preservice elementary school teachers’ conceptual change about projectile motion: refutation text,...

8/9/2019 Preservice Elementary School Teachers’ Conceptual Change about Projectile Motion: Refutation Text, Demonstratio…

1/27


2/27

INTRODUCTION

Conceptual change in science has been of interest to science educators and other edu-cational researchers for more than a decade. Yet, students still have difficulty learning

complex scientific principles, especially if they are counterintuitive, and teachers remainfrustrated in their attempts to help students learn them. Indeed, the interactions among avariety of poorly understood influences on learning are, in all likelihood, responsible for

the difficulty in effecting conceptual change. Therefore, this study is an attempt to ex-plore the interactions of some of those influences. Recruiting preservice teachers as par-ticipants, we combined text with demonstration in one condition, appealed to the

usefulness of learning in another condition, engaged the teachers in thinking about theirpositive and negative experiences, and asked them their opinions about the importance

and usefulness of science and their interest in learning it. We evaluated their conceptualchange by using paper-and-pencil tests and by listening to their explanations of conceptsto an elementary school student. Furthermore, we attempted to trace the teachers changes

in thinking over time. The effect of these conditions were evaluated in light of the re-search and theory regarding the epistemological, cognitive, and affective dimensions of

conceptual change. We begin this article with a discussion of the nature of nonscientificthinking, the concept we wished to teach, and the nature of conceptual change— the cog-nitive task we wished to affect through reading.

BACKGROUND

Nonscientific Thinking

Individuals often hold theories about the way the world works in contrast to those held bycurrent scientific thinking; that is, the theories are considered nonscientific by practicingscientists. In this study, we use the term “nonscientific” to mean unacceptable to today’s

scientists and the term “scientific” to mean currently acceptable. However, we acknowledgethe likelihood that today’s acceptable science will be tomorrow’s naive science. Also, thereare shades of acceptability. For example, Newtonian physics is useful in describing easily

observable phenomena but inadequate for describing phenomena at an atomic level. Be-cause of the changing nature of scientific knowledge, it is often difficult to know what to

call nonscientists’ theories. The way many nonscientists think about the world may havebeen accepted in an earlier time and appears unimportant in their daily lives. They have nodifficulty maneuvering a car, throwing a ball, or doing physical work. Often, it is only the

people who need to be precise who rely on scientific principles. Engineers, for example,must rely on the principles in physics to make accurate calculations. If they did not, their

buildings or bridges might collapse. But even their calculations could be improved throughadvances in understanding (note the collapse of many buildings during the Japanese earth-quake that were thought to be built using sound physical principles). Therefore, it is for

want of a better term that we use “nonscientific.”

The Concept

The scientific idea participants were asked to learn is actually a cluster of ideas that leadone to understand the forces at work when a projectile is put into horizontal motion. If a can-nonball is shot from a cannon at the top of a cliff, it continues in an arced path to the ground.

If an object that is being carried forward is released, it also continues in an arced path to the

2 HYND ET AL.


3/27

ground. The paths are the result of two independent but simultaneous factors. As a cannonballis shot, the force of the shot places the cannonball in a state of forward motion at a constant

rate of speed. Because a carried object is in motion when it is released, it also continues tomove forward at a constant rate of speed. The object’s forward movement is an example of

Newton’s law that an object in motion tends to stay in motion. At the same time, however, theobject is being acted upon by gravity. Gravity, an outside vertical force, pulls the projectile to-ward the earth at a constantly accelerating rate. (It accelerates at a rate of 9.8 m/s2.) The ob-

ject’s movement downward at that accelerated rate is not affected by its movement forward.That is, an object will fall down at the same rate as an object that is simply dropped straightdown from the same point. The forward motion does not slow down or speed up gravity. Like-

wise, gravity does not slow down or speed up forward motion. Although both algebra and cal-culus can be used to solve problems based on these concepts, in this study, we confined our

teaching to these concepts alone.Many people believe that a cannonball will move forward for a while and then begin to de-

viate downward, saying that the cannonball’s forward motion must be “used up” because itoverpowers the effects of gravity until it does. (This theory was accepted thinking beforeNewton’s time.) Those before Newton did not understand that an outside force is necessary to

change a projectile’s motion, believing, rather, that there is a force implanted in the projectilewhen it is placed in motion that somehow dissipates. In addition, many people fail to ascribemovement to a carried object, because it appears to be at rest to the person carrying it. There-

fore, if released, they believe a carried object will fall straight down. They also do not believethat a dropped object will land on the ground at the same time as a horizontally projected ob-

ject, given identical release times. They believe that the forward motion either speeds up orslows down the vertical motion. They often describe the path of an object, not as an arc, but asstraight out and then curved down. These ideas that are inconsistent with Newtonian theory

have been described in other research (Duit, 1991; McCloskey, 1983; Hynd et al., 1994). Priorwork with high school and college students revealed that more than 90% of this population

have nonscientific ideas about motion (Alvermann & Hynd, 1989; Hynd & Alvermann, 1989).

Conceptual Change

For individuals who hold nonscientific ideas about the way the world works, learning ac-cepted scientific concepts can be difficult. Researchers (e.g., McCloskey, 1983; Maria &MacGinitie, 1981; Marshall, 1989) have found that individuals whose ideas conflict with new

information often disregard or discount the new information in favor of existing knowledge,

when it is necessary for their learning to alter existing knowledge. Conceptual change, the al-tering or reorganization of existing schemata to account for new learning, appears to takeplace only under certain conditions. These conditions, as of yet, are poorly understood. Re-searchers have previously described a piecemeal, sawtooth pattern of changes in thinking

(Schymansky et al., 1991; Strike & Posner, 1990; Alvermann & Hynd, 1989) when it does oc-cur. Students do not often completely change their nonscientific theories to scientific ones.

Also, when it does occur, conceptual change has been hypothesized to be the result of severalinteracting factors, epistemological, cognitive, and affective in nature.

Epistemology. There may be epistemological reasons why conceptual change is difficult.

Carey and Smith (1993), for example, say that, developmentally, children move from the ideathat knowledge arises unproblematically from observations to an idea that knowledge acquisi-tion depends on one’s interpretive framework, allowing for two or more interpretations of the

PRESERVICE TEACHERS 3


4/27

same observations. Kuhn et al. (1988) provide evidence that younger students have difficultyinterpreting two conflicting points of view, believing that the “truth” will be found once they

observe the facts. We can understand why people with this less mature epistemology find con-ceptual change so difficult when it comes to counterintuitive science concepts such as con-

cepts about projectiles. Students believe that understandings of the world are direct results of their observations. Students who are asked to explain their hypothesized, but not scientificallyaccurate, path for a projectile often use common sense based on prior observations to explain

them. They have recounted their observations of old World War II movies, for example, whenexplaining that a bomb dropped from a moving airplane travels backwards from the point of

the drop (it actually moves forward). They have said that they have observed a projectile go-ing out for a while and then deviating downward when, in fact, they have merely been unableto detect the slight downward movement of an object at the moment it is placed in motion and

the air resistance that slows the objects’ forward motion (Hynd et al., 1994). Because mostpeople have had much success relying on their observations of the world, they believe that

these observations are accurate. Without observable outside forces, students believe that innerforces must be at work. Observation plus common sense reasoning equals “reality.” Concep-tual change, then, may come about only as people realize that their previous observations and

subsequent beliefs about the world are inaccurate. In this realization, the epistemology of ob-servation leading unproblematically to knowledge is challenged. Other research (Hynd et al.,1994) provides some support for this contention. In that study, students who believed that sci-

ence concepts were complex rather than simple were more likely to learn counterintuitiveconcepts. Strike and Posner (1990) hold that conceptual change is more likely to take place

when students believe that science is logical and useful to them in their everyday dealing withthe world.

Cognition. There are also purely cognitive reasons why conceptual change is difficult. Hew-

son and Hewson (1983) explain that, when students are confronted with conflicting data,rather than undergo conceptual change, they can discount the data, ignore it, or memorize it(compartmentalize it). Chinn and Brewer (1993), in explaining what scientists themselves do

with conflicting data, offer even more ways for scientists to maintain their existing percep-tions. In other words, there are more avenues for maintaining ideas than there are for chang-

ing them. Often, the path of least resistance is maintenance.Posner et al. (1982) hypothesize that there are four essential conditions for conceptual

change. These include: (a) dissatisfaction with one’s current conception, followed by the de-

gree to which the new conception is deemed (b) intelligible, (c) plausible, and (d) fruitful (inthe sense that it will provide a framework for the solution of new problems). Conflict betweenone’s nonscientific ideas and newly introduced scientific concepts is a major component of

this scheme. A meta-analysis (Guzzetti et al., 1993) has documented the effectiveness, at leastin the short term, of strategies believed to produce cognitive conflict. One such strategy is the

use of refutational text—text that refutes common intuitive conceptions in favor on scientificones.

