Students' Scientific Explanations and
the Contexts in Which They Occur
E. David Wong
Michigan State University
Elementary School Journal, 96(5), 495-511. 1996
Abstract
The goal of this analytical essay is twofold. First, I analyze examples of middle school students’ reasoning in science to illustrate (a) that the distinction between students’ and scientists’ reasoning is ambiguous rather than obvious, and (b) that there are good reasons why students do not, cannot, and should not always reason and act as scientists do. These examples are drawn from my 2 years of science teaching with at-risk, inner city students. The second goal of this essay is to develop a useful conceptualization of reasoning in context. As research in this field develops, it is critical that the conception of “context” not become vague or all-inclusive, thereby diminishing its analytical usefulness. To facilitate discussion and analysis, I propose 3 dimensions of context that seem to have an important influence on students' scientific reasoning: (a) the knowledge and technology available to students, (b) the sociolinguistic context and related norms that implicitly define appropriate reasoning, such as the expected purpose, precision, and level of explanations, and (c) the expectations, values, and dynamics of school communities that influence student behavior.
The Difference between Students and Scientists
The goal of science education has always been to develop a more scientifically literate public. Frequently, the image of the scientist has been a model for ideal scientific reasoning and what a science program should strive to develop in K-12 students. In this tradition, important psychological research has emerged from the expert-novice paradigm to highlight differences between students and scientists and to suggest goals or directions for subsequent learning and development.
Comparisons between students and scientists have focused primarily on differences in conceptual knowledge and reasoning. An abundance of psychological and instructional research documents students' naive conceptions (also referred to as misconceptions or alternative frameworks) about scientific phenomena. These studies have concentrated mainly on (a) describing differences between students' and scientifically accepted explanations (Driver & Easley, 1978; Novick & Nussbaum, 1978), and (b) designing and implementing instructional strategies to change students' conceptions (Anderson & Smith, 1986; McCloskey Caramazza, & Green, 1980; Osborne and Freyberg, 1985). Similarly, research by developmental and cognitive psychologists has provided descriptions of the reasoning process of children, adult nonscientists, and scientists. These studies have typically analyzed how individuals make inferences from observation, distinguish between theories and observations, and construct and modify theories (Carey, Evans, Honda, Jay, & Under, 1989; Karmiloff-Smith, 1988; Kuhn, 1989; Schauble, 1990).
A logical consequence of these perspectives that emphasize deficiencies in conceptual knowledge and thinking strategies might be that in order to reason more scientifically young students need to acquire more accurate subject-matter knowledge and develop more sophisticated reasoning skills. Instructional programs designed to foster students' reasoning skills have shown some promise at improving students' performance in certain situations (c.f. Palincsar, Anderson, & David, 1993; Roseberry, Warren, & Conant, 1989). Conceptual change science curricula aimed directly at the diagnosis and treatment of misconceptions have produced promising changes in students’ explanations of scientific phenomena (Anderson & Smith, 1986; Hewson & Hewson, 1983; McCloskey et al., 1980; Osborne & Freyberg, 1985).
The importance of having a solid conceptual knowledge of science is irrefutable: problems, phenomena, or issues are understood as instances of a particular scientific idea or concept. Similarly, in order to solve problems or to develop explanations, individuals must also be able to employ appropriate strategies to bring together what they know, observe, and generate in a logical, sound manner. However, scientists' and students' reasoning is distinguished not only by variations in conceptual knowledge and reasoning skills but also by differences in the conditions or contexts in which reasoning occurs.
The Analysis of Reasoning in Context
When individuals think about scientific phenomena and discuss their ideas and explanations, they do so in a particular context. This assertion, by itself, is a truism: thinking always occurs someplace. However, suggesting that the particular intellectual, cultural, social, and physical character of that context might shape individual and group reasoning is a more interesting and important proposal. Research on scientific and mathematical activity in other ethnic cultures, workplace contexts, and other everyday settings has highlighted important features of learning and reasoning in formal and informal learning settings (Brown, Collins, & Duguid, 1989; Heath, 1982; Lave, 1988; Lave, Murtaugh, & de la Rocha, 1984; Lave & Wenger, 1991; Martin & Scribner, 1991; Resnick, 1990; Saxe, 1989; Saxe, 1991; Scribner, 1984). This research has characterized how learning and understanding are influenced by contextual factors such as social and cultural norms, interactions with other individuals, the nature of the task, and the available physical tools and objects.
