The Demand for Private Tutoring in Turkey: An Analysis of Private Tutoring Participation and Spending

This article explores the process, reasons, and determinants of private tutoring as perceived by the high school students in Şanlıurfa, Turkey. This is a survey study and the quantitative data for the study was collected with a questionnaire from 1329 high school students during the spring semester in 2019. According to the findings, almost half of the participants reported having received private tutoring at private teaching institutions in the last year. The most popular subjects for private tutoring were math, science, and Turkish. Exam-focused learning, poor classroom teaching were reported as the most important reasons behind receiving private tutoring. The individuals who referred most of the participants to private tutoring were the parents. Besides, it was determined that as age, grade, father and mother’s education level, level of income, and parents’ belief in the need for education increases, the likelihood of receiving private tutoring increases; as satisfaction level with the school decreases, students are more likely to participate in private tutoring. Also, it was found out that female students spent on private tutoring more than male students. It is concluded that the demand for private tutoring in Turkey is high, and this may be due to the university entrance system based on high-stakes testing.

technical Anatolian high schools, religious high schools, and private high schools. Any student who completes 8 years of basic education can go to these schools through a high-stake exam. Graduates of high schools can go to higher education institutions upon getting a satisfying score from the nationwide university entrance examination.
Private tutoring (PT), supplement the school teaching in exchange for a fee in academic subjects mainly for usually public examinations, is considered as a problem that can influence the teaching, students, and parents. Despite its prevalent existence, PT, as a parallel education system, remains very firmly in the shadows (Ireson, 2004). The private teaching institutions (dershane) are a highly debated, controversial issue in Turkey because it is believed that their existence and private tutoring practices disturb the equity in education in favor of children of wealthy families. To eliminate these institutions, the Ministry of National Education has made some legal rearrangements and forced them to shut down. However, most of the institutions still exist in different forms under different names such as course center, study center, educational support center, etc., and work like before. Students attend these centers for the same reason, to prepare for high-stake exams. As the closure of the private teaching institutions directly concerns many individuals, the sudden introduction of radical and comprehensive changes and their implementation without sufficient time is regarded as a move that puts the process in trouble.
To address the problem more effectively, it is believed that the possible causes and solutions of the issue need to be determined scientifically and holistically (Baran & Altun, 2014).
As a global phenomenon, PT or "shadow education", which is the process of marketization of education, has long been visible and embedded in Turkish educational life. A vast number of students from different socio-economic backgrounds receive private tutoring in different forms. Although PT is considered a supplement to mainstream education, it poses a serious risk to the educational system with its serious backwash effects on teaching and learning in mainstream schools. However, it is commonly believed to be beneficial and necessary by parents, students, and teachers. PT is considered a means of getting a better education or access to higher education, which enables individuals to move up the social ladder. The reason why PT, as a supplementary strategy, has been a demanded educational service by Turkish families is that the mainstream schools are not assumed to offer quality education in alignment with high-stake examinations, and ambitious parents desire better education for their children regardless of the financial and psychological burden of PT.
