STATISTICAL ANXIETY AND METACOGNITIVE AWARENESS LEVELS OF GRADUATE STUDENTS STUDYING IN MATHEMATICS EDUCATION PROGRAM

Graduate students are faced with some difficulties in determining statistical methods to be used in their assignments and researches and this leads to anxiety in them. In this study, it was aimed to examine the relationships between statistical anxiety and metacognitive awareness levels of graduate students studying in the field of mathematics education and to compare these levels in terms of some variables. The research study group consists of 101 students studying at the master's and doctorate levels in the department of mathematics education and participated in the research voluntarily. As a data collection tool in the research, Statistical Anxiety Scale (SAS) and Metacognitive Awareness Scale (MCAS) were used. In the analysis of the data, descriptive statistics, ttest, ANOVA and Correlation Coefficient were used. According to the research results, it was determined that there is a negative and moderately significant relationship between statistical anxiety and metacognitive awareness levels of graduate students studying in the field of mathematics education. As a result of the analyzes, it was observed that the i Correspondence: email baris.demir@kocaeli.edu.tr Murat Altun, Gül Kaleli Yılmaz, Barış Demir, Hülya Sert Çelik STATISTICAL ANXIETY AND METACOGNITIVE AWARENESS LEVELS OF GRADUATE STUDENTS STUDYING IN MATHEMATICS EDUCATION PROGRAM European Journal of Education Studies Volume 9 │ Issue 1 │ 2022 45 statistical anxiety and metacognitive awareness of the students were high, the statistical anxiety and metacognitive awareness levels differed statistically according to the variables of graduate education level and stage of graduate education, but did not differ significantly according to the variables of gender and taking statistics course.


Introduction
Mathematics is largely the product of the curiosity to know and understand the truth, and some mathematics arose from practical needs. It facilitates life in terms of mathematical results or helps produce thoughts that will make it easier. The value of statistics can also be evaluated in this context as a branch of mathematics (Altun, 2018). In an environment of uncertainty, statistics is seen as a science that uses research, decision-making and estimation mechanisms together and develops them and a technology of other disciplines (Koparan and Güven, 2014). The literature emphasizes the importance and role of statistics in daily life (Guidelines Assessment in Statistic Education [GAISE], 2005[GAISE], , 2006Ben-Zvi & Garfield, 2008;Schield, 2008). In the GAISE (2005) report, which is one of them; it is underlined that our life is surrounded by qualitative and quantitative contexts, and a high school graduate should be equipped with sufficient statistical knowledge and skills to cope with the situations in his/her life and to have a productive and peaceful life. It is also important for a researcher who will conduct research in the relevant field to follow the literature, analyze his/her data and understand the studies conducted (Tan, 2016). For these reasons, statistics is seen as a tool that complements scientific research (Sutarso, 1992). In addition, researchers who conduct scientific studies are expected to know statistical techniques and research methods (Erkuş, 2011). As it is known, the level of education in which scientific studies gain importance is graduate education. One of the ultimate goals of graduate education is to train and gain scientists who have analytical thinking skills, who are creative, sensitive to ethical principles, who produce solutions, who have an academic perspective, who willingly participate in group work and interdisciplinary studies, who produce quality studies and who can meet the needs of society (Aslan, 2016). In this context, having statistical knowledge plays a vital role in graduate education.
A statistics course is a must to take in almost all of the graduate education programs in social sciences, behavioral sciences and educational sciences, where it is valuable to have statistical knowledge. However, statistics course is a negative experience for most graduate students (Collins & Onwuegbuzie, 2007). Anxiety about any lesson negatively affects the achievement of the desired goals of the lesson besides causing the lesson to be perceived as more difficult than it is. In other words, anxiety is a negative condition that affects academic performance (Sutarso, 1992;Onwuegbuzie & Wilson, 2003). In addition, anxiety is a condition that manifests itself with physiological that there is a significant relationship between metacognitive awareness and anxiety in general.

