Comparing Pre-Service Teachers’ Self-Confidence Levels in Technological Pedagogical Content Knowledge in Terms of Several Variables

This study aimed to investigate and compare pre-service teachers’ self-confidence in technological pedagogical content knowledge (TPACK) concerning their gender, department, and owned digital technologies. To achieve this goal, the survey method was conducted as part of a quantitative method design. Participants of the study consisted of 252 pre-service teachers from four different concentrations: physics, chemistry, biology, and german language teaching. TPACK Self Confidence Scale (TPACK-SCS), which was constructed by Graham, Burgoyne, Cantrell, Smith, and Harris (2009) and adapted to Turkish by Timur and Taşar (2011), was used as the data collection tool. TPACK-SCS is a scale with 4 sub-dimensions as Technological Pedagogical Content Knowledge (TPACK), Technological Pedagogical Knowledge (TPK), Technological Knowledge (TK), and Technological Content Knowledge (TCK). The Cronbach’s Alpha internal reliability coefficients of the scale were calculated between .78 and .94. Since the data obtained did not show normal distribution it was analyzed by Mann-Whitney U and Kruskal Wallis H tests. According to the results, significant differences were found in the level of self-confidence and sub-dimensions of students’ TPACK according to gender, department, and owned digital technologies for education.


The Problem and Research Questions
Within the framework of the great technological development and social change that emerged in the twenty-first century (MOE, 2017), the curricula which were updated at various times before were lastly updated by the Ministry of National Education in 2018. The most recently updated curriculum aims to empower students to gather eight key competencies specified in the Turkey Competencies Framework (MOE, 2018). These competencies are: 1) Digital competence.
2) Communication in the mother tongue,

Research Design
In this study, a cross-sectional survey research method was used for a retrospective inquiry. A cross-sectional study is a study that produces a 'snapshot' of a population at a particular point during a retrospective or pre-service inquiry, while also allowing comparison of different groups (Cohen et al., 2018). According to Creswell (2011), one of the cross-sectional designs in educational research is to compare two or more educational groups in terms of attitudes, beliefs, opinions, or groups. As this study intends to compare self-confidence levels of pre-service teachers in TPACK, cross-sectional design is used.

Study Group
There are 252 pre-service teachers in the study group. They were studying at Physics, Chemistry, Biology, and German Language Teaching Programs of a big-sized university in Turkey. The candidates were explained the scope of the research at the beginning of the study, and the volunteers were studied. Data of the study were gained from the participants at spring semester 2019. The distribution of teacher candidates according to their gender and departments is presented in Table 1.

Data Collection Tools
In this study, the TPACK self-confidence scale (TPACK-SCS) was used to determine the self-confidence levels of pre-service teachers in TPACK. TPACK-SCS was developed by Graham et al. (2009) and adapted to Turkish by Timur and Taşar (2011). The original TPACK-SCS is a six-point Likert-type scale consisting of 4 subscales (Technological Pedagogical Content Knowledge (TPACK), Technological Pedagogical Knowledge (TPK), Technological Content Knowledge (TCK), and Technological Knowledge (TK)) and 31 items. The original scale was developed with 50 science teachers. There were no negative items on the scale. The researchers did not investigate the construct validity because the number of samples was not sufficient. Graham et al. (2009) reported that Cronbach Alpha internal consistency coefficients for the sub-dimensions of the scale were α = .951 for TPACK, α = .913 for TPK, α = .971 for TCK and α = .922 for TK (Graham et al., 2009). Timur and Taşar (2011) adapted the scale to Turkish using a 5-point rating. They concluded that the structure of the original 4-factor scale was also preserved in Turkish culture through a confirmatory factor analysis. Cronbach's Alpha reliability coefficient was α = .93 for the whole scale; α = .91 for TPACK dimension, α = .87 for TPK dimension, α = .94 for TCK dimension and α = .94 for TK dimension (Timur & Taşar, 2011).
Besides, the information about pre-service teachers' gender, department, owned digital technologies, and digital technologies' usage level were collected through the Personal Information Form prepared by the researchers.

Data Analysis
The normality test was conducted to determine whether it shows normal distribution or not. Since the data obtained did not show normal distribution it was analyzed by Mann-Whitney U and Kruskal Wallis-H tests. SPSS was used to conduct the tests.
Moreover, effect sizes are calculated in order to interpret the practical meanings of the results obtained. The effect size is a measure of how big the effect is between the two groups, which is something that statistical significance does not tell us. The effect sizes obtained were construed as r = .20 "small", r = .50 "medium" and r = .80 "large" (Cohen, Manion, & Morrison, 2018).
To compare the scores of the pre-service teachers' TPACK and subscales comparatively, the total scores of each scale were divided into item numbers and converted to 6-point ratings. The following score ranges and corresponding confidence levels were used to interpret these mean scores.

