An Investigation of Taiwanese Graduate Students' Beliefs about Scientific Knowledge

(1)

Bu!l. Nat'! Taiwan Nor. Univ. Vo! 40 pp.583-618 1995

師大學報

第 40 期第 583-618頁民國 84年

An Investigation of Taiwanese Graduate Students'

Beliefs About Scientific Knowledge

T zuchau Chang

Graduate Institute of Environmental Education National Taiwan Normal University

Abstract

This research, focused on graduate students' beliefs about scientific knowledge, is guided by Novak池 "Human Constructivism", which addresses a constructive perspective on education with a unity of psychological and epistemological meaning m述ing. N凶s baum's analysis of philosophical perspectives, Lincoln and Guba's distinctions between postpositivism and naturalism, and Edmondson's categorization of logical positivism and constructivism form the base for the discussion of epistemological beliefs in this re-search.

The research design includes a questionnaire survey and the descriptive statistical analysis. Its methodology is qualitatively oriented. It relies on descriptive analyses of survey results, rather than hypothesis testing.

The students' epistemological beliefs about scientific knowledge are revealed through their responses to questions that identify six core factors: observation, inquiry pro臼路， data, inquiry results, knowledge, and the development and growth of knowledge, which are adapted from Edmondson's questionnaire. This research shows that the students express different epistemological beliefs about different ωre factors, so their overall epistemological commitments are not consistent.

Overall, the students hold moderate positions while discussing epistemological issues, but they adopt a frrmly logical positivist view to deal with the issues relative to their re-search work. This suggests the dominan臼 of logical positivism in school education and research settings. However, that their moderate position on epistemological issues also shows their awareness of the constructivist perspective.

(2)

684 師大學報第四十期

Introduction

A theory of education has to encompass four "commonplaces"，品 suggested by Schwab (1978) and modified by Gowin (1981), namely teacher, learner, curriculum, and governance. Any educational event or reform must inevitably be related to these four perspectives: teaching and other pedagogical issues, learning theories, the structure of knowledge and curriculum design, and the social ∞ntexts of educational settings.

However, the emphases of educational rese訂ch have not been equally weighted on each of the ∞mmonpla臼s during the development of educational theories. The struc-ture of knowledge was the central issue during the period that human beings speculated about their universe, while theories of learning were most emph站立ed when psychology dominated the field. Learning theories have even exerted different influen臼s on edu-catìonal practìces. Behavioral learning theories emph部ize observable behavior of learners but intentionally neglect their cognitive structures; in ∞n甘甜t， emerging ∞研討， tive learning theories focus on the development of learners' ∞gnitive structures and some also take into account learners' thinking and feeling and acting.

The research on ∞伊itive learning has gradually challenged the dominan臼 ofbeh前， ioral learning theories since the 1960s. However, only a few studies on the e旺'ects of students' epistemoloical beliefs on their learning have been done 自由ntly. The田 stl肘， ies claimed student epistemological ∞mmitments would exert influen臼 on learning (Edmondson, 1989, Strike, et 祉， 1982, 1992, Confr句， 1980, and Waterman, 1982).

Confrey (1980) studied the influences of students' epistemological beHefs on their ωn臼ptual change while they were learning a transition from algebra / geometry to 叫， culus, Ï.e. a transition of a ∞n臼pt that numbers are discrete to that numbers are ∞心

(3)

研究生的科學知識信念的研究張子超 585

tinuous. She concluded that the student' learning were affected by their beliefs about mathematics which idcluded: (1) the relationship between mathematics and reality; (2) the nature of authority in mathematical truth; (3) rationnality in mathemati臼. Strike has examined students' epistemological beliefs at Cornell university since 1980. He and his colleague (1992) find the correlation between students' epistemological beliefs and their scientific achievement is significant. Edmondson (1989) developed a questionnaire to investigate Cornell students' beliefs about scientific knowledge, and to study the Ín-fluences of students' scientific beHefs on their learning strategies.

四le reserach was designed to investigate graduate students' epistemological ∞m mitments of scientific knowledge. Basically, the research used Edmondson's question-naire to evaluate students' general beliefs about scientific knowledge by thoroughly ex-amining their responses to the issues about the nature and limits of scientific knowledge. Six issus 一 observation， inquiry pro臼ss， data, inquiry resu1ts, knowledge, and the growth and development of knowledge, were proposed after a content ana加is of the 26 statements of Edmondson's questionnaire.

A questionnaire survey was conducted to collect data. The data analyses empha-sized descriptive statistics rather than inferential statistics. In fact, many well known statisticians caution against the misuse of statistical inferen臼恤 socia1 sci閃閃. Freed-man, et al. (1978, p. 407) even su.臨est that scientists might be better 0缸証 they report-ed descriptive statistics, rather than inferential statisti白， from their data. This

sugges-tion needs to be addressed by researchers in social sciences. Because of the nature of the data and the ethical concerns of the practice of social sciences, the data of social studies are usually non-normal, non-random, heavy prior, and small sample. Statistical inferences based on this data can be problematic and even deceiving. Therefore, this research emphasizes more des叮iptive statisti臼 rather than simple hypothesis testing.

(4)

Theoretic

a1

Foundation

1. Epistemological beliefs:

Kuhn's work (Kuhn, 1970) claims that the development of knowledge throughouot histroy c1early shows the interwoven relationships between epistemological beliefs and knowledge construction. In his view, a 臼rtain epistemological perspective guides 部 soicated research methodology and dominates the developmnet of knowledge for a pe-riod of tíme. In fact, research methodology endorsed ín any fíeld ís based on certain epistemological beliefs.

The epistemological belief underlying the development of natural sciences has been logical positivism. Due to the success and achievement of scientific deveIopment, logi-cal positivism has been accepted, albeit passively, in schooI and research settings. This dominance continued untiI recently, when the development of the studies of conceptuaI change among students and the paradigm building of sociaI scien臼s ignited the ree va1-uation of the perspective.

Nussbaum (1989) classified aItemative philosophical frameworks for studying the his-t。可 of science as empiricism / positivism, rationalism, and constructivism. Form his point of view, the distinction between empiricism and rationaIism is the sour臼 and vali-dation of knowledge, while the re∞gnition of human's ∞nstructive role in the devel-opment of knowledge distinguishes constructivism from the two other perspectives.

Lincoln and Guba (1985, 1989) make clear the comparison between positivism and naturalism. They highlight the ∞ntrast -- belief in absolute truth versus belief inωntex tual truth. They also discuss the relationship between researchers and research to point

(5)

研究生的科學知識信念的研究張于超 587

out that the involvement of individual conceptual frameworks in research is inevitable. As to the issue of epistemology, they endorse the importan臼 of the relationship be-tween knowing and the knower, rather than obscuring the knower in the process of knowing.

