Journal of Science Teacher Education

, Volume 23, Issue 8, pp 937–957

Modeling the Interrelationships Among Pre-service Science Teachers’ Understanding and Acceptance of Evolution, Their Views on Nature of Science and Self-Efficacy Beliefs Regarding Teaching Evolution

Authors

    • Faculty of Education, Department of Elementary EducationMiddle East Technical University
    • Aksaray University
  • Ceren Tekkaya
    • Faculty of Education, Department of Elementary EducationMiddle East Technical University
  • Semra Sungur
    • Faculty of Education, Department of Elementary EducationMiddle East Technical University
  • Anne Traynor
    • College of Education, Measurement and Quantitative Methods ProgramMichigan State University
Article

DOI: 10.1007/s10972-012-9296-x

Cite this article as:
Akyol, G., Tekkaya, C., Sungur, S. et al. J Sci Teacher Educ (2012) 23: 937. doi:10.1007/s10972-012-9296-x

Abstract

This study proposed a path model of relationships among understanding and acceptance of evolution, views on nature of science, and self-efficacy beliefs regarding teaching evolution. A total of 415 pre-service science teachers completed a series of self-report instruments for the specified purpose. After the estimation of scale scores using unidimensional IRT models, path analysis suggested that sophisticated views on NOS were associated with higher levels of both understanding and acceptance of evolution, and the higher level of understanding of evolution was related to the higher level of acceptance of evolution. Besides, higher levels of both understanding and acceptance of the theory and naïve views on NOS were found to be associated with stronger self-efficacy beliefs for teaching evolution effectively.

Keywords

EvolutionNature of scienceSelf-efficacyPre-service science teachers

Introduction

Evolutionary theory has long been regarded as the most important and difficult concept to learn and teach (Beardsley 2004; Bishop and Anderson 1990; Jensen and Finley 1995; Finley et al. 1982; Rudolph and Stewart 1998). As a unifying theory in biology, it serves as an explanation of similarities among organisms, biological diversity, and many characteristics of the physical world which are among the most basic characteristics of the world, which in turn help students to realize the biological significance of some phenomena-like reproduction, cell division, and the functioning of ecosystems (Banet and Ayuso 2003; National Academy of Sciences 1998; National Research Council 1996). Hence, evolution is not only one of the important interdisciplinary subject matter in science (Van Dijk and Kattmann 2009), but also an important contemporary global discipline, which should be comprehended by students with scientific explanations. However, for years, research studies have consistently reported the prevalence of the conceptual difficulties that pre-service and in-service teachers experience pertaining to evolution (e.g., Asghar et al. 2007; BouJaoude et al. 2011; Deniz et al. 2008; Graf and Soran 2011; Kim and Nehm 2011; Nehm and Schonfeld 2007; Peker et al. 2010; Rutledge and Warden 2000; van Dijk 2009; van Dijk and Reydon 2010). Researchers attributed these difficulties mainly to religious views (Alters and Nelson 2002; Asghar et al. 2007; Bishop and Anderson 1990), misconceptions and prior ideas about evolution (Gregory 2009; Meir et al. 2007), inadequate level of acceptance of evolution (e.g., Bishop and Anderson 1990; Kim and Nehm 2011; Rutledge and Warden 2000; Peker et al. 2010; Rice and Kaya 2010), and lack of understanding of nature of science (Dagher and BouJaoude 1997, 2005; Kim and Nehm 2011; Rutledge and Warden 2000; Rudolph and Stewart 1998; Scharmann et al. 2005).

Relationships among Understanding and Acceptance of Evolution, and Views on Nature of Science