Results
First of all, the Kolmogorov-Smirnov test was used to determine whether the pre-service teachers had normal distribution characteristics before testing their TPACK self-confidence and subscale scores according to their demographic variables. The result of the Kolmogorov-Smirnov test is shown in Table 3. When we look at Table 3, the Kolmogorov-Smirnov test shows that the distribution of the sample did not meet the assumption of normality for the TPACK self-confidence scale and the subscales. Therefore, non-parametric tests (Mann-Whitney U and Kruskal Wallis-H) were used to determine whether the scores of the TPACK self-confidence scale and subscales differed according to various demographic variables.  According to the results of descriptive analysis, it was seen from Table 4 that the overall average scores of the teacher candidates' TPACK confidence levels were x̄ = 3.82. Accordingly, the average score of the candidates for the overall scale is "I trust a lot". In the sub-dimensions, it was determined that the scores ranged between 3.32 and 4.17 on average. The average of the scores obtained from the sub-dimensions corresponds to "I trust a lot" for all sub-scales.

Pre-Service Teachers' Perceptions of TPACK Confidence Level
When the confidence levels of the pre-service teachers are inspected in detail according to their departments, it can be remarked that physics teaching pre-service teachers trust them in TPACK and TPK dimensions, and German teaching pre-service teachers trust TK and TPK dimensions at the highest level which is "I totally trust".

Pre-Service Teachers' TPACK Confidence Levels Difference According to Gender Variable
The first sub-problem of this research is "Do the pre-service teachers' TPACK self-confidence differ according to gender?
The Mann-Whitney U test was conducted to find the answer to the above problem. The result obtained is shown in the   Table 5 shows that Mann Whitney U results for the first sub-problem. When we look at the p-value of pre-service teachers' TPACK self-confidence scores and TPACK, TCK, and TK sub-scale scores, they did not differ according to gender (p > .05). On the other hand, there was a significant difference between females and males at TPK scores and this difference was in the favor of females (p = .02 and p < .05).
The effect size of the difference between females and males in the TPK dimension was found by the formula r = z / √N for z value and calculated as r = .15. It can be said that gender has a small effect on the confidence levels of the candidates in the TPK dimension.

Pre-Service Teachers' TPACK Confidence Levels Difference According to Department Variable
The second sub-problem of the research is "Do the pre-service teachers' TPACK self-confidence differ according to the department?" The Kruskal Wallis H test was conducted to find the answer to the second sub-problem. The result obtained is shown in the Table 6. When we look at Table 6, Kruskal Wallis H results show that pre-service teachers' TPACK self-confidence scores differ according to the department in TPACK and TK subscales (p < .05). To find out the difference between the departments, the Mann-Whitney U test was performed between the departments. Table 7 shows the results of Mann-Whitney U analysis between departments for TPACK sub-scale scores.  Table 7 shows that Mann Whitney U results for the differentiation of the scores of the candidates on the TPACK sub-scale according to the department. When we look at the p-value of pre-service teachers' TPACK sub-scale scores, there was a significant difference between biology and physics candidates' TPACK sub-scale scores and this difference was in the favor of physics pre-service teachers (p = .029 and p < .05).
The effect size of the difference between biology and physics pre-service teachers in TPACK sub-dimension was found by the formula r = Z / √N for z value and calculated as r = .25. It can be said that the department has a small effect on the confidence levels of the candidates in the TPACK sub-dimension for biology and physics pre-service teachers. Table 8 shows the results of Mann-Whitney U analysis between departments for TK sub-scale scores.  Table 8 shows that Mann Whitney U results for the differentiation of the scores of the candidates on the TK sub-scale related to the department. When we look at the p-value of pre-service teachers' TK sub-scale scores, there was a significant difference between biology and german language candidates' TK sub-scale scores and this difference was in the favor of german language pre-service teachers (p = .029 and p < .05).
The effect size of the difference between biology and german language pre-service teachers in the TK sub-dimension was found by the formula r = Z / √N for z value and calculated as r = 0.28. It can be said that the department has a small effect on the confidence levels of the candidates in the TK sub-dimension for biology and german language candidates.

