An analysis of teacher candidates’ epistemological beliefs: A qualitative study Makalenin Türkçe Başlığı: Öğretmen adaylarının epistemolojik inançlarının analizi: Nitel bir çalışma Yazar(lar):

Bu calismada ogretmen adaylarinin epistemolojik inanclarinin belirlenmesi amaclanmistir. Bu amacla 56 ogretmen adayiyla yari yapilandirilmis gorusmeler yapilmistir. Gorusme kayitlari yazili metinlere cevrilmis ve betimsel analiz yapilmistir. Gorusmelerden elde edilen bulgulara gore, ogretmen adaylarinin birinci boyutla ilgili ogrenmeyi cabaya bagladiklari ve dogustan gelen yetenegin caba ile gelisecegini dusundukleri, ikinci boyutla ilgili cogunlugunun ogrenme surecinde bilgi yapilandirmanin yavas ve goreli bir kavram oldugunudusundukleri, ucuncu boyutla ilgili olarak adaylarin buyuk cogunlugunun bilgi kesinlikle degisir, alanina gore degisir ve degismez seklinde uc goruste olduklari belirlenmistir. Ayrica ogretmen adaylarinin cogunlugu, kelimelerin birden fazla anlama sahip oldugunu, insanlarin deneyimlerine gore kelimelere anlamlar yukledigini ve fen bilimlerindeki problemlerin bilgilerin zamanla degismesi ve bilim adamlarinin problemlere farkli cozumler onermeleri nedeniyle tek bir dogru cevabinin olmadigini belirtmislerdir.


Introduction
The processes of teaching and learning are affected by various cognitive variables. Some of these cognitive variables include students' epistemological beliefs, learning approaches, learning styles and attitudes Hofer, 2001;Schommer & Walker, 1997). According to Schommer (1990), epistemological belief is described as individuals' subjective beliefs on what knowledge is and how knowing and learning take place. Epistemological beliefs are required to be thought of as being more than one independent belief (Schommer, 1990). Individuals are classified as "naïve" and "sophisticated" in terms of adopting these beliefs. Those who have "naïve" epistemological beliefs believe that knowledge is formed by accumulating facts, accuracy of which are confirmed by an authority and are associated with each other by being added into another. Those who have "sophisticated" epistemological beliefs believe that the accuracy of the knowledge relies on the content, continuously formed within social interactions, and that the knowledge is a complex network of facts, theories, and assumptions (Bromme, Pieschl & Stahl, 2010). As seen in Figure 1, Schommer (1990) claims four independent dimensions related to epistemological beliefs: "Knowledge is simple", "Knowledge is exact," "Learning takes place spontaneously" and "Learning ability is determined by birth." The "Knowledge is exact" dimension refers to whether individuals believe that knowledge is precise or temporary truths or falsehoods that can change based on context. "Knowledge is simple" stands for whether individuals believe that they have a simple structure consisting through accumulation of independent, separate pieces or they have a complicated structure through associating the pieces to each other. The "Learning ability is determined by birth" dimension refers to whether individuals believe that learning is genetic and it is ability which cannot be changed or it is something that can be improved through experience. And lastly, the "Learning should take place spontaneously" dimension refers to whether individuals believe that knowledge comes spontaneously or it never comes, or it would come stage by stage in time through experience (Schommer, 1990).

