Chinese students and personal tutorials in a British overseas campus: The strategic choices of emerging adults

Personal tutorials are an essential feature of student support in British universities, and therefore they are duplicated on British overseas campuses. It appears that Chinese students are reluctant to seek help when they experience personal difficulties that affect their engagement with learning and their academic performance. Limited literature explores this phenomenon with relevant studies only focusing on Chinese students’ experiences abroad. Furthermore, these studies mainly refer to cultural factors related to traditional Confucianism to explain why these students do not engage with support structures. Drawing on the theory of Emerging Adulthood, this paper analyses the experiences of students on a Chinese branch campus of a British university. A mixed methods research approach was considered the most appropriate means of engaging with the participants; a quantitative study was used in an exploratory fashion to provide unbiased insight into student opinion and experience, and a qualitative content analysis was used to analyse participants’ comments in the open field questions. The findings reveal an alternative portrayal of the “Chinese Personal Tutee”, distant from the traditional Confucian model still predominantly used as an analytical tool in research on Chinese youth. Chinese students on an international campus strategically select their sources of help and prefer to build symmetrical relationships with personal tutors based on personal goals rather than asymmetrical relationships based on ‘care’ provided by adults. Consequently, ‘transnational’ personal tutorial systems pursuing Chinese students’ successful engagement ought to be conceptualized by considering their emerging adulthoods and by respecting their sense of agency.


Introduction
One feature of the British higher educational system is the personal tutorial system (PTS), which gives students the opportunity to establish a relationship with an academic who will advise them on matters related to their academic and future careers (Earwaker 1992;Ghenghesh 2017;Stork and Walker 2015;Tryfona et al. 2013;Walsh et al. 2009). In this respect, the fundamental objective of personal tutorials is to facilitate students' adjustment to the academic culture by clarifying university expectations whilst providing personal support (Thomas 2002(Thomas , 2006. Research conducted in the UK discusses the challenges for personal tutors (Myers 2008;Ridley 2006;Simpson 2006), the unmet expectations of students (Hixenbaugh et al. 2006), and the relevance of preferences and perceptions on the effectiveness of the PTS (Cahill et al. 2014). Stephen et al. (2008) show that both 1 3 students and staff share a feeling of dissatisfaction in a PTS when their expectations are not negotiated and do not match.
The internationalization of Higher Education led to students studying abroad, enrolling in transnational universities or studying foreign programmes in their own country (Knight 2004;British Council 2013;HE Global 2016); for Chinese students, this yielded experiences in unfamiliar learning environments (Zhang and Brunton 2007;Zhu 2016), which included unknown expectations regarding their studies and learning. Internationalized institutions desire to offer support to Chinese students to help them adjust to unfamiliar education systems and improve the satisfaction of their students, and hence research in this area is necessary.
This research was conducted on the Chinese campus of a British university, which offers the same curriculum provided on the British campus and includes a PTS, delivered by academic staff. The research aimed at understanding the reasons of the dissatisfaction in both personal tutors and students, who appeared to disagree on the nature, purpose and organization of the PTS. In this paper, we will investigate the insights provided by the students. Our initial hypothesis to explain students' differing attitudes was their traditional collectivist culture as opposed to the more individualist British culture (Triandis 1995). We supposed that Chinese students did not engage with academics because the students' role is traditionally constructed in terms of authority, self-discipline and distance (Hui 2005;Zhang et al. 2005). Instead, we shall argue that students' use of the PTS is strategic as related to the emerging of their individual autonomy and the deliberated use of formal and informal support networks available to them. Students prefer a personal tutorial model that allows them to explore their aspirations and identity (Wootton 2013). Therefore, our investigation placed in the context of the theory of 'Emerging Adulthood' (Arnett 2000(Arnett , 2001(Arnett , 2004(Arnett , 2007(Arnett , 2014 and the related empirical research conducted on Chinese university students (Jorgensen et al. 2017;Nelson et al. 2004;Nelson and Chen 2007) challenges the traditional portrayal of the 'Chinese Personal Tutee' and can be used by universities to create efficacious support systems that resonate with Chinese students' strengths.