As far as the association between understanding and acceptance of evolution is considered, related research produced inconclusive results: While some studies found a positive relationship between understanding and acceptance of evolution (Deniz et al. 2008; Rutledge and Warden 2000), others reported no relationship between these two variables (e.g., Bishop and Anderson 1990; Sinatra et al. 2003). Although Deniz et al. (2008) found a low and positive relationship between understanding and acceptance of evolution, they suggested not overstating the significance of the understanding of evolution in expressing the variance in accepting evolution, and they recommended that participants’ religious orientation and views on nature of science should be considered in explaining variance in the acceptance of evolution. In addition, Sinatra et al. (2003) claimed that students with a low level of understanding of evolution may accept its validity, or those who do not accept evolution as a valid scientific theory may have an understanding of it. Bishop and Anderson (1990) attributed the lack of significant association between understanding and acceptance of evolution to the dependence of acceptance of evolution more on “social, religious, and metaphysical commitments than on an analysis of scientific evidence” (p. 426). Consequently, researchers realized that the understanding of evolution is not closely related to the acceptance of evolution and called for studies exploring presence of some other influential constructs. As a result of these studies, nature of science (NOS) emerged as one of the important constructs affecting understanding and acceptance of evolution. Several lines of investigations indicated the importance of having certain level of sophisticated views on nature of science for understanding and acceptance of evolution (e.g., Akyol et al. 2010; BouJaoude et al. 2011; Clough 1994; Dagher and BouJaoude 2005, Dagher and BouJaoude 1997; Johnson and Peeples 1987; Kim and Nehm 2011; Lombrozo et al. 2008; Rutledge and Warden 2000). Besides, the study by Rutledge and Mitchell (2002) revealed that teachers with naïve views on nature of science may not be capable of distinguishing between scientific validity of evolution and religious beliefs. Similarly, in his study, Clough (1994) provided evidence that public debate between evolution and creation largely traces back to misconceptions related to the NOS. On the other hand, Sinatra et al. (2003) found that undergraduates’ acceptance of tentativeness of scientific knowledge was associated with the acceptance of human evolution but not with the acceptance of animal evolution. The authors concluded that approving the tentativeness of knowledge may be particularly significant for accepting human evolution, which is regarded as especially tentative by many nonscientists. In their study, Nehm and Schonfeld (2007) reported that increase in teachers’ knowledge of evolution and the nature of science did not have a great influence on their preference about teaching evolution; most of the teachers still chose to teach antievolutionary thoughts. The authors claimed that although it may be necessary for teachers to have knowledge of evolution and nature of science above a certain level to demonstrate a change in their preference, it is not enough to diminish their antievolutionary thoughts. In another notable study, Evans (2001) demonstrated that children’s evolutionist and creationist beliefs developed as a function of beliefs of the community and their intuitive beliefs about origins of species. In their review article addressing the distinction between knowledge and beliefs, Southerland et al. (2001) mentioned that the distinctions between knowledge and belief comprise both philosophical and empirical factors and individual’s epistemological position (i.e., foundationalist, objectivist, nonfoundational fallibilist, or radical constructivist) has an influence on the definition of knowledge and construction of belief. Specifically, Southerland et al. (2001, pp. 339–340) stated that “In terms of evolution education, an educator that employs a fallibilist, nonfoundational epistemology would understand descriptions of speciation by natural selection to be part of a wider knowledge framework. In contrast, the notion of a spontaneous production of species by a creator would be seen as a component of a larger belief system. The former is based on a rational evaluation of scientific evidence and the latter is a matter of faith, an extraempirical conviction” (see also Nehm and Schonfeld 2007; Smith and Scharmann 1999). Hence, comprehension of evolution and nature of science may not guarantee accepting evolution as a scientifically valid theory. The picture that emerged from these studies is that in order to understand and accept evolution, it is important to appreciate that scientific theory is reliable since it is validated through accumulation of overwhelming evidence from various methods, that scientific theory necessitates scientists’ inference, imagination, and creativity, that scientists are influenced by social factors, their personal beliefs, and previous studies, and that scientific theory is subject to change as the result of new research and perspectives, but it does not develop into a law (see Lombrozo et al. 2008; McComas 1998). On the basis of these results, in the current study, we proposed a path model suggesting that views on nature of science are directly related to both understanding and acceptance of evolution. In addition, a link was hypothesized between understanding of evolution and acceptance of evolution (see Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs10972-012-9296-x/MediaObjects/10972_2012_9296_Fig1_HTML.gif
Fig. 1

The proposed path model of views on nature of science, understanding and acceptance of evolution, and self-efficacy beliefs for teaching evolution

Self-efficacy Beliefs for Teaching Evolution

As highly highlighted by many organizations in the field (e.g., American Association for the Advancement of Science 2006; National Academy of Sciences 1998; National Association of Biology Teachers 2008; National Science Teachers Association 2003), available research has demonstrated that evolution has not been given adequate emphasis during instructions (e.g., Aguillard 1999; Rutledge and Mitchell 2002; Moore 2007; Moore and Kraemer 2005; Shankar and Skoog 1993; Trani 2004). Studies attempted to identify the possible reasons for avoidance of teaching evolution in science classes listed several concerns raised by pre-service and in-service teachers. Among them are their own inadequate content knowledge on evolution (Asghar et al. 2007; BouJaoude et al. 2011; Griffth and Brem 2004; Nadelson and Nadelson 2010; Sanders and Ngxola 2009), their insufficient knowledge about specific instructional strategies to teach evolution (Asghar et al. 2007; BouJaoude et al. 2011; Sanders and Ngxola 2009), conflicting views between evolution and religious beliefs (Asghar et al. 2007; BouJaoude et al. 2011; Sanders and Ngxola 2009), and type of school (Asghar et al. 2007). As a result of these studies, one can conclude that teachers’ inadequate knowledge of both evolution and instructional strategies to teach evolution, as well as conflicting views between evolution and religious beliefs appeared significant concerns that may have an influence on their instructional decisions about evolution. For an effective evolution teaching, teachers should be comfortable and confident in their knowledge and abilities to teach evolution. Such beliefs have been called self-efficacy (Bandura 1977), and perceived self-efficacy has been defined as “beliefs in one’s capabilities to organize and execute the courses of action required to produce given attainments” (Bandura 1997, p. 3). In the theory of self-efficacy, Bandura (1977, p. 193) suggested that behavior is based on “person’s estimate that a given behavior will lead to certain outcomes” (outcome expectancy) and “conviction that one can successfully execute the behavior required to produce the outcomes” (efficacy expectation). Based on Bandura’s self-efficacy theory, it is assumed that teachers’ self-efficacy beliefs are in relation to their teaching behaviors (e.g., Enochs and Riggs 1990; Riggs and Enochs 1990). On the basis of these research studies, in this study, we examined pre-service science teachers’ self-efficacy beliefs regarding teaching evolution in an effort to gain insights about their probable teaching behaviors in their future classes with regard to evolution.