Waterman (1982) assessed college biology sutdents' beliefs abuout scientific knowl-edge by using two contrastive theories of scientific knowlknowl-edge: hypothetico-deductive (discovered) and conceptual change (∞nstructed) to design her questionnaire. Analyzing the students' comments, she summarized three themes about the nature of and growth of knowledge: "1) Knowledge is based on observed facts (dis∞vered)， it accumulates over time. 2) Knowledge is discovered 品 in (1), but it changes with new evidence and better techniques. 3) Knowledge is constructed based on theories which guide and limit observations月Waterman， 1982, aþstract)

1. Nussbaum's distinctions between constructivism, empericism, and rationalism:

According to Nussbaum (1989), there are two major perspectives on the nature of knowledge: one is the rationalist view, which emphasizes the activities of mind in spec-ulating on the origin of the universe; and the other is the empiricist view, highlighting sensory experien臼s as the base of reliable knowledge. PhilosophicaI literature illustrat-ed a pendulum swang between the two perspectives before 20th 臼ntury. A ∞nstruc tivist perspective on scientific knowledge has gradually emerged because of the skepti-cism in confirmable or provable knowledge, the recognition of the influences of socio-logical and psychosocio-logical factors on the development of scientific knowledge, and the students of the change of conceptual structures among students in 2伽h 臼ntu可﹒

Based on the differences in some basic assumptions, Nussbaum classified various philosophical frameworks iilto empiricism / positivism, rationalism, and constructivism (Figure 1, Nussbaum, 1989). The division between empiricism and rationalism is their

(6)

a$sumes

yM立面 's constructing the best current knowledge

assumes

hat (should) determirie the selection of best

prrent knowledge

Figure 1: Nussbaum's classification of various philosophical frameworks accord

.

ing to differences in some basic assumptions (Nussbaum 89, Intema-tional Journal of Science Education, Vol. 11, p. 532).

(7)

differences on the most substantial assumptions about the source of, and the methods for validating. human knowledge. Empiricism stresses the importance of the evidence of the senses as the sour臼 and validation of knowledge, while rationalism emphasizes the power of the intellect. Regarding the nature and the growth of knowledge, they share the same beliefs. Both of them believe the nature of knowledge is provable and confirmable, so it is absolute and true; and the growth of knowledge results from Man's finding out the existing true knowledge. However，∞nstructivism has distinctive per-spectives on the two issues. A constructivist perspective stresses Man's active role in the pro臼ss of knowledge construction, and adopts a relative view on the nature of knowl-edge, assuming that knowledge is changeable and never confirmable.

Since the nature of knowledge is relative and changeable, the concern about what should determine the selection of best current knowledge has inevitably be∞me the fo-cus of debate among researchers. Nussbaum pointed out two determinant criteria: in-ner disciplinary and outer disciplinary criteria, and accordingly, distinguished among the positions held by Popper, Kuhn, Lak泌的， and Toulmin. Inner disciplilna可叮iteria

stress the consideration of the logical, rational, and empirical structure of knowledge during the selection of best current knowledge. Popper is a good representative. He proposed deductive logic as a criterion to select knowledge. While outer disciplinary criteria highlights the influences of sociophychological factors on the selection of knowl-edge. Kuhn is a good example of this approach. The perspectives of Lakatos and Toulmin represent moderate positions.

2. Lincoln and Guba's naturalism vs. positivism:

Lincoln and Guba make a clear comparison between naturalism and positivism in their book, " Naturalistic Inquiry" (Lincoln & Guba, 1985). They show the contrast of the two views on five axioms, namely the nature of reality, the relationship of konwer to

(8)

the known, the possibility of generalization, the possibility of causal linkages, and the role of values (Lincoln & Guba, 1985, pp. 36-38).

Regarding the nature of reality, in Lincoln and Guba's view, the positivist paradigm professes a "single, tangible, and fragmentable" reality, while the naturalist paradigm looks at reality as "multiple, constructed, and holistic." As to the issue of the relation-ship of knower to the known, they think the positi討st paradigm has a dualistic view, while the naturalist paradigm believes in an inseparable and interactive relationship be-tween knower and known. For the issue of generalization, Lincoln and Guba emphas泣e that the positivist paradigm asserts the possibility of time- and context-台ee generaliza-tion, but the naturalist paradigm insists generalization must be time- and context-bound. The fourth axiom is causal linkages. They argue that these are assumed by the positivist paradigm, but questioned by the naturalist paradigm. Furthermore, they point out that the naturalist paradigm agrees with ∞rrelation， rather than causation. Regarding the role of an individual value system in inquiry pr。但ss白， they claim that the value-free inquiry endorsed by the positivist paradigm is impossible; by contrast, they agree with the claim of the naturalist paradigm that a value bound inqui可 is inevitable.

3. Edmondson's constructivism 嗯. logical positivism:

Edmondson (1989) classified epistemologicaI beIiefs into constructivism and Iogical positivism. She pointed out a major distinction between the two beliefs -- the way in which knowledge is constructed. From the perspective of logical positivism, knowledge is discovered, and it is an "entity which is external to human experience", while in the view of constructivism, knowledge is constructed by human beings, and it is "always changing and being revised by people who incorporate new information with that which already exists" (Edmondson, 1989, p. 17).

(9)

dif-研究生的科學知識信念的研究張于超 591

ferent explanations. Constructivists emphas泣e scientis臼， interpretation of events or ob-jects, and thus acknowledge that the construction of knowledge is influenced by 屯onceptual goggles", i.e., theories and epistemological commitments. However, logical positivists believe knowledge exists by itself and is separate from human experience; therefore, knowledge is discovered, tested, and constant.

According to Edmondson, the foundation of logical positivism is the belief in the unity of science and an absolute truth. Therefore, logical positivism employs the appli臼 tion of logic, mathemati臼 and theoretical physi臼 to pursue the final truth of the uni-verse, and avoids meaning, personal feelings, and values in the inquiry processes. In ∞ntr品t， ωnstructivism asserts that knowledge ∞nstruction results from people's inter-actions with the world, and th山 knowledge ∞nstruction has contextual and construc-tivist natures (Edmondson, 1989, pp. 27-29).

