Summarization provides cognitive shopping bags: Students who group ideas into schemata and label the schemata and the relationships among them have a reasonable number of schemata to keep in mind instead of an unmanageably large aggregation of discrete ideas.

Novices often try to judge the importance of ideas for studying the way they would judge the importance of personal experiences. Personal experiences are stored in episodic memory. Novelty, emotion, and drama are cues to importance in episodic memory. In content area reading these elements may be used to add interest, but they are usually not important. In fact Garner (Garner & Gillingham, 1989) labeled them as seductive details, because they draw students' attention away from the important ideas, which are usually more abstract. How can students judge importance if personal experience and idiosyncratic personal values don't help them judge what is important to the author?

Cues to text-based importance

In working with college and community college students on how to write summaries, I found two essential cues to text-based importance: repeated reference and generalization (Morton & Hosey, 1984). 1 was delighted to find that the research of van Dijk and Kintsch (1983) emphasized the same cues to provide a theoretical explanation of textbased importance. Van Dijk and Kintsch told us that expert readers learn the cognitive processes of summarization gradually through immersion. We can speed up the process for our students by presenting the essential cues to importance through direct instruction and guided practice. A discovery approach would leave students trying to apply personal criteria of importance or standards for episodic memory instead of standards for the semantic memory used for content area learning (Garner, 1987; Nordin, 1996). After students have learned the basic strategy, cooperative learning groups can be helpful in reinforcing the strategy as students practice applying it. They must practice the new strategy and learn to apply it so that the strategy becomes familiar. Otherwise applying the new strategy requires students' attention, reducing their concentration on the content they are trying to learn.

To exemplify students' own strategies for summarization and their responses to instruction, I shall present some summaries of "Native Americans," a passage I adapted from News for You (1990, December 12). (See Sidebar "Native Americans.") I have used this simple passage at different points in instruction with different classes over the years. I often begin teaching summarization by asking students to write a summary "the way you learned before." A student using personal criteria of importance or responding to the most dramatic ideas may even miss the point of the passage. In this example a weak student wrote, Many years ago Native Americans were living on this continent. Their land was destroyed by Europeans when they had came to their homeland. The land was taken away from the Native Americans by Europeans. The United States government had set aside land for different tribes. They not only have land problems but job problems as well the unemployment rate was very high. Some tribes started their own business but were taken away from their customs and their way of life, especially language. Their way of life was changed by Europeans.

Students may identify with others who experienced discrimination, or they may have been so strongly influenced by the opening of the passage that they did not fully acknowledge the contrast the author set up (Sorrells & Britton, 1998). Other students ramble or include specific details that would enliven an essay, but should not be in a summary, for instance, "The Choctaw tribe built auto-parts factories." Strategy instruction will enable these students to determine what is most important to the author and to transform the information into a terse summary.

Repeated references. The first cue to importance found in all content area reading is what van Dijk and Kintsch (1983) called argument repetition. The more an idea is referred to by other ideas, the more important it is. It is primed most frequently, and it has more connections with other ideas than the less central ideas. Argument repetition is part of the surface structure of written material. In teaching I call it repeated references. Students find this cue easy to apply. Kintsch and van Dijk (1978) employed a research paradigm in which the number of references was counted, but our students do not have to count. They just have to ask themselves, "Is this idea central or on the edge?" or "Does the author keep talking about it?"

To introduce repeated references, I wrote a little exercise:

Johnny Lee has a good imagination and loves playing tricks on people, so his mother tries to keep an eye on her little darling. Yesterday Ms. Lee had a lot to do, so she left Johnny eating a snack while she started her work. Suddenly something cold and slimy slid down her back. She screamed. After a moment she realized that her son was up to more mischief.

I ask the students, "What does the author keep talking about all the way through? What other words or phrases does the author use to refer to Johnny?" It is important for students to recognize that the idea may be referred to in many different ways; it is an idea we are following, not a word or phrase.

Next I give each student a set of written guidelines to follow as the class learns how to write summaries (see Sidebar "Guidelines"). I also post or hand out the diagram shown in Figure 1. I have the students read a one-page passage, and I model the processes that I use to follow the guidelines to write a summary. Next I lead a whole-class session in which the class constructs the summary of a second short passage using the guidelines' four steps to summarization.