Sophisticated Epistemological Beliefs (-) Naive Epistemological Beliefs (+)
Knowledge is simple Knowledge is exact Learning takes place spontaneously Learning ability is determined by birth It cannot be improved afterwards Knowledge is complex Knowledge is not exact Learning may take place in time Learning ability can be improved  (Deryakulu, 2004, p.267).
Following Schommer's development of a scale to determine epistemological beliefs easily, the studies related to the effective factors on these beliefs and on learning and teaching processes, and their relationship with other individual differences, have gained speed (Deryakulu, 2004, p.267). These studies tried to determine what students' epistemological beliefs are, the effects of gender, level, the department on epistemological beliefs, and its relations with academic success, studying attitudes and the relationship with learning-teaching styles (Belet & Güven, 2011;Deryakulu, 2002;Deryakulu & Büyüköztürk, 2005;Dursun Sürmeli & Ünver, 2017;Enman & Lupart, 2000;Eroğlu & Güven, 2006;Karabulut & Ulucan, 2012;Koç Erdamar & Bangir Alpan, 2011;Sapancı, 2012;Schommer, 1990;1993;Schommer, Calvert, Gariglietti & Bajaj, 1997;Schommer-Aikins, Duell & Hutter,2005;Terzi, 2005;Topçu, 2013;Tümkaya, 2012;Yeşilyurt, 2013). According to the studies examining the relationships between epistemological beliefs and study strategies, the students possessing naïve epistemological beliefs do not try different strategies for studying and do not make an effort to learn Rodriguez & Cano, 2006;Schommer-Aikins et al., 2005). Researchers who study the relationship between epistemological beliefs and academic performances reported that students who believe that learning does not take place quickly and requires effort demonstrate higher levels of academic performance and that students with sophisticated epistemological beliefs develop more positive attitudes towards education and studying (Koç Erdamar & Bangir Alpan, 2011;Schommer, 1993;Schommer & Walker, 1997;Önen, 2011). In some studies, examining the effect of gender on epistemological beliefs, it was found that females believed that learning happens spontaneously, and that it is a stable and inherent ability, which is less frequently occurring than male teacher candidates (Schommer, 1993). Whereas in some studies, it was deducted that female students possessed more sophisticated beliefs than male students, in that learning is dependent on ability (Chai, Khine & Teo, 2006;Öngen, 2003).
In research, it was also reported that the naivebeliefs related to simple knowledge, precise knowledge and learning fast decreases as the class levels rise (Koç Erdamar& Bangir Alpan, 2011;Schommer, 1993;. Schommer and others (1997) and Sadıç, Çam and Topçu (2012) argued that students in higher grades have more developed epistemological beliefs. On the other hand, Schommer and others (1997) observed that students who believe that learning rely on effort tend to be more successful. Jehng, Johnson and Anderson (1993) stated that the epistemological beliefs of students will be formed with the education level of the individual and that as the education level increases, the students will believe that the nature of the knowledge becomes indefinite, the act of learning does not occur in a gradual process as a whole, and independent learning is more important. On this matter, Spiro, Coulson, Feltovich and Anderson (1988) argued that the systematical arrangement of introduction courses and working system of students with well-structured problems that are solved following well-known algorithms and procedures in these courses prevent the epistemological beliefs of students to improve (Jehng, Johnson & Anderson, 1993). Besides, researchers argued that epistemological beliefs of students are related to the fields they have learned and that students attending social sciences and the fields of art possess more sophisticated epistemological beliefs than those studying sciences and engineering (Chai et al., 2006;Deryakulu & Büyüköztürk, 2005;Enman & Lupart, 2000;Jehng, Johnson & Anderson, 1993). Jehng et al. (1993) expressed that the structures of knowledge in engineering and the natural sciences tend to be systematic and sequential. As a result of this, students tend to think that the nature of the knowledge is absolute and solutions to problems are accessible within a certain period of time. In their studies, Deryakulu and Büyüköztürk (2005), Enman and Lupart (2000) expressed that the students studying social sciences possess more sophisticated epistemological beliefs than those studying sciences. Chai et al. (2006) reported that teacher candidates studying in the fields of sciences and technology believed that knowledge does not change more than it does for students attending humanistic fields, language and business.
Unlike these studies, Topçu (2011) investigated whether there is any relationship between teacher candidates' epistemological beliefs and moral reasoning by making use of mixed method research design. Topçu (2011) did not find any relationship between epistemological beliefs and moral reasoning in his study. In a different study, Topçu (2013) examined whether teacher candidates' epistemological beliefs differ on the basis of departments (Physics, Chemistry and Biology) following the same research design. The researcher reported that the teacher candidates thought that the knowledge in Biology is more temporary than knowledge in the fields of Physics and Chemistry and that knowledge in Physics is more precise than in Chemistry and Biology. Moreover, the teacher candidates were observed to think that personal assessments and research rules in justifying the knowledge in Biology are more important than justifying the knowledge in Physics and Chemistry, and it was revealed that they attributed biological learning to their own constructions of knowledge while they associated learning in Physics and Chemistry fields with external sources such as books and teachers.
Once students regard knowledge to be complex, complicated, and inter-correlated rather than categorical, they start to think in a way that requires evaluation and criticizing (Brownlee, Petriwskyj, Thorpe, Stacey & Gibson, 2011). Prospective teachers, who have such epistemological beliefs will be able to train their students as individuals who can evaluate and question knowledge they are given, and individuals who can approach knowledge from different perspectives. The beliefs of teachers regarding education have a significant impact on their ability to achieve successful teaching (Scholtz, Braund, Hodges, Koopman & Lubben, 2008). Similarly, their beliefs for knowledge and learning will have an impact on their ability to teach (Chan, 2004). Prospective teachers will form learning environments in which complex discussions take place and adverse opinions are defended and which enables students forming arguments with interactive dialogs between students rather than monologues. In such learning environments, students will learn how to conduct, question, and justify knowledge while understanding how science developed by living the process of reaching knowledge with discussions (Yıldırır, 2013), and improving beliefs regarding knowledge and learning. As Jehng et al. (1993) emphasized in such environments, students become more aware of the fact that the knowledge is not absolute, the reality is open to interpretation, and that learning is a process where the learner needs to arrange different knowledge parts from different perspectives. In the constructivist epistemology idealized for science education, it is pointed out that students believe that scientific knowledge is formed by going through many different processes which are open to discussion and interpretation as teachers will ensure (Koç & Memduhoğlu, 2017). Epistemological beliefs of teachers are very important for the development of epistemological beliefs of students. It is also observed that students and teachers who have epistemological beliefs are very successful academically (Brownlee et al., 2011). On the other hand, it can also be observed that epistemological beliefs have an impact on all variables regarding education and learning, considering that epistemological beliefs of students affect students' problem solving beliefs (Schommer, 1990;Schommer-Aikins et al., 2005), academic performance (Deryakulu, 2002;Schommer, 1993), learning and studying strategies Önen, 2011;Rodrigues & Cano, 2006), motivation (Başbay, 2013), behavior towards education (Schommer & Walker, 1997), understanding of teaching and learning (Chan, 2004), grasp of learning and teaching styles (Kaleci, 2013), critical thinking tendencies (Başbay, 2013) and metacognition awareness (Başbay, 2013;Belet & Güven, 2011). Thus, it is quite important to determine teacher candidates' epistemological beliefs, the significant factors playing role, and their opinions concerning epistemological beliefs. When literature is reviewed, it is seen that studies on epistemological beliefs are mostly quantitative research towards determining epistemological beliefs and factors affecting these beliefs; however, studies using qualitative research and mixed methods are limited in number (Kaleci & Yazıcı, 2012). For this reason, this research is important in terms of determining the epistemological beliefs of teacher candidates in more detail with qualitative data contributing to the literature. Aiming to find out epistemological beliefs of teacher candidates through interviews, the study adresses the following research questions: • Do candidate teachers associate learning with innate ability or effort?
• Do candidate teachers think that knowledge is simple and that there is only one correct answer to the problems?
• What do candidate teachers think about the speed of knowledge construction process?
• What do candidate teachers think about the changeability of the knowledge?