The first studies on PT in Turkey were conducted by Tansel and Bircan (2004, 2005, 2006. They claim that PT is an important, widespread phenomenon in Turkey and the main purpose for the demand for PT in Turkey is preparing for the competitive university entrance exam. PT is considered as a necessity to provide better education to the children for higher future incomes because it is generally believed that receiving PT increases the probability of being placed in a university program. Besides, it is believed that the content of the examination is not in alignment with the subject matters taught in the last year of high school, therefore, the senior students stop going to the school before the exam and focus on their studies in PT institutions. Thus, there is a tendency among Turkish parents to have their children receive PT to increase their probability of success at the exam. The main form of PT is receiving education at private teaching institutions. Compared to public schools, these are profit-oriented school-like organizations with better facilities such as under crowded classrooms, better class materials, and better teacher-student interaction. These institutions have become a large industry and PT expenditures are a very significant item in the budgets of most households. Tansel and Bircan (2004, 2005, 2006 determined several factors that affect the probability of receiving PT in Turkey. The expenditure on PT is found to be strongly related to the family income. In other words, the expenditure on PT increases with family income. Another important factor that increases or decreases the probability of receiving PT is the parents' educational level. With the increase of the parent's education level, the chances of receiving PT increases. The mother's education level and their being wage earner increases the likelihood of receiving PT rather than that of fathers. There is also a relationship between academic achievement and the probability of receiving PT. The students' self-reported or actual academic achievement is also an important determinant of resorting to PT. The students with high motivation, the ones who graduated from high school with honor or satisfactory degrees are more likely to receive PT. Females are statistically less likely to attend PT than males, but the gender difference does not seem to be much. Transitions between the levels of education based on results of the high-stake exams cause a competitive environment, which makes students feel a need for more support. The students seek educational support at private tutoring centers, study centers, and courses, as well as taking private tutoring. The TED (2010) report reveals that the level of private tutoring of students falls to the first grade in primary education. Despite the high prevalence of PT, the number of studies conducted on the issue is considered to be very low (as cited in Turkan & Çeliköz, 2018). Also, due to lack of official statistics and documentation (Tansel & Bircan, 2006), it has appealed to little academic interest and studies usually focus on the issues related to its nature, scope, scale, and motivation.

Importance of the Problem
Existing studies in Turkey (Tansel & Bircan, 2004, 2005, 2006 mainly examined the matters such as expenditures, the effect on the university entrance exam, demand for PT, and PT in general. Furthermore, the studies were big scale, but not recent. According to the findings of the study conducted by Tansel (2013), the characteristics and distribution of private tutoring vary significantly from province to province. In provinces located mostly in the east and southeast of the country, they are of lower quality. Beyond this, the lower number of secondary school students attending dershanes may be considered as an indicator of significant inequalities. Compared to other provinces in Turkey, Şanlıurfa has a smaller share of private teaching institutions (Tansel & Bircan, 2007). Therefore, the study is conducted in Şanlıurfa. To sum up, it is generally believed that today schools fail to fulfill their functions adequately and this leads to alternative educational practices that support academic success in many countries around the world. The most common of these is the PT, which is considered as an alternative to support for school lessons and preparation for exams. When the literature is examined, it is seen that the studies on private lessons mainly focus on equality of opportunity in education, the competitive environment created by education systems, and the cost of education. The research is considered important in terms of giving an idea about why private lessons are needed, the prevalence of PT, the reasons behind it, and the factors that affect the process. It is thought that the place of the private lesson in our education system and its reflections will contribute to the field, draw attention to the private lesson applications and give an idea for the new studies to be done.

Method
The main aim of the research is to analyze the practice and determinants of private tutoring participation in Şanlıurfa, Turkey. This research is a descriptive study designed in the survey model. Descriptive studies focus on a specific characteristic of a population so that some inferences can be made about some characteristics, attitudes, or behavior of this population (Babbie, 1990 as cited in Creswell,p. 137). A survey study provides a quantitative or numeric description of trends, attitudes, and opinions of a population, or tests for associations among variables of a population, by studying a sample of that population. Survey designs help researchers answer three types of questions: (a) descriptive questions; (b) questions about the relationships between variables; or in cases where a survey design is repeated over time in a longitudinal study; (c) questions about predictive relationships between variables over time. From sample results, the researcher generalizes or makes claims about the population (Creswell, 2008, p. 156). For the current study, the survey model was employed because it aims to answer descriptive questions (e.g., how many participants receive PT, why they receive PT, in what form of PT they receive, etc.) and questions about the relationships between variables (e.g., if there is a statistically significant relationship between the gender, age, parents' education level, involvement in education etc. and receiving PT or not, etc.). The detailed information regarding the sampling method, participants, their demographics etc. are given below.