Emerging Adulthood and contemporary Chinese university students
To understand university students' help seeking preferences, it is useful to consider their psychological development. In 2000, Jeffrey J. Arnett proposed the theory of Emerging Adulthood, which identifies specific features of the identity development of individuals between 18 and 25 (later extended to 29). This time of life course is culturally constructed, not universal, as it specifically regards individuals who live in industrialized countries, are part of the middle or upper classes (Arnett 2016) and study at university. In fact, during their university years these young people live in a condition of semi-independence. They do not live with their families, but they continue to rely on their families' financial support. Despite their 'daily independence', university students do not define themselves as adults because they do not feel able to make major life decisions, to take responsibility of themselves, nor to support themselves financially. On the other hand, they feel that this stage of their life is a period of personal growth and exploration in intimate relationships, work and worldviews (Arnett 2000(Arnett , 2001(Arnett , 2004. Studies conducted on Chinese students in China (Nelson et al. 2004;Nelson and Chen 2007) or comparing Chinese and American students in the US (Badger et al. 2006) show that the theory of Emerging Adulthood can be used to understand Chinese university students' identity construction, with some limitations and changes due to the different cultural and social backgrounds. Financial independence, the ability to make life decisions and to take responsibilities are important criteria of adulthood for both Chinese and US students; however, Nelson et al. (2004) note that Chinese students add the ability to control their own emotions, a greater sense of consideration for others and the ability to support their parents financially. These criteria are related to Chinese traditional cultural values emphasizing the responsibility towards the group and the family (Markus and Kitayama 1991;Triandis 1995). Nelson and Chen (2007), on the other hand, note that the rapid social and economic change has created a 'market-oriented society' whose growing focus on the individual's personal growth, independence of opinion and self-confidence may integrate with the values of the Chinese collectivist culture in unexpected ways. For example, filial piety and loyalty to the group combined with the new focus on personal development and fulfilment may result in greater pressure on students to achieve (in economic and social terms) to meet their family and group expectations. Therefore, Chinese university students may experience different types of instability and stress. Since personal tutorials are envisaged to support students' academic and personal growth, it seems important to also consider in which situations and how they would ask for help when difficulties arise.

University support and Chinese students in foreign environments
Research on Chinese students has focussed on their learning habits and preferences, analysed their studying strategies, and debunked some 'myths' about their passivity and lack of learning autonomy (Gao 2006;Gu and Schweisfurth 2006;Jin and Cortazzi 2006;Biggs 1996, 2001;Wu 2015). However, the literature on their habits and preferences on support seeking is sparse. Given the limited number of British campuses abroad, and in particular within China, the lack of specific literature on the support offered and used by students is understandable. However, there are also limitations in the literature available on support used by Chinese students when studying abroad. The main limitation is that the existing studies mainly focus on cultural and/or academic adaptation and support received. However, they may give us some useful insights. Chinese students studying abroad prefer not to ask for help when experiencing difficulties and they attribute this to cultural reasons: they are taught not to disclose personal issues to individuals external to their family or immediate peer circle. Wei et al. (2005) on Chinese students in the United States illustrates that for academic matters, Chinese students prefer to seek help from their friends when in China or from members of the Chinese community when in the US. In this context, academics are trusted as reliable, but they are not the students' first choice when seeking support (Bertram et al. 2014).
In his research on Chinese and American students coping with academic failure, Mortenson (2006) found that whilst both groups of students shared the same definition of effective social support (based on emotional closeness and focused on problem-solving strategies), Chinese students would seek refuge in emotional avoidance and refrain from asking for help.
These studies highlight cultural dimensions, such as respect for authority and fear of stigma, that hinder Chinese students' engagement with support provided by academics in one-to-one settings during their studies abroad (Yan and Berliner 2011). Conversely, Wei et al. (2007) show how Chinese students abroad do have an ability to make use of their own parallel sources of support and to differentiate between the types of help that they ask for.