Research Design
In qualitative research, the main purpose is to present the survey in detail in a realistic way. For this reason, it is important to present the data as detailed and direct as possible (Cohen &Manion, 1997). In this study, qualitative research method was chosen in order to provide detailed and in-depth knowledge of the teacher candidates' epistemological beliefs through semi structured interviews.
Interviews are one of the most common data collection methods used in qualitative research. Briggs (1986) argues that the reason for this is that the interview method is a very effective method for collecting data about individuals' experiences, behaviors, opinions, complaints, emotions, or beliefs (Yıldırım & Şimşek, 2013). In a semi-structured interview, researchers prepare interview questions in advance but they can be rearranged by allowing flexibility to the participants during the interview (Ekiz, 2015).

Participants
Participants of the study consist of 56 teacher candidates currently studying at the Science Education (SE), Primary Grade Mathematics Education (PGME), Biology Education (BE), Physics Education (PE), Chemistry Education (CE) and Mathematics Education (ME) Departments at Necatibey Education Faculty, Balıkesir University. Two teacher candidates were chosen from each class level of each department and interviews were held with the selected 56 teacher candidates. Teacher candidates volunteered to participate in the study. As it was determined that gender has an impact on beliefs, it was aimed to reveal how male and female prospective teachers think according to their beliefs in studies on epistemological beliefs (Belet & Güven, 2011;Chai et al., 2006). Considering this, a particular attention was given to pick teacher candidates from both genders for the study that was made on a volunteer basis. Table 1 demonstrates the distribution of teacher candidates based on department.

Data Collection Tools
The study made use of the Epistemological Beliefs Interview Form for the aim of determining epistemological beliefs of teacher candidates. Each of the questions used during these interviews were prepared by Topçu (2011) including all of the epistemological belief dimensions suggested by Schommer (1990). During the process of drawing up the interview form that was developed by Topçu (2011), the researchers had 3 pilot interviews with prospective teachers as a priority. In the light of these interviews, the interview questions were revised accordingly. Topçu (2011) benefited from the studies of Schommer (1990) and Cheng et al. (2009) in creating an epistemological beliefs interview protocol. There are four questions in the interview form. The first question is related to innate ability; the second one is related to simple knowledge, the third one is related to quick learning and the fourth one is related to certain knowledge. The questions in the interview form are presented below: 1. When you are learning a subject, what percentage will you attribute to your innate ability and to your learning effort? Why? Do you believe that perseverance and hard work can overcome difficulties in learning?
2. Do you think most words have one clear meaning and most problems in science have only one right answer? Why?
3. Do you think learning process is a slow or a fast process of building up knowledge? Why?
4. Do you think knowledge changes, or is it something that does not change? Why?

Data Analysis
The interviews were made with 56 teacher candidates, two teacher candidates from each grade from each department. When the study data is presented, the teacher candidates were coded by specifying the departments and classes. For example, second year Chemistry Education Department's students were coded as "C2". Departments are coded as Science Education, SE, Chemistry Education CE, Physics Education PE, Primary Grade Mathematics Education PGME, Mathematics Education ME and Biology Education BE.
The interview records were transcribed and were analysed through descriptive analysis, because descriptive analysis allows the data to be organized according to the themes that the research questions reveal, and to be presented considering the questions or dimensions used in the interview (Yıldırım & Şimşek, 2013). In this study, the data is analysed according to the categories determined by the questions asked in the interviews. Therefore, the categories include four main categories such as "innate ability", "knowledge is simple", "quick learning", and "knowledge is certain". Sub-categories were divided into groups according to the answers given by the teacher candidates.
If a participant's opinion is included in more than one idea, this opinion is placed in both ideas. Thus, considering the frequency and percent distributions in the tables, the total frequency and percent values related to an opinion can be more than the number of the participants. With the aim of ensuring the reliability of the study data, the analysis was conducted by the researchers who are experts in the field separately, and then results of the analysis were compared, which indicated a 90.00% consistency.

Results for First Sub-Problem (Innate Ability)
During the interviews, with the aim of determining teacher candidates' opinions related to the "Innate ability" dimension, the teacher candidates were first asked to what extent they attribute learning to innate ability and to learning effort. The teacher candidates indicated that while learning a topic, both innate ability and effort were of significance explaining their opinions under three statements: "Nothing can be achieved without making an effort," "Innate ability can only be improved through effort" and "Hereditary features affect learning". The results of the first sub-problem related to the dimension of "innate ability" are shown in Table 2. All of the teacher candidates agreed on "Nothing can be achieved without making an effort" and they attributed learning a topic to making an effort rather than inborn ability, as it is understood from the sample statements below.
"Innate intelligence is quite important. However, our own resolutions are important, as well. No matter how much intelligent you are, it is not without studying. Namely, we aren't born with formulas in our minds. I believe that, through hard work and resolution, anything can be achieved. Intelligence just helps us." (PGME3).
"Hereditary features affect learning" was shared by 5 (8.90%) of the teacher candidates (2 CE, 1 SE, 1PGME, 1 BE). 9 (16.10%) of the teacher candidates (4 BE, 3 ME, 1 SE, 1CE) agreed on "Inborn ability can only be improved through effort" opinion: "I think innate abilities are around 30.00-40.00%. Not everybody can have ability in every field. However, the person who makes an effort and does hard work can learn something about many topics. Although inborn abilities are important, hard work and effort are needed to reveal that ability," (ME2).
While the teacher candidates explain their ideas about whether learning take place by birth or through making an effort, they also shared their thoughts about learning. 12 (21.40%) of the teacher candidates (5 CE, 2 PGME, 2 PE, 1 SE, 1BE, 1 ME) stated that if an individual possesses multiple intelligences such as visual, mathematical, verbal etc., it facilitates the learning process. Another idea related to learning was "Willingness facilitates learning," which was agreed on by 16 (28.60%) of the teacher candidates (5 PE, 4 SE, 4 BE, 2 CE, 1ME). This idea means that if a person has interest, desire and motivation to learn a topic, he strives to learn it, and he learns it more easily. According to the results obtained from the interview data of the first question regarding whether the learning is innate or comes from effort, the majority of the teacher candidates (98.00%) have the opinion that "Nothing can be achieved without making an effort". As shown in the Table 3, first grade participants gave the answer "Nothing can be achieved without making an effort" at the rate of 92.00% and participants of other grades gave the same answer with the rate of 100.00%. Also, first grade participants believe that natural ability can be improved by effort at the rate of 33.00%. Fourth and fifth grades gave the same answer with the rates of 25.00%. Teacher candidates who have the opinion that heredity features affect learning are first and fourth grade participants with the rate of 17.00%. An average 21.00% of teacher candidates have the opinion that "possessing related multi intelligence facilitates learning". 46.00% of participants who are at the second grade think that "being willing facilitates learning." According to the examination of the results obtained from the first question with regard to gender, as seen in Table 4, "Nothing can be achieved without making an effort" have approximately same ratios. The opinion "Innate ability can only be improved through making an effort" is more common in male teacher candidates compared to female teacher candidates with the rate of 23.00%. Compared to the female teacher candidates; more male teacher candidates have the belief that inherited features affect learning. Similarly, the opinion "possessing related multi intelligence facilitates learning" is at the rate of 38.10% in male teacher candidates which is more than the female teacher candidates.