In the evolution education literature, the work of Scharmann and Harris (1991), which examined the effect of a 3-week institute providing an understanding of nature of science, enhanced content, and an environment to talk about problems about teaching of evolution on secondary biology and earth science teachers’ self-confidence concerning teaching of evolution indicated a significant increase in teachers’ acceptance of evolution, and a high significant reduction in their anxiety concerning teaching evolution. In their study, Griffith and Brem (2004) demonstrated the biology teachers’ lack of confidence in their knowledge of evolution. The authors pointed out that workshops on the most recent information about evolution and how to teach it may enhance biology teachers’ confidence in teaching evolution. In this aspect, we believe that examining the factors contributing pre-service and in-service teachers’ self-efficacy beliefs regarding teaching evolution is crucial for improving teaching evolution in science classes. It is clear that teachers with thorough understandings of evolution and nature of science are more likely to feel confident about their abilities to teach evolution effectively. Indeed, it is difficult to expect teachers to be confident in their teaching of the concept whose scientific validity they do not accept. Therefore, in an attempt to further our understanding of the factors that predict self-efficacy beliefs regarding teaching evolution, this study examined pre-service science teachers’ understanding of evolution, acceptance of evolution, and their views on nature of science. More specifically, this paper seeks to address following research question: In what ways are the understanding of evolution, acceptance of evolution, and views on nature of science related to self-efficacy beliefs regarding teaching evolution? Along these lines, we proposed that the understanding of nature of science would be linked to self-efficacy beliefs for teaching evolution directly and indirectly through its effect on understanding and acceptance of evolution.

Method

Participants

The participants of the current study were 415 pre-service science teachers (133 males and 282 females, mean age = 22.02 years, SD = 1.32 years) from seven public universities located in different regions of Turkey. Of the 415 participants, 309 were junior and 106 were senior pre-service science teachers. The universities were selected based on convenience due to the limitations of travel, time, and cost. Given the fact that the present study focused on evolution, juniors and seniors were selected purposively from the universities since they completed general biology courses.

Majority of participants reported to have unemployed mothers (81.7 %) and employed fathers (98.2 %). In terms of education level, 8.9 % of mothers and 1 % of fathers were illiterate. In addition, while 58.4 % of mothers and 41.2 % of fathers had elementary school degree, 23.4 % of mothers and 31.6 % of fathers had attained high school education. Moreover, about tenth of mothers (9.4 %) and one-quarter of fathers (24.8 %) had graduated from university, whereas relatively small percentages of mothers (0.2 %) and fathers (1.5 %) had master or PhD degrees.

Participants were also asked for a self-evaluation of both their knowledge about and level of interest in evolution. Of the participants, 59.4 % claimed to have “a little” of interest in evolution and 17.8 % claimed to have “a great deal” of interest in evolution, whereas 9.3 % rated themselves as not being interest in evolution at all. In addition, while more than half of the participants reported knowing “a little” about evolution (60.9 %), few reported having “a lot” knowledge about evolution (2.7 %). In terms of prior experience, considerable amount of participants reported that they had not been taught about the evolution in elementary and high school (36.5 %) and at university (16.7 %). In relation to the questions about necessity of introducing evolution in elementary science classes, 68.2 % of the participants viewed the teaching of evolution as important as teaching of the rest of the science topics; however, 20.3 % of them disagreed with this view. Besides, while majority of the participants believed that teaching evolution was worth the effort and time (74.7 %), 14.3 % did not agree with this opinion. Participants were also requested to indicate the sources they obtain their knowledge in evolution. Findings revealed that their main sources of information were school (81.1 %), followed by newspapers and magazines (75.4 %) and internet (75.4 %). Television (65.5 %) and friends (43.9 %) were the next. A relatively small percentages (17.9 %) of the participants identified family as their main source. However, 7.6 % stated that they had not obtained any knowledge concerning evolution.

In Turkey, higher education system is centralized under the supervision of the Council of Higher Education. In order to register in an undergraduate program in Turkey, individuals are required to have a high school diploma and a necessary score on the student selection and placement examinations (The Council of Higher Education 2010). Among the undergraduate programs, an elementary science education program is a four-year program that trains future elementary science teachers (Grades 6–8). Individuals completing all requirements of the program are awarded the degree of Bachelor of Science in Elementary Education. Due to centralized higher education system in Turkey, all elementary science education programs need to follow similar coursework suggested by the Council of Higher Education. Pre-service science teachers are required to take a number of courses related to science (e.g., general biology, human anatomy and physiology, genetics and biotechnology, evolution) as well as several courses related to teaching profession.