11. The development and growth of knowledge:

The studies of historical development of scientific knowledge have shed light on the issue of the nature of knowledge and promoted discussion of epistemological beHefs in human knowledge construction. Based on the switch of important theories throughout the development of scientific knowledge, Popper highlighted the crucial role of this dramatic switch in the growth of knowledge, while Kuhn emphasized ordinary scientific practices in the development of scientific knowledge. The theme of their debate was "normal science" vs. "revolutionary scien白" (Lakatos & Musgrave, 1970). Regarding the development of knowledge, Toulmin (1972) proposed the contrasting view that the de-velopment of knowledge resulted from the evloution of human ∞n臼ptual knowledge.

1. Thomas S. Kuhn (1970, 2nd edition): 'Paradign" and 'The Structure of Scientific Revolutions"

(10)

592 師大學報第四卡期

Kuhn's work on the structure of the scientific revolution set a landmark in the study of the development of scien臼. He denied that the development of science has been an accretion of theories, noting that, as a matter of fact, the crucial periods of scientific development involved the pro∞ss of theories competing and replacing each other. For this, he used the term "scientific revolution". However, Kuhn did not think scientific revolutions (or extraordinary science) were the m吋or part of scientific development. On the contrary, the described normal science as the real momentum in the development of science. He used puzzle solving under a set of aα:epted scientific rules, or a "paradigm" to describe normal science.

According to Kuhn's perspective (1970), the nature of the practice of scientisits is puzzle solving. Under certain accepted "paradigms", scientists work on assumed solvable puzzles. If a puzzle can not be solved, scientists usually attribute its cause to their lack of ingenuity, instead of doubting the validity of their paradigms. Due to the beHef in the solvability of puzzles, scientists have continuously worked on them. As a resuIt,

sci-entific development has repidly progressed. Kuhn obviously puts more emphasis on the process of normal science than on "paradigm shift" when describing the development of scientific knowledge.

Kuhn's observation on scientists' practices in the period of normal science finds that the criterion for a "puzzle" is not its intrinsic value at all, but the assured existence of its solution. He points out that "a cure of cancer or the design for lasting peace are often not puzzles at all, largely because they may not have any solution" (Kuhn, 1970, p. 37). This view may have some criticism among scientists, but it does indicate the general perception about science that individual value systems should be excluded from its pr缸，

tlce.

(11)

scientific knowledge contains some major switches of paradigms. This process of "paradigrn shift", described by Kuhn, begin with the appearance of an "anomaly". Then, the accurnulation of "anornalies" results in the "crisis" of a paradigm, and, finally, through the competition of different paradigrns, a newly accepted paradigrn emerges (Kuhn, 1970).

2. Stephen Toulrnin (1972): The evolutionary developrnent of Concepts

Toulrnin explains the phenornenon of hurnan conceptual change and the evolving content of intellectual disciplines by exarnining the history of the development of sci-ence. He uses the ∞ncept of evloution to describe these processes in ths book "Hurnan U nderstanding" (Toulrn恤， 1972).

Instead of isolating hurnan factors from the development of science, Toulmin weaves the conceptualization of hurnan attitudes and scientific subject-rnatter by pointing out that "an adequate definition of a 'discipline' must refer not only to the subject-matter on which its disciplinary activities are focused, but also to the professional attitudes by which those activites are guided" (Toulrnin, 1972, p. 149).

Toulrnin sees concepts as transits of historically developing disciplines; therefore, previous scientific findings rnay be p品sed to suc∞ssive scientists. Aαordingly， he ∞n cludes that any scientific developrnent is based on previous endeavors, and that the pro-自ss of conceptual change is progressive. In addition, he thinks ∞n臼pts can be commu-nicated, rnanipulated, and learned; "scientific ∞n臼pts are, by their ve可 nature， capable of being transrnittable, handed on, and learned, in the processes by which the discipline rnaintains its existence beyond the lifetirne of its original creators" (Toulrnin, 1972, p.

158).

Toulrnin agrees that ∞n臼pts 臼n ∞nstituteωn臼ptual structures to represent intel-lectual disciplines. Although he did not explain exact1y how this happens, he greatly

(12)

emphasizes the functions of ∞ncepts in his elaboration on the evolution of intellectual disciplines. For instance, he says "speaking less portentously, we might rather say that the task of a science is to improve our ideas about the natural world step by step, by identi句ring problem areas in which something can now be done to lessen the gap be-tween the capacities of our current concepts and our reasonable intellectual ideals" (Toulmin, 1972, P. 150). Obviously, in this passage he uses the capacities of our current ∞ncepts to represent human scientific endeavors.

11 1. Edmondson's questionnaire and its validity and reliability

Only a few studies have been conducted to evaluate student epistemological com-mitments. Edmondson's work was finished recently. She developed a questionnaire to evaluate Cornellωllege stud凹的， belifs about scientific knowledge(1989). She conduct-ed a questionnaire survey to examie the epistemological commitments of the college students enrolled in a college Biology course, and then she interviewed 19 students to study their epistemological positions and their subsequent influences on learning. There are 26 items in the questionnaire. The Ïtems were adapted from four sour臼s， name妙， the Cornell Assessment of Scientific Belifs(CASB)(Strike et aI. 1981), Waterman's qu臼 tionnaire for assessing college students' epistemological commitments (Waterman, 1982), the Study Behavior Questionnaire constucted by Biggs(1976, 1979), and the Approaches to Studying Inventory devised by Entwistle(1983). As Edmondson daimed, her ques-tionnaire aimed at students' epistemological commitment in the general sense.

τbe scoring method of the questionnaire was a five - point Likert scale. Each item had five choices, enabling students to express strong agreement, agreement, uncertainty, disagreement, or strong disagreement. Each choice was coded with scores of 1 to 5. Then the sum of the scores of the 26 Ïtems were ∞mputed to represent each student's

(13)

epistemological disposition.

The reliability and validity of the questionnaire were discussed by Edmondson (1989). She used the questionnaire to obtain pretest and posttest scores for a group of students enrolled in a biology course. The sample size was 680. The values of Cronbach's alpha for pretest aÍ1d posttest were .3410 and .3057 respectively. The Pear-son's correlation coefficient of the two tests was .47668.

The Pearson's correlation coefficient and Cronbach's alpha of the two tests are not good enough to claim the questionnaire is a reliable and valid evaluation scale for epis-temological commitments. However, most scales for attitades, emotions, and cognitive capacities do not have high coefficients for reliability and validity.

Undoubtedly, the evaluation of epistemological disposition is at least as difficult as the evaluations of attitudes, emotions, or cognitive capacities. Therefore, as Edmondson described her questionnaire, "There seemed to be some internal consisten可 between items, Although none of these tests indicated that the reliability was particularly high" (Edmondson, 1989, p. 88).