Participant (Subject) Characteristics
The data used in the research are collected from 1329 high school students in Şanlıurfa, Turkey during the spring semester, 2019. Participants were selected among 81216 high school students using two-stage sampling. In the first stage, 5 high schools were selected randomly among 143 high schools in the city. As can be seen in Table 1, the study was conducted with the participation of 1351 high school students. 46.5% (n = 618) of the participants are female; 51.5% (n = 685) of them are male. The percentages are the participants in terms of the grade they attend to are namely 9 graders (30.9%; N: 418), 11 graders (25.4%; N: 343), 10 graders (18.6%; N: 251), and 12 graders (14.7%; N: 199). Most of the participants go to Anatolian High Schools (83.2%; N: 1124), Science high School (16.8%; N: 227). Table 2 presents background information related to the parent's education level, profession, family income level, parents' involvement in their children's education, and their belief regarding the necessity of education.  Table 2 shows the demographics of the participant's parents. As can be seen in the table, in terms of the education level of the parents, the fathers are namely primary school graduates (24.1%; N: 325), high school graduates (22.1%; N: 298), undergraduates/BA (21.7%; N: 293). As for the mothers, most of them are (24.9%; N: 337) are illiterate, or primary school graduates (23.4%; N: 316), secondary school graduates (15.5%; N: 209). The level of income of the majority is medium (60.7%; N: 820). Most of them come from middle-class families. As understood, the family participation in education is quite high (90.7%; N: 1225) and they have a positive attitude towards the necessity of education (94.9%; N: 1282). Table 3 presents background information related to the participants' school selection (voluntary/involuntary), overall satisfaction with their school (how satisfied the students are with their school, what they think about the quality of the instruction, whether the school alone is sufficient for high-stake exams or not, etc.)  Most of the participants think that their school is not sufficient to be successful at high-stake exams alone (47.1%; n: 636). They think that they need extra educational support (44%; n: 595). The percentage of the participants who think their school is not successful at high-stake exams (56.6%; n: 764) is higher than the ones who think it is (42.5%; n: 574).

Sampling Procedures
For the study, the cluster sampling method was used for school selection. Cluster sampling is a method in which the population is divided into clusters and some of these clusters are randomly selected as the sample of the study. As for this study, the schools were clustered based on their type, as Anatolia high school, Science high school. The reason why these school types were chosen is that these schools admit their students based on the scores of high nationwide standardized test (LGS). As for the sampling of the participants, the same technique-cluster sampling-was employed. First, the students were clustered based on the grade they attend, and then random groups were chosen with a simple random sampling technique for data collection.

Data Collection Tool
The data for this study were collected with a questionnaire. The questionnaire consists of 4 main sub-sections. The first subsection is composed of questions about the demographics of the participants (their sex, age, school type, parents' professions, education levels, income level, etc.) The second subsection involves questions concerning the opinions of the participants about their school (their selection of the school, how content they are with the quality of the instruction, being a student at their school, school's sufficiency for the high-stake exams, etc.). The third sub-section involves questions regarding PT (whether they receive PT, the reasons why or why not, etc.) The fourth sub-section has questions who reported receiving one form of PT (how they receive PT, which courses, why they receive PT, who referred them to PT).
The stages of developing a questionnaire are described by Anderson (1990) are as follows: a. defining the problem, b. writing the questions, c. seeking for expert opinion, d. piloting and finalizing the form (as cited in Büyüköztürk, Kılıç Çakmak, Akgün, Karadeniz, & Demirel, 2012). While developing the data collection form-questionnaire-these steps are followed. Following the related literature review, the questions in the form (Yes/No questions, Likert-scale questions, marking questions, etc.) are developed, and the first draft of the form was developed. Having formed the first draft of the data collection tool, to ensure the validity of the form, expert opinions were sought. For that purpose, 3 expertise in the field were consulted and asked to fill in the expert opinion form developed by the researcher. The level of agreement of the experts is expected to be 90−100% at the point where each problem is valid (Büyüköztürk, Kılıç Çakmak, Akgün, Karadeniz, & Demirel, 2012). The agreement rate on the items was found to be 92%. Upon revising and making necessary adjustments, the final form of the questionnaire was formed. The pilot study was conducted with the participation of 73 students at a high school who was not involved in the actual study. Having analyzed the returned questionnaires and the feedback, some small changes were made, and the development of the data collection tool stage was finished, and the form was made ready to apply, the questionnaire was distributed to all (2061) students in the 5 high schools. Since the participation was voluntary, 1373 forms were filled. Inaccurate and incomplete forms were eliminated. The remaining 1330 forms were analyzed employing the analysis of variance, chi-square, and t-tests.