Measures

The research was carried out using a direct administration of (1) Evolution Content Knowledge Test (Rutledge and Warden 2000), originally developed by Johnson (1985), (2) Measure of Acceptance of the Theory of Evolution (MATE, Rutledge and Warden 1999), (3) Nature of Science as Argument Questionnaire (NSAAQ, Sampson and Clark 2006), and (4) Personal Science Teaching Efficacy Beliefs regarding Evolution (PSTE) adapted from the existing instrument (Enochs and Riggs 1990). Besides, information related to participants’ gender, age, year in the elementary science education program, and self-perceived knowledge about and interest in evolution was collected.

Evolution Content Knowledge Test (ECKT)

A 21-item Evolution Content Knowledge Test with one correct answer and four distracters, originally developed by Johnson (1985) and modified by Rutledge and Warden (2000), was used to assess pre-service science teachers’ understanding of evolution. The test covers following content areas: natural selection, extinction processes, homologous structures, coevolution, analogous structures, convergent evolution, intermediate forms, adaptive radiation, speciation, evolutionary rates, fossil record, biogeography, environmental change, genetic variability, and reproductive success. Rutledge (1996) reported that the test items were evaluated in terms of the contribution of each item to the measurement of the intended construct by four university professors who are experts in the fields of evolutionary biology, science education, and the philosophy of science and that the test had satisfactory reliability (α = 0.78). The test was translated and adapted into Turkish by Deniz et al. (2008). Deniz et al. reported that internal consistency estimate for the test computed by Cronbach’s alpha was 0.98. However, in the current study, the reliability coefficient computed by Cronbach’s alpha for the final ECKT including 12 items, which were appropriate for computing scale scores based on unidimensional item response theory (IRT) models, was 0.43, which suggests that effects in the current study may be somewhat underestimated due to the lower-than-desired level of reliability, yet are comparable to the estimates in other studies (e.g. Athanasiou and Papadopoulou 2011; Kim and Nehm 2011).

Measure of Acceptance of the Theory of Evolution (MATE)

The MATE (Rutledge and Warden 1999) was used to assess participants’ acceptance of evolutionary theory. It consists of 20 items (10 positively phrased and 10 negatively phrased items) about the process of evolution, scientific validity of evolutionary theory, evolution of humans, evidence of evolution, scientific community’s view of evolution, and the age of the Earth. The items are scored on a 5-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). In computing the MATE score of each participant, the negative items were reversed to have higher scores reflecting higher levels of acceptance of evolutionary theory. Rutledge (1996) reported that the instrument was assessed in terms of the contribution of each item to the measurement of the intended construct by four university professors who are experts in the fields of evolutionary biology, science education, and the philosophy of science and that the test had satisfactory reliability (α = 0.84). The MATE was translated and adapted into Turkish by Tekkaya et al. (2010) and examined by a group of experts in science education. For the present study, the reliability coefficient computed by Cronbach’s alpha for the final MATE including 13 items, which were appropriate for computing scale scores based on unidimensional item response theory (IRT) models, was found to be 0.89, suggesting satisfactory reliability.

The Nature of Science as Argument Questionnaire (NSAAQ)

The NSAAQ is a 5-point Likert-type scale developed by Sampson and Clark (2006) used to assess participants’ understanding of nature of science. The questionnaire consists of 26 contrasting alternatives items (one of the viewpoints indicates a more naïve ideas about the nature of science, and the other viewpoint is in line with the idea of science as a process of explanation and argument) that are divided into 4 dimensions about the nature of scientific knowledge (6 items); the methods that can be used to generate scientific knowledge (6 items); what counts as reliable and valid scientific knowledge (7 items); and what role scientists play in the generation of scientific knowledge (7 items). In computing the NSAAQ score, the negative items were reversed to have higher scores reflecting a more sophisticated understanding of nature of science. Sampson and Clark (2006) reported that the NSAAQ had high content, translational, face, convergent, discriminant, and concurrent validity and reasonable reliability (α = 0.70). The NSAAQ was translated and adapted into Turkish by Tekkaya et al. (2010) and examined by a group of experts in science education. For the present study, the reliability coefficient computed by Cronbach’s alpha for the final NSAAQ including 13 items, which were appropriate for computing scale scores based on unidimensional item response theory (IRT) models, was found to be 0.67, suggesting satisfactory reliability.

Personal Science Teaching Efficacy Beliefs Regarding Evolution (PSTE)

A five-item scale that was adapted from Enochs and Riggs’ (1990) science teaching efficacy belief instrument was used to assess participants’ Personal Science Teaching Efficacy Beliefs Regarding Evolution. The items are scored on a 5-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). In computing the PSTE score of each participant, the negative items were reversed to have higher scores reflecting a higher level of confidence about abilities to teach evolution. The science teaching efficacy belief instrument was adapted and translated into Turkish by Tekkaya et al. (2004). For the present study, the reliability coefficient computed by Cronbach’s alpha was found to be 0.74, suggesting satisfactory reliability.