Attitude/Motivation. Conceptual change has also been linked to attitudes and motivation.

Gilovich (1991) argues that erroneous perceptions of the world can arise from motivational as

well as cognitive sources. Affective conditions influence what kind of and how much evidenceis considered in cognitive decisions and will presumably effect what one learns from reading.

Strike and Posner (1990) state: “A wider range of factors needs to be taken into account in at-tempting to describe a learner’s conceptual ecology. Motives and goals and the institutional

4 HYND ET AL.


5/27

and social sources of them need to be considered” (p. 10). The idea of a “conceptual ecology”that includes motives and goals provides a focus on the underlying factors that interact to

cause individuals to maintain a certain view of the world—a view that may be regarded as er-roneous by scientists. Pintrich et al. (1993) claim that students are not “child-scientists” and

have ego-involved or extrinsic motivations for learning that may not be conducive to concep-tual change, in that conceptual change appears to require considerable effort. In one study, wefound that students who believe that physics knowledge is instrumentally useful to them are

more likely to exhibit intrinsic motivation to learn physics and, hence, are more likely to un-dergo conceptual change (Hynd et al., 1994). Usefulness, to them, was used more in the sense

of being useful to their future careers than it was to being useful in the sense of helping themto more accurately explain physical phenomena, although both were important. In contrast,students who did not undergo conceptual change could think of no use for physics, were not

planing careers where they would use physics, and described their motivations for learningphysics in ways that were more extrinsic than intrinsic. Unfortunately, for the understanding

of counterintuitive physics principles, many students in high school and beyond lack intrinsicor instrumental motivation to learn science.

Instruction

With all of the difficulty involved, what sort of instruction is helpful in inducing conceptualchange in students? In another study (Hynd et al., 1994), high school students who read refu-

tational text about the targeted physics principles learned, in the long term, more than studentswho participated in demonstrations and who talked to each other in groups about the princi-ples. In that study, talking to each other actually appeared to have a debilitating effect on con-

ceptual change. Students were talked out of scientifically viable explanations of phenomenaby their peers. Demonstration produced improvements on immediate but not delayed achieve-

ment tests, if combined with reading. In a study by Marshall (1989), a combination of demon-stration and reading produced the greatest change in subjects’ understanding of the causes of seasonal change. The importance of reading in science instruction has been a debated issue.

Although textbooks appear to guide the curriculum and be the mainstay of science instruction(Schymansky et al., 1991; Yore, 1991), their use is often discouraged by science educators,

who prefer that teachers engage students in the process of learning science through discoveryand hands-on experiences (Lloyd, 1990; Newport, 1990; Osborne et al., 1985). In a meta-analysis, however (Guzzetti et al., 1993), text that refutes common nonscientific ideas proved

effective at helping students learn scientific principles that seemed counterintuitive.

The Study

Because one goal was to study complex interactions previously mentioned in the literaturethat were influences on learning, we collected several types of data, some experimental andsome exploratory. For the exploratory data, our intention was to generate hypotheses. We

asked five questions.

(a) Will combining demonstration and reading enable preservice teachers to overcome their

prior conceptions? This question was in response to previous research with younger popu-lations showing that text and demonstration interacted to induce short-term learning, but

that only text produced conceptual change over time (Hynd et al., 1994). Deriving similar

results from an older, more conceptually mature population would add veracity to our other



6/27

studies. More robust or different results would provide fodder for discussing the develop-

mental nature of conceptual change.

(b) Will telling preservice teachers that they will be expected to teach a concept provide the

necessary motivation for overcoming prior conceptions? This question was asked because

of Posner et al.’s (1982) caveat that a new theory must appear fruitful in order for learners

to reorganize their existing schemata to include it and because of our other research with

high school students showing that relevance and usefulness were influences in conceptual

change (Hynd et al., 1994). Although we had documented student’s comments about the

usefulness of physics in other studies, we had never actually manipulated a usefulness con-

dition.

(c) What influences do prior experiences and attitudes have on conceptual change from read-

ing? We chose this question because of researchers (Strike & Posner, 1990; Gilovich,

1991; Pintrich et al., 1993) calls for studies dealing with affective factors such as motiva-

tion and epistemology. Another study with high school students found that students’ atti-

tudes toward physics topics and their attitudes toward the structure of their physics classespredicted the learning of counterintuitive information (Hynd et al., 1994). Again, similari-

ties to other studies would provide validity; differences with other studies would allow dis-

cussion of developmental effects.

(d) What changes in thinking do preservice teachers make as they proceed from being taught a

physics principle to actually teaching the principle themselves? Conceptual change is still

poorly understood and hypothetical in nature. We were looking for evidence that concep-

tual change did, in fact, occur, and wanted to document the form it took.

(e) What interactions among variables help explain why some learn counterintuitive informa-

tion from text and others do not? We asked this question because it is important to describe

conditions in which conceptual change would be likely to take place—conditions that in-

cluded each student’s background, epistemology, and attitude as well as cognition. More

mature, preservice teachers were especially interesting as a contrast to students in other

studies.

While the first two questions were investigated experimentally, the others were not. Obser-vations made concerning the last three questions were exploratory.

The differences between this and other studies are threefold. First, we wanted to involvepreservice teachers as participants. The reasons for involving preservice teachers was be-cause it is important for teachers to understand the scientific principles they teach and be-

cause we wanted to test the idea that they would be more motivated to learn information

due to their need to teach it—a type of instrumental motivation (relevance) similar to thatfelt by an engineering student. Furthermore, preservice teachers are older than the highschool students we had involved in studies previously. If Carey and Smith (1993) are right,then, these students should have a more mature epistemology about learning—one that

might allow them to change their conceptions stemming from observation of physical phe-nomena. Hence, they may more successfully undergo conceptual change than high school

students would.Second, in this study we asked participants to describe their attitudes toward physics

and physics textbooks and the influences in their thinking as a result of their schooling.

Although other studies have explored these influences (e.g., Hynd et al., 1994), the factthat these participants were practicing to become teachers meant that they had been

primed to think about the effects of schooling on learning as a result of the educationcourses they were taking. Also, they had taken science, including physics, in college andhad more experiences to bring to bear on the questions we asked.

6 HYND ET AL.


7/27

Third, this study limited the discourse in order to prevent the preservice teachers from talk-ing to each other. The researcher initially talked to each individual preservice teacher. Later,

the preservice teacher talked to an elementary school child. Although the learning environ-ment was somewhat artificial, we wanted to limit the solidification or learning of nonscientific

information that has been found to take place after students talk to other students. Therefore,in this study, researchers interacted with participants and the participants interacted with achild, but they did not interact with each other.

METHODS

In this section, we describe the participants in the study, the materials and methods used tocarry out the study, and the methods used to analyze the findings of the study.

Participants

The participants were drawn from a pool of 94 fourth-year elementary education majors en-rolled in the first of two reading methods courses at a large, state-supported university in the

southeastern United States. The second reading methods course was held in conjunction withstudents’ preservice teaching experience. Hence, the course we chose was their last fully in-

class experience before practice teaching. Close to 95% of the subject pool was made up of white, middle class females between the ages of 19 and 25 who lived in small towns and citieswithin the state. Participants reported taking as few as two and as many as seven science

courses throughout high school and college. Approximately 72% reported having taken aphysical science course, and they had taken a science methods course including physics as

part of their teacher education program. Of the 94 participants, four were dropped at the be-

ginning of the study when a pretest failed to reveal nonscientific conceptions about projectilemotion and 17 chose not to be part of the study. Seventy-three teachers, then, actually partic-

ipated.

The Conditions

We taught principles of projectile motion while manipulating two conditions. The first was

a demonstration– text condition. Teachers either participated in a demonstration before read-ing or merely read a text. The demonstration– text condition was created to bring about dissat-

isfaction with one’s current conception by asking participants to make a prediction about theoutcome of a demonstration, view and explain the demonstration, and then read. The second

condition was a usefulness condition. In this condition, participants were either told or nottold that they were to teach a lesson to a student about the information they were learning.