Research design
To investigate Chinese students' attitudes towards student support, we created a quantitative study surveying their opinions of the strengths and weaknesses, and expectations of the institution's PTS. The questionnaire was designed based on Saris and Gallhofer's work on questionnaire design (2007). The questionnaire was checked for validity by inviting critiques from six academics with experience in questionnaire design. Based on this feedback, the questionnaire was improved by simplifying the language, changing question types to improve the validity of the results and streamlining the questions. After asking questions regarding respondent demographics, the questionnaire consisted of four parts: (1) five questions asked for students' experiences with the current PTS focusing on the meeting arrangements and their tutor's accessibility; (2) six questions asked students to describe their meeting experiences based on meeting duration, meeting pacing and categories of topic discussion; (3) five questions asked students to describe their relationship with their personal tutor by investigating gender and identifying whom students would initially approach when they have a problem, and finally (4) five questions asked the students for suggestions on how to improve the personal tutorial scheduling and meeting format. A total of 16 nominal questions, 5 multiple nominal, 3 rank order, 2 Likert and 11 open field questions were used to collect the students' opinions and retain their interest during the questionnaire.
After obtaining ethical clearance from the University Ethics Committee, the on-line questionnaire was made available during 2 weeks of April 2015. The survey was created using the commercial platform Qualtrics and distributed to all students over the email system by requesting voluntary participation; no incentives were given. A total of 408 out of 6676 students attempted the questionnaire, which was completed by 291 students. The reduction of responses was considered to be due to normal questionnaire attrition. Table 1 shows the demographics of the respondents. Table 1 indicates an uneven distribution of responses occurring, most notably, within the gender distribution of the respondents with only 28% of the respondents being male. This research did not investigate the motivation behind the respondents' engagement with the questionnaire. After the responses had been collected, the data were analysed using the commercial software IBM SPSS version 21. The nominal data and Likert questions were investigated using categorical principal component analysis (CATPCA), which was used as an unbiased exploratory data analysis tool for the purposes of discovering latent opinions in the data set. To complete the analysis, simple descriptive statistics of the multiple nominal questions were used to support the findings from the CATPCA and a Qualitative Content Analysis was carried out on the responses to the open question fields (Schreir 2012). CATPCA was only conducted on completed questionnaires.

Using categorical principal component analysis (CATPCA)
The first version of a CATPCA was carried out by Guttman (1941) and there have been several contributions to the literature since; the interested reader is directed to the work of Gifi (1990) for a historical overview. CATPCA is different from a linear principal component analysis because it can handle mixed measurement levels (e.g. nominal, ordinal and numerical) and it is able to establish relationships between responses that may not be linearly related to each other. The responses were analysed as nominal variables (unless stated otherwise), which places the fewest constraints on the analysis hence the greatest quantity of variance can be accounted for in the results. A key part of the analysis is to assign a number to each of the nominal responses, which is not assigned arbitrarily and its mathematical calculation is not within the remit of this paper. The procedure of assigning a number to each response is called quantification. For each component, eigenvalues are calculated and directly related to the 'variance accounted for'. CATPCA is different from traditional linear principal component analysis (PCA) at this point. In linear PCA, the varianceaccounted-for is maximized over each component in a consecutive manner. That is to say, the maximum variance is found for component 1, then the maximum variance of the remaining responses is found for component 2 and this process is continued until the maximum variance is accounted for over all of the desired components. In CATPCA, the maximum variance is simultaneously found for a number of pre-specified components. Consequently, changing the number of pre-specified components alters the CATPCA scree plot and may change the number of components necessary to capture the majority of the variance in the data.
In this investigation, the number of components was chosen by inspecting the scree plot and the variance captured in the results for different numbers of components. Several different preliminary investigations were conducted on a varying number of components by both simultaneously inspecting the variance-accounted-for and the component loadings. With a higher number of components, it was seen that the number of questions loading onto the higher components significantly reduced. Whilst 6 components had eigenvalues greater than 1.0, the 6th component had two item loadings, little meaning and was consequently disregarded. A table of the components and the variance they describe within the results is presented in Table 2.
Once each nominal response has been quantified, CAT-PCA is similar to standard linear principal component analysis in that correlations between responses are discovered. The CATPCA process determines how many variables (called components) to look for. Each of the questions 'loads' on to a component and the strongest component loadings can be used together with the category quantifications to identify each component. The component loadings represent Pearson correlations between the quantified responses and the component, and the components can only be understood by simultaneous inspection of both the quantifications and the loadings. As shown in Table 2, five different components were derived and the meaning of each component will be described where appropriate. These components highlighted two correlations in student opinions on the strengths and weaknesses of the PTS and three correlations on student expectations of the PTS.