Data Collection and Analysis

The study was carried out during 2009–2010 spring semester at seven public universities of Turkey. Because of constraints about travel, time, and cost, the surveys were posted to the instructors who studied in the related universities. Pre-service science teachers in the universities were given necessary information about the study and invited to participate in the study voluntarily. The instruments were administered to the pre-service science teachers who wanted to participate in the study. On average, it took respondents about 30 min to complete the instruments. The completed surveys were posted to the authors by the instructors.

In the present study, scale scores for response data from the four instruments were estimated from unidimensional item response theory (IRT) models using the software Mplus. Considering the related literature, a path model (see Fig. 1) was proposed to explain how understanding and acceptance of evolution as well as views on nature of science are related to pre-service science teachers’ self-efficacy beliefs regarding teaching evolution. The path analysis was utilized to assess the proposed model through LISREL 8.30 program.

Results

Preliminary Analysis

Before path analysis, scale scores for response data from the four instruments were estimated from unidimensional item response theory (IRT) models. Since the estimation method that was to be used to generate IRT scale scores performs poorly when pairwise contingency tables, or “cross-tabulations,” of the item data contain any cell frequencies of zero, upper or lower score categories for some items in the PSTE, NSAAQ, and MATE were collapsed. If the sample size was larger, more data would be anticipated in the extreme categories for the items (i.e., “1” and “5”), eliminating the need to collapse response categories; however, the obtained sample size was adequate to perform the analysis after consolidating the categories for some items.

Initially, item analyses were conducted for each instrument to examine item functioning. The distractor analysis of the NSAAQ identified seven items on which high-scoring examinees tended to perform poorly, with their responses often evenly distributed across several response options. The marginal distribution of ECKT sum scores suggested that this test was generally difficult for the examinees; out of 21 items, the maximum sum score was 18, and the median sum score only 8. NSAAQ sum scores were approximately normally distributed, with a median of 89 on a scale from 26 to 130. The distractor analysis of the ECKT identified four items for which high-performing examinees were more likely to select a wrong response than the correct response. Because the item response theory (IRT) model that was to be used to produce scale scores assumes that the probability of correct response to a test item increases monotonically with examinee ability, these four items were excluded from further analyses.

Following the distractor analysis of the ECKT, for all four instruments, polyserial correlations between each item and the total sum score from its instrument were computed (e.g., McDonald 1999). Items with polyserial correlations less than .3 were excluded from all further analyses unless their content suggested they should be retained in the scale. This preliminary analysis flagged 3 items from the MATE and 2 from the ECKT with response data that were not sufficiently related to its corresponding total sum score. Since theory suggested that each instrument should be measuring either a single-latent construct or multiple-related constructs, these items were excluded from further analyses.

Although prior research utilizing the MATE (Rutledge and Warden 1999; Rutledge and Sadler 2007) and PSTE (Riggs and Enochs 1990; Bleicher 2004) suggested response data from these instruments formed unidimensional scales; that is, all items measured the same latent construct “acceptance of the theory of evolution” or “personal science teaching efficacy beliefs,” respectively, previous research had been conducted in different populations than that sampled in this study, so scoring of the MATE and PSTE data began with a nonlinear exploratory factor analysis (EFA), appropriate for the ordinal items composing the MATE. Scoring of the NSAAQ and ECKT also began with an EFA, given that previous research on the NSAAQ suggested it tends to generate highly multidimensional data (Sampson and Clark 2006) and that the factor structure of the ECKT had not previously been examined. EFA solutions were considered for up to three factors on the MATE, and five factors on the NSAAQ and ECKT, but only a single factor on the five-item PSTE. To select factor structures that were as parsimonious as possible, and substantively plausible given each item’s content, factor loading patterns were examined, and the chi-square test of model fit statistic, root mean square error of approximation (RMSEA), Tucker–Lewis index (TLI), and comparative fit index (CFI) were considered qualitatively. Items with loadings that failed to exceed .3 in any of the possible factor solutions were dropped from the scales, and the EFA repeated. Factors that had large loadings for only one or two items were noted, and, based on the consideration of item content, a few items that loaded primarily on these minor factors were excluded from subsequent scoring of the instruments. Exclusion of items that loaded poorly on all factors, or loaded heavily on factors that appeared to be only weakly related to the main construct measured by each instrument, removed 4 additional items from the MATE, 2 from the ECKT, and 9 from the NSAAQ.

Since the results of the final EFA indicated that a single underlying factor could be posited for the reduced PSTE, MATE, and ECKT item sets, scale scores for response data from the four instruments were estimated from unidimensional item response theory (IRT) models via robust weighted least squares estimation, implemented in the software Mplus (Muthén and Muthén 2010). To estimate the scores on the ECKT, composed of binary items, a two-parameter normal ogive model (e.g., Lord 1980) was utilized. To estimate the scores from the ordinal data produced by the MATE and PSTE, graded response IRT models (Samejima 1969) were used.