Demonstration. Our demonstration technique was in line with science educators’ notions

that there must be some cognitive conflict before nonscientific conceptions can be changed.We demonstrated the physics principle in a way meant to cause cognitive conflict. We asked

participants to predict where an object carried at shoulder-height would fall if dropped. Be-cause of the pretest they had taken and previous research findings (e.g., McCloskey, 1983), wewere confident that most preservice teachers would predict that the object would fall straight down or backwards. Then, using a piece of tape on the floor as a reference point, we demon-strated that a carried object falls in front of the release point. We asked the teachers if their

predictions were correct and, if not, to explain what really happened, encouraging them to de-scribe the simultaneous effects of vertical and horizontal motion through a type of scaffold-ing. We also asked them to predict the path of cargo dropped from a moving airplane; then we



8/27

showed a film that depicted the cargo falling forward in a curved path. Finally, we had partici-pants make predictions about the path of a penny being shoved off a table and a bullet being

fired from a gun. After each demonstration, we had them explain why the demonstration pro-ceeded as it did, helping them to emphasize the effects of forward motion and gravity.

In this demonstration procedure, other students were not involved. As previously men-tioned, other research indicated that students who talk to other students often learn or solidifynonscientific information rather than scientific, and we wished to limit that influence. How-

ever, the social-constructionist nature of learning is important. Therefore, we used Leontev’s(1981) notion of appropriation. That is, the researcher appropriated students’ comments to

provide them with the scaffolding needed to reason through the demonstration explanation.For example, if a preservice teacher described the path as an arc, the researcher would ask what was responsible for the arced path. If the preservice teacher could not answer, the re-

searcher would ask what was making the object move toward the ground. (Preservice teacherscould all describe the effects of gravity.) Then the researcher would ask what was making the

object move forward. These and other questions helped preservice teachers appropriate scien-tific explanations. The scaffolding that was provided may have also served to help the preser-vice teachers understand the plausibility of scientific explanations.

Although the demonstration procedure took only 15–20 minutes, it involved the teachers inrather complex processing of the targeted principle and had the potential to induce cognitiveconflict. It also involved participants in using the demonstration procedure often recom-

mended by science educators (Anderson & Smith, 1987; Champagne et al., 1983). Guzzetti etal.’s (1993) meta-analysis of science studies in conceptual change revealed that approaches in-

cluding demonstration appeared to be helpful, possibly because they produced dissatisfactionwith previous predictions, meeting Posner et al.’s (1982) first condition for conceptual change.Therefore, demonstration should help students learn from reading about a counterintuitive sci-

entific principle.

Usefulness. The other variable, usefulness, was chosen in deference to Posner et al.’s(1982) idea that, for conceptual change to occur, the new concept must appear fruitful (help

them explain future scenarios, solve future problems) and because of our findings that stu-dents who find physics useful (relevant to their lives) are more likely to undergo conceptual

change (Hynd et al., 1994). Half of the preservice teachers read a statement indicating thatthey should pay close attention to the information to be learned from the demonstration and/orreading because they would be teaching it to an elementary school student. A time was sched-

uled to teach an elementary school student for eight of the students who were told they wouldhave that experience. We also scheduled a time for eight preservice teachers who were not

told they would teach an elementary school child. We were interested in documenting differ-ences in their lessons and their understanding of the targeted physics principles that were dueto their being told or not told of the experience.

Reading. A 606-word expository passage that had been adapted from an article in Scientific American (McCloskey, 1983), titled “Newton’s Theory of Motion,” was given to all students.The adaptation had been checked for accuracy by a professor of physics at the university

where the study was conducted. The passage was considered refutational, in that it contra-dicted the commonly held theory of impetus, a pre-newtonian explanation of projectile mo-

tion asserting that objects had internal forces which dissipated over time. The text wasdesigned to elicit cognitive conflict, make the concept understandable and plausible, and toshow its usefulness. Other students who have read the text have been asked for their com-

8 HYND ET AL.


9/27

ments about it (Hynd & Guzzetti, in press). They say that it is effective because it engagedtheir prior knowledge (they found that impetus theory was what they believed in), told them

they had been inaccurate, and helped them to realize that others had also had the same nonsci-entific conceptions about motion. It also gave them concrete examples to explain the new the-

ory in a way that they could understand from their daily lives. Therefore, we consider the textto be understandable and plausible and to offer a fruitful theory—one that could be used byits readers to explain the phenomena they came into contact with on a daily basis. It was cal-

culated to be at the tenth grade readability level according to the Fry readability formula (Fry,1977). Refutational text has previously been found to exert a positive influence on conceptual

change even if not combined with other variables. Guzzetti et al. (1993) discovered that allforms of refutational text, when considered together, were superior, producing learning of counterintuitive concepts to all kinds of nonrefutational text across grade levels.

Test Materials. Three pretests were used to measure participants’ prior knowledge aboutprojectile motion. The first, a test of relatedness, was adapted from materials validated byValencia et al. (1987). It assessed participants’ ability to distinguish between vocabulary

terms that were either related or not related to the concept of motion. “Gravity,” “growth,”and “velocity” were three of the ten terms on the test of relatedness. A second pretest was

a shortened version (n 10 items) of an experimenter-constructed 21-item true/false test(test/retest reliability coefficient 0.71). A true item reflected Newton’s theory; a falseitem reflected impetus theory. A third pretest, an application task, required the preservice

teachers to study a diagram of a projectile shot from a cannon. Participants had to label thepath the projectile would take and explain the reason for their choice of paths. These tests

had been used in other studies (e.g., Hynd & Alvermann, 1989; Hynd et al., 1994).

Two of three posttests were administered immediately after the treatment and, then, againafter a 2-month delay. The first of these was the 21-item true/false test from which the ten-

item pretest was derived. The second was the application task described earlier. The thirdposttest, an eight-item short-answer test, was administered only one time, immediately afterthe treatment. Examples of items from the short-answer test include the following:

1. A person is walking forward at a brisk pace carrying a stone at shoulder height. Explain, ac-

cording to Newton’s theory, where the stone would fall in relation to the point where it was

dropped.

2. Why would this happen?

All of the aforementioned items—the text and the tests—had been used in other studies.Anyone desiring copies of these instruments may write to the first author and receive copies

of them.

Questionnaire. We investigated preservice teachers’ attitudes and their formal and informal

learning experiences in science. Preservice teachers completed a 16-item questionnaire de-signed to elicit responses to questions regarding: (a) the number of science courses taken;(b) beliefs about the importance of science, in general, and physics, in particular; (c) ratings

of their own knowledge; and (d) attitudes toward and experiences with teachers, textbooks,demonstrations, formal instruction, and informal learning experiences. Preservice teachers

were directed to rate issues such as importance and attitudes on a five-point Likert scale, butwere also asked to write explanations for each item rated. This questionnaire was evaluatedinformally (as we would an interview), using descriptive techniques that assessed the



10/27

responses to individual rather than pooled questions. In addition, the 16 items were meant totap separate areas of influence, so we believed the correlation between items to be of little im-

portance. Because of this, reliability is not reported. However, the questionnaire had a greatdeal of construct validity. Note our earlier discussion of the importance of affective feelings

about physics as a topic. We believed that many preservice teachers would think that physicswas important, but would not be able to define why it would be useful to them personally. In-deed, another study has pointed to the idea that usefulness is an important variable to consider

apart from perceived motivation to learn physics, which could be due to extrinsic factors suchas getting a good grade or getting the approval of peers (e.g., Hynd et al., 1994). We also be-

lieved that past experiences with instruction is a determining factor in students’ attitudes to-ward instruction in physics. In the study cited earlier, for example, we found that positiveattitudes about the structure of the classroom and the topics covered in physics accounted for

a significant amount of variance in postlesson achievement tests about counterintuitivephysics concepts. Finally, we were struck with students’ attitudes toward physics text in our

previous studies. Generally, attitudes toward text (in most cases, physics textbooks) are nega-tive. Yet, researchers have found consistently positive results from reading refutational text(Guzzetti et al., 1993). For these reasons, the questionnaire tapped what we believed were im-

portant constructs.