CATPCA analysis
A categorical principal component analysis of the following 16 nominal questions was performed. The components were evaluated post hoc to see if they described a strength-andweakness or an expectation of the PTS.
(ST1) What year of study are you in? (ST2) Are you male or female? (ST3) Are you a domestic or international student? (ST4) What school are you in? (ST5) What was your average mark last semester? (ST6) How many scheduled Personal Tutorials have you attended? (ST7) What type of meeting do you prefer? (ST8) Have you had additional meetings with your personal tutor outside of those scheduled by the university? (ST9) What language do you use to speak to your personal tutor? (ST10) How long did your last personal tutorial last? (ST11) During the meeting(s), did you feel that you had enough time for discussion with your personal tutor? (ST12) Do you have a personal tutor of your gender? (ST13) How many personal tutorials would you like to have per semester? (ST14) When would you like to have your first meeting with your personal tutor? (ST15) Should Personal Tutorials be compulsory? (ST16) Should students be allowed to request to change their personal tutor?
It was found that the responses to ST11 and ST15 did not correlate with any of the other responses and hence these were removed from the CATPCA. A CATPCA was performed for the remaining 14 questions and it was found that 5 components explained 51.5% of the variance in the answers ( Table 2). The reliability analysis gave a Cronbach's Alpha of 0.928 which indicates excellent internal consistency. The five component loadings are given in Table 3.
The bold and italicized fonts indicate which questions were used to define each respective component. Numbers in italics indicate that the quantifications correlate positively with the component, whilst numbers in bold indicate that the quantifications correlate negatively with the component. The factor loadings together with the category quantifications allowed for the definition of each component and these will be explained in the following sections.

Component 1: Student engagement
The quantifications for the responses that were positively correlated with Component 1 are presented in Table A1 of the Appendix and the corresponding quantifications for the responses that were negatively correlated with Component 1 are presented in Table A2. Furthermore, a histogram of the Component 1 scores is found in Appendix Fig. A1. The histogram has a skewness of 0.225 and an excess kurtosis of 0.296 revealing that the data are mildly skewed with a longer tail on the positive side and moderately centred on the mode. Component 1 identified a correlation between student academic achievement and their attendance in personal tutorials for two particular categories of students: international or domestic students in their later academic years and domestic students in their early academic years. The histogram (Appendix, Fig. A1) shows that domestic students who were both in their earlier years and attended tutorials achieved well. Conversely, both domestic and international students who were in the later years of study and did not engage with the PTS did not achieve well. The CATPCA analysis cannot determine the causative effect; it cannot be determined if the students did poorly in their exams during their early years and hence they chose not to attend their personal tutorials or if not attending their personal tutorials caused them to achieve poorly in their later years.
Using a multiple nominal response question, the students were asked for their discussion topics during the tutorials. The most common responses were improving my academic performance (67%), issues relating to exams (57%), improving my study skills (51%) and gaining further academic support from the university (28%).
The university has allowed group meetings to take place to strengthen students' cross-year peer support and information sharing about university procedures. However, students were particularly critical towards this practice. Some felt that this practice was simply used to save time, whilst others indicated that this setting was not conducive to discussions about crucial topics: I think I am not comfortable because I am afraid to talk [about] some personal issues, such as grades or This sentiment was also expressed in the responses to nominal questions where 82.5% of the responses indicated a preference for individual meetings as opposed to group meetings.
When the students were asked why they felt uncomfortable in a one-to-one tutorial meeting, the most selected responses indicated that students felt discomfort when they were unable to express themselves clearly in a foreign language (31%), when they thought that the tutor could not help them (26%), when they felt that their problem should not be a matter for discussion in a personal tutorial (19%) or they felt embarrassed about their poor academic performance (18%).
When giving their opinion of their personal tutor, 71.6% of students felt that their personal tutor was approachable, helpful and knowledgeable, they liked talking with their personal tutor and felt that they were cared for.
We talked about recent life, the study and the future. My personal tutor always gave me good advice and encouraged me.
However, there was a smaller body of students who felt that their personal tutor was not approachable or helpful, they did not know if their personal tutor was knowledgeable; consequently, they did not like talking to their personal tutor and did not feel that their tutor cared about them.
Some students complained when they were not invited to personal tutorials, or that their tutor did not respond to their meeting proposals. Other students complained about the behaviour of their personal tutors during the meetings such as speaking English without making sure that they were being understood, rushing the meeting, not listening and only providing generic advice. These are perceived as signs of a tutor's superficiality, lack of care and commitment, and students elaborated on these feelings in their responses to the open questions: Please do not only follow the question sheets, which makes us feel the answer related to the question is more important than ourselves.
He kept talking that I have no chance to discuss with him. It seems that he wants to finish the meeting quickly.
Very arrogant and never repl[ies] your emails! Students feel these members of staff do not embrace their responsibility and this is why many asked for an evaluation system of the PTS and the possibility of changing personal tutor. For example, students commented that: It can be better if the quality of [the] personal tutor [ial] can be ensured.
[The] Personal tutor should be trained to some extent and have some basic knowledge on how to help students.
I think that we should have system of student evaluation of [the] personal tutor each year.
I strongly advise there to be a feedback system for we students to make complaints about the irresponsible tutor.
Some students felt that meetings were not effective and suggested they could be improved if the personal tutor sent them an 'agenda' ahead of time. Another student suggested that the meetings would be more effective if tutees sent tutors a list of their questions prior the meeting so that tutors could gather the necessary information to answer the students' questions, as articulated clearly by a student in their final year: For my experience, I have prepared for some questions of academy and internship. So, there was lots to discuss. So I needed more time. In addition, my PT of year two is not as proactive as the one [from] year three. So, when my year 3 PT communicated with me, he/she shared lots of information for helping, especially providing empirical experience of other students which I think is extremely helpful, rather than just provide general advice.
In conclusion, it can be seen that there is a correlation between personal tutorial attendance and academic achievement; however, it is not possible to determine a causal effect. When students attend, the conversation tends to focus on academic matters, either current issues or future plans. Whilst it cannot be determined who chooses the meeting topics, students will approach their tutors when they need help on academic matters, and they are very critical if personal tutors are not able to address their specific issues. Students will not engage with the system when they feel discomfort or if they do not a have functional relationship with their tutor. For this reason, they suggest an evaluation system that could help personal tutors to become more effective and improve their 'supporting' skills.