Results
The descriptive and inferential findings based on the analysis are given in this section. As for the descriptive analysis, the number of participants who reported receiving PT, whether they would like to receive PT, the reasons why they wish to receive PT, the reasons why they do not receive PT, the types of PT, the subjects, and the reasons that the participants reported behind receiving were analyzed. Table 4 shows the number of the participants who reported receiving PT.  Table 4 shows the number and percentages of the participants who reported receiving one form of PT in the last educational year. Only 1330 out of 1373 participants stated whether they received PT or not. As can be seen in the table, almost half of the participants (n: 622; 46.8%) reported receiving at least one form of PT. Table 5 presents information concerning the number of participants who wish to receive PT.  Table 5 shows the number and percentages of the participants who wish to receive PT in the following year. Only 916 participants replied to this question. As can be seen in the table, more than half of the participants (n: 636; 69.4%) express their wish to receive PT in the following educational year. Table 6 presents information related to the reasons why participants do not receive PT.  Table 6 shows the number and percentages of the participants who do not receive or wish to receive PT. Only 625 participants replied to this question. As can be seen in the table, almost one-third of the participants (n: 303; 48.5%) stated that the reason why they do not receive PT is because of financial impossibility, whereas another one third (n: 280; 44.8%) think that it is not a necessity and few participants (n: 42; 6.7%) reported not having an academic goal, which can be interpreted as it is considered futile. However, almost one-fourth of the participants marked the other option but since they failed to express what they mean by others, this finding is difficult to interpret.   Table 8 presents information concerning the type of PT practices of the participants who reported receiving PT. The descriptive analysis was conducted on 556 participants who reported having received one form of private tutoring in the last year. As can be seen in Table 4, although 556 participants reported having private tutoring outside the school, only 506 of them marked the form of PT they received. The percentage of the participants who reported receiving PT in form of one-to-one is 12.9% (n: 80); small group is 3.5% (n: 22); foreign language courses is 1.4% (n: 9); online courses is 7.7% (n: 48); other form is 1.3% (n: 8). The vast majority (n = 310, 61.3%) reported having private tutoring at a private teaching institution (Dershane); 18.6% (n = 94) reported having private tutoring at study centers (Etüt Merkezi) and 15.8% (n = 80) reported having one-to-one private tutoring. Small-group tutoring was the least common type of private tutoring (4.3%). Table 8 shows the subjects that the participants reported receiving.  Table 8 presents information related to the subjects that participants got as private tutoring. As can be seen in Table 8, most of the participants reported having received some sort of private tutoring in Math (82.9%; n: 515), in Science (71.3%; n: 443), and in Turkish (50.4%; n: 313). The less favorite subjects were social (21.4%; n: 133), English (10.8%; n: 67) and not specified (5.2%; n: 32). The reasons why the participants resort to PT are given in Table 9 below.  Table 9 shows the reasons why the participants wish to receive PT. As can be seen in the table, 511 participants (28.8%) stated that the reason why they wish to receive PT is to get better grades at school. Likewise, a vast majority (n: 560; 31.6%) stated that it is because of high-stake exams. The other reasons behind wishing to receive PT are as follows: poor classroom instruction (n: 292; 16.5%); not having someone at home to help with schoolwork (n: 181; 10.2%); for homework (n: 86; 4.9%); family pressure (n: 31; 2.1%); to make new friends (n: 31; 1.7%); everybody receives (n: 23; 1.3%) and other (n: 50; 2.8%). These findings show that the participants do not think the education they receive at school is not sufficient to be successful not only at the high-stake exams but also at exams at school alone, which means they need extra support to be academically successful. They also reported that they did not understand well at the school and therefore they needed PT. Table 10 below shows the ones who referred the participants to PT.  