Videotaped Teaching and Postteaching Interview. In order to have qualitative documenta-tion of preservice teachers’ ideas about Newtonian principles after instruction, 16 preservice

teachers were videotaped as they taught concepts of motion to a fifth grade child. We also de-signed a ten-item structured, audio-recorded, postsession interview in which participants wereasked to rate and explain their level of knowledge, comfort in teaching, and success in teach-

ing the targeted physics principle.

Procedure

The study was conducted in four phases. In phase 1, the test of relatedness,

the ten-item true – false test, and the application task were used as pretests todetermine preservice teachers’ levels of prior knowledge about Newton’s first law of mo-

tion, their ability to apply the law, and their nonscientific conceptions about it. Onlythose who held nonscientific notions were retained in the study. Teachers were consideredto have nonscientific notions if they chose the wrong path of the projectile in the applica-

tion pretest and/or gave the wrong explanation for the projectile’s path. If the correct pathwere chosen, and a borderline explanation given, key items reflecting general principles

were checked on the multiple-choice test.In phase 2, the preservice teachers were assigned to one of four groups representing the

two levels of demonstration (Demo– Text/Text only) and two levels of usefulness

(Told/Not Told). Participants were required to attend a one-on-one (researcher and partic-ipant) session that lasted between 50 and 60 minutes. In the Demo – Text/Told condition,preservice teachers were told (in writing) about the forthcoming videotaped lesson they

would teach an elementary school child as they entered the session. Next, they partic-ipated in several demonstrations of Newton’s first law of motion in which they made

predictions and then compared those predictions to the outcomes of the demonstrations.

They also read the one-page refutational text on Newton’s theory of motion, workedon a buffer activity to control for the effects of short-term memory, and completed a

short-answer test, a 21-item true/false test, and an application task. Except for not beingtold about the videotaped lesson that would follow in phase three, those in the

10 HYND ET AL.


11/27

Demo–Text/Not Told group participated in the same activities as those in theDemo – Text/Told group. Participants in the Text only/Told and Text only/

Not Told groups worked on word search puzzles that contained words taken from thestimulus passage in lieu of the demonstrations. They, like the other two groups, then read

the stimulus passage and completed the posttests. Finally, all preservice teachers weregiven the attitude questionnaire to complete after the session.

In phase 3, eight randomly selected participants who had been told they would use

their newly acquired information in an actual lesson and eight who had not been toldwere videotaped as they individually taught a fifth grade student. Four teachers in the

Told and four in the Not Told group had been exposed to the Demo– Text condition. Asthe participants entered the room where they would teach, they were provided with a setof materials that they could use if they liked, but they were not given time immediately

before the lesson to prepare. We placed this constraint of no preparation time on the pre-service teachers because we wanted to reduce the possibility that some teachers would

prepare elaborate lessons reflecting nontreatment information rather than information re-called from the treatment. Participants were reassured that no one else was allowed toprepare and that their level of preparation would not be judged. They were also told that

they could have as much time as they wished to explain the concept. Audiotaped inter-views were held with each of the 16 preservice teachers following their lessons. In phase4, approximately 2 months after the initial lesson, the true/false test and the application

task administered in phase 2 were readministered as delayed posttests. Only two testswere given because of time constraints.

Scoring and Interpreting Data

Pre- and Posttests. We scored each of the pretests and posttests without knowledge of thegroup membership of the preservice teachers. Scores on all measures except the short-answertest, the application task, and the questionnaire were obtained by comparing participants’ re-

sponses to the responses on a prepared answer key. On the six-item, short-answer test, weawarded one point to each correctly answered question except for questions 2 and 4, whichwere two-part answers and awarded one point for each part. The application task scores

ranged from zero to two. Full credit was awarded if participants correctly labeled the path theprojectile would take and gave the correct explanation for their choice of paths. One point

was awarded if participants either correctly labeled or explained the path the projectile wouldtake. Zero points were awarded for a completely nonscientific answer.

Questionnaire. Preservice teachers’ beliefs about the importance of science in general

and physics in particular, their knowledge level, their attitudes, and their feelings about theinfluence of teachers, text, formal instruction, and informal experiences were tabulated us-ing data from the 16-item questionnaire. In the case where rating scales were not used (on

items about teachers, text, formal instruction, and informal experiences), each written an-swer to a question was rated as negative, neutral, or positive and assigned a score of 1, 2,

or 3, respectively. In addition, the 16 videotaped teachers’ questionnaires were separatedfrom the other questionnaires and the explanations that the participants supplied were ana-lyzed qualitatively.

Videotaped Lessons. Two researchers viewed, transcribed, and coded the 16 videotapedlessons and the postteaching interviews, looking for evidence of an overall correct explanation



12/27

of the physics principle that had been taught, noting the length of the lesson, and noting theself and fifth graders’ ratings of the teacher. The data are shown in Table 4.

Another researcher viewed each of the videotapes and read the transcriptions in order todiscover patterns not previously identified. She noted that the teachers often appeared to hold

seemingly contradictory ideas. With that observation in mind, she then analyzed the demons-trations, the text, and all test items for discrete concepts of motion, concepts that could beheld independent of others. Four concepts were identified. The first was the idea that a hori-

zontally propelled projectile’s path will form an arc on its way to the ground. The second wasthat something that is carried (in motion) and released will maintain its forward motion. The

third idea was that horizontal motion and gravity are both independent influences in the pathof a projectile, thus explaining why the object moves in an arc rather than first going out andthen down. The last idea was that forces are external to the projectile: that is, changes in mo-

tion are brought about by external forces rather than internal ones. All of these ideas were in-troduced in both the demonstration and in the text and were tested as well. Once these

discrete concepts were identified, all data records of the 16 videotaped students (pretests,posttests, videotaped lesson, structured interview, and delayed posttests) were coded for evi-dence that preservice teachers either had or lacked these concepts. In the case of the

true–false and other forced-choice items, only a lack of the concept could be documented, be-cause a preservice teacher could choose a scientific answer by chance. In the open-endedquestions, the videotaped lessons, and the interviews, both scientific and nonscientific princi-

ples could be noted, if mentioned. In this way, we could trace an identified nonscientific con-ception from pre- to posttest to delayed posttest. The contents of the videotaped lesson and

interviews provided informal opportunities to view preservice teachers’ thinking about the tar-geted concepts and to note if (and sometimes why) these concepts changed during the courseof the study.

To interpret the data, we arranged the coded items on a matrix for each teacher (Miles &Hubermann, 1984) and included scores from the pre- and posttests. From these individual ma-

trices, we analyzed: (a) how many categories of nonscientific concepts were held by teachersat the start and end of the study; (b) which nonscientific concepts were more prevalent;(c) which nonscientific concepts seemed to be replaced by scientific ones; and (d) what be-

haviors seemed to explain changes either from nonscientific to scientific thinking or visaversa. The Appendix shows examples of these matrices. Finally, all the data on two of the pre-

service teachers who were videotaped were analyzed qualitatively. We attempted to describethe interaction of variables that resulted in conceptual change at the end of the study.

RESULTS AND DISCUSSION

In this section, we present quantitative findings and discussion, then the descriptive and

qualitative observations and discussions. In a subsequent section, we discuss, in a generalway, the cumulative findings of the entire study.

Quantitative Analysis

The quantitative parts of our study were undertaken to help us answer the first two questionswe posed earlier: (a) Will combining demonstration and reading enable preservice teachers toovercome their prior conceptions? (b) Will telling preservice teachers that they will be expected

to teach a concept provide the necessary motivation for overcoming prior conceptions?To see if possible group differences existed prior to instruction, three one-way analyses

of variance were run on each of the pretests. None of the groups differed significantly on anyof the tests. The posttests were then analyzed using either analysis of covariance or analysis of

12 HYND ET AL.


13/27

covariance with repeated measures. A 2 (Demo–Text/Text only) 2 (Told, Not Told) com-pletely crossed design was used. The relatedness pretest, which resulted in the greatest reduc-

tion of error variance, was the covariate for all posttests. An analysis of covariance was run onthe 73 observations obtained for the short-answer posttest. Separate analyses of covariance

with repeated measures were conducted for the immediate and delayed true/false and immedi-ate and delayed application task posttests. Absenteeism at the time of the delayed tests re-sulted in dropping 6 of the 73 subjects from the analyses. Consequently, immediate and

delayed posttest measures for the true/false and application tasks were analyzed using datafrom 67 subjects. Tables 1, 2, and 3 present the adjusted means and standard deviations bygroup for each of the posttests.