Edmondson discussed epistemological dispositions from two perspecitives: logical positivism and constructivism. She cIassified her subjects into logical positivists， ωn structivists and a mid - range group. High scoring students on the questionnaire were indicated to be more constructivism - oriented, while the students with low scores had a pOSlllVlStíC perspective.

While discussing her survey questions, Edmondsion pointed out two perspectives: one was regarding the questionnaire, and the other, students' epistemological ∞mmit ments. She admitted the questionnaire items did not discriminate well. Most of her respondents clustered around the mean. During her interviews of students, she found that her subjects switched their epistemological positions in answering different

(14)

ques-596 師大學報第四十期

tions.The instability of students' epistemological ∞mmitments ∞n甘ibuted to the low re-liability value r. It also indicated that college students did not have firm epistemologi-cal commitments.

Research Design

Edmondson's(1989)questionnaire was used to evaluate the epistemological commit-ments of Cornell graduate students from Taiwan(Appendix). Basically, the questionnaire survey probed the students' overall perceptions of the nature of scientific knowledge, and their responses were analyzed furhter to examine their specific epistemological be-liefs about the issues of observation, data, inquiry, inquiry results, knowledge, and the growth and development of knowledge.

Sin自 the questionnaire did not delineate students' epistemological belief忌 well，部 al so notes by Edmondson(1989), this research used a different approach to，面前 investi

gate the core factors of the questionnaire, and then, to detemine the students' attitudes toward each factor, iostead of using ∞mposite scores. Obviously, the analysis of stu-dents' commitments to each ∞re factor of the questionnaire would offer more informa-tion than the evaluainforma-tion of their attitudes toward a whole ∞nception， epistemological commltments.

To search for certain factors, a factor analysis w部∞nducted to investigate underly-ing constructs of this questionnaire, but no distinctive factors emerged. Thus, content analysis became a major tool to find out certain core factors. Six relevant ∞re factors were concluded. They are observation, inquiry process, data, inquiry pro臼ss， data, in-qui可 results， knowledge, and the development and growth of knowledge. The first four factors are generally conceived 品 the major procedures of scientific knowledge ∞n struction. The other two factors are directly related to the issue of epistemology(Table 1). Although Edmondson did not elaborate certain ∞re factors of her questionnaire in

(15)

Table 1: The six core factors of the questionnaire and their statements

Core Factor

Statements

Observations "Theories guide observations."(Statement

14)

"Theories limit observations."(Statement

19)

"Scientists' observation are affected by the ideas they

have about their subjects."(Statement

25)

Inquiry

"conceptual frameworks guide scientific inquiry"

Processes

(Statement

12)

"The scientific method of testing hypotheses in

Rigorous

,

controlled experiments ensures that scientific

theories are objectively based and free from bias."

(Statement

5)

Data

"What counts as data is determined by the accuracy of

the measurement

1

make."(Statement 8)

"What counts as data is determined by the theory

1

use

to interpret i

t.

"(Statement

16)

Inquiry

"Results of scientific investigation are true only if

Results

they can be repeated."(statement

22)

Knowledge

"Scientific knowledge is discovered."(Statement

20)

"scientific knowledge is constructed from discovered

facts."(Statement

21)

"Knowledge in the humanities is more subject to

change than scientific knowledge."(Statement

15)

The

"Scientific knowledge is a changing and evolving body

Development

of concepts and theories."(Statement

17)

And Growth

"Knowledge changes as our experience changes."

of Knowledge (Statement

23)

"Which of the fOllowing is closest to your view of how

the growth of scientific knowledge occurs?

(A)Through the

discover、y

of new facts and the

formulation of generalizations from these facts:

(B)Through the reformulation of ideas and the

development of new ideas to solve problems previous

theories could not solve adequately.(Question

2)

her research

,

the Cornell

As

sessment of Scientific Beliefs(CASB) written by Strike et a

l.

and Waterman's questionnaire which are

m吋or

sources of

Ed

mondson's questionnaire

,

(16)

na-598 師大學報第四十期

ture of presuppositions, and the objectivity of scien也爬出峙， et al., 1981); and the fac-tors of the latter are the nature of scientific knowledge, the objectivity in science, and the growth of knowledge (Waterman,

198月-The questionnaire was conducted in winter, 1993. The survey sample consisted of Cornell graduate students from Taiwan, with students enrolled in the Intensive English Program and professional degree seekers in architecture, business, and law schools 眩， cluded. The students' addresses were obtained from the directory of the Federation of Alumni From Taiwan (F AFT). Ninety five questionnaires were sent out through the cornell campus mail service. Fi的 seven questionnaires were returned giving a response rate of 60 % . ln contrast to Edmondson, who asked her subjects not to deliberate on the questions and statements, 1 asked my subjects to think about each item carefully and to define any terms as he or she would like. Finally, they were asked to put their comments on the questionnaire.

Research Findings

甘le questionnaire includes three multiple choice questions and twenty - three statements. Half of the 26 items probe students' general and broad belìefs about scien-tific knowledge, and half of them can be specifically classfied into six categories. There are three statements in the questionnaire regarding the issue of observation, two state-ments for inquiry process, two statements for data, one statement for inquiry results, three statements for knowledge, and two statements for the development and growth of knowledge.

(17)

研究生的科學知識信念的研究張子起 599

their epistemological beliefs about each factor respectively, instead of undertaking a general and integrative, but uninformative, analysis.

Observation:

Three statements in the questionnaire are related to the issue of observation. Statement 14, "theories guide observation", statement 19, "theories limit observation", and statement 詣， "scientists' observations are affected by the idea they have about their subjects'\The responses to these three statements should reveal Taiwanese graduate students' epistemological beHefs concerning observation.

Statements 14 and 15 were designed to contrast the constructivist and logical posi-tivist views on observation. Edmondson thought the construcposi-tivist would agree more with the statement that theories limit observation, while logical positivists would endorse the positive 記lation between theory and observation, agreeing that theories guide ob-servatlOn.

Statement 25 probes each student's opinion on the relation between researchers' preconceived idea about objects and their observations. 1ρgical positivists are assumed to emphasize the objectivity of observation and to disclaim any interference of individu-al preconceived ideas. Howev甘， constructivists would admit the inevitable influences of individual prejudice on the objects of observation.