Objective and Importance of the Study
The inclusion of statistical information in every field in daily life attracted the attention of program developers, and then reform movements began in curricula (NCTM, 2000). For this reason, the achievements in the field of statistics have an important place in mathematics teaching programs. In addition, the achievements in the secondary school mathematics curriculum are discussed at each grade level and in more detail. Along with the mentioned reform movements, it is expected that statistics education is of great importance and that the teachers who perform this education should have some skills and field knowledge. In terms of teaching mathematics, metacognition skills have an important place in increasing the success of their students for teachers. For this reason, it is important for students to notice, model and internalize the expressions and behaviors of teachers towards their metacognitive awareness (Schunk, 1989). Therefore, it can be said that teachers' metacognitive awareness levels are very effective in improving students' metacognitive skills. It is expected that the person will have less anxiety about the situations when metacognitive awareness is characterized as a system that controls the information one has. It is expected that an individual who knows to what extent his/her knowledge and skills are, and which method s/he will use to apply it, and most importantly, is aware of his/her weaknesses, will be less anxious than an individual who does not have this awareness. For this reason, although anxiety is a negative attitude, it is thought that the factors affecting it may also directly affect anxiety. For this reason, it is desirable that teachers' metacognitive awareness is high and their anxiety levels are low during the teaching processes in statistics courses. Because the first goal of statistics education should be to reduce the statistics anxiety in students (Blalock, 1987). Since the teacher's anxiety in teaching processes will increase his/her academic performance, it is important for the students to structure the lesson better and be more productive because otherwise, it reduces the quality of the lesson. At the same time, it is thought that this negative attitude towards the statistics course, which is passed from the teacher to the student, will also affect the students' future lives. In addition to the level of anxiety in learning, learning paths supported by the individual's metacognitive awareness skills or the control and planning of these paths are also important for the realization of effective learning. A high level of anxiety control is also a factor that makes learning difficult. The extent to which individuals possess metacognitive skills and characteristics and to what extent these skills are related to statistical anxiety arouses curiosity. Therefore, the research is important to reveal the relationship between statistical anxiety and metacognitive awareness of graduate students studying in the mathematics education program. In addition, when the literature was examined, no studies were conducted with graduate students and in which the variables of statistical anxiety and metacognitive awareness were discussed and examined together. For these reasons, it was necessary to carry out this study.
In this study, it was aimed to examine the relationships between statistical anxiety and metacognitive awareness levels of graduate students studying in the field of mathematics education and to compare these levels in terms of some variables. In addition, the research will seek answers to the following sub-problems: 1) What are the statistical anxiety levels of the graduate students studying in the mathematics education program? 2) What are the metacognitive awareness levels of the graduate students studying in the mathematics education program? 3) Do the statistical anxiety levels of the graduate students studying in the mathematics education program differ significantly according to the variables of gender, education level, stage of graduate education and taking statistics course? 4) Do the metacognitive awareness levels of the graduate students studying in the mathematics education program differ significantly according to the variables of gender, education level, stage of graduate education and taking statistics course? 5) Is there a significant relationship between statistical anxiety levels and metacognitive awareness levels of the graduate students studying in the mathematics education program?

Material and Methods
A relational screening model was used in the research. The relational screening model was used to determine the relationship between two or more variables and the level of this relationship (Creswell, 2003). No intervention was made to the group with which the research was conducted, and the existing situation was tried to be described as it was.
The study aimed to examine the relationships between statistical anxiety and metacognitive awareness levels of students and to compare these levels in terms of some variables.

Study Group
The research participants consisted of 101 students studying at master's and doctorate levels in the department of mathematics education affiliated to the Educational Sciences, and Science Institutes of the universities in the Central Anatolia, Eastern Black Sea and Marmara regions voluntarily participated in the research. The study group was selected by a convenient sampling method. The demographic information of the graduate students studying in the mathematics education program participating in the study is given in Table 1. According to Table 1, 73.3% of the students participating in the research were female, and 26.7% were male. 69.3% of the students were studying at a master's degree, and 30.7% at PhD degree, 56.4% of the participants were in the master's course stage, 12.9% in the master's thesis stage, 18.8% in the PhD course stage and 11.9% in the PhD thesis stage. 56.4% of the students participating in the research stated that they took courses related to statistics only at the undergraduate level, and 43.6% of them both at undergraduate and graduate levels.