Short-Answer Posttest. The analysis of covariance on the short-answer posttest revealed astatistically significant main effect for Demo–Text, F

(1,68) 7.34, p 0.01, in favor of the

group that participated in the prediction/Demo–Text, but not for Told, F (1,68)

0.68, p 0.42. No statistically significant interaction was found between Told and Demo–Text, F

(1,68)

2.80, p 0.10. Effect sizes were calculated by expressing mean differences between the

Demo–Text and Text only groups in standard deviation units (see Glass et al., 1981). On themain effect for Demo–Text, R2 0.14, with an effect size of 0.64.

Immediate and Delayed True–False Posttests. The analyses of covariance with repeated

measures on the true/false posttests revealed a statistically significant main effect forDemo–Text, F

(1,62) 5.85, p 0.02, in favor of the group that participated in the

Demo– Text, on the immediate true/false posttest, but not on the delayed true/false posttest,

F (1,62)

0.43, p

0.52. There were no statistically significant main effects for Told on eitherthe immediate, F (1,62)

1.11, p 0.30, or the delayed, F (1,62)

0.79, p 0.38, true/false

posttests. On the main effect for Demo–Text, R2 0.11, with an effect size of 0.48. Therewas no statistically significant interaction between Told and Demo–Text on either the imme-

diate, F (1,62)

0.41, p 0.53, or delayed, F (1,62)

0.67, p 0.42, true/false posttests.

Immediate and Delayed Application Task. The analyses of covariance on the applicationtask revealed a statistically significant main effect for Demo–Text, F

(1,62) 10.09, p

0.005, in favor of the group that participated in the Demo–Text condition on the immediateapplication task, but not on the delayed application task, F

(1,62) 1.87, p 0.18. There were

no statistically significant main effects for Told on either the immediate, F (1,62)

0.25, p

0.62, or the delayed, F (1,62)

0.00, p 0.99, application tasks. On the main effect for


TABLE 1

Adjusted Means (M ) and Standard Deviations (SD ) by Group on Short-Answer Posttest

Group Number M (SD )

Demonstration 39 6.65 (1.38)

No Demonstration 34 5.76 (1.39)Told 36 6.07 (1.37)

Not Told 37 6.34 (1.53)

Highest possible score 8.


14/27

Demo–Text, R2 0.18, with an effect size of 0.70. There was no statistically significant in-teraction between Told and Demo–Text on either the immediate, F

(1,62) 1.07, p 0.31, or

the delayed, F (1,62)

0.79, p 0.38, application tasks.As mentioned, the quantitative part of the study was initiated to answer the first two ques-

tions. These are discussed next.

Will Combining Demonstration and Reading Enable Preservice Teachers to Overcome

Their Prior Conceptions? Preservice teachers who made predictions and saw demonstra-

tions before reading a text did significantly better than the read-only groups on all three of theimmediate outcome measures. This finding supports Marshall’s (1989) study of preservice el-ementary teachers. In that study, the combination of demonstration and reading produced the

greatest change in subjects’ understanding of the causes of seasonal change.After 2 months had elapsed, however, it was impossible to distinguish between participants

in the Demo–Text and Text only groups, at least as measured by their performances on thedelayed true– false and application tasks. The lack of posttest differences after 2 months doesnot mean, however, that participants maintained their nonscientific ideas in the long term. In a

post hoc 2 (Demo–Text/Text only) 2 (posttest/delayed posttest) split-plot ANOVA (theDemo–Text and Text only groups were contrasts of different individuals and the posttest—

delayed posttests were contrasts of the same individuals), where test consisted of applicationpretest and application delayed posttest, we found that although the effects of Demo–Text

14 HYND ET AL.

TABLE 3

Adjusted Means (M ) and Standard Deviations (SD ) by Group on Immediateand Delayed Application Posttests

Immediate Delayed

Group Number* M (SD ) M (SD )

Demonstration 33 1.70 (0.57) 1.58 (0.56)

No Demonstration 34 1.17 (0.76) 1.38 (0.70)Told 34 1.48 (0.74) 1.48 (0.66)

Not Told 33 1.39 (0.69) 1.48 (0.62)

*In repeated measures analysis, observations with missing values are not used.

TABLE 2

Adjusted Means (M ) and Standard Deviations (SD ) by Group on Immediateand Delayed True/False Posttests

Immediate Delayed

Group Number* M (SD ) M (SD )

Demonstration 33 18.56 (1.74) 16.78 (2.56)

No Demonstration 34 17.48 (2.26) 16.39 (2.32)

Told 34 17.78 (2.33) 16.32 (2.65)

Not Told 33 18.27 (1.78) 16.85 (2.20)

*In repeated measures analysis, observations with missing values are not used.


15/27

were nonsignificant, there were significant differences between pretest and delayed posttestscores, F

(1,65) 109.75, p 0.001, on the application task. Therefore, preservice teachers

did change their previous ideas about motion, regardless of whether they were in theDemo–Text or Text only group. Furthermore, there was no significant loss of concepts for ei-

ther the Demo–Text or Text only groups on the true–false delayed posttest when compared tothe immediate posttest, (F

s2). This finding supports our other research with high school stu-

dents, which suggested that reading produced long-term gains while demonstration did not

(Hynd et al., 1994).Our interpretation of these results is that some long-term conceptual change occurred. Be-

cause the only factor experienced by all groups was the text, we speculate that the text, rather

than the demonstration, may have helped solidify concepts. The effect of demonstration less-ened over time. The long-term benefits of text versus demonstration should be the focus of

subsequent research.

Will Telling Preservice Teachers That They Will Be Expected to Teach a Concept

Provide the Necessary Motivation for Overcoming Prior Conceptions? In our study,

attempting to manipulate the usefulness of the information to be learned had no effect on sub-sequent conceptual change. We had anticipated that giving elementary education majors ad-

vance information of an impending teaching assignment would increase motivation to learnthe physics principle. That it did not produce the desired effect might be explained in severalways. Perhaps telling preservice elementary teachers they would have to teach the physics

principle produced anxiety, counteracting any potential increase in motivation. Perhaps theteachers did not believe they would really be called on to teach a physics principle to an ele-mentary school student. Another possibility, however, is that the participants in both the Told

and Not Told groups were already motivated, and any attempt to increase motivation by mak-ing the information useful was superfluous. They may have been at least extrinsically moti-

vated to achieve under any circumstance, and increasing a topic’s usefulness, in the sense of merely being able to explain a concept, may have had little bearing on their motivation. Itseems plausible to assume that these teachers-in-training were conditioned to want to do well

on tests and to present a good “face” when explaining the concepts to a child while beingvideotaped. But it was Posner et al.’s (1982) original idea that usefulness implied the ability to

help one solve future problems rather than the ability to explain information. In addition,teachers-in-training may not have the sense of urgency to learn physics that engineers-in-training have, explaining why this usefulness condition did not appear to be as good at ex-

plaining the participants’ behavior as that in a previous study where high school honorsphysics students (who were, by and large, planning on careers that used physics) found learn-

ing physics useful (Hynd et al., 1994). Teachers merely have to teach physical principles. En-gineers must rely on them for their livelihood. And individuals become elementary schoolteachers because of their love of children or perhaps their love of other subjects rather than

their love of physics. But individuals do not become engineers or engage in other physics-related careers unless they love physics. Therefore, our usefulness condition may have been

inadequate to test the notion we proposed to test. In the other research, cited previously, wefound that usefulness appeared to be an important variable in determining whether studentsunderwent conceptual change about physics topics. Students who believed that physics was

relevant to their lives in either a substantive or aesthetic way (i.e., need it for career, want to

appreciate physics in daily life) were more interested in truly understanding counterintuitivescientific explanations of ideas. Teachers-in-training may not feel that physics is especiallyrelevant to their lives. That would be unfortunate, because their teaching of physical princi-ples would be enhanced by a sense of relevance.



16/27

Descriptive/Qualitative Analysis

Analysis of Questionnaire. From the questionnaire we found that:

1. Participants’ attitudes about science were somewhat neutral when the field was consid-ered as a whole ( M 5.38 on a ten-point scale). When other science courses andphysics were separated, however, participants reported liking the other sciences more

than physics ( M for physics 3.6).2. Most participants felt uncomfortable with their knowledge about the sciences in gen-

eral and physics in particular. When other sciences and physics were considered to-

gether, students rated their knowledge as being somewhat low ( M 4.28 on aten-point scale). They rated their knowledge of physics lower still ( M 2.2).