Pre-Service Teachers' TPACK Confidence Levels Difference According to Used Digital Technologies for Education Variable
The third sub-problem of the study is "Do the pre-service teachers' TPACK self-confidence differ according to used digital technologies for education?" The Mann-Whitney U test was conducted to find the answer and the result obtained is shown in the table 9.  Table 9 shows that Mann Whitney U results for the differentiation of the scores of the candidates on the scales related to used digital technologies for education. When we glance at the p-value of pre-service teachers' TPACK self-confidence and TPACK, TCK, and TK sub-scale scores, they did not differ according to mobile phones and desktop computers (p > .05). However, there were significant differences for laptop computer variable at TPACK self-confidence and TPACK, TK, and TPK sub-scale scores (p < .05). The pre-service teachers who use their laptops for their classes have higher scores at TPACK self-confidence and TPACK, TK, and TPK sub-scale than those who do not have. Again, there are similar results for tablet computers. There was a significant difference for the tablet computer variable at TPACK self-confidence and TPACK and TK sub-scale scores (p < .05). The pre-service teachers who use their tablets for their classes have higher scores at TPACK self-confidence and TPACK, and TK sub-scale than those who do not have.
However, the effect sizes calculated for the above significant differences are between .17 and .24 and so they are all small effect sizes. This means that even preservice teachers who use laptops and tablets for their classes have a significant and positive difference than those who do not, it can be said that using laptops and tablets for classes has a small effect on the confidence levels of the pre-service teachers.

Discussion and Conclusion
In this study, it was aimed to investigate and compare pre-service teachers' self-confidence in TPACK according to the variables of gender, department, and used digital technologies. A total of 252 preservice teachers attending at physics, chemistry, biology, and german language teaching programs participated in the study. Data were obtained through a TPACK self-confidence scale constructed by Graham et al. (2009) and translated and validated into Turkish by Timur and Taşar (2011). The results pointed out that the candidate teachers have high self-confidence (I trust a lot) level in all domains. This finding is similar to the findings of Raman (2014) and Açıkgül and Aslaner (2015), Al-Abdullatif (2019), and Rahmadi, Hayati, and Nursyifa (2020). They indicated in their studies that pre-service teachers have enough confidence in themselves in all domains. This result can be interpreted as positive since teachers need to trust themselves to use technology in their courses.
However, Akyıldız and Altun (2018), Altunoğlu (2018), Bağrıyanık (2015), and Karadeniz and Vatanartıran (2015) in the dimension of TPACK self-confidence and Koh, et al. (2010), Gömleksiz and Fidan (2011), and Kazu and Erten (2011) found significant differences in some sub-dimensions according to gender. For example, Koh et al. (2010), in their studies with 1185 pre-service teachers, stated that gender is an effective factor in terms of technology knowledge, content knowledge, and technological pedagogical knowledge. In the TPACK self-confidence dimension, Bağrıyanık (2015) and Altunoğlu (2018) were in favor of male teachers; Akyıldız and Altun (2018) found a significant difference in favor of women. These different results show that gender is not a dominant independent variable in terms of TPACK dimensions (Tuncer & Dikmen, 2018).
The results also showed that the department had a significant effect on pre-service teachers' TPACK scores only in TPACK and TB self-confidence sub-dimensions. The reason for this difference may be due to the experiences of the pre-service teachers in the related departments in using technology in teaching processes. For instance, Jang and Tsai (2012) stated that pre-service mathematics teachers had lower TPACK scores than science teachers, and that science teachers used technological tools and methods more frequently in their classes. Also,  found that there were differences between TPACKs of pre-service teachers in different departments.
When self-confidence scores of the pre-service teachers in TPACK according to used digital technologies for their classes were examined, a statistically significant difference was seen between the pre-service teachers who had laptop or tablet computers. In the related literature, Açıkgül and Aslaner (2015) and  stated that students who have computers have higher TPACK self-confidence.
Based on the study findings, the following policy recommendations will be proposed: . Pre-service teachers' high level of confidence in TPACK shows that they have the basic competencies to use technology in their education. So, it may be important to train teachers who can adapt to new technologies that may arise in the future. Otherwise, it will be inevitable that the investments to be made in this context will be wasted.
. To teach today's digital natives, it is important that teachers should be educated as digitally competent teachers.
. Similar studies in other teaching departments are recommended for the purpose of generalization of findings.
. In this study, the pre-service teachers' beliefs on confidence levels in TPACK were only investigated, in future studies, how they implement their beliefs into practice should also be explored.
. TPACK confidence levels of pre-service teachers were investigated using only quantitative method. In order to fully understand TPACK confidence levels, qualitative methods should be used also.