Data Collection Tools
As a data collection tool in this research; the "Statistical Anxiety Scale (SAS)", which was developed by Faber et al., (2018) for graduate students and adapted in Turkish by Güler et al., (2019), and the "Metacognitive Awareness Scale (MCAS)", which was developed by Fırat-Durdukoca and Arıbaş (2019) for teacher candidates were used. Data were collected online via www.surveey.com between 20 February 2021 and 30 June 2021. The construct validity of the Statistical Anxiety Scale was examined by applying exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). Both onedimensional and three-dimensional structures were obtained similarly to the results obtained in factor analysis in SAS's original form. Additionally, since the original form of the SAS was prepared with a three-dimensional structure in anxiety, avoidance and emotionality, it was tested in CFA. In the reliability analysis, the Cronbach Alpha internal consistency coefficient was 0.96 for the overall scale; for the anxiety, avoidance and emotionality dimensions, it was found as 0.91, 0.83 and 0.91, respectively. These results reveal that the Turkish version of the SAS is a valid and reliable measurement tool (Güler et al., 2019). This scale was developed as a four-point Likert scale; I strongly disagree, I agree somewhat, I quite agree, and I completely agree. There is no reverse-scored item.
The highest score that can be obtained from this scale is 68, and the lowest score is 17. The Cronbach-Alpha coefficient for the results of the reliability analysis conducted in this study was calculated as 0.80 for the anxiety sub-dimension, 0.79 for the avoidance subdimension, 0.75 for the emotionality sub-dimension, and 0.90 for the overall Cronbach's alpha coefficient of the scale. A low score can mean that students have low anxiety about statistics, and a high score can mean high anxiety levels about statistics. Considering that the gap of this scale was calculated with the formula of "range/number of groups to be performed" (Tekin, 1996), the arithmetic mean score gaps based on the evaluation of the data obtained in the research were interpreted as 1.00-1.75; "low anxiety", 1.76-2.50; "moderate (normal) anxiety", 2.51-3.25; "high anxiety", 3.26-4.00; "very high anxiety".
The Metacognitive Awareness Scale, developed by Fırat-Durdukoca and Arıbaş (2019) to determine the metacognitive awareness levels, consists of 18 items and 3 subdimensions: self-awareness, organizational awareness and judgmental awareness. In the reliability analysis of the scale's sub-dimensions, the Cronbach-Alpha coefficients were found to be 0.79 in the sub-dimension of personal awareness, 0.72 in the sub-dimension of organizational awareness, and 0.62 in judgmental awareness. The scale is a five-point Likert type, and the degree of agreement with the items is classified as 1 "never", 2 "rarely", 3 "often", 4 "usually" and 5 "always" (Fırat-Durdukoca & Arıbaş, 2019). In the evaluation of the mean scores of the scale sub-dimensions; 1.00-1.80 "very low", 1.81-2.60 "low", 2.61-3.40 "medium", 3.41-4.20 "high", 4.21-5.00 "very high" values will be taken as criteria. As a result of the reliability analysis made in this study, the Cronbach alpha coefficients were calculated as 0.81 for the sub-dimension of personal awareness, 0.82 for the sub-dimension of organizational awareness, 0.86 for the sub-dimension of judgmental awareness, and 0.92 for the overall Cronbach alpha coefficient of the scale.

Data Analysis
The program SPSS v22 was utilized in data analysis. Normality assumptions were tested to determine the data analysis tests to be used. In the data analysis, descriptive statistics, t-test, ANOVA and Pearson Moments Correlation Coefficient were used. The Scheffe test was used to determine between which groups the difference was in the case of the significance of the F value obtained. Effect sizes were found with eta squared in ANOVA and Cohen's d in t-tests. To interpret the effect value result obtained 0.01: small effect, 0.06: medium effect and 0.14: large effect gaps suggested by Cohen (1988) were used. In the statistical analysis, the level of significance was taken as 0.05 and 0.01.