3. Generally, participants felt that science was important to study ( M 7.14 on a ten-point scale). This was also true of physics. They rated the importance of physics just ashigh ( M 7.15) as the other sciences. While participant ratings were high, however,

their comments revealed less enthusiasm. While only one person made negative com-ments about the importance of science, there were eight such comments about the irrel-

evance of physics.4. The preservice teachers disliked science textbooks. The mean rating on a three-point

scale was 1.67, with a one being negative, two being neutral, and three being positive.

They thought science textbooks were boring, hard to understand, irrelevant, and unnec-essary.

5. The teachers liked demonstrations and experiments. The mean rating on a three-pointscale was 2.82.

6. Most of the participants had neutral to negative formal experiences in science classes.

The mean rating on a three-point scale was 1.57. Students cited the teacher, the text,and the assignments for their negative feelings.

We contrasted the delayed posttest results of those who rated themselves high or low inknowledge, attitude, importance, and usefulness. These data are presented in Table 4. In every

case except importance, those who rated themselves higher did better on the delayed multiple-choice and application test than those who rated themselves lower.

What Influences Do Prior Experiences and Attitudes Have on Conceptual Change

From Reading? Many preservice teachers had endured negative experiences with

physics and disliked science texts, preferring demonstrations. Despite these negative feel-ings, however, they still felt that science, including physics, was important to learn. This

16 HYND ET AL.

TABLE 4

Means of Delayed Multiple-Choice and Application Tests for High and Low Ratingof Knowledge, Attitude, Importance, and Usefulness

Knowledge Attitude Importance Usefulness

5


17/27

belief in the importance of science lends credence to the previous explanation that the par-ticipants in this study were already motivated to learn the scientific concept. It appears

that, despite motivation to learn, most of the preservice teachers we worked with found theconcept we taught uninteresting and difficult. Their overall neutral attitude about science

and more negative attitude about physics may be partially a result of the field’s perceiveddifficulty as well as the result of negative experiences. We note, for instance, the partici-pants’ lack of confidence in their knowledge of science, especially physics, and the lack of

a relation between ratings of importance and performance on the delayed posttests. The at-titudes of these participants corroborates findings from Pratt (1981) and Lederman and

Gess-Newsome (1991) that teachers’ lack of comfort in teaching science results from feel-ings of unpreparedness and lack of understanding. It also corroborates our findings usinghigh school students. Hynd et al. (1994) found that high school students’ positive attitudes

toward physics (liking physics topics) and the structure of their physics’ classes predictedtheir learning of counterintuitive physics topics.

Furthermore, the type of motivation for learning that we observed some students having (“Iknow I should learn it”) reflects extrinsic rather than intrinsic motivation. Although a few par-ticipants expressed the sentiment that they believed physics would help them to deal with and

understand their physical world, the majority of participants did not. Perhaps many of thesepreservice teachers were not really convinced of usefulness of the concept in the sense of helping them to solve future problems, an idea that might lead to more intrinsically motivated

behavior, nor were they concerned that they learn physics because their careers depended onit. Lee (1991) has noted that those who are motivated intrinsically behave differently when

learning science than those motivated by only extrinsic factors, and suggests that those whohave intrinsic as well as extrinsic motivation have a better chance at conceptual change thanthose who are not. These descriptive findings support the quantitative findings discussed ear-

lier that the type of usefulness we tried to instill in these participants—usefulness in beingable to explain the concept to an elementary school student—was inadequate at inducing in-

trinsically motivated behavior.On a positive note, the fact that some long-term conceptual change occurred from reading,

as evidenced by delayed posttest results, despite only neutral attitudes and negative experi-

ences, is encouraging. This finding of change in the face of some negative attitudes supportsthe findings of other researchers who discovered that attitudes explain only a small to moder-

ate part of the variation in science learning (Schibeci, 1989; Talton & Simpson, 1987). Ourfinding is significant in that it documents this effect with conceptual change and not just withscience achievement as in previous studies.

Analysis and Discussion of Videotaped Teaching

The analysis of the videotaped teaching and interviews was initiated to help answer the

last two questions: What changes in thinking do preservice teachers evidence as they aretaught a counterintuitive scientific principle? What interactions among variables help ex-plain why some learn counterintuitive information from text and others do not? Table 5

presents data gathered from the 16 preservice teachers who taught a videotaped lesson(8 who were told about the videotaped teaching and 8 who were not). Interestingly, the

students who participated in the Demo–Text condition seemed to outperform those in the

Text only condition on the posttest and delayed posttest, and do better on their teaching as-signment. It is possible that their teaching the lesson allowed them to rehearse the con-

cepts learned from the demonstration, leading to improved retention of those concepts anda better ability to explain them.



18/27

1 8

TABLE 5

Summary Data for Videotaped Teaching Lesson

Subject Length Self- Self- Correct Correct

Number of Science Rated Rated Fifth Concept Concept Short- T/

and Video Class Knowledge Success Grader’s in in Answer Imme

Name Group Lesson Nos. Level Level Rating Lesson Interview Posttest Post

#14 T.O. T/D 3 4 7 7 81 ⁄ 2 NO YES 6 17

#15 D.L. T/D 6 6 6 8 81 ⁄ 2 YES YES 8 18

#16 S.K. T/D 7 5 5 8 81 ⁄ 2 YES YES 5 19

#13 E.T. T/D 4 4 7 9 81 ⁄ 2 YES YES 6 21#33 E.Z. T/ND 6 5 3 7 61 ⁄ 2 NO NO 5 18

#34 K.M. T/ND 4 4 2 6 81 ⁄ 2 NO NO 4 19

#35 E.H. T/ND 6 4 1 1 51 ⁄ 2 YES YES 4 15

#36 B.C. T/ND 4 4 8 5 71 ⁄ 2 NO NO 5 12

#37 L.R. NT/D 7 7 8 5 81 ⁄ 2 YES YES 8 20

#38 P.B. NT/D 1 6 0 2 61 ⁄ 2 NO NO 6 13

#39 M.M. NT/D 3 6 7 5 61 ⁄ 2 YES YES 8 19

#40 J.S. NT/D 4 6 2 5 81 ⁄ 2 YES YES 3 18

#57 C.S. NT/ND 4 7 3 6 71 ⁄ 2 YES YES 8 20

#59 L.F. NT/ND 3 6 1 1 61 ⁄ 2 NO NO 6 14

#61 C.C. NT/ND 4 6 7 8 91 ⁄ 2 YES YES 6 19

#58 S.W. NT/ND 2 4 8 2 3

1 ⁄ 2

YES NO 6 19


19/27

What Changes in Thinking Do Preservice Teachers Make as They Proceed from Being

Taught a Physics Principle to Actually Teaching the Principle Themselves? In order totrack students’ changing conceptions, individual answers from pre and posttests, transcripts of

videotaped teaching assignments, and postteaching interviews were analyzed in a matrix (dis-cussed earlier) for evidence of changes in thinking. Discrete concepts were tracked frompre- to posttest. From this analysis, it was found that some concepts were more difficult than

others. (The results of the matrix data are presented in Tables 6 and 7.) The easiest concept forpreservice teachers was that the path of a projectile is an arc. Only 9 of the 16 missed itemsabout the arc at the beginning of the study. By the end of the study, no one missed these items.

The ideas that a carried object maintains its forward motion if dropped and that forward mo-tion and gravity operate simultaneously were difficult. Everyone missed items about forward

motion and simultaneity at the beginning of the study. Ten of the 16 (4 Demo –Text and6 Text only) still missed items about forward motion by the end of the study, and five(1 Demo–Text and 4 Text only) still missed items about simultaneity. The last concept was

also hard. Fourteen of the 16 missed external force items at the beginning of the study. By theend, 10 (5 Demo–Text and 5 Text only) were still missing those items. These findings con-

firm the sawtoothed nature of conceptual change.