Component 2: Gender difference
The quantifications for the responses that were positively correlated with Component 2 are presented in Table A3 of the Appendix and a histogram of the Component 2 scores is found in Appendix Fig. A2. The histogram has a skewness of −0.844 and an excess kurtosis of 0.513 revealing that the data are moderately skewed with a longer tail on the negative side and moderately centred on the mode.
Component 2 of the CATPCA identified a gender (im)balance existing between staff and students in some schools and departments. A positive component 2 score indicated that the student was in the later part of their studies, they were female (72%) and that they did not have a personal tutor of their own gender (55%). A histogram of the component scores indicated that this situation predominantly exists in the Schools of Business and International Communications; however, the responses to a separate Likert scale question indicated that the students did not consider this an issue.

Component 3: Meeting format and organization
The quantifications for the responses that were positively correlated with Component 3 are presented in Table A4 of the Appendix and a histogram of the Component 3 scores is found in Appendix Fig. A3. The histogram has a skewness of −0.126 and an excess kurtosis of −0.077 revealing that the data are mildly skewed with a longer tail on the negative side and not unusually centred on the mode.
Component 3 identified a correlation between students' engagement within the PTS and their preferences in terms of meeting format and timetabling. There was a group of students who appeared to strongly engage with the PTS. They preferred one-to-one meetings, they sought their personal tutor's advice when they needed it and they preferred to have their first tutorial earlier in the semester. On the other hand, there was another group of students who did not mind which meeting type they had, never sought their tutor's advice outside of prearranged tutorials and preferred to have their first meeting later in the semester. A histogram of responses (Appendix, Fig. A3) indicated that different variations of this opinion were distributed throughout the student population and was not dependent upon the students' year or programme of study; however, students achieving well seemed to display more apathy towards the PTS.
A rank order question was used to find out how students would like to interact with their personal tutor. It was found that 45% of students chose 'scheduled meetings', 30% chose 'email' and 21% of students selected 'appointment' as their primary choice; the students were not in favour of 'drop by meetings'. Furthermore, in a separate multiple nominal question, students indicated that they approached their personal tutor when they needed advice about their academic performance (58%), their study skills (54%), issues related to exams (43%) and future employment (36%). This indicates that students like to hold formal meetings, either scheduled by the personal tutor or arranged in advance, and will approach their personal tutor when they feel that they need support on academic matters either related to their current experience or to future plans.