Finding
In this part of the research, depending on the data obtained in the sub-problems, firstly, the statistical anxiety levels and metacognitive awareness of the graduate students in the mathematics education program were determined. Then the statistical anxiety levels and metacognitive awareness were examined according to the variables of gender, education level, stage of graduate education and taking a statistics course. Finally, the findings regarding the relationship between statistical anxiety and metacognitive awareness of graduate students in the mathematics education program were included. As can be seen in Table 2, the mean scores of the students studying in the mathematics education program regarding statistical anxiety were at the level of "high anxiety" (X ̅ =2.57). The mean scores of the sub-dimensions related to the statistical anxiety of the students were respectively; "high anxiety" in the anxiety dimension (X ̅ = 2.71), "normal anxiety" in the avoidance dimension (X ̅ =2.28) and "high anxiety" in the emotionality dimension (X ̅ =2.59). As it can be seen in Table 3, the mean scores of students studying in the mathematics education program regarding their metacognitive awareness were at the "high" level (X ̅ =3.60). The mean scores of the sub-dimensions related to the metacognitive awareness of students were respectively; "high" in the dimension of self-awareness (X ̅ = 3.63), "high" in the organizational awareness dimension (X ̅ =3.61) and "high" in the judgmental awareness dimension (X ̅ =3.51). When Table 4 was examined, a statistically significant difference was found in the avoidance sub-dimension when the mean scores of statistical anxiety depending on the gender variable of the graduate students studying in the mathematics education program were examined (p=0.024<0.05). The effect value showing the size of this significant difference was calculated as (η 2nd = 0.050) and determined as a small level effect. When the mean scores were examined, it was seen that this difference was in favor of female students and female students had lower anxiety levels. No statistically significant difference was found between the mean scores of the other sub-dimensions and the gender variable (anxiety p=0.122>0.05, emotionality p=0.475>0.05). When the total mean scores were examined, it was seen that there was no statistically significant difference between the statistical anxiety of the graduate students studying in the mathematics education program and the gender variable (p=0.866>0.05). When Table 5 was examined, a statistically significant difference was found in the avoidance sub-dimension when the mean scores of statistical anxiety depending on the gender variable of the graduate students studying in the mathematics education program were examined (p=0.000<0.05). The effect value showing the size of this significant difference was calculated as (η 2nd = 0.079) and determined as a small level effect. When the mean scores were examined, it was seen that this difference was in favor of master's degree students, and these students had lower anxiety levels. No statistically significant difference was found between the mean scores of the other sub-dimensions and the gender variable (anxiety p=0.122>0.05, emotionality p=0.241>0.05). When the total mean scores were analyzed, it was seen that there was a statistically significant difference between the statistical anxiety of graduate students studying in the mathematics teaching program and the variable of education level (p=0.046<0.05). The effect value showing the size of this significant difference was calculated as (η 2 = 0.037) and determined as a small level effect. When the mean scores were examined, it was seen that this difference was in favor of master's degree students, and these students had lower anxiety levels. When Table 6 was examined, a statistically significant difference was found in the avoidance sub-dimension when the mean scores of statistical anxiety of the graduate students studying in the mathematics education program depending on the gender variable were examined (p=0.001<0.05). When the results of the Scheffe test performed to determine the difference between the groups were examined, it was seen that this difference was in favor of the group that was at the master's thesis stage in graduate education between the group that was in the master's thesis stage and the PhD course stage and between the group that was in the master's thesis stage and the PhD thesis stage. No statistically significant difference was found between the mean scores of the other sub-dimensions and the variable of the students' graduate education stage (anxiety p=0.180>0.05, emotionality p=0.342>0.05). When the total mean scores were examined, it was observed that there was a statistically significant difference between the statistical anxiety of graduate students studying in the mathematics teaching program and the variable of the graduate education stage (p=0.042<0.05). When the results of the Scheffe test performed to determine the difference between the groups were examined, it was determined that this difference was in favor of the group in the master thesis stage between the master's thesis stage and the PhD course stage. It was determined that the effect size values of the dimensions with a significant difference regarding the variable of the graduate education stage were at the high effect level in the avoidance dimension (η2 = 0.152) and at the medium effect level in the total of the scale (η2 = 0.069). When Table 7 was examined, a statistically significant difference was not found in all subdimensions and in total when the statistical anxiety mean scores of the students studying in the mathematics