 |
Asia-Pacific Forum on Science Learning and Teaching, Volume 17, Issue 1, Article 7 (Jun., 2016) |
This study was conducted to develop an instrument with content-embedded items for determining science-specific epistemological beliefs of prospective science teachers. For this purpose, the validity and reliability of the instrument were examined.
Participants
The participants of the study were 364 (Male=82, Female=282) prospective science teachers enrolled in science teacher education program of a middle scale university. The participants were freshmen (n=94), sophomore (n=87), junior (n=98), and senior students (n=85). When looked at their history about epistemology, 148 of them took course about epistemology and 58 of them participated into conferences about epistemology before. Prospective science teacher’s epistemological beliefs shape their learning (Taskın Sahin, 2012). Moreover, their future science teaching environment from the methods used to their students’ epistemological beliefs is also affected by these beliefs (Alpan, 2013; Boz & Boz, 2014; Elby, Scherr & Redish, 2005; Torres & Vasconcelos, 2015). Consequently, starting from the teacher education, determining their science-specific epistemological beliefs in a reliable and valid way is the first and important step for forming their scientific epistemological belief.
Development process of the instrument
In this study the purpose of the instrument (Science-specific Epistemological Beliefs Inventory) was to determine scientific-epistemological beliefs of the prospective science teachers by focusing on five dimensions of science-specific epistemological beliefs: speed of knowledge acquisition, tentativeness of knowledge, structure of knowledge, source of knowledge and control over knowledge acquisition. These five dimensions were suggested by Schommer (1990) and her model of five-dimension epistemological beliefs system or sub-components of the model was used to study epistemological beliefs of prospective teachers (Sinatra & Kardash, 2004). The instrument involved 15 items. Considering with the five dimensions, the instrument involved 15 science-specific items (three per the dimension) were selected from the literature (Pomeroy, 1993; Schomer, 1990). The items were embedded into three different texts about well-known science topics and scientists (Action and reaction forces, Isaac Newton, Law of dominance, Gregory Mendel, Conservation of mass, Antoine Lavoisier) from physics, biology and chemistry. These topics and scientists were known enough by the prospective science teachers, by this way we decreased cognitive load of the texts. Just two participants did report that they did not know about Newton and Mendel. However Lavoisier were known less than Newton and Mendel, 110 of the participants did not know about Lavoisier. But after recalling conservation of mass, they indicated their unfamiliarity was confusion.
In the text the items were inserted into short descriptions of the topic and following story of a prospective science teacher. Each text involved five different groups of items about each dimension of the beliefs; hence 15 items were involved in three texts. Each item had five choices representing sophisticated and unsophisticated beliefs. In answering the items one prospective teacher could select no choice or more choices than one. In scoring prospective students’ answers were categorized as “blank (0)”, “unsophisticated belief (1)”, “mixed belief (2)” and “sophisticated belief (3)” by considering model of Schommer (1990). The inventory can be seen in the appendix. Completing the inventory took 30 minutes.
In development of the instrument following stages were conducted. (1) Determining purpose of the instrument, (2) Researching for related literature, (3) Deciding about dimensional structure of the instrument, (4) Writing items from the literature, (5) Deciding about science content in which the items were embedded, (6) Deciding about embedding design of the items (7) Preparing draft of the instrument, (8) Asking three prospective science teachers about their ideas on understandability, easiness to answer and reading load (9) Taking two science education experts’ ideas about the draft, (10) Making revisions based on the ideas (11) Applying the instrument to prospective science teachers, (12) Making confirmatory factor analysis, (13) Making reliability analysis and (14) Preparing final form of the instrument.
Research Method
This study was conducted to develop an instrument with content-embedded items for determining science-specific epistemological beliefs of prospective science teachers. For this purpose, the validity and reliability of the instrument were examined.
Content Validity
For content validity, table of specifications were used. Following Table 1 involves content and dimensions of the instrument.
Table 1. Specifications for alignment of Items of the Instrument and Dimensions of Science Epistemological Beliefs
Dimensions of Science Epistemological Beliefs
Items of the Instrument (Numbers = Item numbers) Physics (Newton, Action and reaction forces) Biology (Mendel, Law of dominance) Chemistry (Lavoisier, Conservation of mass) Speed of science knowledge acquisition
1 1 1 Tentativeness of scientific knowledge
2 2 2 Structure of scientific knowledge
3 3 3 Source of scientific knowledge
4 4 4 Control over science knowledge acquisition
5 5 5 Construct Validity and Reliability
Before making confirmatory factor analysis, multivariate normality, missing data and outlier analyses were conducted. Multivariate normality value was found as 2.06, it was in acceptable range (Raykov & Marcoulides, 2008). Also there was no missing data and outliers. The results of the confirmatory factor analysis indicated that data on scientific epistemological beliefs of prospective science teachers supported five dimensional epistemological beliefs structure. The fit indexes and non-fit indexes for model confirmation were Chi-square ratio (CMIN/DF), Goodness-of-fit index (GFI), Comparative fit index (CFI), Root mean square error of approximation (RMSEA) and root mean square residual (RMR) (Arbuckle, 1997). Proposed model for confirmatory factor analysis results are represented in Figure 1.
Figure 1. Proposed factorial structure model for measurement on scientific epistemological beliefs test D1: Dimension 1 (Learning Time), D2: Dimension 2 (Tentativeness), D3: Dimension 3 (Structure of Knowledge), D4: Dimension 4 (Source of Knowledge), D5: Dimension 5 (Learning Ability) P: Physics, B: Biology, C: Chemistry 