Table 10 presents information about the individuals who directed students to receive private tutoring. As can be seen in Table 10, mostly the parents direct their children to receive private tutoring (63.3%; n: 321). An interesting finding is that some teachers (17.8%; n: 90) also refer them to PT, which may be interpreted as some of the teachers also think that education at school needs to be supported. İndividuals such as siblings (9.7%; n: 49), relatives (6.5%; n: 33) except for friends (14.8%; n: 75) do not have a big impact on the decision of the participants receiving PT. To understand what factors may affect receiving PT, the Pearson chi-square test was run. Table 11 presents chi-square test results. As can be seen in Table 11, according to the Pearson chi-square test results, as age and grade, which shows that students are nearing high-stake exams, increase, they are more likely to receive PT (age: χ2 = 357.365, df = 2, p < 0.001; grade: χ2 = 36.224, df = 2, p < 0.001). Whether the parents' education level effect receiving PT, it was determined that as the parents' education level increases, the likelihood of receiving PT increases (father's education level: χ2 = 236.399, df = 6, p < 0.001; mother's education level: χ2 = 147.3755, df = 6, p < 0.001). The finding shows that mothers are more influential in their children's education. Parents' involvement is determined to be statistically significant in the participants' receiving PT (parent involvement: χ2 = 447.696, df = 1, p < 0.001). The more parents believe in the necessity of education, the more likely the students receive PT (χ2 = 566.096, df = 1, p < 0.001). There was also found a relationship between the satisfaction with school and receiving PT (χ2 = 1057.063, df = 2, p < 0.001).
To sum up, according to the Pearson chi-square test results, age, grade, school type, father's and mother's education level, parents' belief in the need for education, satisfaction with the school may affect participation in private tutoring. As age, grade, father's and mother's education level, and parents' involvement, their belief in the need for education, dissatisfaction with the school increase, students are more likely to participate in private tutoring. As satisfaction with the school decreases, students are more likely to participate in private tutoring. Besides, Science high school students are more likely to participate in private tutoring.

Discussion
According to the findings, it was determined that most of the participants express a preference of receiving PT at private teaching institutions and study centers whereas quite a few of the participants reported receiving one-to-one and small-group PT. This may be since one-to-one and small-group tutoring are the most expensive types of PT. Since the Şanlıurfa city has lower income levels compared to the rest of the country, it makes sense that families on a budget send their children to low-cost types of PT. According to Turkish statistical institute reports, in 2019, Istanbul ranked first in GDP per capita with 86 thousand 798 TL. Kocaeli was followed by Istanbul with 81 thousand 228 TL and Ankara with 71 thousand 27 TL. In the GDP per capita calculations at the provincial level, Van with 18 thousand 708 TL, Şanlıurfa with 17 thousand 465 TL, and Ağrı with 16 thousand 727 TL took the last three places (TUİK, 2021).
The most popular subjects of PT were determined to be math, science, and Turkish. These are considered to be the most important subjects affecting central exam scores. Therefore, there is no coincidence in students' favoriting these subjects. The reason behind this could be to get higher exam scores and gain entrance to the best schools. This finding is also in line with the finding that the most important reasons for receiving PT are to prepare for exams, to get better grades at school, and to understand school subjects better. Liu and Bray (2016) state that there is a positive and significant relationship between students' future academic expectations and receiving PT and students who are not satisfied with their academic success are more likely to receive PT. Also, Tan (2017) states that the majority of parents who want to give their children a competitive advantage in exams lead them to PT. In some studies, it is claimed that private tutoring increases students 'self-esteem, interest in learning, and academic performance (Hajar, 2018;Ghosh & Bray, 2018;Ha & Park, 2016). On the other hand, Chan and Mongkolhutthi (2017) states that a competitive education system negatively affects students' social development.