education program were examined depending on the variable of taking statistics course before (anxiety p=0.150>0.05, avoidance p=0.742>, 0.05 emotionality p=0.326>0.05, total p=0.269>0.05). When the mean scores were examined, it was seen that the students who took statistics courses at both undergraduate and graduate levels had a higher level of anxiety. When Table 8 was examined, a statistically significant difference was not found in all subdimensions and in total when the mean scores of metacognitive awareness of the graduate students studying in the mathematics education program were examined depending on the gender variable (self-awareness p=0.152>0.05, organizational p=0.819>0.05, judgmental p=0.128>0.05, total p=0.341>0.05). When Table 9 was examined, a statistically significant difference was found in the subdimensions of self-awareness and organizational awareness when the metacognitive awareness mean scores of the graduate students studying in the mathematics education program were examined depending on the education level variable (Self-awareness p=0.006<0.05, Organizational awareness p=0.014). <0.05). When the mean scores were examined, it was seen that this difference was in favor of students at the master's degree level. No statistically significant difference was found between the mean scores of judgmental awareness and the education level variable (p=0.271>0.05). When the total mean scores were analyzed, it was seen that there was a statistically significant difference between the metacognitive awareness of graduate students studying in the mathematics teaching program and the variable of education level (p=0.018<0.05). When the mean scores were examined, it was seen that this difference was in favor of the master's degree students. It was determined that the effect size values related to the education level variable of the dimensions with a significant difference were at the medium effect level in the self-awareness dimension (η2 = 0.063), at the small effect level in the organizational awareness dimension (η2 = 0.053), and the small effect level in the total (η2 = 0.051).  When Table 10 was examined, a statistically significant difference was found in the subdimensions of self-awareness, organizational awareness, and judgmental awareness when the metacognitive awareness mean scores of the graduate students studying in the mathematics education program were examined depending on the variable of the graduate education stage (self-awareness p=0.000<0.05, organizational awareness p=0.000<0.05, judgmental awareness p=0.000<0.05). When the results of the Scheffe test performed to determine the difference between the groups were examined, it was seen that this difference was in favor of the group that was in the master's thesis stage in all sub-dimensions, between the master's thesis and the master's course, the master's thesis and the PhD course, and the master's thesis and the PhD thesis.

Findings Regarding the Analysis of Metacognitive Awareness Levels of Graduate Students Studying in the Mathematics Education Program According to Some Variables
When the total mean scores were examined, it was seen that there was a statistically significant difference between the metacognitive awareness of the graduate students studying in the mathematics education program and the variable of the graduate education stage (p=0.000<0.05). When the results of the Scheffe test performed to determine the difference between the groups were examined, it was determined that this difference was in favor of the group that was at the master's thesis between the master's degree thesis and the master's course, the master's thesis and the PhD course, and the master's thesis and the PhD thesis. The effect size values of the factors with significant difference related to the graduate education stage variable were determined the high effect level in the self-awareness factor (η 2 = 0.235), high effect level in the organizational awareness factor (η 2 = 0.296), high effect level in the judgmental awareness factor (η 2 = 0.170) and high effect level in the total of the scale (η 2 = 0.262).  When Table 11 was examined, a statistically significant difference was not found in all sub-dimensions and in total when the mean scores of metacognitive awareness of the graduate students studying in the mathematics education program were examined depending on the taking statistics course variable (self-awareness p=0.876>0.05, organizational p=0.072>0.05, judgmental p=0.144>0.05, total p=0.230>0.05). When the mean scores were examined, it was seen that the students who took statistics courses at both undergraduate and graduate levels had a higher level of metacognitive awareness.  Table 12 was examined, it was seen that there was a moderately significant negative correlation between all dimensions of the metacognitive awareness scale and all dimensions of the statistical anxiety scale applied to graduate students studying in the mathematics education program. When the results were examined in detail, the correlation values between the anxiety sub-dimension of statistical anxiety and selfawareness, organizational awareness and judgmental awareness were found respectively (r = -0.500**, -0.495**, -0.495**), correlation values between avoidance sub-dimension and self-awareness, organizational awareness and judgmental awareness (r =-0.443**,-0.419**, -0.482**), the correlation values between emotionality sub-dimension and selfawareness, organizational awareness and judgmental awareness (r = -0.436**, -0.450**, -0.424**). It was determined that there was a moderately significant negative relationship between their total statistical anxiety and their total metacognitive awareness (r=-0.564, p=.000).