Most of these studies have focused on the learning and practice of mathematics. Saxe (1992) observed that mathematics may be particularly well suited for this kind of analysis because it is a well delimited domain, the nature of students’ problem solving is manifest in their actions, and mathematics allows a relatively clear specification of what counts as more and less complex forms of mathematical operations. By contrast, scientific understanding and practice have received significantly less research attention. Perhaps within the dominant comparative paradigm, cases of informal, workplace, or cross-cultural scientific learning and reasoning are difficult to identify, characterize, and contrast. Furthermore, a distinct description or definition of scientific literacy may be difficult to produce. In research about both mathematical and scientific reasoning, conceptions of what constitutes legitimate literacy remain open to constant clarification or revision. Comparative studies consistently suggest that powerful and legitimate ways of reasoning about mathematical tasks can be found in informal as well as traditional disciplinary settings.
The study of students' scientific practices and the context in which they occur must be approached with prudence. First, since the definition of scientific practice or literacy is nebulous, the study of such practices is challenging. Second, the definition of context is equally, if not more, ambiguous. Although the surrounding environment may influence the substance and process of scientific reasoning, context can quickly become a vague or all-encompassing term, thereby diminishing its analytical usefulness.
In this article I focus on one activity of scientific practice: the construction of explanations about natural or technological phenomena. Other activities of scientific practice (e.g. designing careful investigations, constructing a logical argument, appreciating beauty) are related and important, but I leave their analysis to others. In addition, I avoid assertions about the influence of context as a general factor. Instead, one goal of this article is to sharpen and promote future analysis of the contextualized nature of scientific reasoning by highlighting three features of the environment that seem to play an important role:
(a) the knowledge and technology available to students,
(b) the sociolinguistic norms inherent in a context that
implicitly define what "counts" as reasoning, and
(c) the expectations, values, and dynamics of school
communities.
A second goal of this article is to encourage a closer examination of the relation between students’ and scientists’ thinking. Two assumptions are implicit in the much of the research on science learning and science instruction: (a) students generally do not reason like scientists, and (b) students should learn to reason as scientists. A growing and compelling body of research, however, suggests that students' existing knowledge and ways of reasoning are rarely arbitrary or senseless. Instead, reasoning strategies used by students and scientists alike are frequently learned, functional ways of thinking in particular contexts (e.g., Brewer & Samarapungavan, 1991; Karmiloff-Smith, 1984). Further research on reasoning in various contexts can reveal both ways that students are like scientists and ways that scientists are like students. Such research is also likely to reveal compelling reasons why students do not, cannot, and should not always reason and act as scientists do.
Classroom Context
For 2 years, I worked as a regular science teacher, within an "alternative education" program, for a low-income, urban, middle school science class. Students in this class had been identified by their seventh- and eighth-grade teachers as severely disruptive, academically at-risk, or both. Sixteen students were enrolled in the class, with slightly more boys than girls. All students were African-American; the school’s population was 95% African-American. For 18 weeks, these students spent their entire day (with the exception of lunch and elective classes) in this class covering all subjects with the alternative education teacher and a small team of assistants such as myself. After two quarters in this program, they returned to their regular classes. Frequently, such pull-out programs are little more than glorified in-school suspension. In this case, however, the alternative education program was explicitly intended to provide a supportive environment suited to these students’ particular academic and social needs (e.g. small classes, home visits by the teacher, special couselors, employment opportunities).