For statement 25, the majority chose either "agree" or "strongly agree". Only 7 % of the students disagreed with this statement(Table2). Obviously, most of the studets had a constructivist perspective. From their point of view, the observations of scientists are not completely objectiv它，

Compared with the response to statement 25, a greater pcrcentage of the students reponded to the statement 14 with "uncertain". 21

%

of the students were uncertain about the statement (Table 3). By comparison with the 8.8 % who chose "uncertain" for

(18)

Table 2 The response to the statement 25: Scientists' observations are a何'ected by the ideas they have abouttheir subjects.

Response Frequency Percentage

Strongly Disagree 2 3.5 Disagree 2 3.5 Uncertain 5 8.8 Agree 40 70.2 Strongly Agree 8 14.0 」

statement 話， the relationship between theories and obselVations is obvio山ly more ∞n

fusing them than that between researchers' ideas about their objects and their obsetva-tions. More strikingly, 57.9

%

of the students chose either "disagree" or "strongIy

dis-agree" regarding this statement. Apparently, a greater per臼ntage of the students dis-agreed with the statement that theories guide obselVations than otherwise.

Table 3 The response to the statement 14: Theories guide observations.

Strongly Disagree 9 15.8 Disagree 24 42.1 Uncertain 12 21.1 Agree 8 14.0 Strongly Agree 4 7.0 一一

Very interestingly, the students were more decided about the statement that theories limit obselVations than the statement that theories guide obselVations. Only 10.5 % of

the students chose "uncertain" for statement 19, compared to 21.1 % for statement 14 (Tables 3 & 4). Moreover, a greater percentage of the students (52.6 % )disagreed or strongly disagreed that theories limit obselVations. Remember that aα刀rding to

(19)

be-研究生的科學知識信念的研究張子超 601

tween theories and observations.

Table 4 The response to the statement 19: Theories limit observations. '

Strongly Disagree 6 10.5

Disagree 24 42.1

Uncertain 6 10.5

Agree 19 33.3

Strongly Agree 2 3.5

A

comparison between the two responses(Table 5) shows that more student(20 studetnts) either disagreed or strongly disagreed with both statements, that is, they did not believe theories would either guide or limit observations. This shows that the stu-dents held a positivist view on the relationships between theories and observations.

Table 5 Cross frequency of statement 14 and statement 19 statement 19

Statement 14 itrong1y disagree uncertain agree strongly

1 sagree agree strongly agree

_。

2

_。

agree

_。

3 2 3 l uncertain 2 5 1 4

_。

disagree 3 11 3 6 1 strongly disagree 5

_。

3

_。

百le Taiwanese students' r自ponses to the three statements generally reflect a waver-ing position on the issue of observations. They showed a strong constructivist view on the influence of researchers' ideas about their subjects on their observations, while demonstrating a logical positivist perspective on the relationships between theories and observations.

(20)

There are two statements in the questionnaire related to the factor of inquiry pro-cess. One statement probes the relationship between a researcher's conceptual frame-work and his research inquiry. This statement is "∞nceptual framework guides scientific inquiry." The other, statement 5 一 "the scientific method of testing hypotheses in rigor-ous, controlled experiments ensures that scientific theories are objectively based and free from bias," examines respondents' beliefs about the objectivity of the scientific

in-qm可 process.

The students with a constructivist perspective would be willing to accept the role of cognitive framework and to admit the involvement of personal value systems in inquiry processes. In ∞ntrast， students with logi臼1 positivist views would insist on the plausibil-ity of value free inqui可 processes to avoid any involvement of individual cognitive framework.

82.5 % of the students endorsed the statement that 可onceptual frameworks guide scientific inquiry"(Table 6). The response was very decisive. Only 7 % of the students disagreed with this statement. 四lis result obviously represents an important r吋ection of the broadly well con臼ived ∞n臼ption of scientific inquiry that emphas泣es the avoid-ance of individual frameworks in order to obtain objectivity. Above all, the response reveals an acceptance of the personal role in the pro臼ss of scientific inqui可﹒

Table 6 The responses to statement 12: Comceptual frameworKs guide scientific inquiry.

Strongly Disagree

_。

Disagree 4 7.0

Uncertain 6 10.5

Agree 36 63.2

(21)

personal frameworks on scientific inquiry. Statement 5 goes futher to argue that the scientific theories that result from rigorous and controlled experiments are objectively based and free from bias(Table 7). Surprisingly the students' positions on the statement were quite neutral. The premise of statement 5 has been overwhelmingly supported by researchers in almost eve可 field. However, about 40

%

of the students disagreed with it; apparently, the assurance of scientific theories and the confidence in the scientific inquiηpro臼ss are gradually diminishing.

Data:

Table 7 The response to the statement 5: The scientific method of testing hypotheses in rigorous, controlled experiments ensures that scientific theories are objectively based and free from bias.

Strongly Disagree 4 7.0

Disagree 20 35.1

Uncertain 9 15.8

Agree 21 36.8

Regarding the factor of data, two statements are inc1uded in the questionnaire: statement 8 一 "what counts as data is determined by the accura可 of the measurement 1 make"叭". and statement 1師6 一

i泊nt臼erpre前t i江t." They were designed to contradict each other. One statement emphasizes the measurement of data, representing the perspective of logical positivism, and the other stresses the role of theories in interpreting data, with the view of constructivism.

A general comment on the two statements is their vagueness of meaning especially that of the latter.

(22)

dorsed the importance of accurate measnrement as a determinant of data. Apparent1y, their attitudes toward the statement were clear. Only about 15

%

of them disagreed.

Table

8

The response to the statement

8:

What counts as data is determined by the accu阻cy of the measurement I make.

Strongly Disagree l 1.8

Disagree 8 14.0

Uncertain 12 21.1

Agree 29 50.9

In contrast to the logical positivist orientation shown by the response to statement 8, the overall response to statement 16 was close to neutral. 45.7 % of the students reject-ed the idea that the theory usreject-ed to interpret data plays a role in the pro臼SS of its sele-tion(Table 9) 四le percentage of the students who ac臼pted interpretative theory as a determinant of data was much smaller than that that aαepted accurate measurement 部 one. The comparison between the respon田s to the two statements shows that 出e stu-dents held a more logical positivist orientation toward the issue of data.

Table 9 The response to the statement 16: What counts as data is determined by the theory I use to interpret i

t.

Strongly Disagree 3 5.3 Disagree 23 40.4 i Uncertai n 10 17.5 Agree 21 36.8 Strongly Agree

_。

h血!旦巨型lts:

(23)

statement 22: "results of scientific investigation are true only if they can be repeated." The statement underscores the important ∞ncern of any inquiry result -- whether it is correct or not -- and proposes the criterion of repeatability to answer this concern.

