Development and validation of a Professional Ethics Scale for pre-service teachers

DOI: 10.31704/ijocis.2018.013 The purpose of this study was to develop a scale to investigate Turkish pre-service teachers' perceptions of professional ethics in teaching. This survey research consisted of two studies. In study 1, an exploratory factor analysis (EFA) was performed in SPSS 23 in order to investigate the scale’s factorial structure. Study 1 involved a convenience sample of 220 senior pre-service teachers studying at two Turkish public universities. The EFA yielded a single factor structure which accounted for 39.36% of the total variance and included 43 items. In study 2, a confirmatory factor analysis (CFA) was conducted in LISREL 9.30 to test and confirm the unidimensional structure of the 43-item scale obtained in study 1 on a data set of 724 participants from four public universities. Overall, the results prove that this 43-item scale is a valid and reliable tool to test pre-service teachers’ perceptions of professional ethics. Article History: Received Revised Accepted Online 23 April 2018 23 July 2018 19 October 2018 27 December 2018


Introduction
The development of teacher training processes has recently been the subject of much discussion within educational research. Maxwell and Schwimmer's (2016) research emphasizes teachers' role as moral models for their students from the beginning of formal teacher education. In the international literature, a recent trend has been to place greater emphasis on establishing ethical standards for the teaching profession (Barrett, Headley, Stovall & Witte, 2006;Barrett, Casey, Visser & Headley, 2012;Campbell, 2000;Weil, 2005) and integrating course(s) developed in accordance with these standards into teacher training programs.
Ethical standards and ethical principles for the teaching profession will help teachers make more consistent, transparent and rational decisions in different situations they encounter. However, only a limited number of studies have been conducted on this topic. The Turkish Higher Education Council's (HEC) proposal on teaching profession ethics is noteworthy within the national literature on teaching profession ethics. The HEC (2007) proposed establishing "elective courses for the training of professional ethics in teacher education programs" (p.4) like those in Europe and the US in its explanation of the programs to be implemented in HEC institutions. However, examining the courses in these higher education institutions' teacher training programs, "teaching profession ethics" is not among the compulsory courses. Instead, topics related to teaching profession ethics are generally discussed for one or two weeks as part of other education courses. Thus, standards for teaching profession ethics are needed and lessons must be developed on this topic in order to increase the quality of teacher education and thus also of students' education. In 2015, Turkey's Ministry of National Education (MoNE) established six professional ethics principles for educators, which are as follows: (1) ethical principles in relation to students, (2) ethical principles for the education profession, (3) ethical principles in relation to educators, (4) ethical principles in relation to parents, (5) ethical principles in relation to school management and society, and (6) ethical principles related to school administrators' relationships with teachers, students and parents.
Professional ethics is defined as "the totality of behaviors that the parties have to obey or avoid" (Turkish Language Association, 2016). Ethics norms shaped by values, experiences, culture and philosophy help people decide what behavior is right or wrong (Miller, 2000;Singer, 1993). Ethical studies are used to determine the standards of "good", "wanted", "loyal", "accepted", "should be supported", "right or wrong", and "acceptable and unacceptable" (Gözütok, 1999, p. 85). As in other professions, education has its own ethical principles. McKelvie-Sebileau (2011) set out principles for teaching profession ethics in 24 countries and examined the use of them. The most prominent themes were values (respect, equality, honesty, dedication, etc.); relationships with others (colleagues, students, parents, etc.); gender issues (sexual discrimination, harassment, etc.); and occupational competences (knowledge, pedagogy, etc.) (p.19). Campbell (2000) similarly emphasized the importance of teachers' actions regarding ethical principles in relation to the environment and suggested that teachers had different relationships in their working lives: with their students, with other teachers, with school administrators, with ministries, and with parents. They must act in keeping with their responsibility to all of these stakeholders. Teaching profession ethics are also referred to in Turkish higher education law. Higher Education Act No. 2547 lists among the purposes of higher education training students who are "free and respectful of scientific thinking and human rights" (Article 5) and "balanced and healthy in spirit, morality and feeling" (Article 6) (p. 5350).