Component 4: Meeting efficacy
The quantifications for the responses that were positively correlated with Component 4 are presented in Table A4 and the quantifications for the responses that were negatively correlated with Component 4 are presented in Table A5 of the Appendix. A histogram of the Component 4 scores is found in Appendix Fig. A4. The histogram has a skewness of −0.416 and an excess kurtosis of 1.107 revealing that the data are moderately skewed with a longer tail on the negative side and highly centred on the mode.
The fourth component identified an inverse correlation between student attendance in personal tutorials and their academic achievement. Positive loadings indicated that respondents had low achievement, they may have had additional meetings with their personal tutors outside of the prescheduled meetings and had longer than average meetings. Conversely, negative loadings indicated that respondents had shorter meetings and did not seek additional meetings with their personal tutor; however, they were performing well in their summative assessments.
The histogram of component scores (Appendix, Fig. A4) indicates that there are students who have longer tutorials and seek extra help when they feel it is necessary, but achieve poorly in their exams. On the other hand, there are students who have shorter personal tutorials and do not seek extra personal tutorials, but still achieve well. 40% of the students did not fit into either of these extremes.
A Likert scale question was posed to examine reasons why students might approach their personal tutor. The most common responses were related to improving their academic performances (66.9%) and issues concerning graduate employment (26.9%). Personal issues, issues with specific tutors or with university requirements and procedures were not common reasons for the students to approach their personal tutor.
A small group of students indicated that they felt helped when talking about postgraduate study, internships and future employment, but displayed little motivation to discuss academic concerns.
Language difficulties were also acknowledged by students, but their views about the impact on their communication were more nuanced. The following quote seems to suggest that perhaps the problem is not in the use of English for discussing opinions and personal feelings, but rather a general language limitation.
Language is a problem [for me] to explain my ideas clearly, but I think even if we communicate in my mother tongue this problem would remain. The prime reason for this, I think, is that I don't expect a longterm relationship with my tutor and I resolve problems by myself.
Rank order questions were used to explore the issue of student confidants on both academic and personal issues. It was found that when students have a personal problem affecting their studies, they prefer to discuss it with their friends (68.5%), senior students (3.2%) or course mates (3.2%). However, when students do decide to approach a tutor, they would prefer to approach the lecturer or seminar tutor who is teaching a relevant course, rather than their personal tutor. Finally, it was found that some students would prefer to approach their parents (21.7%) before approaching their personal tutors. The issue of the problem being personal or academic in nature had little influence on this result.

Component 5: Meeting preferences
The quantifications for the responses that were positively correlated with Component 5 are presented in Table A7 and the quantifications for the responses that were negatively correlated with Component 5 are presented in Table A8 of the Appendix. A histogram of the Component 5 scores is found in Appendix Fig. A5. The histogram has a skewness of −0.818 and an excess kurtosis of 1.202 revealing that the data are strongly skewed with a longer tail on the negative side and highly centred on the mode. Component 5 identified student preferences regarding the personal tutorial format and the option of changing their personal tutor. The histogram of component scores indicates that the students want to have both individual meetings (81%) and the option to change their tutor (93%), but they believe that it is more important to be able to change their tutor than to be able to change their meeting type.
An analysis of the open question responses revealed that students focus on becoming familiar with their tutor. 'Familiarity' is the condition for them to start a dialogue and also to listen to their tutor's suggestions. Hence, students place importance on keeping the same tutor for their whole academic career once the relationship is established; for example: In most cases, if I am familiar with the tutor I feel comfortable [enough] to share my problems. However, if the tutor is one who I totally don't know, I will feel embarrassed and doubt whether he or she will be helpful or not.
Better not change the personal tutor for students in four years because they have been familiar [with] each other, furthermore, [to get to] know a new tutor might need time.
Familiarity for students can be established only through frequent and meaningful contact, which is characterized by the students as the tutors' kindness, patience and approachability; their capability for providing sound information and to offer individualized (not generic) advice.
The friendship with my personal tutor is a great treasure to me. Having compared with the experiences of my friends and their personal tutors, I feel that I have the best personal tutor, because my personal tutor is approachable, willing to help, and never misses any meeting schedule.
However, the student comments revealed some dissatisfaction with the fixed time allocated by the university for the tutorials and this affected the quality of the relationship with their personal tutor.
In order to build and foster 'familiarity', the survey results show that the majority of respondents suggested holding two personal tutorials per semester (36%) and comments in the open question fields suggested increasing the length of the meetings from 15 to 20 min.