1ρgical positivism holds that science is human endeavor to discover the truths of the universe, and, thus, the results of science must be replicable. Constructivism doubts the existence of an absolute truth, and emph品zes the connection between understanding and interpretation in the construction of knowledge. Therefore, the responses of "strongly agree" or "agrl白" indi，臼te the disposition of logical positivism, while the re-sponses of "disagree" or "strongly disagree" represent an inclination toward

∞nstruc-tlVlsm.

64.9 % of the students chose "agree" or 可trongly agree" with this statement(Table 10). The students' responses demons甘ate a s甘ong endorsement for the αiterion of re-peatability in the validity of inquiry results. Obviously, the students were more willing to accept the logical positivist view on this factor.

Table 10 The response to the statement 22: Results of scientific investigation are ture only if they can be repeated.

Strongly Disagree 3 5.3 Disagree 9 15.8 Uncertain 8 14.0 Agree 20 35.1 Strongly Agree 17 29.8 knowledgβ:

The questionnaire contains three statements that investigate the students' epistemo-logicaf beliefs about knowledge. Two of them are designed to discuss the substance of

(24)

scientific knowledge. The assertion that "scientific knowledge is discovered" (statement 20), highlights the discove可 of facts; this view neglects the role of scientists' ∞gnitive contributions in the development of knowledge. ln contrast, the other statement, "scientific knowledge is constructed from discovered facts" (statement 21), emphasizes sdentists' cognitive role in the process of manufacturing knowledge. Another statement, 屯nowledge in the humanities is more subject to change than scientific knowledge" (statement 15), is to examine students' opinion on the differences between scientific know1edge and the knowledge of humanities.

四le students' response to statement 20 was very decisive discovered. Less than 20 % of students doubted this statement (Table 11). Most of them agreed that scientific knowledge is discovered. Perhaps, the picture of scientists conducting experiments in a lab to discover facts has been implanted into the students' minds leading them to agree with the statement.

Table 11 The response to the statement 20: Scientific knowledge is discovered.

Strongly Disagree l 1.8 Disagree 10 17.5 Uncertain 11 19.3 Agree 30 52.6 Strongly Agree 5 8.8 L 可

The students' response to statement 21, "scientific knowledge is constructed from discovered facts," shows their acknowledgment of the weight of scientists'ω伊itive func-tions in the process of knowledge ∞nstruction (Table 12). Considering the responses to the previous statement and to this statement, the students seem to have ∞nflicting posi-tions on the substance of scientific knowledge. However, based on their comments on

(25)

the questionnaire statements, clearly, this is not the case. Rather, they had separate views on discovered facts and construction from discovered facts in the process of knowledge construction. They did think both were needed. Accordingly, they agreed 叫th each statement when the two statements were asked separately. They thought that scientific knowledge is discovered and agreed that scientific knowledge is constructed from discovered facts, too.

Table 12 The response to the statement 21: Scientific knowledge is constructed from discovered facts.

Response

I

Frequency

I

Percentage

Strongly Di sagree

I

0

I

0

Disagree

I

8

I

14.0

lJncertai n I 5 I 8.8

Agree

I

35

I

61.4

Strongly Agree I 9 1 5 . 8

The students' response to statement 15, "knowledge in the humanities is more sub-ject to change than scientific knowledge," represents a ∞mmon belief that the knowl-edge in humanities is dynamic, while scientific knowledge is changeless. There were 64.9% of the students who either agreed or strongly agreed with the statement, and no-body strongly disagreed with it (Table 13).

Table 13 The response to the statement 13: Knowledge in the humanities is more subject to change than scientific knowledge.

Strongly Disagree

_。

Disagree 13 22.8

Uncertain 7 12.3

Agree 29 50.9

(26)

甘le development and growth of knowledge:

Two statements of the questionnaire are connected to the factor of the development and growth of knowledge --. statement 17, "scientific knowledge is a changing and evolv-ing body of concepts and theories," and statement 23, ''knowledge changes 品 our e:耳b rience changes." Basically, the two statements emphasize the infIuences of people's ex-periences on the change of knowledge, and the changing and evol討ng features of the development and growth of knowledge. Obviously, the two statements are oriented to-ward constructivism.

86 % of the students agreed or strongly agreed with statement 17 (Table 14), which states dearly that ∞ncep臼 and theories of scientific knowledge are changing and evo卸﹒ ing. If the response represents the students' epistemological commitments ∞rrectly， then they must doubt the ∞mmon belief that scientific theories and ∞n臼p的 are true and changeless.

Table 14 The response to the statement 17: Scientific knowledge is a changing and evolving body of concepts and theories.

strongly disagree

_。

disagree 3 5.3

uncertain 5 8.8

agree 33 57.9

strongly agree 16 28.1

Like the response to the previous statement, the students' r臼pon田s to the state-ment that "knowledge changes as our experien臼 change，" w部 very decisive. 86 % of the students either agreed or strongly agreed with this statement (Table 15). 四le students' responses to the two statements show well that they were more constructivism oriented

(27)

toward the issue of the growth and development of knowledge.

Table 15 The response to the statement 23: Knowledge changes as our experience changes.

strongly disagree

_。

disagree 3 5.3

uncertain 5 8.8

agree 33 57.9

strongly agree 16 28.1

Students' comments on the questionnaire

Since research on epistemological beHefs has not been well developed and clarified, any questionnaire designed to probe students' epistemologica1 beliefs is inevitab1y∞n﹒ fusing. The complaint about the confusion over some terminology and statemellts reached a ∞nsensus. Moreover, students' epistemologica1 beliefs 訂e su吋ect to vacilla-tion, because their beHefs are a sort of philosophical disposition instead of an exact knowledge recitation.

The confusing terminology and statements raised by the students are as follows: 1. Three statements in the questionnaire designed to evaluate students' opinioos 00

their learning of chemis訂y， biology, and mathemati白， "the (mathemati白 biology，

chemistry) 1 have studied has had little to do with everyday experience," are not c1ear in two ways: first, what content of each su吋ect is, such 品 calculus or geometry in mathemat悶， and what level of each subject is discussed, such 品 high school

(28)

ev-610 師大學報第四十期

eryday experience of conducting experiments, or daily life experien白， or something else.