The results of IRT modeling indicated that the fit of a graded response model to the MATE data was adequate: χ2(63) = 149.46, p < .0001, RMSEA confidence interval = (0.046, 0.069), TLI =0.98, and CFI = 0.98. The coefficient alpha reliability of data for the final MATE scale was 0.89. Fit of a graded response model to the PSTE data was good: χ2(4) = 9.62, p = .05, RMSEA confidence interval = (0.006, 0.107), TLI = 0.99, and CFI > 0.99. The coefficient alpha reliability of the PSTE data was 0.74. Fit of a two-parameter normal ogive model to the ECKT data was adequate: χ2(65) = 93.19, p = .01, RMSEA confidence interval = (0.016, 0.046), TLI = 0.80, and CFI = 0.83. The coefficient alpha reliability of the final ECKT scale data was 0.43. To some extent, the reliability estimate, CFI, and TLI may have been low because a two-factor model, rather than a unidimensional model, would have been preferred to fit the ECKT data. Items formed distinct clusters loading on each of two highly correlated factors that might be interpretable as “evolution knowledge” (i.e., factual recall) and “evolution understanding.” However, for simplicity, the best-fitting unidimensional IRT model was used to generate ECKT scores. Fit of a graded response model to the NSAAQ data was marginally adequate: χ2(35) = 120.79, p < .0001, RMSEA confidence interval = (0.062, 0.092), TLI = 0.88, and CFI > 0.85. The coefficient alpha reliability for the final NSAAQ scale data was 0.67 (The item parameter estimates for response data from the four instruments were given in Appendices 1, 2, 3, and 4).

Interrelationships Among Understanding and Acceptance of Evolution, Views on Nature of Science, and Self-Efficacy Beliefs for Teaching Evolution

The conceptual model presenting the relationships among pre-service science teachers’ understanding and acceptance of evolution, their views on nature of science, and their self-efficacy beliefs for teaching evolution was tested through path analysis. The goal of path analysis is to determine how well a proposed model with a set of specified relationships among variables, explains the observed relationships among these variables. Accordingly, path analysis has the ability to reveal the relations among several variables simultaneously and show the relative contribution of each variable to the observed variance. Thus, this technique is different from separate regressions approach in that path analysis provides both direct and indirect effects and a test of the overall model fit (Savalei and Bentler 2006).

In the present study, the proposed model suggested that pre-service teachers’ views on nature of science were directly linked to their understanding and acceptance of evolution and their self-efficacy beliefs for teaching evolution, and also indirectly to their self-efficacy beliefs through its effect on pre-service teachers’ understanding and acceptance of evolution. Moreover, a link was specified from the understanding of evolution to the acceptance of evolution (See Fig. 1). Because the reliability coefficient for the ECKT was quite low and low alphas can have a biasing effect on path coefficients, which may result in relations that switch signs of path coefficients (Bollen 1989; Cohen et al. 1990), sensitivity analysis was conducted to correct for measurement error. In sensitivity analysis, the score for each construct was used an indicator of its latent variable fixing factor loading to 1. Then, the error variance was adjusted for the scores considering the alpha coefficient for the related construct. Finally, the fit of the model was compared with that of the original model (see Table 1). The fit indices of the two models were exactly the same indicating a perfect model-to-data fit. However, the two models explained different amounts of variance in the self-efficacy beliefs for teaching evolution: While the original model explained only 10 % of the variance, the variance explained by the adjusted model was 26 %. Because the adjusted model accounted for a greater portion of variance, interpretations were made based on this model.
Table 1

Comparison of original and adjusted models

Model

GFI

NFI

CFI

SRMR

RMSEA

R2

Original

1.00

1.00

1.00

.00

.00

0.10

Adjusted

1.00

1.00

1.00

.00

.00

0.26

Results revealed that the model explained the data well. Indeed, fit indices were indicative of perfect model-to-data fit: GFI = 1.00, NFI = 1.00, CFI = 1.00, SRMR = 0.00, and RMSEA = 0.00. Since fit indices implied a theoretically sound model, standardized path coefficients for direct, indirect, and total effects were examined (Table 2). The standardized path coefficients for direct effects were also graphically displayed in Fig. 2.
Table 2

Path coefficients of understanding of evolution, acceptance of evolution, views on nature of science, and self-efficacy beliefs regarding teaching evolution

Effect

Direct effects

Indirect effects

Total effects

Standard errors of the estimates

t

R2

On acceptance

     

0.17

 Of NOS

0.32

0.05

0.37

0.08

4.88

 

 Of Understanding

0.18

0.18

0.16

2.18

 

On Understanding

     

0.08

 Of NOS

0.28

0.28

0.06

3.18

 

On self-efficacy

     

0.26

 Of acceptance

0.21

0.21

0.06

2.98

 

 Of understanding

0.45

0.04

0.49

0.16

4.37

 

 Of NOS

−0.17

0.21

0.04

0.08

−2.11

 
https://static-content.springer.com/image/art%3A10.1007%2Fs10972-012-9296-x/MediaObjects/10972_2012_9296_Fig2_HTML.gif
Fig. 2