Discussion
The results of our research conducted with Chinese students studying in a British campus abroad show that Chinese students have similar expectations and opinions regarding the PTS as the British students (Cahill et al. 2014;Owen 2002;Stephen et al. 2008). These findings rule out the existence of fundamental cultural differences in Chinese and British students' conceptualisations of the PTS. Results of Component 1 (Student engagement) and Component 4 (Meeting efficacy) show that our students favour a mentoring/educative model, whereby personal tutees look for support to help improve academic achievement and to make informed choices about their future careers. Tutee and tutors, like mentors and mentees, are in a symmetrical relationship where the tutor encourages the student to reflect on personal goals, experience and options so they can decide their best course of action (Wootton 2013). In this situation, students start taking responsibility for their own support network, exploring their usefulness in assisting their goals. Hence, they leave adolescence and begin to emerge as adults by learning to take responsibility of their academic experience, to organize their time and obligations, and ultimately growing both on a personal and professional level (Arnett 2001(Arnett , 2004.

Engagement
Results in Component 1 (Student engagement) and Component 3 (Meeting format and organization) show that students who engage with the PTS try to build a relationship with their tutor through individual, frequent and meaningful interactions and once the relationship is established, students do not wish to abandon it. Components 5 and 1 (Meeting efficacy and Student engagement) show that this relationship is important for the students; however, the focus is to improve their academic performance, not their personal growth and development. Therefore, students discuss personal issues only when they impact on their academic performance. This choice may be based upon cultural norms, emphasizing emotional self-control, as they are integrated into the definition of adulthood of Chinese emerging adults (Nelson et al. 2004;Yan and Berliner 2011). Given this premise and when tutors appear ineffective in the students' eyes, the findings of component 1 (Student engagement) show that tutees will disengage from the relationship. Furthermore, component 4 (Meeting efficacy) reveals that students articulate the reasons of their focus on academic matters in the open question fields; they indicate that if they do not have a close relationship with their tutor, they do not feel that they can discuss any personal issues, but they also appear conflicted because they do not want to develop a relationship that they know will not last. Hence, students are readily willing to abandon non-functional relationships. If disengaged tutees decide to attend their tutorials, they are likely to ask for post-course advice or reference letters.

Self-disclosure preferences and autonomy
Despite the good relationship students may have built with their personal tutor, the results of Component 4 (Meeting efficacy) show that when they have an academic problem, they prefer to discuss it with friends, (senior) course mates and family, and only approach their personal tutors when other sources of support are not useful or appropriate. Similarly, Cahill et al. (2014) showed that students prefer to talk firstly with friends and only lastly with their personal tutors. Not only is this a matter of convenience, but this behaviour seems understandable because self-disclosure is a complex relational process influenced by the nature and the variety of topics shared between individuals over a period of time (Greene et al. 2006). What is certainly missing in the comments provided by our students is the difference between personal issues in general and personal issues affecting the students' academic performance. In any case, the choice of sharing their difficulties firstly with friends and members of their family also indicates that the students are trying to solve their problems by themselves, by drawing on their own personal resources and strengths, rather than turning to teachers. Tao et al. (2000) found that this choice is positively related to an ability to cope with the stress related to the new university life and is common both for Chinese and Western students. It is also a typical trait of emergent adults' habits to seek support from peers and friends for social-emotional adjustment to university life (Arnett 2001;Hopmeyer and Medovoy 2017). The fact that some of the respondents are achieving well demonstrates that they are capable of selecting sources of help and solving their own problems.