2. The meanings of question 1 and 2 are ωnfusing， and the students were forced to choose either "the discovery of new facts and the formulations of generalizations from the facts," or "the reformulation of ideas and the development of new ideas to solve problems previo凶 theories could not solve adequately" 的 describe their views on the growth of scientific knowledge. Some of the students pointed out that the two choices were not complete, and they were forced to choose the one that was least far away from their views. Some other students complained that both were needed and interdependent, and were reluctant to throw away any one factor. Another comment about the issue of the growth of scientific knowledge is its dependency on the field of study. One respondent pointed out that in the biological field, the dis∞Ivery of facts and the formulation of generalizations from these facts would be more important than the reformation of ideas and the development of new ideas, while in the field of physics, the answer would be the opposite. Her description of the contribution of newly developed methods to the biological 日eld s甘esses the importan臼 of the devel-opment of methodology on the growth of knowledge in molecular biology.

"In biology, most new discovries and findings are due to newly developed methods. More thean 10 years ago, there was no such field 品 'Molecular Biology', now almost every lab in the biological field uses a 臼rtain degree of molecular tech-nique to study whatever their topic of interest" (A comment of a respondent in the field ofmi口obiology).

3. Some frequently used terrninology turned out confusing to the respondents, such as truth about our universe, scientific methods, universal 甘uths in science, scientific the-ories, scientific truths, and scientific law. Even the terms "data" and "observation"are confusing. One student in the field of electrical engineering pointed out that they

(29)

研究生的科學知識信念的研究張子超 6 呵呵

were usually considered synonyms in the fields of engineering and statisti臼.官le fun-damental concept in the questionnaire -- science -- was marked for clarification in some students' comments too. The confusion over the term 、cience" was whether it should be distinguished from a pure logical discipline like mathemati白， according to ome student.

In spite of the comments on the confusion of terminology and statements in the questionnaire, the question of the real cause of the development and growth of knowl-edge laid out by questions 1 and 2 attracted some students' deliberations on the issue and prompted their explanations. A student in the field of botany described the devel-opment of scientific study as a cyclical pro臼ss: observation.> conducting an experiment

.>results(data) ->formulation of theories or new ideas .> new experiments. Another description about the growth of knowledge proposed by a student in the field of elec剖 cal engineering is self-renewing also: observation ﹒ >modeling ﹒ >the。可﹒ >testing -> observation. Obviously, the two models are quite similar. Both of them are cyclical processes that include observation of facts and formulation of new ideas or theories through observation. The reason why questions 1 and 2 were difficult for the students to answer were that both the choi臼s are included in the well aαepted cydic pro臼ss of knowledge construction, and that most students'research is dedicated to the pro臼ss of observation to formulation of new ideas or theories, paying less attention to the whole process of the development and growth of knowledge.

Finally, students from different fields had different definitions of some terminology in the questionnaire, and had different opinions on the development and growth of sci-entific knowledge.

(30)

Conclusion

In short, the students held di旺'erent epistemological beliefs on different epistemolog-ical issues. Their responses to the questionnaire were internally in∞nsistent. For in-stance, they held a constructivist view on the development and growth of knowledge, while they showed logical positivist dispositions regarding the issue of inquiry results.

The analyses of the questionnaire survey showed the students'epistemological beliefs about the issue of observation were inconsistent across questions. They were aware of the influences of researchers' ideas about their subjects on their observations, but they were not able to clari句 the relationships between theories and observations. Generally speaking, they held a ∞nstructivist view on the influences of researchers' existing ideas on their observations, while they adopted logical positivist perspective that theories guide observations rather than limit them.

The students re∞伊ized the influen臼s of ∞n臼ptual frameworks on inquiry prlα:es8-es, and, surprisingly, half of the students were skeptical that the results of scientific methods of testing hypotheses in rigorous and ∞ntrolled experiments were objective and free from bias. Based on the philosophy under1ying the questionnaire design, the students app訂ently held a constructi、rist perspective on the issue of inquiry pro臼sses. However, they had strong beliefs in logical positivism regarding the issue of the deter-minant of data. They obviously believed that the determinant of data shoule be accurate measurement rather than the inte中retative the。可 ofthe data.

The con臼ph that inquiry results must be repeatable has been endorsed by the ma-jority of the students. However, they agreed that the knowledge in the humanities was more subject to change. Moreover, most of them agreed that scientific knowledge is a changing and evolving body of ∞n臼pts and theories, and knowledge changes 品 human experiences change. App訂閱旬， their epistemological beliefs were inconsistent.

(31)

References:

Biggs, J. B. 1979 "Individual Differences in Study Processes And the Quality of Learning Outcomes." Higher Education 8381-394.

Biggs, J. B. 1976. "Dimensions of Study Behavior: Another 1ρok at A.T.I." British

Jour-nal o[ EducatioJour-nal Psychology 46 68-80.

Confrey, Jere. 1980. "Con臼ptual Change, Number Concepts And The Introduction to Calculus." Ph.D. Thesis, Cornell University, Ithaca, NY.

Edmondson, Katherine McAdoo. 1989. 叮be Influen臼 of Students' Con臼ptions of Sci-entific Knowledge and Their Orientations to learing on Their Choices of Learning Strategy in a College Introductory Level Biology Course." Ph.D. Thesis, Cornell University, Ithaca, NY.

Edmondson, Katherine McAdoo. 1985. "CoUege Students' Con∞ptions of Their Re-sponsibilities for Learning." Master Thesis, Cornell University, Ithaca, NY.

Entwistle, Noel and Paul Ramsden (eds.). 1983. Understanding Student Leaming. New York: Nichols Publishing Company.

Freedman, D. A., R. Pisani, and R. Purv臼. 1978, 1980. Statisti臼. New Yor

k,

NY: Nor-ton Inc.

Kuhn, Thomas S. 1962, 1970. The Structure o[ Scientific Revolutions. Chicago: The Uni-versity of Chicago Press.

Kuhn, Thomas. 1970. "Logic of Dis∞Ivery or P可chology of Re臼缸ch?" in Imre Lakatos Alan Musgrave (eds), Criticism And The Growth o[ 品的wledge. New York, NY: Cambridge U niversity Press.

(32)

i-6 可 4 師大學報第四十期

fornia: Sage Publication, Inc

Novak, Joseph D. 1987. "Human Constructivism: Toward A Unity of Psychological and Epistemological Meaning Making". J. D. Novak (ed.)，抖。ceedings 01 the Second

In-temational Seminar: Misconceptions and Educational Strategies 的 Science and

Mathe-matics. Ithaca, New Y ork: Dept. of Education, College of Agriculture & Life

Sci-en間， Cornell University.

Nussbaum, Joseph. 1989. "Classroom Conceptual Change: Philosophical Perspectives."

Intemational Joumal 01 Science Education. 11. pp. 530 - 540.

Schwab, J. Joseph. 1978. Science, Curriculum, and Liberal Education. Ian Westbury and Neil J. Wilkof, (eds). Chicago: University of Chicago Press.