Step 1 in writing a summary is to think of the passage as a whole. We are learning to judge importance with respect to an entire essay or article or a section of a book chapter. Students must learn to read the entire passage and think about how the parts add up to a whole. They must find out how the authors resolve conflicting data or use straw men. Students must give the end of a chapter as much weight as the beginning.

Step 2 is to determine the thesis: What is the central idea (main idea) that the whole passage refers to? If students state only the topic, it is essential to teach them to tell what the author has to say about the topic. A content area summary must give information.

Step 3 involves determining the major supporting ideas. I have the students group consecutive paragraphs that refer to the same aspect of the topic and actually draw a bracket in the margin to identify each group of paragraphs. Then we write the central idea of each group of paragraphs as determined by repeated references. As students suggest major supporting ideas, we use repeated references for judging importance: Do the other ideas refer to this idea? Is it central or at the edge? Does the author keep talking about it?

Step 4 in writing a summary is checking your work. Students must be sure their thesis states the idea to which the rest of the article refers. A literary summary may refer to topics or discuss the author's craft, but a summary for content area study cannot. Every topic must have its comment: what does the author say about it? Do not refer to the author, and avoid the words about, how, or the way. Don't say "It is about Native Americans," but say "Native Americans are overcoming serious problems."

Following a whole-class construction of a summary, I ask each student to use repeated references to construct an individual summary. At this point a student summary of "Native Americans" usually includes the major supporting ideas, but many students still don't begin the summary with the thesis, and some may have trouble writing a single sentence to sum up several paragraphs. Typical summaries are as follows:

Native Americans were stripped of their culture and when the Europeans arrived in the United States. Because of the two different views of life, the Native Americans were pushed off their land. They were then forced to live on reservations. Being used to living on natural instinct, Native Americans are having a hard time finding jobs. For those who are not, they have built many different businesses. There are many tribes of different names, and some have become well successful. They are now being educated the American way and the Native American way. Even though the Native Americans were treated badly, they still don't forget where they came from. Each year they hold gathering for all of their people. It is called powwows. With this gathering it shows the people that their culture will last a long time.

Native Americans were forced to live on reservations across the U.S. The majority of them without jobs. Different tribes try and form jobs on the reservation. The children are deny the teaching of their culture. Celebration of religious ceremony are held each year which are called powwows.

Generalization. Repeated references are valuable as cues to importance, because they are part of the surface structure of the text. Frequently the gist of a text is not part of the surface structure, but is constructed by the reader. The second cue to importance that I teach, generalization, is a cognitive process that students have used in other contexts. Expository text is usually a hierarchical arrangement of ideas. Generalization enables students to identify the ideas at the top of the hierarchy.

Van Dijk and Kintsch (1983) found that while the reader is building a cognitive representation of the surface structure of text, simultaneous processes construct a brief representation in long-term memory called "gist." They broke down the process into three rules that are applied recursively:

1. Deletion: Leave out specific details and background information unless an idea is necessary for interpreting the main ideas.

2. Generalization: For each group of ideas, paragraph, or group of paragraphs substitute a sentence that "pulls together" the specific ideas. It must be general enough to cover the details, but it must not go beyond the details to become vague.

3. Construction: In forming your generalization, include the inferences you drew from the ideas in the paragraph or group of paragraphs. (Adapted from van Dijk & Kintsch, 1983, p. 190)

These three rules can be taught as generalization. When a series of specific ideas are replaced by a general idea, deletion has occurred. The theorists distinguish between generalization and construction, but for the sake of teaching construction can be considered one way of making a generalization.

I introduce generalization after using repeated references to write summaries for a day or two. This cognitive process is very familiar from many other uses. When I introduce it, I recite a list like "carrots, spinach, tomatoes, string beans." Before I can finish the list, students are saying, "vegetables." I give them a second list, more abstract, but still familiar, "Washington, Lincoln, Roosevelt, and Bush." With weak students I have them compare pairs of ideas such as flavor/chocolate, poodle/dog, fingerprint/clue. Which is general and which is specific?

The little paragraph I made up to teach generalization says, "After work, I have to buy groceries and make dinner, and then wash the dishes. I'll have to do the laundry and then dust and vacuum. If I have time, I ought to wash the kitchen floor." I ask students to write the main idea in a sentence in their own words. If they write topics without comments, I use the cue "Who did what?" specifying that "who" may be an organization or an idea.