Conclusion and Discussion
In this study, it was aimed to examine the statistical anxiety and metacognitive awareness levels of graduate students studying in the mathematics education program and the relationship between them. In addition, it was also examined whether the students' statistical anxiety and metacognitive awareness differed according to the variables of gender, education level, graduate education stage, and taking a statistics course.
Considering the mean scores obtained from the statistical anxiety scale in the study, it was seen that the graduate students studying in the mathematics education program had a high level of statistical anxiety in the anxiety and emotionality subdimensions, normal level in the avoidance sub-dimension, and generally high level in statistical anxiety. This result supports the studies in the literature (Baloğlu, 2003;Baloğlu & Zelhart, 2004;Primi & Chiesi, 2018;Roberts & Saxe, 1982;). When the mean scores of the students obtained from the metacognitive awareness scale were examined, it was seen that they had a high level of metacognitive awareness in all sub-dimensions and in general. This result obtained from the research is similar to the studies in the literature (Aykut et al., 2016;Özsoy et al., 2010).
The study examined the statistical anxiety levels of graduate students studying in the mathematics education program in terms of variables such as gender, education level, graduate education stage, and taking a statistics course. As a result of the research, it was concluded that the statistical anxiety levels of the students did not change according to the gender variable. A statistically significant difference was found only in the avoidance sub-dimension. When the mean scores were examined, it was seen that this difference was in favor of female students and female students had lower anxiety levels. Different findings were obtained in the studies performed to determine the effect of gender variable on statistical anxiety. In Sutarso's (1992) study, it was determined that there was no significant difference between the statistical anxiety of male and female students; in the studies conducted by Benson (1989), Baloğlu (2003), andRodarte-Luna andSherry (2008), it was determined that the statistical anxiety of female students was significantly higher than that of males.
In this study, students' statistical anxiety levels differed statistically significantly in the avoidance sub-dimension and total, according to the variables of graduate education level and graduate education stage. This finding is similar to the result obtained from the studies of Büyüköztürk (1997) and Saracaloglu et al., (2005). When the results were examined, it was seen that PhD students had higher anxiety levels than graduate students, and students at the PhD thesis stage had a similarly higher level of statistical anxiety than the students at other stages. This situation can be interpreted as the increase in statistical anxiety as the academic level rises, and the intensity of statistical knowledge and methods on the subjects studied increases the anxiety.
According to the findings of this study, there was no statistically significant difference in all sub-dimensions and in total when the statistical anxiety mean scores of the graduate students studying in the mathematics education program were examined depending on the variable of taking a statistics course. In this case, it can be said that statistics anxiety and taking a statistics course are not related. When the findings were examined, it was seen that the students who took a statistics course at both undergraduate and graduate levels had a higher anxiety level, unlike the students who only took a statistics course at the graduate level. This result is similar to the result of Saracaloğlu (2008) and Aslan and Karagül's (2016) studies that anxiety levels about doing research do not differ significantly according to taking courses.
No statistically significant difference was found in students' metacognitive awareness levels according to gender in the study. This result is similar to some studies in the literature (Memnun and Akkaya, 2009;Öztürk and Kurtuluş, 2017). However, there are studies in the literature stating that gender is an important predictor of metacognitive awareness levels contrary to this situation and in some studies in the literature, there are research results showing that female students (Alcı andAltun, 2007 Gül et al., 2015) or male students (Gürefe, 2015;Memiş and Arıcan, 2013) have a higher level of metacognitive awareness.