What Interactions among Variables Help Explain Why Some Learn Counterintuitive In-

formation from Text and Others Do Not? Everyone at the beginning of the study held

nonscientific conceptions, and, of the 16, only 2 answered all items correctly at the end of the study. Both of these two began the study with nonscientific ideas about all four of thetargeted concepts but with high perceived knowledge ratings. If these participants had

rather solidified intuitive concepts (in that they began the study with naive notions about all

four targeted concepts), it is intriguing to speculate what allowed them to change their con-cepts. Perhaps they experienced more conflict between previous and new notions, and weretherefore more impelled to overcome cognitive dissonance. We looked at the means of thedelayed multiple-choice and application posttests to see if those having more nonscientific

notions would end up evidencing more conceptual change. Participants who missed itemsreflecting four concepts did better on those posttests ( N 7; mean of multiple-choice 18.3; mean of application test 1.7) than those who missed items reflecting three or fewerconcepts ( N 8; M 16.4; M 1.1).

Several preservice teachers adopted scientific concepts but also retained some seemingly

incompatible ideas. Four of the 16 (2 Demo–Text and 2 Text only) believed in both externaland internal forces as a result of the study. Three of the four relegated internal forces to forces

that did not cause movement, however. The fourth became confused when her demonstrationdid not go as planned. She then explained that a force goes “up, in the ball” so that it cannotbe seen. Another believed that the weight of the object would somehow determine whether or

not the path of an object formed an arc. This teacher, during the videotaped teaching lesson,dropped a bottle during a demonstration, and it did not go forward as she had predicted. (Shetwisted her hand backward as she released it.)

A possible explanation for the appearance of new nonscientific conceptions is that theseteachers were merely trying to assimilate the new notion of external force into their prior

notions of internal force and did not restructure their thinking. This explanation is in linewith Hewson and Hewson’s (1984) idea that, when confronted with conflicting information,

students compartmentalize and memorize information rather than changing their existingconcepts. However, their explanations did reveal some restructuring, in that they relegatedthe notion of internal force to something that had little bearing on motion. These teachers

seemed to become confused in their attempts to reconcile previous notions with new ones



20/27

and to reconcile what they had learned from reading and demonstration with the incompati-ble (and erroneous) observations they made while teaching. Although four of the five teach-ers had lower knowledge and attitude ratings than others, a high level of perceivedknowledge and a good attitude did not guarantee that the fifth teacher would gain only sci-

entific understanding.The participants in the videotaped teaching seemed to exchange one nonscientific concept

for another, adopt new nonscientific conceptions, and relinquish nonscientific conceptions

throughout the study, not just as an immediate result of instruction. Because we could detectno pattern, we contrasted the knowledge, attitude, and importance ratings of the four students

who gained in scientific knowledge, the three students who lost scientific knowledge, and the

20 HYND ET AL.

TABLE 7

Number of Nonscientific Conceptions of Individual Prospective Teachers

Pretest Posttest Videotape/Interview Delayed Posttest

E.T. 4 0 1 0*

T.O. 4 3 2 1*

D.L. 3 2 0 1*

S.K. 4 2 1 2*

E.Z. 3 2 2 2*

K.M. 3 0 1 1*

E.H. 3 3 1 3*

B.C. 4 3 1 3*

L.R. 4 0 0 0*

P.B. 3 3 0 2*

M.M. 4 1 1 2*

J.S. 3 2 0 3*

C.S. 4 1 1 1*

L.F. 3 3 1 3*

S.W. 2 1 0 1*C.C. 3 1 0 0*

*Switched from one nonscientific conception to another between posttest and delayed posttest.

TABLE 6

Summary of Matrix Data: Number of Students Evidencing Nonscientific Conceptions

Videotape/ Delayed

Pretest Posttest Interview* Posttest

Arced path 9 4 1 0

Forward motion 16 11 2 10

Simultaneity 16 6 3 5

External forces 14 6 4 10

*Prospective teachers may have maintained nonscientific conceptions during this phase without

their being detected.


21/27

seven students who seem to gain and then lose scientific knowledge between the immediateposttest and delayed posttest. For the students who gained in knowledge, their mean ratings

were 3.5, 4.0, and 5.6 for knowledge, attitude, and importance, respectively. All but one of these students had made comments on the questionnaire about the usefulness of physics for

understanding the world (this student also had low knowledge, attitude, and importance rat-ings that brought down the average). For the students who lost knowledge, their mean ratingswere 2.33, 4.0, and 3.33, respectively. One of the students mentioned the usefulness of

physics and the other two did not. Of the seven who appeared to gain then lose scientificknowledge, their mean ratings were 2.0, 3.6, and 6.7 for knowledge, attitude, and importance,

respectively. Only one student mentioned that information about physics was useful. We hy-pothesize, therefore, that a combination of knowledge, attitude, importance, and usefulnessfactors are responsible for the fluctuation we noted. The nature of this combination should be

tested in more controlled conditions with more participants.

Qualitative Analysis

Two In-Depth Analyses. We analyzed two cases in-depth in order to better answer the lastquestion we asked: What interactions among variables help explain why some learn counter-intuitive information from text and others do not? We chose two participants for in-depthanalyses, one from a Demo–Text treatment group and one from a Text only group. Both had

been members in the group of 16 participants randomly selected to teach the videotapedlessons. D.L. appeared to have a better understanding of Newtonian explanations for physicsprinciples at the time of the delayed posttest, while B.C. showed almost no movement in ideas

from pre- to delayed posttest. We analyzed the behavior of these two contrasting participantsto examine more closely how their differences would be manifested.

It is possible that being in a Demo–Text group was beneficial to D.L. When she taught thelesson, she used the examples from the demonstration and commented on its helpfulness. D.L.also had the advantage on several factors other than instruction that are believed to be impor-

tant in producing conceptual change. For one, D.L. had a firm grounding in science, havingtaken six courses. It is possible, then, that she had internalized a scientist’s view of the nature

of scientific knowledge. Epistemological variables are believed to account for some of the vari-ation in understanding content (Schommer, 1990; Strike & Posner, 1990). D.L. also had a posi-tive attitude toward physics and believed that physics was important and useful. Strike and

Posner (1990) observed that students who learn physics principles do so because they believethat physics gives them “reliable and objective knowledge of the real world” (p. 11). Songer

and Linn (in press) call D.L.’s view of science dynamic, and say that this dynamic approach toscience may cause students to expend more energy in learning counterintuitive concepts. Fi-nally, D.L. recognized that she had previously held naive conceptions about motion. In her in-

terview, she discussed how she had initially been wrong and commented on how thedemonstrations helped clarify her misunderstandings. If, as Posner et al. (1982) argue, students

must first be dissatisfied with their current conceptions for conceptual change to occur, then,D.L. demonstrated this dissatisfaction. Only two other people in the study reported experienc-ing conceptual conflict. Both of these students also evidenced considerable change.

B.C. had none of the advantages D.L. had. For example, she lacked a firm grounding in sci-ence or physics and exhibited a neutral attitude toward both science and physics. She did not

participate in the demonstration group, and she learned little from reading. During the inter-view, B.C.’s stated motivation for learning was to get the right answers, a fairly low-levelform of motivation. During our discussions with her, we saw no evidence that B.C. was ever

confronted with the notion that her previous understandings were not scientific. In fact, she



22/27

rated her knowledge as high after incorrectly teaching about Newton’s law of motion. We be-lieve that B.C. chose not to confront her existing notions, which would have led to conceptual

change, and either rejected or tried to memorize the new ideas, despite reading refutationaltext designed to foster conceptual conflict.

In summary, D.L. represents a case in which conditions are ripe for conceptual change.B.C. represents a case in which conditions are ripe for the maintenance of one’s existing con-ceptions. Their cases suggest that individuals who are intrinsically motivated, believe that

learning physics principles can be useful to them, consider themselves knowledgeable, andopenly experience conceptual conflict are more likely to undergo conceptual change.

CONCLUSIONS

In this study, we investigated a variety of conditions, cognitive and affective, under whichpreservice elementary teachers might be expected to relinquish nonscientific conceptions

about projectile motion after reading. After a summary of findings, we will refer to epistemo-logical, cognitive, and affective influences on conceptual change mentioned in the Introduc-tion.