To teach generalization as a basis for summarization, I use the same procedure I used with repeated references: guidelines (see Sidebar "Guidelines"), a diagram (see Figure 2), modeling, and guided practice. The four steps are the same, but students now have a cue to importance, generalization, which also enables them to transform a list of detailed sentences into a single "higher level" sentence. In using repeated references students' attention is directed to the surface structure of the passage, but using generalization emphasizes that the thesis or major supporting idea may not be written on the page. The reader can put two and two together, make sense of the passage, and find the thesis and major supporting ideas "in your head."

Enlarge 200% Enlarge 400%

When I have students write an individual summary after a couple of days of work with repeated references and a group lesson on using generalization, many more students can figure out the thesis and are able to group paragraphs together. Typical summaries would be as follows:

In the early years Native Americans were faced with problems by the Europeans; nevertheless, they continued with their traditions. Finding a job is a big problem for Native Americans. In a way of finding a solution, Native Americans have started their own business. Education has become an important part of their lives. Their traditions have not changed throughout the years.

Native Americans were forced out of their land and culture but are finding a way to get back their traditions and to live. Europeans set them aside on reservation. Most of them are jobless and were forced to create their own businesses. Their children are learning their traditions. They use powwows for religious ceremonies and keep united.

By the end of the semester, everyone who has done the assignments conscientiously has learned to write an adequate summary.

To begin practicing summarization I use short passages in which it is easy to discern which paragraphs would be grouped together. Then I use longer passages and finally chapters from content area textbooks. In working with textbooks I usually recommend previewing the chapter as a whole, and then summarizing a section at a time, writing one sentence to sum up each subsection. While there are some courses such as anatomy in which every detail is important, in most courses determining the general idea that sums up each subsection and checking that the details all refer to that idea help students understand the material

formal research

A formal test of these classroom procedures showed that repeated references and generalization were both effective in assisting students to summarize (Friend, in press). In this study repeated references and generalization were taught separately to illustrate theoretical differences in processing. At three colleges of a large urban university, 149 freshmen who had failed the writing placement test were randomly assigned to summarization instruction using repeated references, summarization instruction using generalization, or instruction in relating the text to readers' personal judgments of importance (control). The study was conducted during two 90-minute sessions of the writing classes to which these students were assigned. The experimenters, three experienced teachers, delivered the instruction using scripts to assure consistency. To avoid experimenter effects, the condition taught by each experimenter was rotated at each college.

Enlarge 200% Enlarge 400%

The procedure (discussed in detail in Friend, in press) used the definition, guidelines, and figures described earlier. The passages to be summarized, taken from Day (1980) or adapted from adult literacy materials, were comparable to "Native Americans," which was one of the passages used in the study. Experimenters modeled the strategy, led the class in developing a summary as a group, and then had students apply the strategy to construct a summary individually while the experimenter circulated and answered individual questions. Papers were collected, corrected, and returned in the next class section with a model summary for experimental conditions and a model self-response for the control group. The experimenter led the whole class in using the strategy to construct another summary, then each student was asked to write his or her own summary. This product served as the dependent measure.

Each summary was scored for inclusion of predetermined important ideas, exclusion of predetermined unimportant ideas, construction of a thesis statement, sentence transformation (number of sentences of the original text summed up by each sentence in the summary), and stating the full main idea, not the topic (how...). Summaries were typed and copied; each was scored by two experimenters achieving a .80 interrater reliability.

Sentence transformations were not significantly affected by the treatment, but participants who were taught to summarize using either generalization or repeated references did significantly better than the control group on the other measures: including predetermined important concepts in their final summaries, F (2, 144) = 4.1032, p < .019, excluding unimportant concepts, F (2, 144) = 26.1525, p < .001, constructing the thesis of the entire article chi^sup 2^ (2, N = 129) = 10.04, p < .006, and stating full ideas, not just topics chi^sup 2^ (2, N = 147) = 26.64, p = .000.

Informal feedback supports these results. My students say that using generalization and repeated references removes the mystery from summarization and helps them understand what the author means. They are using this strategy for their content area classes. Students with A's and B's in courses as different as biology, economics, music, and criminal justice say, "Now I know what to do!"