Strike, A. Kenneth and George J. Posner. 1992. "Re吋sionist Theory of Conceptual Change" in Richard A. Duschl and Richard J. Hamilton (eds). Philosophy 01 Science

Cognitive Psychology, And Educational Theory And Practice. Albany, NY: State

Uni-versity of New Y ork Press.

Strike, A. Kenneth and George J. Posner. 1982. "Epistemologi臼1 Assumptions of Co I-lege Students: An Initial Report." Paper presented at the Annual Convocation of the Northeastern Educational Research Association, Ellenville, NY.

Strike, Keneth, William Gertzog, Peter Hewson, George Hoagland, Joseph Nussbaum, and George Posner. 1981. The 臼rnell Assessment of Scientific Beliefs (CASB): A Report on Ouestionnaire Development. Research Report No. 10, C叮討ωlum Se-ries. Department of Education, Cornell University.

Toulmin, Stephen. 1972. Human Understanding, Vol.I: The Collective Use and Evoluation

01 Concepts. Prin臼ton， New Jersey: Prin臼ton University Press.

Waterman, Margaret Ann. 1982. College Students' Beliefs About Scientific Knowledge: Foundation for Study of Epistemological Commitments in Con臼ptual Change.

(33)

Ph.D. Thesism, Cornell Universi句1.

Appendix

The questionnaire has been 凶ed by Katherine McAd∞Edmondson (1989) to exam-ine Cornell undergraduate students' beHefs in scientific knowledge. 1 will appreciate your careful deliberation on each question; however, don't hesitate to put your answers, since there are no right or wrong answers. Y ou may define any terms and interpret the questions as you Iike. So, your answers would show your inclination to the development of scientific knowledge. For the 1 to 3 questions, ple品e write down your answer before the number of each question.

1. Some people argue that rapid progress in science occurs primarily because of a change in methodology, from one emphasizing speculation about nature to one em-phasizing observation of nature. Other people argt時， however, that rapid progress in

science is not due to a change in method 部 much as it is the result of the develop-ment of powerful new theories or ideas which evolve to the point where rapid changes are possible.

Which of these statements do you agree more?

(A) The former statement (speculation to observation). (B) The latter statement (powerful new ideas).

2. Which of the following is dosest to your view of how the growth of scientific knowl咽 edge occurs?

(A) Through the discovery of new facts and the formulation of generalizations from these facts;

(34)

(B) Through the reformulating of ideas and the development of new ideas to solve problems previous theories could not solve adequately.

3. When scien臼 ωnflicts wi曲 your ∞mmon 記nse， b品ed on persnoal e月Jenen臼 or ob-servatIon,

,' . ..

(A) Science is prob~bly right, common sense is probably wrong. (B) Scien臼 isprobably wrong, common sense is probably right. (c)而e conflict is due to the interpretation you gave the data.

(0) There shouldn't be any conflict between the two; go back and check for missing information.

For the remanining items in this section, you will be given five choi臼s: (A) Strongly Agree

(B) Agree (C) U ncertain (D) Disagree

(E) Strongly Disagree

Please write down your answer before the number of each question.

4. Our scientific ideas will likely appe訂協 much in error to people living in the year 3前約 as the scientific ideas of the Middle Ages to us

5. The scientific method of testing hypotheses in rigorous，∞ntrolIed e哥兒riments en-sures that scientific theories are objectively based and f記e from bi品.

6. As new scientific theories and dis∞veries en∞mpass an everwider range of phenom-ena, the natural laws thougth to regulate the universe will probably 切∞me in帥部﹒ ingly unified.

7. The chemistry 1 have taken has had little to do with eve可daye可erience. 8. What ∞unts 部 data is determined by the aαura可 ofthe measurement 1 make. 9. Scientific truths are discovered by a fewexperts.

(35)

10. There are no universal truths in science.

11. The mathematics 1 have studied has had little to do with everday experien臼. 12. Conceptual frameworks guide scientific inquiry.

13. Through science 1 can know the truth about the wor1d, and 1 can veri句 my claims through the scientific method.

14. Theories guide observations.

15. Knowledge in the humanities is more subject to change than scientific knowledge. 16. What counts as data is determined by the the。可 1 use to interpret it.

17. Scientrfic knowledge is a changing and evolving body of concepts and theories. 18. The biology 1 have studied has had little to do with eve可day experience. 19. Theories limit observations.

20. Scientific knowledge is discovered.

21. Scientific knowledge is constructed fom discovered facts.

22. Results of scientific investigation are true only if they can be repeated. 23. Knowledge changes as our experience changes.

24. We can never be sure that we know the final truth but it is a goal of science to keep trying to find it.

25. Scientists' observations are affected by the ideas they have about their subject. 26. A scientific law is an exact report of a truth about out universe.

Background Information Genders: M/F

Field of study:

Degree: Master / Ph.D.

If you are interested in or have comments on the questionnaire, I'd appreciate that you would let me know.

(36)

研究生的科學知識信念的研究

張子趙

國立台灣師範大學環境教育研究所

摘要

諾瓦克 (Novak) 的『人類建構論 J (Human Constructivism) 指出學習是

一種意義建構的過程，而這意義須包含心理學和哲學認識論的範疇;據此

，本研究在於探討研究生對於科學知識的信念及態度。對於認識論的定義及類剔，本研究是根據努茲係 (Nussbaum) 對哲學派別的分析、林肯及辜巴

(Lincoln & Guba) 對後實證哲學和自然哲學所給的區別、及艾得默生 (Ed

mondson) 對邏輯實證主義和建構主義的描述。本研究設計以問卷調查法為資料收集的工具及以描述性統計分析為資料分析的方法，這研究較偏重質的分析，以調查結果的描述及闡釋代替假說的測試。本研究分析研究生對於問卷土與觀察、研究過程、數據資料、研究結果、知識本質、及科學知識的建構與發展等主題所作的反應，以暸解他們的科學知識信念。研究結呆指出研究生對於不同的主題會採取不同的立場和信念，因而他們的科學知識信念常相互衝突而前後不一致。整體而言，在討論與科學知識有關的認識論議題時，研究生的態度較為中庸，在一些議題上贊成建構論的觀點，但當論及與實際所從事研究工作相關的議題時，貝'J 採取絕對的邏輯實證論觀點。可見邏輯實證論觀點仍鳥學校教育及研究單位的主流，不過研究生能以中庸的態度討論認識論的議題，也說明建構論的觀點漸受注意。關鍵詞:科學知識信念，建構論，邏輯實誼論