Results for Second-Sub Problem (Knowledge is Simple)
In relation to the second sub-problem, with the aim of determining teacher candidates' opinions about "Knowledge is simple", the teacher candidates were asked whether words have only one meaning, and whether scientific problems have only one correct answer. The results of the second sub-problem related to the dimension of "Knowledge is simple" are shown in Table 5. The analysis results show that all of the teacher candidates stated that words can have more than one meaning, and they presented two different ideas about this matter. The first idea referred is that "words' meanings can change based on sentences they are used in" was shared by 43 (76.80%) of the teacher candidates (10 ME, 9 BE, 9 PE, 6 CE, 5 SE, 4 PGME): "Each word has its meaning in its own topic. For example, the word "seed" refers to dried fruits in daily life, to nucleus in biology, and to core construct including proton and neutron in chemistry." (SE2).
The second opinion was that people attribute meanings to words because they interpret the words in terms of their experiences, so words can have more than one meaning, which was shared by 13 (23.20%) teacher candidates (4 PGME, 4 CE, 3 SE, 1BE, 1PE). 41 (73.30%) of the teacher candidates stated that scientific problems don't have only one solution while 15 (26.70%) of them indicated that there is only one solution to scientific problems. Following the analysis of the teacher candidates' responses who think that scientific problems can have only one solution, two different perspectives were revealed. In the first perspective, 11 (19.60%) of the teacher candidates (5 ME, 4 PGME, 2 BE) stated that scientific knowledge is proven and based on scientific facts, that proven knowledge cannot change, thus there is only one truth in scientific problems. The second opinion, "Knowledge can change based on its field, was agreed on by 4 (7.10%) of the teacher candidates (3 CE, 1 PE). These teacher candidates approached the issue from different perspectives. Some of them stated that truth can change since social sciences are dependent on interpretation, while some other claimed the opposite and since the events happened in the field of history, they asserted that the truth cannot change. Another related opinion was that the truths in the science field do not change, as they are based on formulas and numbers while the truths in social sciences can change in terms of the different perspectives in the field.
Following the interview analysis of the teacher candidates thinking that scientific problems cannot have only one truth, three different points of views were found: "Knowledge changes in time" "Scientists bring different perspectives to problems" and "Physical science is an inter-disciplinary field". 24 (42.90%) of the teacher candidates (6 SE, 6BE, 4 PGME, 4 ME, 3 PE, 1CE) stated that knowledge can change in time through renewed technology and scientific researches: "Many scientific problems don't have only one right solution. Today science is established on already accepted truths. However, when a problem is explained via a different logic, or observed with experiments, that fact can change. Anything which was not observed through experiments or found through technologies of those times can be observed and found with today's technologies. For example, the previously accepted atom model is not accepted today owing to the discovery of a new model." (SE4) 14 (25.00%) of the teacher candidates (6 PE, 4 CE, 3 BE, 1 SE) agreed that scientific problems cannot have only one truth since scientist approach to problems with different perspective stemming from their different knowledge backgrounds and experiences. 3 (5.40%) of the teacher candidates (2 CE, 1 SE) stated that scientific problems cannot have only one truth because problems are solved with the contributions of many science fields such as physics, chemistry, biology and mathematics, so they asserted that only one answer or explanation cannot be the solution. Scientists present different perspectives to problems.
Science is interdisciplinary field.
The truth of knowledge is proven through scientific facts.
Knowledge changes based on fields. Interview findings related to "Knowledge is simple" dimension according to the class level are shown in Table 6. After the analysis of the data of the "knowledge is simple" aspect of the second interview question with regard to the class level, the opinion that words have different meanings depending on the place they are used is determined to be 100.00% among second grades and 83.00% among first grades. It was also determined that the rate of teacher candidates who believe in the changeability of knowledge increases as the class level increases, which means that teacher candidates believe more that problems have multiple answers as the class level increases. According to this finding, higher level teacher candidates have more improved beliefs. The belief that problems have a single answer is more common amont the first and the second graders. This shows that lower grade participants have less developed beliefs compared to higher levels with regard to the opinion "knowledge is simple". Scientists present different perspectives to problems.
Science is interdisciplinary field.
The truth of knowledge is proven through scientific facts.
Knowledg e changes based on fields. Interview findings related to "Knowledge is simple" dimension according to the gender are shown in Table 7. After the analysis of the dataregarding "knowledge is simple" aspect of the second question based on gender, the opinion that words have different meanings depending on the place they are used was shared at the rate of 68.60% by females and 90.00% by males. The male teacher candidates believe that knowledge can change over time at the rate of 66.70%. This ratio is lower for the females (28.60%). Compared to the female teacher candidates, male teacher candidates think that problems have multiple answers so the male teacher candidates have more developed beliefs regarding scientific problems having multiple answers.