Pre-service teachers should be able to define ethical and unethical behaviors and attitudes in the teaching profession and act accordingly. Unethical behavior by teachers can lead students to interpret these behaviors as appropriate or correct even though they are not. For this reason, teachers should demonstrate ethical behavior and serve as role models. Carrying out studies on professional ethics in teacher education is of great importance, since teachers are role models for their students through their attitudes and behaviors and influence their students' ethical development indirectly or directly. Unfortunately, ethical rules for teachers are not generally written down. The development of professional ethics principles can be seen as a sign of the profession's maturation (Lovat, 1998, p. 4) and can guide professional practitioners (Haynes, 1998, p.42). Barrett et al. (2006, p. 422;2012, p. 891) addressed the lack of standardized professional ethics principles and criticized the existence of different ethical principles in different US states. The need for additional studies in this regard was also mentioned by Campbell (2000). Culture has been observed to shape people's behaviors and interactions. McKelvie-Sebileau (2011) stated that the codes of the teaching profession developed in different countries, regions or provinces take on different tones and address different phenomena (e.g. social media use), but have similar general qualities.
A number of scales measuring teachers' perceptions of professional ethics exists (Aydoğan, 2011;Barrett et al., 2006;2012;Forsyth, 1980;Gözütok, 1999;Özbek, 2003;Öztürk Aynal, Kumandaş & Ersanlı, 2013;Poisson, 2009;Sevim, 2014;Tabachnick, Keith-Spiegel & Pope, 1991;Yılmaz, 2005). These scales mainly have a multi-dimensional factor structure except one (Aydoğan, 2011). To illustrate, Barrett et al. (2006) developed a scale to detect teachers' perceptions of professional ethics with a threedimensional structure: boundary violations, non-professional inattentive (behavior), and subjective evaluation. Barrett et al. (2012) further refined this three-factor structure and applied it to social media. The three factors identified were as follows: boundary violations, carelessness in teaching, and subjective assessment. Öztürk Aynal, Kumandaş and Ersanlı (2013) applied a four-dimensional structure in their ethical scale for preschool teachers: the teacher's responsibilities to himself/herself and the child, the teacher's responsibilities to the profession and to his/her colleagues, the teacher's responsibilities to families and to society, and the teacher's responsibilities to children with special education needs. In another study,  developed a three-dimensional scale for teacher candidates encompassing relations to students, to the teaching profession (job-related responsibility) and to colleagues.
Currently, international scales are usually developed for teachers working in different countries and cultures (Barrett et al., 2006;2012;Keith-Spiegel & Pope, 1991;Poisson, 2009;Tabachnick, Keith-Spiegel & Pope, 1991). The scales developed in Turkey are generally related to specific teaching fields (Aynal, Kumandaş & Ersanlı, 2013;Özbek, 2003) and/or do not contain material related to technology and/or the changing nature of media use (Gözütok, 1999;Yılmaz, 2005). The relevant literature shows that the use of technology and social media is an important point to consider in the context of teachers' professional ethics (Barrett et al., 2006;2012). For example, Barrett et al. (2012) consider the use of social media to be an ethics violation in their comprehensive field work on teacher ethics. They refer to studies in different countries finding out that the unethical use of social media causes teachers to be fired or suspended from their work. This highlights the importance of addressing social media and technology in ethical rules.
Examining the national and international literature makes clear the need for a scale measuring pre-service teachers' perceptions of professional ethics, considering social, cultural and technological changes. Thus, the aim of this study was to develop a scale to determine pre-service teachers' perceptions of professional ethics in faculties of education in Turkey.

Method
Survey research study design which is one of the mostly used designs in social sciences as well as psychology and health was employed in this study. Data were collected from a large group of pre-service teachers in Study 1 and in Study 2.