There are several educational, economic, and cultural factors that affect the demand for PT. The high-stake tests are the most widely given reason to receive PT (Azam, 2016;Bray et al., 2014;Kim & Lee, 2010;Stevenson & Baker, 1992;Tansel & Bircan, 2005). Hajar (2018) states that PT is perceived as a powerful tool that will provide an advantage in school entrance exams; increases the interest of students in learning, and help them gain an advantage in the competitive environment that creates pressure on their parents and themselves in addition to receiving efficient education. It is argued that there is a close relationship between the quality of teaching in schools, the type of school, and the demand for PT. The perceived low quality of teaching in schools seems to the driving force behind the extensive demand for PT. It is argued that teachers' inadequacy (low quality) is the main factor behind this educational failure. (Hamid et al., 2017;Akdemir & Kılıç, 2020). However, Bray et al. (2014) didn't find a significant relationship between the quality of teaching in schools and the demand for PT. Compared to other grades, the senior students seem to resort to PT more, which is understandable given the fact they will face a high-stake test soon enough (Dang, 2014). It would be wrong to assume that the extensive demand for PT is a sign of educational failure (Bray, Zhan, Lykins, Wang, & Kwo, 2014), but receiving PT might be an important obstacle to social mobility (Bray, 2011;Buchmann, 2002;Dawson, 2010). According to the findings of several studies (Bray & Kwok, 2003;Zhang & Xie, 2015;Psacharopoulos & Papakonstantinou, 2005), there is a significant positive relationship between the family income and the probability of receiving PT.
The wealthier families spend more on PT and provide their children with better quality tutoring. The higher the family income, the more is spent on PT (Azam, 2016;Bray et al., 2014;Bray & Kwok, 2003;Zhang & Xie, 2015;Dang, 2014;Kang, 2011;Tansel & Bircan, 2006). The richer families seem to get their children to receive more one-to-one type of PT compared to the poorer families who resort to less expensive forms of PT or none. Some studies found a positive significant relationship between the education level of the mothers and demand for PT while the education level of fathers was not found to be effective (Bray et al., 2014). There seems to be a consensus concerning the reasons behind PT. Generally speaking, the reasons such as low quality of mainstream education, poor classroom instruction, teacher inadequacy, etc. may lead to families seeking alternative education which gives rise to the prevalence of PT (Pallegedara & Mottaleb, 2018;Tsiplakides, 2018;Sieverding, Krafft, & Elbadawy, 2019;Akdemir & Kılıç, 2020).
In most situations, parents direct their children to receive private tutoring. Since parents understand the importance of education and high-stake tests better, they want their children to enter prestigious universities and find high-paying jobs upon graduation. Therefore, parents may see PT as a way of achieving these goals. Melese and Abebe (2017) state that parents and friends are the driving force for students to receive PT and that there is widespread demand, especially among students who will take national exams. Accordingly, the main reason why students resort to PT is to improve their academic scores.
According to findings of this study, age, grade, school type, father's and mother's education level, income level, parents' belief in the need for education, satisfaction with the school may affect participation in private tutoring.
As age, grade, father's and mother's education level, income level, and parents' belief in the need for education increase, students are more likely to participate in private tutoring. As satisfaction with the school decreases, students are more likely to participate in private tutoring. Science high school students are more likely to participate in private tutoring.
In some studies, it is investigated if there is a relationship between the students' academic achievement, and receiving PT (Dawson, 2010;Bray et al., 2014). It is generally accepted that self-estimated academic level and grade are two of the important variables in demand for PT. Bray et al. (2014) found that students whose self-estimated academic achievement is low or fair probably receive more PT compared to students who consider themselves as higher academic achievers. Besides, higher achievers could be more active in looking for assistance from their schoolteachers and might have higher self-confidence in learning themselves (Ghosh & Bray, 2018). However, there are no clear findings of the relationship between receiving PT and academic achievement (Liu, 2012;Zhang, 2013;Zimmer, Hamilton, & Christina, 2010). The driving force behind receiving PT might change from person to person and society to society. While some students actively seek tuition, others may be forced by their parents or other individuals such as their friends, relatives, and so on (Barrow & Lochan, 2012).