Results for Third-Sub Problem (Quick leArning)
In relation to the third sub-problem, with the goal of learning about teacher candidates' opinions concerning "Learning takes place spontaneously" dimension, they were asked if learning process was slow or fast in construction of knowledge. The results of the third sub-problem related to the dimension of "Quick learning" are shown in Table 8.As seen in Table 8, the teacher candidates referred to two different perspectives in relation to the speed of knowledge construction in the learning process. These perspectives include "It is slow" and "It changes based on agents.".Other participants, 16 (28.60%) teacher candidates (4 PGME, 4 PE, 3 BE, 2 SE, 2 CE, 1ME) indicated that in the learning process, knowledge construction can change based on agents. The teacher candidates emphasized that knowledge construction can be fast or slow in relation to an agent's readiness, motivation, need and interest in the topic: "I think it is a relative concept. It can change based on students' readiness levels, I mean pre-existing knowledge, mental skills and learning motivations. Also, students' needs and interests affect this process as well. The learning process can be slow or fast in knowledge construction. It is dependent on agents. Comprising knowledge in the brain is just like establishing a bridge between two cliffs; it is hard but after establishing that bridge, it becomes easier. It becomes a path for other knowledge." (CE5)

Opinions
In the learning process, knowledge construction… …Changes based on agents.
Agents' pre-knowledge, interests and needs affect knowledge construction process.
Making meanings through associations takes time.
There is a specific relation of progress.
Reinforcing and repetition take time. The teacher candidates thinking that knowledge construction is slow in the learning process asserted four reasons for it: "Making meanings through associations takes time," "There is a specific relation of progress," "Reinforcing and repetition take time" and "An agent learns how he has learnt." The first reason was asserted by 24 (42.90%) of the teacher candidates (6 CE, 6 ME, 4 SE, 4 PE, 2 PGME, 2 BE). The teacher candidates indicated that in order to acquire knowledge through construction, there should be established associations between old and new knowledge and there should be a correction of incorrect knowledge, both of which are slow processes. The second reason was agreed on by 12 (21.40%) of the teacher candidates (3 SE, 3 BE, 3 ME, 2 CE, 1PE).
"It is a slow process, because real learning takes place when new knowledge is established over the old one, and when elaborate associations are made in different times and topics. It should be slow, each step should be thought out, understood and learnt separately, and then the next step should be taken, which shows that learning is slow but progressing." (CE4).
The third reason was emphasized by 5 (8.90%) of the teacher candidates (2 PGME, 2 BE, 1PE). In this sense, the teacher candidates stated that knowledge is dependent on time and repetition, and thus it becomes permanent in memory, which is a slow process. In addition, the fourth reason was asserted by 2 (3.60%) of the teacher candidates (1 SE, 1CE). Teacher candidates thought that the knowledge construction process is slow at the beginning, but after an agent learns how he acquires knowledge, this process is sure to speed up: "Construction of knowledge is difficult until you discover how you learn and until you discover it's systematic. However, after an agent determines how he learns, placing new knowledge in memory and making sense of it would speed up the knowledge construction process. It becomes easier." (CE5) After the examination of the findings obtained from the third questions regarding the aspect "Quick learning", it was observed that fifth grade participants have the opinion that "making meaning through association takes time" at the rate of 75.00%. This opinion is shared by the fourth and the second grade participants at the rate of 50.00%. Some of the participants are specified under multiple sub-categories as they think that learning is slow and can change depending on the person.
As seen in Table 9, the opinion that the learning will take more time as the class level increases is more dominant. This indicates that participants have more developed beliefs as the class level increases regarding whether the learning occurs immediately or not. Table 9. Interview Findings Related to "Quick learning" Dimension According to the Class Level.

Opinions
In the learning process, knowledge construction… …Changes based on agents.
Agents' pre-knowledge, interests and needs affect knowledge construction process.
Making meanings through associations takes time.
There is a specific relation of progress.
Reinforcing and repetition take time.

Opinions
In the learning process, knowledge construction… …Changes based on agents.
Agents' pre-knowledge, interests and needs affect knowledge construction process.
Making meanings through associations takes time.
There is a specific relation of progress.
Reinforcing and repetition take time. The aspect whether the learning occurs immediately or not is analysed based on genders and shown in Table 10 and it was observed that there is no difference between the speed of the general learning process and the related beliefs. In sub-categories, male teacher candidates think that giving meaning by association takes time at the rate of 57.10%. At the rate of 25.70%, the teacher candidates have the belief that existence of a certain progressivity cause the learning process to be slower.