Item generation
An item pool regarding pre-service teachers' perceptions of professional ethics in three successive and complementary steps was generated. A thorough review of the relevant literature, examination of codes of ethics in teacher education in different countries (Australia, Canada, France, Netherlands, UK, USA) as well as in Turkey, and analysis of the existing scales currently used to assess teachers' perceptions of professional ethics were carried out initially. In the second stage, structured brainstorming sessions on the "key characteristics of good teachers" and "acceptable/unacceptable behaviors in the teaching profession" with pre-service teachers from different departments at two different major Turkish universities were held. In the final stage, the item pool was generated on the basis of (a) the basic concepts in the literature, (b) existing scales on teachers' professional ethics (Barrett et al. 2006(Barrett et al. , 2012Gözütok, 1999;Keith-Spiegel & Pope, 1991;Poisson, 2009;Sevim, 2014;Tabachnick, Ei & Bowen, 2002;Yılmaz, 2005), (c) the results of the brainstorming sessions held with pre-service teachers, (d) codes of professional ethics for teachers established by the MoNE (2015), (e) cultural values and norms, and (f) our professional experience as teacher educators. After eliminating parallel items, a pool of 67 items on a Likert Scale ranging from 1 (Definitely unethical) to 5 (Definitely ethical) was generated.

Content and face validity
After the items were written, expert feedback from four university faculty members in Measurement and Evaluation, two faculty members in Curriculum and Instruction, one faculty member in Turkish Language and Literature (total seven experts), and three teachers was obtained in order to ensure content and face validity.
The experts focused on the clarity of the instructions and items as well as the items' appropriateness in measuring pre-service teachers' perceptions of professional ethics. In addition to the experts, we asked two pre-service teachers to review the scale using a think-aloud protocol. Just like the experts, the teacher candidates commented on the clarity of the instructions and items as well as the items' appropriateness for measuring the construct and suggested an estimated duration for the scale. The scale was modified based on the experts' and students, reviews, and had 60 items at the end of the revision process.

Procedure
The data collection process took place from April 2016 to February 2017. The scale was administered to 220 pre-service teachers in their final year of teacher education at two major public universities in Ankara, Turkey, during class hours and with the permission of course instructors. The pre-service teachers were given information about the study and participated in the study on a voluntary basis. They were asked to provide demographic information on their gender and department and to mark the extent to which they found the statements in the scale to be ethical.

Participants
The study's sample encompassed 220 pre-service teachers (175 female, 45 male) in their final year of university (seniors) enrolled in different majors at two major public universities in Ankara, Turkey. The reason seniors were selected as participants is that they had already completed all educational courses and thus likely had more awareness of professional ethics in teaching than students in earlier years. Convenience sampling was used to recruit the participants in two different public universities. In this study, we tried to obtain the largest possible number of participants, as suggested by Henson and Roberts (2006). Table 1 presents the demographics of the study participants. The students were majoring in classroom teaching (n=92), English language teaching (n=41), mathematics education (n=15), computer education and instructional technologies (n= 14), early childhood education (n=13), guidance and psychological counseling (n= 13), science and technology teaching (n=11), physical education and sports, (n= 9), German language teaching (n=7), French language teaching (n=4), and physics education (n=1). Data analysis SPSS 23 for Windows was used to analyze the data. An exploratory factor analysis (EFA) with maximum likelihood extraction and a varimax rotation was performed to investigate the factorial structure of the scale.

Results
Kaiser-Meyer-Olkin (KMO) test was used to measure sampling adequacy and Bartlett's test of sphericity to measure the data's suitability for factor analysis. As can be seen in Table 2, the KMO measure of sampling adequacy was .94 (>.60), suggesting that the sample size was adequate for the analysis (Tabachnick & Fidell, 2007). Moreover, Bartlett's test of sphericity was significant at the .00 level (χ2 = 6332.89, DF = 903, p=.00), indicating the suitability of the data for factor analysis (Büyüköztürk, 2004). As can be seen in Table 3, the results of the EFA with maximum likelihood extraction and a Varimax rotation showed that there were nine factors with an eigenvalue larger than 1. The factorial structure of the scale was determined on the basis of both the conceptual framework and empirical results. The first factor accounted for 39.36 of the variance, with an eigenvalue that was 5 times larger than the eigenvalue of the second factor (2.46). This indicates that a single factor was strongly predictive of pre-service teachers' perceptions of professional ethics. The scree plot showing the eigenvalue components (see Figure 1) revealed that there was one obvious dimension after the bending point. Including the second and third factors in the analysis revealed that the items did not cluster within these factors in a conceptual way. Thus, factor loadings below .45 were excluded from the analysis, and consecutive analyses were conducted using Varimax rotation, forcing the items to load onto a single factor. Table 4 presents the items' factor loadings, item-total correlations, and the variance they accounted for. The final scale consisted of 43 items with a uni-dimensional structure which accounted for 39.36 of the variance. Cronbach's alpha coefficient of the scale was calculated to examine the internal consistency reliability. The Cronbach's alpha coefficient of the scale was .96, above the suggested threshold level of .70 in the social sciences (Murphy & Davidshafer, 2005). This revealed the high internal consistency reliability of this measurement tool.