Summary of Findings

The present study’s findings are important in that they replicate earlier work on the effec-tiveness of combining demonstration with reading to help students learn counterintuitive sci-

ence concepts. The absence of a statistically significant retention effect for the instructionaltreatment corroborated the results of the Hynd et al. (1994) study, in which demonstration in-teracted with text in the short term but not over longer periods. Text was the only factor that

produced long-term conceptual change. In this study, although we did find long-term concep-tual change, the effects of Demo–Text were no greater than the effects of merely reading the

text after 2 months. The fact that conceptual change did occur just by having students read arefutational text, however, is encouraging, and supports the effectiveness of this kind of textdocumented by Guzzetti et al. (1993).

Telling preservice teachers in advance of a forthcoming teaching assignment may be an in-adequate motivator. Perhaps experimental conditions that are more intrinsically rewarding are

needed to motivate preservice elementary teachers to relinquish their previously held notionsabout complex science concepts.

Our analysis of the videotaped teaching and the in-depth analysis of two teachers support two

ideas: (1) Conceptual change proceeds in a piecemeal, sawtoothed fashion. Although partialconceptual change may be evident, complete change of a conceptual set is unlikely given the

level of instruction implemented in this study. (2) A complex interaction of factors is responsiblefor conceptual change. The data suggest that perceived knowledge and belief factors account forwhat is not explained by instruction, and that there are differences between students who learn

physics because they believe it will help them understand the world and those who want to getthe right answers. We also observed that teachers who evidenced more nonscientific ideas weremore likely to change their ideas than those evidencing fewer nonscientific ideas.

Epistemology

This study was different from our other conceptual change studies because we used preser-vice teachers rather than high school students. Carey and Smith suggest that more mature

epistemologies enhance the chances that science learning will occur. Few of the preservice

22 HYND ET AL.


23/27

teachers we studied evidenced the more mature thinking that reality did not proceed directlyfrom observations. In fact, because their erroneous observations sometimes led to the produc-

tion of nonscientific theories, we believe that many of the preservice teachers operated withless mature notions about the world.

Strike and Posner (1990) say that conceptual change is more likely to occur with individualswho believe science can help them logically make sense of the world, providing them with “reli-able and objective knowledge.” Only a minority of students (5 of the 16 who were videotaped)

mentioned the usefulness of scientific knowledge. However, three of those five evidenced signifi-cant gains in conceptual learning from pre- to posttest. Developmental/epistemological issues in

conceptual change are important to understand. Can more mature epistemologies be taught? Isthere a way to convince students of the relevance and usefulness of scientific knowledge?

Cognition

Although researchers believe cognitive dissonance is a necessary factor in conceptualchange (e.g., Posner et al., 1982), teachers’ apparent attempts to make sense of the differ-ence between the targeted scientific concept in the test and their own prior and present ex-periences sometimes led them to other nonscientific conceptions. In other words, the

preservice teachers we observed may have restructured their previous notions, just not inthe direction we had hoped. This dynamic nature of conceptual change and the role of cog-

nitive conflict should be explored further. Whether or not conflict is a necessary prerequi-site to conceptual change is still not clear. Although participants who expressed cognitiveconflict and those with more naive notions about motion were successful in adopting sci-

entific principles from reading, unreported cognitive dissonance may have led others to

adopt new nonscientific understandings. Creating cognitive conflict is a common part of many instructional techniques, but documenting that it has occurred is problematic. In thisstudy, only 3 of the 16 teachers volunteered that they had experienced conceptual conflict.Were these students more metacognitively aware? Is awareness of conflict as necessary as

the conflict itself? These questions, along with those raised previously, need to become thefoci of future investigations into conceptual change.

Attitudes/Motivations

Positive attitudes seem to be at least partially responsible for conceptual change as well

as more intrinsic motivations for learning. This finding corroborates the findings of otherstudies (e.g., Hynd et al., 1994). Students who like physics and students who are more in-trinsically motivated are more likely to undergo conceptual change. Although these results

seem logical and hardly surprising, we must consider the fact that most of the preserviceteachers reported neutral to negative experiences and neutral to negative attitudes about

physics, even though they would have to be teaching physics principles to their students aspracticing teachers. The task of teacher educators to change the attitudes of preserviceteachers seems daunting but important if we would like their students to maintain positive

attitudes.In this study, we found that a combination of demonstration and text in the short-run and text

in the long-run helped preservice teachers to learn counterintuitive physics concepts. The effect

sizes suggest that our instruction was successful. However, only a handful of teachers learnedall of the concepts we wished them to, and some invented naive conceptions they did not previ-

ously have. Furthermore, although we tried to elicit cognitive conflict through our text and



24/27

demonstrations, we had little evidence that cognitive conflict actually occurred. Besides our in-struction, success appeared to be related to influences that we did not control, such as episte-

mology and attitude. Many of the teachers had not had positive experiences with physics andmay have been approaching the tasks we set for them with beliefs thought to be incompatible

with conceptual change. The implications of this study for the teaching and learning of coun-terintuitive physics is that physical science instruction may need to focus on the complex na-ture of scientific observation, the usefulness of understanding physical principles, and the

evolutionary nature of students’ thinking.

24 HYND ET AL.

APPENDIX

Examples of Matrices for Subject P. B.

Condition #: 3B Nonscientific Scientific

(Demo/Not Told) Conceptions Conceptions

Pretest info. and scores

H.S. and college science

courses 6

True/false test 3/10

Application 1/2

Post-demo info. and scores

True/false test 13/21Application test 0/2

Short-answer test 6/8

Minilesson 1/10

Attitude questionnaire Knowledge 3;

attitude 2;

importance 8.7;

usefulness—

Video lesson info. and

scores

Knowledge 0/10Teaching 2/10

Student rating 8/10

Interview

Delayed posttest info. and

scores

True/false 12/21

Application 0/2

Symbols: > forward motion; simultaneity;• inner force.

Lack of forward

motion (carried,

tossed) >;

simultaneity ;

inner force

(launched) •.

Lack of forwardmotion >;

simultaneity ;

inner force

(launched) •.

“Just guess, ’cause

I don’t have theanswer either.”

(Didn’t pay attention

to treatment.

Didn’t know she’d

be called on.)

Lack of forward

motion >; inner

force (launched)

•.

Two directionsout and down

(launched)

Two forces,

forward motion

and gravity

cause things

to move to

earth


25/27

REFERENCES

Alvermann, D. E., & Hynd, C. R. (1989). Effects of prior knowledge activation modes and text struc-

ture on nonscience majors’ comprehension of physics. Journal of Educational Research, 83, 97–102.

Anderson, C. W., & Smith, E. L. (1987). Teaching science. In V. Richardson-Koehler (Ed.), Educators’

handbook: A research perspective New York: Longman, pp. 84–111.

Carey, S., & Smith, C. (1993). On understanding the nature of scientific knowledge. Educational Psy-

chologist, 28, 235–251.

Champagne, A. B., Gunstone, R. F., & Klopfer, L. E. (1983). Naive knowledge and science learning.

Research in Science & Technological Education, 1, 173, 183.Chinn, C. A., & Brewer, W. F. (1993). The role of anomalous data in knowledge acquisition: A theoreti-

cal framework and implications for science instruction. Review of Educational Research, 63, 1–49.


Examples of Matrices for Subject D. L.

Condition #: 15 Nonscientific Scientific

(Demo/Told) Conceptions Conceptions

Pretest info. and scores

H.S. and college science

courses 6


Application 1/2

Post-demo info. and scores


Application test 2/2

Short-answer test 8/8

Minilesson 6/10Attitude questionnaire Knowledge 3;

attitude 8;

importance 7.7;

usefulness

Video lesson info. and

scores

Knowledge 6/10

Teaching 8/10

Student rating 8/10

Interview

Delayed posttest info. and

scores

True/false 19/21

Application 2/2

Symbols:

Arced path; > forward motion;

simultaneity;• inner force.

Lack of forward

motion (carried,

tossed) >; arced

path ;

simultaneity .

Lack of forward

motion > (tossed);

arced path

(launched)

.

Lack of forward

motion > (tossed)

Idea of two

forces

(launched).

Simultaneous

external force

(launched,

carried) •.

Simultaneous

combination of

two forces

External forces

act like

pressure •(treatment

caused her to

realize

previous ideas

were wrong).

Simultaneous

combination of

two forces


26/27

Duit, R. (1991). On the role of analogies and metaphors in learning science. Science Education, 75,

649–672.

Fry, E. (1977). Fry’s readability graph: Clarifications, validity, an

preservice elementary school teachers’ conceptual change about projectile motion: refutation text,...

Documents