Results for Fourth Sub-Problem (Knowledge is Certain)
During the interviews, the teacher candidates were asked about whether knowledge changes through the "Knowledge is certain" dimension. The results of the fourth sub-problem related to the dimension of "Knowledge is certain" are shown in Table 11.
According to the analysis results in Table 11, three different perspectives were revealed: "Knowledge absolutely changes," "Knowledge changes based on field" and "Knowledge never changes". 47 (83.90%) of the teacher candidates (9 BE, 9 ME, 8 PE, 8CE, 7 PGME, 6 SE), expressed that through the development of science and technology and through the changes in peoples' interpretations of knowledge and the world, knowledge absolutely changes: "It would be wrong to claim that knowledge doesn't change, because as science develops, it is impossible to escape from change. For instance, in the Physics field, Galileo asserted that 'Mechanical physics principles are valid within passive observation frameworks,' but years later Einstein stated that 'All physics principles are valid within passive observation frameworks' and improved the validity of the idea proposed by Galileo." (SE2) "Knowledge changes, because day-to-day technology and science develop as well. There are many things we still don't know. It would be cheating us to possess absolute judgments without knowing them. We just try to content ourselves with the current ones in our hands." (CE2). 5 (8.90%) of the teacher candidates (2 SE, 2 CE, 1PGME) claimed that knowledge can change based on the field. The sources of teacher candidates' opinions about this issue were seen to differ. While Chemistry Education teacher candidates think that knowledge in social sciences cannot change, Science Education teacher candidates express that it can change in social sciences, but it cannot change in physical sciences. A teacher candidate from Primary Grade Mathematics Education additionally claimed that except the religious knowledge, all other kinds of knowledge change: "Verbal knowledge doesn't change. History was experienced and left in the past. We can't have any effect on past events, but in time, unknown facts can be revealed. However, the physical science fields such as Chemistry and Physics can undergo changes in any time. All theories can be confuted, and new ones can be replaced." (CE3) "The change of knowledge is dependent on its field. In fields such as history, anthropology knowledge can change, as new documents can be found, but in physical science fields such as physics, I think, knowledge doesn't change, as formulas don't change." (SE1) 4 (7.10%) of the teacher candidates (2 PE, 1BE, 1ME) indicated that knowledge cannot change. These teacher candidates underscored the fact that the pathways to knowledge and its sources can change, but the knowledge itself does not change.
The results of the dimension of "Knowledge is certain" with regard to the class level are shown in Table 12. According to the Table 12, it can be seen that teacher candidates share the belief that knowledge will absolutely change as their class level increases.In lower class levels, more first grade participants believe that knowledge cannot change compared to other class levels (16.70%). Thus, as the class level increases, it can be said that the teacher candidates have more advanced beliefs about the exchangeability of knowledge.
According to the analysis results in Table 13, when the findings regarding the "knowledge is certain" aspect, it can be seen that male teacher candidates (84.70%) believe more in the change ability of knowledge compared to female teacher candidates (82.90%). Furthermore, 8.60% of female teacher candidates think that the knowledge will definitely not change. Based on these ratios, it can be said that male teacher candidates have more developed beliefs regarding the changeability of the knowledge.