Discussion
In study 1, a uni-dimensional 43-item scale to investigate pre-service teachers' ethical perceptions was developed. In order to do so, 17 items were eliminated as a result of the EFA; the remaining items were reordered and renumbered, and the instrument was finalized. Among the excluded items were "accepting gifts from the students", "not meeting dress code", and "collecting money from students". The final instrument was titled the "Professional Ethics Scale for Pre-Service Teachers". Some of the scale items are: "using school supplies for non-school purposes"; "Spending course time on unrelated activities"; and "being indifferent to the injustices students are exposed to". Responses were given a 5-point Likert-type scale: "I definitely don't find it ethical", "I don't find it ethical", "Not sure", "I find it ethical", "I definitely find it ethical".

Study 2: Confirmatory Factor Analysis
In study 2, a confirmatory factor analysis was employed using a data set of 724 participants to test and confirm the uni-dimensional structure of the 43-item scale obtained through exploratory factor analysis in study 1.

Procedure
The data collection process took place from October to December 2017. The scale was administered to 724 senior pre-service teachers enrolled in different departments at four major public universities in Ankara, Turkey, during class hours and with the permission of the course instructors. Pre-service teachers were furnished with information about the study and were asked to participate on a voluntary basis. They provided demographic information on their gender and department, and were asked to mark the extent to which they found the statements in the scale to be ethical. Participants 724 senior pre-service teachers (596 female, 127 male) enrolled in different majors at four major public universities, including the two public universities in Study 1, in Ankara, Turkey, were involved in the study. Convenience sampling was used to recruit the participants. Table 5 presents the demographics of the study participants. The students were majoring in English language teaching (n=197), early childhood education (n=117), science and technology education (n=103), classroom teaching (n=93), mathematics education (n=73), computer education and instructional technologies (n=69), Turkish language teaching (n= 38), and primary school mathematics education (n= 33).

Data analysis
A confirmatory factor analysis was applied to a data set of 724 participants to test and confirm the unidimensional structure obtained in the exploratory factor analysis in study 1. The analysis was conducted in LISREL 9.30. Note: One participant did not provide gender and department information. Table 6 presents the maximum likelihood estimates for the 43-item Professional Teaching Ethics Scale; squared multiple correlations (R2) of the observed variables and t values for the factor loadings showing the significance of the relationship. Figure 2 presents the confirmatory factor analysis model.

Results
The t-values of all items were significant (p<0.05), indicating that the relation between the latent variable and each item was significant and that no items needed to be excluded from the scale. Fit indexes demonstrated the data's suitability to the uni-dimensional model established in Study 1. Table 7 shows the fit indexes of the 43-item model of the Professional Teaching Ethics Scale.
The model fit was assessed using Hu and Bentler's (1999) two-index strategy. Standardized root mean square residual (SRMR) was used and supported by the root mean square error of approximation (RMSEA), comparative fit index (CFI), non-normed fit index (NNFI), normed fit index (NFI), and relative fit index (RFI). As can be seen in Table 7, the ratio of chi-square to degrees of freedom was X2(860)=3338.32, p<0.01; the root mean square error of approximation (RMSEA) = .06; standardized root mean square residual (SRMR) = .06; comparative fit index (CFI) = .95; goodness-of-fit index (GFI) = .69; non-normed fit index (NNFI) = .94; normed fit index (NFI) = .94; relative fit index (RFI) = .93. The results indicated that the model fit the data, as all fit indexes except GFI were acceptable and suggest valid results. Figure 2 shows the path diagram of the scale items' lambda (factor-loading) values.
The Cronbach's alpha coefficient of the scale was computed to assess the internal consistency reliability. The Cronbach's alpha coefficient of the scale was .96 in study 2 as well, showing that the scale had high internal consistency reliability as a tool to assess pre-service teachers' perceptions of professional teaching ethics.