The unique and flexible nature of this program provided an opportunity to teach science in nontraditional ways. In particular, I was interested in understanding and facilitating students' efforts to construct their own explanations for scientific phenomena. In my role as the teacher, I made a concerted effort to encourage students to generate, elaborate, share, evaluate, and modify their own ideas. The students themselves were responsible for the substance and direction of our class discussions. I saw myself as a facilitator of their thoughts and energy rather than as an authoritarian source of correct scientific knowledge (for other examples of studies of context in instructional settings see Cobb, Wood, & Yackel, 1991; Schauble, Klopfer, & Raghavan, 1991).
The Effect of Knowledge and Technology on Reasoning
In the following example, students' explanations might be judged as scientifically incorrect when evaluated in light of standard textbook or canonical explanations. However, considering the conceptual and empirical information available to the students reveals the reasonableness of their explanations. This example also illustrates how the construction of explanations - by scientists and students alike - is enabled and constrained by contextual features of situations in which explanations occur, such as available knowledge and technology.
Explaining Why a Candle Goes Out
One scientific phenomenon that my science students examined involved baking soda, vinegar, and a candle. A 3-inch utility candle is mounted inside a 5-inch tall, 800 ml glass beaker. The candle is lit and a teaspoon of baking soda is carefully placed at the bottom of the beaker. Then, about 50 ml of vinegar is poured slowly down the side of the beaker. The mixture bubbles gently and the candle soon goes out. When attempting to relight the candle, one finds that the match goes out before it can reach the wick of the candle (See Fig. 1).
Students were quite intrigued by this phenomenon and were eager to provide an explanation. The three most popular explanations students offered stated that the flame was extinguished by (a) the "wet mist" from the fizzing mixture, (b) moisture that "seeps up the candle wick", or (c) a "puff of wind" produced by popping bubbles. None of these accounts corresponds to the accepted "scientific" explanation that maintains that the vinegar/baking soda reaction produces carbon dioxide that fills the beaker and extinguishes the flame. By many criteria, the students’ explanations would be labeled as misconceptions, incorrect, and unscientific. Is it appropriate, then, to conclude that the students' explanations are unscientific?
During class I persistently asked students to explain their ideas. As they attempted to justify their ideas, it became evident that although these three explanations did not correspond with that of a physicist or chemist, they were reasonable in light of (a) students' prior knowledge, and (b) the evidence available. For example, as middle school students, they undoubtedly understood that flames can be extinguished by putting water on them or by blowing them out. Students’ explanations suggested that they were attempting to apply this knowledge. The students' speculations were neither capricious nor arbitrary: no one suggested "magic" or invented a fire-extinguishing "force".
In essence, students constructed their explanations by integrating an idea or concept from their prior knowledge with empirical evidence associated with the phenomenon. This process is both scientifically authentic and logically sound; it is what scientists do. In fact, it is the only means by which anyone can generate rational explanations. Also, most students clearly sought observational data to develop and support their explanations. Proponents of the "wet mist" explanation cited sizzling noises from the candle and visible droplets of water on the walls of the beaker as evidence that moisture was extinguishing the flame. Students who suggested that moisture "seeps up the wick" also cited the sizzling noises as evidence that the wick was wet. The "puff of wind" explanation, although advocated by more than one student, was not supported by observational evidence.
Available Conceptual and Empirical Information
Examining the students' explanations in the context of their prior knowledge and the available empirical data complicates the traditional distinction between students' and scientists' explanations. Judging from prior and subsequent discussions, I inferred that these students did not have a well-developed understanding of carbon dioxide, its properties, or how it can be produced. It is difficult to imagine how they could possibly have generated an alternative to their explanations without this knowledge.
It can be argued that the students in this situation were generating explanations that were rational rather than "correct." Contemporary philosophical perspectives also contend that since most scientific explanations have eventually proven fallible, the notion that scientists have a claim to "correct" theories is untenable. Therefore, like the students' explanations, scientific theories can, at best, only strive for verisimilitude, or what Popper (1972) described as an ever-improving approximation of reality.
Both students' and scientists' explanations are also affected by access to empirical information and use of technological equipment. For example, the scientifically accepted explanation for why the candle goes out is founded on the assumption that carbon dioxide is produced by the reaction between the baking soda and vinegar. How does one come to realize or verify that carbon dioxide is produced?