Discussion & Conclusion
According to Schommer (1990), epistemological beliefs have direct and indirect effects on learning and they can change in time and through teaching methods. Teachers' epistemological beliefs, affect which teaching methods and techniques they apply, how they manage classroom, what they focus on while teaching, and which teaching activities they make use of in the classroom (Öngen, 2003). In this sense, the sophistication level or shaping of epistemological beliefs and teachers' opinions about teaching and learning processes affect students' epistemological beliefs (Kazu & Erten, 2015;Yılmaz & Şahin, 2011).
In relation to the dimension of "Innate ability" (Table 2), in this study, all of the participants thought that both effort and ability are important and without making any effort, nothing can be achieved. Some of the participants stated that inherent ability can only be improved through effort while some other emphasized the effects of genetic factors on learning. During the interviews, the participants' thoughts about learning were also revealed. While some of the students thought that individuals' possessing multi-intelligence about the topic they tried to learn facilitated the learning process, other students believed that if individuals had interest, desire and motivation to learn a topic, they would strive to learn that topic, and they would learn it more easily. In this sense, teacher candidates possess less sophisticated beliefs that learning is dependent on effort, which shows parallel results with the related literature findings (Aypay, 2011;Belet & Güven, 2011;Çam, 2015;Deryakulu & Büyüköztürk, 2002;. For example, Aypay (2011) reported that teacher candidates were disposed to believing that making an effort was important in learning, that the knowledge acquiring process was also significant in learning and that expert knowledge should be questioned, the participants were also found to be irresolute in the issues surrounding whether ability is inherent, or whether it is stable or changeable and whether knowledge is certain and exact. Following the review of innate ability across the class levels (Table 3), it was revealed that the first classes had less developed beliefs than the higher classes although almost all the classes had developed beliefs. Additionally, among all the class levels, the first classes mostly believed in that innate abilities can be improved with efforts. In literature, there is opposite evidence; in that, as the class level rises, there is no difference in beliefs concerning whether learning is based on effort or ability (Biçer, Er & Özel, 2013). In other words, it was concluded that class level did not have any effect on development level related to "innate ability" which generates parallel results with literature (Meral & Çolak, 2009;Yeşilyurt, 2013). In contrast to the findings of the studies in which no difference was found in "innate ability" dimension across class levels, some studies indicated that as the class level rises, epistemological beliefs improve (Önen, 2011). Also, other studies pointed out that as the class levels rise, there is no difference in teacher candidates' beliefs related to learning concept (Belet & Güven, 2011;Eroğlu & Güven, 2006). Similarly, other studies in literature remark that as the class levels rise, students/teacher candidates possess more developed beliefs about "innate ability" dimension (Balantekin, 2013;Schommer, 1990).
Upon examining the innate ability dimension in terms of gender (Table 4), both male and female teacher candidates agreed on the idea of "Nothing can be achieved without making an effort." with similar rates. In this study, the idea of "Innate ability can only be improved through making an effort" was mostly believed by the male teacher candidates, which differs from females. In literature, there are specific studies indicating that female students shared the belief that learning is dependent on making efforts (Biçer et al., 2013;Deryakulu & Büyüköztürk, 2005;Önen, 2011) Among the findings of the study data, regarding "the simple knowledge" dimension (Table 5), the students stated that words' meanings can change according to the sentences they are used in, and that individuals attribute different meanings to the same words based on their personal experiences. Regarding the thought "scientific problems have only one right answer" in the simple knowledge dimension, most of the students stated that there cannot be only one true answer while some of the students, except those in the Science Education departments, stated that there can be only one answer. In a study which compared Physics, Chemistry and Biology fields in terms of simplicity of knowledge, the students stated that the words in Biology possessed only one meaning while the words in Chemistry and Physics might possess more than one meaning, and that the students attending Physics, Chemistry and Biology fields adopted the same perspective in the problem-solving process (Topçu, 2013).
Among the findings of the study, teacher candidates who thought that the scientific problems did not have only one true answer explained it with three reasons ( Table 5). The teacher candidates stated the scientific problems did not have only one true answer, because in the light of renewing technology and scientific research, knowledge changes in time and scientists approach scientific problems through different perspectives as a result of various background knowledge and experiences. In addition, scientific problems can be solved through contributions from various fields such as Physics, Chemistry, Biology and Mathematics. The students who believed in the existence of only one truth also expressed two reasons for it. Among these students, those from the Science Education, Primary Grade Mathematics Education and Mathematics Education departments stated that scientific knowledge is proven through scientific facts, and that proven facts cannot be changed while the students from Physics and Chemistry departments implied that the changes in knowledge depends on the field. Primary grade mathematics and mathematics teacher candidates were mostly found to believe that there cannot be a different answer since all the knowledge has already been proven. That especially mathematics teacher candidates shared this idea might be result of deductions they make during courses. All of the students think that facts in science do not change since they depend on numbers and formulas while some students expressed that facts in social sciences can change as they are based on comments, and others stated that knowledge related to the past events as in the field of History cannot be changed since they all happened and finished in the past. Other surveys conducted on this issue reported that students think that the problems in the fields of science and social science have multiple solutions when compared to the problems in the Mathematics and Religious fields (Chai et al., 2006;Topçu, 2011;. Considering the studies in the literature, students are observed to possess less sophisticated beliefs. Although a constructive approach is adopted in preparing the curriculum programs, the courses are still delivered with traditional methods. Furthermore, teachers are active during the courses while students passively accept what teachers say without any pre-existing knowledge about the topic and without making use of their own experiences or assessing what is told to them. At this point, through providing a student-centered teaching approach and composing small cooperative groups, students can be enabled to take charge of their own learning tasks. Hence, students can assess their own learning tasks together with their pre-existing knowledge, experiences and characteristics, can find opportunities to discuss with their peers and can realize the existence of multiple and temporary truths rather than acceptance of only one absolute truth . Otherwise, the ones adopting naïve epistemological beliefs regarding the existence of only one truth would demonstrate adoption of others' thoughts and would present only one answer to complicated problems (Aksan & Sözer, 2007). Additionally, as Öngen (2003) stated, the reason why teacher candidates possessed naïve epistemological beliefs might be that the instructors at universities are accepted as the only authority and the exam systems require students to give an absolute true response. Teaching methods which center around students' epistemological beliefs should be applied to improve students' epistemological beliefs (Sandoval & Çam, 2011). In this sense, especially from the point of view of the teacher candidates, teacher training programs should focus on students' epistemological beliefs (Çam, 2015;Vinolainen et al., 2014). When teacher candidates receive an instruction activates and contributes to their epistemological beliefs during their lifetime, they realize the importance of them when they become a teacher, and they would arrange their programs for students accordingly. After examining the data related to "Knowledge is simple" dimension across the class levels, it was seen that teacher candidates' beliefs towards the variability of knowledge increased as the class levels rose. In other words, the teacher candidates who did not agree on it, believed in that there was not only one right answer to the problems. It presents that the teacher candidates had more developed beliefs as the class levels rose. Among the first and the second-class candidates, the rate of "Problems have only one right answer" was higher than the other class levels. In general, it was pointed out that as the class level decreased, teacher candidates had less developed beliefs about "Knowledge is simple" dimension (Table 6). This is in accordance with the findings of Sadıç et al. (2012). The findings related to "Knowledge is simple" dimension in terms of gender demonstrated that male teacher candidates believed that knowledge could change in time more than female teacher candidates. Moreover, the rate of male teacher candidates thinking that problems did not have only one right answer was higher than the females' (Table 7). In this sense, male teacher candidates were found to possess more developed beliefs than the female teacher candidates. However, certain studies in literature referred that females had more developed beliefs than males (Kurt, 2009;Sadıç et al., 2012;Topçu & Yılmaz-Tüzün, 2009).