Path coefficients of views on nature of science, understanding and acceptance of evolution, and self-efficacy beliefs for teaching evolution

In the model (see Fig. 2), pre-service science teachers’ views on nature of science together with their understanding of evolution was accounted for 17 % of variance in the acceptance of evolution. Parameter estimates showed that pre-service teachers’ understanding of evolution (β = 0.18) and views on nature of science (β = 0.32) were significantly linked to their acceptance of the theory. This finding implied that both higher levels of the understanding of evolution and sophisticated views on nature of science were associated with higher levels of acceptance. Concerning the relationship between pre-service teachers’ views on nature of science and their understanding of evolution, the results revealed that pre-service teachers’ views on nature of science (β = 0.28) were significantly related to their understanding of the theory. This finding indicated that pre-service teachers having sophisticated views on NOS are likely to have a better understanding of evolution.

In addition, the results demonstrated that pre-service teachers’ views on nature of science and their understanding and acceptance of evolution explained 26 % of variance in their self-efficacy for teaching evolution. More specifically, parameter estimates revealed significant positive relationships between pre-service teachers’ both understanding (β = 0.45) and acceptance (β = 0.21) of evolution and their self-efficacy for teaching evolution. On the other hand, there was a significant negative association between participants’ views on NOS (β = −0.17) and their self-efficacy for teaching evolution. The largest total effect on self-efficacy was from the understanding of evolution (see Table 2). The total effect of views regarding NOS on self-efficacy was 0.04, while its indirect effect was 0.21, which could be attributed to the positive direct effect of this variable on acceptance of evolution (β = 0.32). Besides, the total effect of acceptance of evolution on self-efficacy was 0.21. Overall, these findings suggested that, for pre-service science teachers, higher levels of both understanding and acceptance of the theory and naïve views on NOS were associated with stronger self-efficacy beliefs for teaching evolution effectively.

Discussion and Conclusion

The present study provides further support to the growing body of literature reporting that nature of science can be a critical factor in understanding and acceptance of evolution. The path analyses suggested that the more sophisticated views on nature of science were associated with higher levels of both understanding and acceptance of the theory. This finding is not surprising considering that individuals with sophisticated views on nature of science believe that (1) scientific theory is reliable since it is validated through the accumulation of overwhelming evidence from various methods, (2) scientific theory necessitates scientists’ inference, imagination, and creativity, (3) scientists are influenced by social factors, their personal beliefs, and previous studies, and (4) scientific theory is subject to change as the result of new research and perspectives, but it does not develop into a law (see Lombrozo et al. 2008; McComas 1998). Accordingly, it is reasonable to assume that thorough knowledge related to nature and process of a scientific theory can enhance the understanding of the theory of evolution and foster the acceptance of its scientific validity. The present findings concur with those revealed by the previous research that understanding of NOS is linked to the understanding of evolution (e.g., Rutledge and Warden 2000; Kim and Nehm 2011) and the acceptance of evolution (e.g., Akyol et al. 2010; Johnson and Peeples 1987; Kim and Nehm 2011; Lombrozo et al. 2008; Rutledge and Warden 2000). As far as the relationship between understanding and acceptance of evolution is considered, the present study revealed that pre-service science teachers’ understanding of evolution was significantly related to their acceptance of the theory. This implied that pre-service teachers having sound understanding of evolution are likely to accept scientific validity of evolutionary theory. The present result is also consistent with studies pointing to a positive correlation between understanding and acceptance of evolution (e.g., Deniz et al. 2008; Rutledge and Warden 2000).

Results of the present study also showed that pre-service science teachers with high levels of understanding and acceptance of evolution appeared to be confident about their knowledge necessary to teach evolution to their students effectively, their abilities to use various teaching strategies to deal with evolution, and to develop teaching and learning materials about evolution. The present findings are not surprising considering the research suggesting that conceptual understanding level has a positive impact on one’s abilities to teach science effectively (e.g., Tekkaya et al. 2004). In addition, it is difficult to expect teachers to be confident in their teaching of the concept whose scientific validity they do not accept. As far as pre-service and in-service teachers’ concerns about teaching evolution in elementary and high school classes are considered, these concerns seemed to generally stem from their low sense of self-efficacy beliefs regarding teaching evolution, such as inadequate knowledge of evolution and of pedagogical techniques to teach evolution (Asghar et al. 2007; BouJaoude et al. 2011; Griffth and Brem 2004; Nadelson and Nadelson 2010; Sanders and Ngxola 2009), and their personal beliefs about evolution (Asghar et al. 2007; BouJaoude et al. 2011; Sanders and Ngxola 2009). Therefore, it can be inferred that teachers who do not have thorough knowledge of evolution and do not accept evolution are not confident in their knowledge and abilities to teach evolution. Concerning the relationship between pre-service science teachers’ views on NOS and their teaching self-efficacy, on the other hand, the present study revealed that pre-service teachers’ views on nature of science were negatively linked to their self-efficacy beliefs regarding teaching evolution. However, when the indirect effect of NOS views on teaching self-efficacy through their effect on understanding and acceptance of the theory was examined, positive associations were found. This finding is interesting that sophisticated views on nature of science were directly associated with lower sense of self-efficacy beliefs regarding teaching evolution, but when the relationship of the NOS views with understanding and acceptance of evolution is taken into consideration, sophisticated views on nature of science were found to be indirectly associated with stronger self-efficacy beliefs. Thus, this result may suggest the presence of some other factors (e.g., social-cultural factors) affecting the relationship between NOS views and self-efficacy beliefs regarding evolution, which is a controversial issue. Therefore, further research is needed to better understand these observed relations and their possible causes.