As for the expenditure on PT, it was found that gender may affect spending on private tutoring. Female students tend to spend more than male students. In several studies, the relationship between gender and demand for PT was analyzed and controversial findings were obtained. In some countries, it is found that male students are supported more than females given that they are expected to look for jobs that necessitate educational qualities (Aslam & Atherton, 2014;Ghosh & Bray, 2018;Nath, 2008). The opposite is also true. Kim and Lee (2010) found that money spent on education for females is higher than that of males. This could be because girls study harder, or they strive for academic achievement more strongly than boys. However, the subjects of PT could change according to gender. For example, Zhang (2013) found that more female students receive tutoring in English and Chinese, while male students prefer Math and Science. To sum up, it can be concluded that there is no significant gender difference that determines the demand for PT (Dang, 2007). The difference in findings could be because of the cultural values, social gender expectations, and different approaches to the education of the genders (Bray et al., 2014).

Conclusions
It is interesting to see that the determinants of participation in PT (age, grade, school type, father's and mother's education level, income level, parents' belief in the need for education, satisfaction with the school) do not determine the spending on PT. This may be because most of the students participate in private tutoring institutions that compete with each other. Therefore, the prices of these institutions cannot differentiate much.
Since most of the parents of the students who participated in this research have middle-level income, their spending preferences may be similar.
As it was stated in the report of TUIK (2021), although Şanlıurfa has a low socioeconomic status compared to other cities in Turkey, almost half of the students who participated in this research reported that they had received PT in the last year. This is an important result that reveals the common trend in Turkey. This may be a sign that the education system in Turkey does not give students what they need. So, the students seek help from PT as a shadow education system and spend their parents' limited income. This is a side-effect of high-stake testing that is seen in other countries. To overcome this problem, Turkey should decrease the use of high-stake test scores in school registration processes. Also, an education system that is more aligned with student needs should be established. However, further research is needed to prove these assertions.
This study also concludes that the new PT policy of the Ministry of National Education seems to be ineffective. Students are still eager to attend private teaching institutions that are forced to exist under different names. This may be due to the need for PT to excel at high-stake exams. This implies that adopting such policies to prevent PT without providing a quality education tailored to the needs of students is destined to fail. The Ministry should eliminate the demand for PT, not the PT itself. This may be achieved by embracing the recommendations given previously.

Recommendations
Based on the findings of the study, given that almost half of the students who participated in this study reported receiving PT in the last year, the schools alone seem not to provide good quality education for the success at high-stake exams as well as school exams. So, the students seek help from PT as a shadow education system and spend their parents' limited income. This is a side-effect of high-stake testing that is seen in other countries. To overcome this problem, in Turkey, it may be considered to decrease the use of high-stake test scores in school registration processes. Also, the education system needs reforming to be more aligned with student needs should be established. However, further research is needed to prove these assertions. Besides, the new PT policy of the Ministry of National Education seems not to be effective. Students are still attending private teaching institutions that are forced to exist under different names. This may be due to the need for PT to excel at high-stake exams. This could be interpreted as that adopting such policies to prevent PT without providing a quality education tailored to the needs of students is likely to fail. Therefore, the Ministry of Education might consider eliminating the demand for PT, not the PT itself. This could be achieved by embracing the recommendations given previously.

Limitations
Since this research has limitations, some caution is needed to generalize the results generated from the population. First, weightings by gender, grade, and school type were not calculated. Second, data is based on students' self-reported perceptions. Last, this is a quantitative research that describes the overall practice, but not the underlying reasons. Qualitative research may reveal deep meanings of the reasons for the practice presented in this research. Also learning about what happens in PT classes and gathering the views of parents, teachers, tutors, school administrators, and other stakeholders would benefit all parties.