In this study, a statistically significant difference was found in the metacognitive awareness levels of the students, in the sub-dimensions of self-awareness, organizational awareness and total according to their graduate education levels. When the mean scores were examined, it was seen that this difference was in favor of students at the master's degree level. This result is in line with the result of the study of Irak et al., (2015) that the level of metacognitive awareness differs according to the education level. A statistically significant difference was found in all sub-dimensions and in total when the metacognitive awareness mean scores of the graduate students studying in the mathematics education program were examined depending on the variable of the graduate education stage. When the results were examined, it was seen that the students in the master's thesis stage had a higher level of metacognitive awareness. This is an unexpected situation contrary to the thought that metacognitive awareness will increase as the education level increases. Why PhD students do not have higher metacognitive awareness levels than graduate students should be examined in detail in a separate study. The relationship between statistical anxiety and metacognitive awareness of graduate students studying in a mathematics education program was examined in the study. It was determined that there was a negative and moderately significant relationship between these variables. In other words, it can be said that as the metacognitive awareness of the students increases, their statistical anxiety decreases or as their statistical anxiety increases, their metacognitive awareness decreases. In their study, Öztürk and Serin (2020) stated that as the metacognitive awareness levels of teacher candidates increase, their anxiety towards mathematics and, therefore, their anxiety towards teaching decreases. These results show parallelism with the research findings. Kaçar and Sarıçam (2015), Gökbulut and Akdağ (2016) stated in their studies that there is a positive relationship between mathematics anxiety and metacognitive awareness levels of primary school teacher candidates. Although the type of anxiety examined in our research is statistical anxiety, statistics can be evaluated as a branch of mathematics in this context. Everson et al., (1994) stated that individuals with low anxiety levels could use their metacognition positively and thus perform better than individuals with high anxiety, and in times of high anxiety, metacognition affects more negatively, and thus results in poor performance. Contrary to this situation, Wells and Matthews (1996), Spada et al., (2010), Irak (2012), Gargabi et al., (2014) and Savaşan (2019), saw in their studies that there are significant positive relationships between anxiety and metacognitive awareness. Everson et al. (1994) suggested that individuals with high metacognitive awareness, who pay close attention to their thinking processes, show high anxiety and cause a decrease in their performance. They stated that high anxiety and high metacognition show the worst performance, so if an individual is too aware of metacognition, his/her excessive metacognition (negative metacognition) may lead the individual to failure due to a high anxiety level in a given task.
When we think of metacognitive awareness as a system that controls what a person knows and does not know, it means that the person does not have anxiety or is less anxious about the situation s/he is aware of. The more people are aware of their knowledge, and the better they use it. Most importantly, if they are aware of their deficiencies, their negative attitudes, such as anxiety towards the situation or the lesson, are expected to decrease. From this point of view, it can be said that anxiety is also a negative attitude and situations that affect this attitude can directly affect anxiety. Therefore, it can be concluded that increasing metacognitive awareness will reduce anxiety; in other words, individuals with high levels of metacognitive awareness have less anxiety. Studies have shown that having metacognitive awareness skills improves attitudes towards the lesson in a positive way (Akdur, 1996;Gelen, 2004;Küçük-Özcan, 2000;Manning and Payne, 1996).
Future studies can determine whether there are significant relationships between the statistical anxiety of students who continue their graduate education in different fields and variables that are not addressed in this study. In addition, quantitative data can be supported by the findings obtained from the analyzes by conducting more detailed and in-depth qualitative research to determine the reasons for students' anxiety towards statistics. And most importantly, why PhD students have less metacognitive awareness and more anxiety than graduate students should be examined in detail.

Conflict of Interest Statement
The authors declare that they have no conflict of interest related to the study or preparation of the manuscript.
About the Authors Murat Altun has been working as a Professor Doctor at Department of Mathematics Education Faculty of Education at Bursa Uludağ University in Turkey. Gül Kaleli Yılmaz has been working as an Associated Professor Doctor at Department of Mathematics Education Faculty of Education at Bursa Uludağ University in Turkey. Her research is mainly on technological pedagogical content knowledge, mathematics and geometry education. She has journal articles published in international and national indexes, book chapters and papers presented to international meetings. Baris Demir is lecturer at Kocaeli University Hereke Omer Ismet Uzunyol Vocational School and PhD student at Bursa Uludag University Mathematics Education. His academic interests are mathematics education and mathematical modeling. He has articles and presentations on these topics. Hulya Sert Çelik is a mathematics teacher at a secondary school affiliated to national education and a doctoral student at Bursa Uludag University. Her academic interests are mathematics and geometry education. She has articles and presentations on these topics.