In closing, this study has revealed insights about the relationships among pre-service science teachers’ understanding and acceptance of evolution, their views on nature of science, and their self-efficacy beliefs regarding teaching of evolution. Nevertheless, the present study also has some limitations that future research can address. One limitation involves the reliance on self-reported measures such as MATE, NSAAQ, and PSTE to measure participants’ acceptance of evolution, views on nature of science, and self-efficacy beliefs about teaching evolution, respectively. Self-reported instruments may not capture participants’ actual views so that in-depth interviews with participants are required to verify the consistency and accuracy of the present findings. Moreover, in the present study, the understanding of evolution was measured by Evolution Content Knowledge Test with only 12 multiple choice questions that were appropriate for computing scale scores based on unidimensional item response theory (IRT) models. Hence, although the original test consisted of even 21 items, future studies can develop and use more comprehensive tests or can use qualitative approaches to determine the understanding of evolution. Furthermore, the reliability coefficient for Evolution Content Knowledge Test was low. However, it should be noted that the reliability coefficient is comparable to the estimates in other studies (e.g., Athanasiou and Papadopoulou 2011; Kim and Nehm 2011). In addition, in the current study, to correct for measurement error, the results from the adjusted model were interpreted. Still, the present study points out a need for future investigations to evaluate psychometric properties of the test for Turkish sample. Moreover, preliminary data analysis results suggested that while the NSAAQ’s items might individually be useful to determine views on nature of science, many items had to be omitted from the scale to produce a coherent item set that could yield an overall “views on nature of science” score.

In the current study, Turkish version of the MATE, NSAAQ, and PSTE were used as data collection instruments. However, semantic equivalence of Turkish version of the instruments was not established by cross-cultural testing that involves the administration of both original and Turkish versions to individuals fluent in both languages using a cross-over design (Jones 1986). Although, to overcome this problem, bilingual experts were involved to review the instruments for semantic equivalence, it is suggested that future studies examine the translated versions of the MATE, NSAAQ, and PSTE by using stringent methods such as cross-cultural testing. In addition, the psychometric properties of the instruments in source and target languages can be compared in future studies. Additionally, in the present study, the path analysis was conducted to investigate the relationships among pre-service science teachers’ understanding and acceptance of evolution, their views on nature of science, and their self-efficacy beliefs for teaching evolution. However, the path analysis does not give information related to the direction of causation. Hence, future studies can use longitudinal or experimental designs to determine the direction of causation among the variables. Given these limitations, there may be some recommendations for further research to illuminate the results of the present study. Findings of this study contribute to our understanding of factors related to Turkish pre-service science teachers’ self-efficacy beliefs about teaching evolution; however, conclusions and implications drawn from the current study may not be applicable to pre-service science teachers in other cultural contexts. Hence, this study should be replicated with subjects in different cultural contexts to get more valid conclusions. Besides, since the study was limited by its reliance on self-reported data, in-depth interviews with participants are required to verify the consistency and accuracy of the present findings. Furthermore, future research can include other constructs into the present path model to develop a more informed model explaining the factors related to self-efficacy beliefs for teaching evolution with in-depth interviews. Moreover, although this study provides support to the growing body of literature showing that the nature of science is a critical factor in understanding and acceptance of evolution; present findings do not explain causal relationships among these variables. Therefore, subsequent research is needed to clarify the direction of causation between correlated variables in the present study.

The findings of the present study can have practical implications for teacher education programs and evolution education in Turkey. Since evolution is a central and unifying theory of science, teachers should be confident in their knowledge and abilities to teach it effectively. According to the findings of this study, stronger sense of efficacy beliefs was associated with higher levels of understanding and acceptance of evolution. Thus, we suggest that teacher education programs should be designed to assist pre-service science teachers to have sound understanding of evolution and to accept scientific validity of evolution. Such understanding and acceptance could be fostered through facilitating sophisticated understanding of nature of science as implied by the present findings, suggesting that nature of science can be a critical factor in understanding and acceptance of evolution. Hence, we advocate that nature of science should be addressed while teaching of evolution as proposed by other researchers (e.g., Farber 2003; Nickels et al. 1996; Scharmann and Harris 1992). However, because the direct link proposed between views on nature of science and teaching self-efficacy was found to be negative, the current study calls for future studies to better understand this finding and its possible causes.

Copyright information

© The Association for Science Teacher Education, USA 2012