Discussion
Study 1 produced a uni-dimensional 43-item, 5-point Likert-type scale. We conducted study 2 to test the uni-dimensional structure established in study 1. The results of study 2 confirmed that the uni-dimensional scale measured pre-service teachers' perceptions of professional ethics in teaching. The results further suggested that the scale is a valid and reliable instrument.

General Discussion
This research developed and validated a scale to investigate pre-service teachers' perceptions of professional ethics in teaching. To this end, a 43-item uni-dimensional scale was developed in accordance with the relevant literature, existing scales assessing teaching professional ethics, the results of the brainstorming sessions held with the pre-service teachers, codes of professional ethics for teachers established by the MoNE (2015), cultural values and norms, and our professional experience as teacher educators. .06 Note: GFI = goodness-of-fit index; CFI = comparative fit index; NFI = the normed fit index; NNFI = non-normed fit index; RFI = the relative fit index; SRMR = standardized root mean square residual; RMSEA = root mean square error of approximation The development and validation of the scale was carried out in two studies. Study 1 produced a single factor scale that accounted for 39.36 % of the variance with an eigenvalue that was 5 times larger than the eigenvalue of the second factor. Among the items eliminated were lending students money, accepting gifts from the students, collecting money from students, and being friends with students on social networking sites such as Facebook. These results might be explained in part by the relationship between culture and ethics (Sorokin, 2017). As  defines culture as "collective mental programming" and cultural norms can affect and shape ethical codes. Thus, it seems possible that the eliminated items mostly did not work in the Turkish context as Turkish pre-service teachers might have found the items mostly right. These results further support the idea of McKelvie-Sebileau (2011) who suggests that codes of teaching profession might take on different tones and different phenomena in different countries.
In study 2, the uni-dimensional scale produced in study 1 was tested and confirmed via confirmatory factor analysis. The results indicate that the model fit the data, as all fit indexes except GFI were acceptable. According to the results of study 1 and study 2, the 43-item scale is a valid and reliable tool for examining preservice teachers' perceptions of professional ethics. The Cronbach's alpha coefficient of the scale was .96 in both studies 1 and 2, indicating high internal consistency reliability.
In the literature, there exist many scales exploring teachers' ethical perceptions (Aydoğan, 2011;Barrett, 2006;2012;Özbek, 2003;Öztürk Aynal, Kumandaş & Ersanlı, 2013;Poisson, 2009;Sevim, 2014;Tabachnick, Keith-Spiegel & Pope, 1991;Yılmaz, 2005) with a limited focus on preservice teachers (Gözütok, 1999;Yılmaz & Altınkurt, 2009). In contrast to the earlier studies which suggest a multi-dimensional structure to assess pre-service teachers' ethical perceptions, this study proposes a uni-dimensional structure to investigate pre-service teachers' ethical perceptions, which is consistent of that of Aydoğan (2011) who collected data from a Turkish context. This scale can be used to develop interventions and/or training programs to increase pre-service teachers' awareness of professional ethics. It could be initially used to identify pre-service teachers' perceptions of teaching professional ethics, with the results used to develop specific interventions and/or trainings. A limitation of this study is that participants in Study 1 and Study 2 were selected using convenience sampling, which limits the generalizability of the results. In future studies, the scale can be administered to larger samples using random sampling techniques.

Acknowledgement
A part of this study was presented at the 4th International Congress on Curriculum and Instruction (ICCI-EPOK, 27-30 October 2016, Antalya, Turkey).