Among the findings, in terms of the "Quick learning" dimension (Table 8) and that the knowledge construction process in learning can be fast or slow, the teacher candidates were revealed to think that learning takes place through making an effort. The majority of the teacher candidates stated that the knowledge construction process in learning is slow while others stated that this is a subjective process. Most of the teacher candidates who think that the knowledge construction process in learning is slow indicated that making meaningful associations among knowledge is a time-consuming process. Other teacher candidates thought that the knowledge construction process should be progressive, that reinforcement and rehearsals, and individuals' learning processes were the most time-consuming factors. In terms of subjective characteristics of the knowledge construction process, teacher candidates expressed that individuals' prior-knowledge, interests and needs also influence learning processes. As can be, the teacher candidates possessed sophisticated beliefs in this sense and they individualized it because students construct their own understandings and perspectives about knowledge (Schunk, 2009), and knowledge is gained through individuals' experiences rather than imposing. With a similar result, Topçu (2011) reported that teacher candidates believed that learning is not a fast but a progressive and complex process, and that fast learning is just memorizing. In the mixed study they conducted with science and physics prospective teachers, Koç and Memduhoğlu (2017) revealed that prospective teachers have less developed beliefs regarding the opinion that learning occurs immediately in quantitative results, whereas they think learning is a cumulative and gradual process in qualitative results, the reason being that teachers' metacognitive activities and planning of learning process, and the strategies they choose affect epistemological beliefs of students.
In relation to this dimension, Science education and Chemistry education department candidates emphasized individuals learning how to learn and that this affects individuals' learning processes, because if a student can criticize his or her own learning process, that student could develop more sophisticated beliefs (Wyne, 2007;cited: Başbay, 2013). If an individual considers how they think, taking all possibilities into consideration, their self-assessment indicates that they have meta-cognitive characteristics. Meta-cognition provides individuals insights about their own thinking processes, and it encourages independent learning (Sapancı, 2012). Thus, that teacher candidates possess meta-cognitive characteristics is of quite significance in that they will be future teachers. The teachers' metacognitive activities affect the epistemological beliefs of the students (Chan, 2004). Considering the "Quick learning" dimension across the class levels, all of the five graders thought that learning developed slowly in time (Table 9). In general terms, as the class levels rise, that learning takes time becomes a common idea. Thus, within the "Quick learning" dimension, it was revealed that as the class levels rose, teacher candidates adopted more developed beliefs. Reviewing this dimension across genders, no difference was found between female and male teacher candidates' beliefs regarding the speed of learning process (Table 10). However, the findings in literature indicate that females have more developed beliefs than males (Belet & Güven, 2011).
In relation to the dimension of "Knowledge is certain" (Table 11), the majority of participants expressed that knowledge changes absolutely; some claimed it does not change at all, and others stated that it changes based on the field. The teacher candidates emphasized that together with the development of science and technology, people's way of interpreting the world and knowledge has changed, as has knowledge itself. Some of the candidates studying at the Physics, Chemistry and Mathematics departments assert that knowledge absolutely does not change, but the source and methods of investigation do change. As seen in the dimension of simplicity of knowledge, some candidates from the Science Education, Primary Mathematics Education and Chemistry departments indicated that the certainty of knowledge is based on the field. The teacher candidates who agree on these issues were observed to give different explanations. Chemistry education candidates think that knowledge in social science never changes while Science Education candidates assert that knowledge in social science changes and those in Physical Sciences never change. On the other hand, the students in the Primary Mathematics Education department think that all kinds of knowledge would change except for religion. Since Science Education teacher candidates perceive the physical sciences as just based on formulas, they might be disposed to think that scientific knowledge would never change. Similarly, the students think that social science knowledge can change due to the possibility of finding new documents. While Chemistry education teacher candidates attribute their view that social science knowledge does not change to the idea that the core of social science is based on past events, Primary Mathematic Education teacher candidates attribute the idea that religious knowledge never changes to the thought that the source of religious knowledge is not humanity. In the literature, similar findings have been detected, as well (Topçu, 2011;. Topçu (2011) reported that students think that religious knowledge never changes, which is attributed to the notion that Turkish society is mostly Muslim, and that religious beliefs influence students' epistemological beliefs. In his other study examining the differences of epistemological beliefs based on different disciplines, Topçu (2013) revealed that when teacher candidates were asked to compare knowledge in Physics, Chemistry and Biology fields, all the participants stated that knowledge in Physics is more certain than the knowledge in Chemistry and Biology; and that Chemistry knowledge is more certain than Biological knowledge. The reason for teacher candidates' opinion about the uncertainty of Biological knowledge is that it is a new field and open to development (such as genetic cloning). Additionally, Jehng et al. (1993) studied with the students about their beliefs towards the certainty of knowledge. In this study, the researchers concluded that the students studying in social sciences, art and human sciences tend to think and believe that knowledge is not precise, that learning is not a steady process and that they mostly relied on their reasoning skill when they are compared to the students studying and working at engineering departments. The researchers inferred that the students at engineering or science departments believed that the structure of knowledge is systemic and limited, so they tend to believe that knowledge is precise and a problem can be solved in specific lime limits, which might be reason for the findings above. In relation to the "Certain Knowledge" dimension across the class levels, as the class levels of the teacher candidates rise, their beliefs towards the changeability of knowledge get strength, too (Table  12). The first-class teacher candidates were found to adopt the most naive epistemological beliefs regarding the change of knowledge. In the study, it was determined that as the class level rose, the teacher candidates adapted more developed beliefs regarding the variability of knowledge, which is in accordance with the findings in literature (Balantekin, 2013;Rodriguez & Cano, 2007). When "Certain knowledge" dimension is examined according to the gender variable, male teacher candidates believed in the variability of knowledge more than females (Table 13). In other words, the rate of female teacher candidates' believing that knowledge absolutely does not change, is higher than the male ones, which refers to that male teacher candidates possess more developed beliefs related to the certainty of knowledge. These findings do not show parallelism with the literature evidence. In literature, the studies generally remark that females adapt more developed beliefs about knowledge notion than males (Balantekin, 2013;Kurt, 2009).

Implications
 Since epistemological beliefs play an important role in learning and teaching environments, and since they differ on the basis of different variables, future studies can focus on examining how they affect students' learning approaches and self-reflective thinking levels in learning and teaching environments, which in turn provides us with valuable knowledge about this issue and enables us to examine these processes in various contexts.
 As this study was carried out with prospective teachers who study in more scientific areas, the variety of epistemological beliefs can be researched by comparing the epistemological beliefs of prospective teachers studying in social and scientific departments.
 Development of epistemological beliefs of prospective teachers depends on the education programs and teachers (Chan, 2004). Teacher education programs should focus on epistemological beliefs more (Viholainen, Asikainen & Hirvonen, 2014). Therefore, it is recommended that education programs are adjusted accordingly and teachers adopt teaching and learning strategies which focus on the development of such beliefs. Moreover, focusing more on courses about history and nature of science where they can learn about knowledge philosophy and ways to reach knowledge may help improve epistemological beliefs of prospective teachers.
Considering that epistemological beliefs of teachers have a significant impact on teachingand education activities within the class, the impacts of epistemological beliefs of teachers on the epistemological beliefs of students can be researched conducting longitudinal studies. In light of the information above, learning environments should be adjusted and improved to enrich epistemological beliefs of students.