Expectations and opinions of personal tutors
As shown by Stephen et al. (2008), students have precise opinions and expectations of personal tutors. Both Hixenbaugh et al. (2006) and Cahill et al. (2014) describe that when tutors respond to emails, are flexible in arranging meetings and are friendly, students are willing to engage in the tutorial relationship. Evidence of this is shown also in Component 1 (Student engagement) and Component 5 (Meeting preferences), which confirms the conclusions of Simpson (2006) that tutees will only discuss problems with their tutor if a meaningful relationship has been established. In the open fields, our respondents indicated the need to feel cared for and the need to develop 'familiarity' with a personal tutor who should be knowledgeable, available, approachable and a good listener. When students feel cared for and that their personal tutors are friendly and committed, they are happy with their tutors. Conversely, students feel uncared for when their tutor was unhelpful, unapproachable or did not answer emails. In these cases, students became particularly critical of their tutor and completely disengaged from the PTS. The personal tutor's availability and supporting skills are factors that determine students' engagement with the personal tutorials when they feel the need to talk with a personal tutor rather than with a member of their family or a peer, thus providing further examples of strategic volition (as shown by results of Component 4, Meeting efficacy). Finally, Component 5 (Meeting preferences) shows that given the choice, students would prefer to change their tutor rather than change their meeting type, further confirming that it is the relationship with their personal tutor that is most important to them. However, students would like the option to both specify their meeting type and change their tutor if they find that the relationship is not working for them. Staff-student gender imbalances do not appear to concern the students and it is likely that any cultural and gender issues that may arise are easily mitigated by giving the students the option of changing tutor.

Conclusion
This study analysed Chinese students' experiences and opinions of personal tutorials as they are provided on a British campus in China. The aim of this research was to investigate the factors that explain students' apparent lack of engagement in personal tutorials. The existing literature yielded a hypothesis suggesting a link between disengagement and cultural factors such as traditionally strong family ties and the power divide between students and academics. However, our study rejected this hypothesis because our Chinese students' decision making is revealed to be highly strategic. The theory of Emerging Adulthood, which acknowledges that university students are not yet fully adults, but individuals exploring their identities, and learning to take decisions, offered a useful analytical framework to understand our respondents' opinions and preferences. These contrast with the current literature on Chinese students' use of support systems as greatly influenced by traditional cultural factors limiting their willingness to ask for support (McDonald 2014;Yan and Berliner 2011).
In this study, we conclude that the two major factors influencing Chinese students' engagement with personal tutorials are (1) the perceived tutor's competence and their commitment to the relationship; (2) personal tutorials are only one of the sources of help among the many available to students.
Our participants' responses showed that they are selective and strategic in their relationship with personal tutors and disengage from personal tutorials when tutors are not competent or committed. Our respondents are very explicit on the kind of relationship they expect to establish with their personal tutors. They do not like to talk about personal matters with tutors, and they do not like the traditional teacher-pupil relationship whereby the adult provides guidance on life matters. A closer analysis of the topics students like to discuss with their tutors showed that they prefer to focus on academic matters such as study skills, exam preparation, university academic support and further academic career opportunities. These findings indicate that our respondents prefer a mentoring/coaching model (Wootton 2013) of personal tutoring.
Students also showed very clear expectations towards personal tutors' attitudes and behaviours. They engage with tutors who are knowledgeable, competent, approachable and interested in the students' concerns. Conversely, students are critical of tutors who do not pay enough attention or are dismissive of them, even asking to create a feedback system on the personal tutors and to be allowed to change them if their relationship does not work. Our respondents engage with personal tutorials, ask for longer and more frequent meetings, and appreciate tutors' engagement. Emerging adulthood includes one's desire to explore life choices as well as the ability to analyse and evaluate their suitability. Our respondents are demonstrating an ability to critique their situation and implement change if it does not help them.
Finally, the personal tutorial is just one among many possible formal and informal resources at students' disposal and they demonstrate agency when they select it for specific purposes. Our respondents indicate that they have different sources of help at their disposal and are confident using them. Similar to other studies in the UK (Cahill et al. 2014) or in the US (Mortenson 2006), they prefer to talk firstly with peers and family members and secondly with their personal tutors or academics provided they have previously developed a meaningful relationship with them.
This study has two main limitations. Firstly, the sample is limited to only one international campus and its PTS. It could be interesting to compare students' use of PTSs on different campuses where they may have a different structure, involve different types of staff or focus on different topics. Following this direction, it will be possible to understand the PTS within the wider context of student learning experiences in transnational universities and programmes in China and in the different Asian countries. The second limitation is that it was not possible to discuss student experiences in one-to-one interviews or focus groups to gain a more indepth understanding of their experience of the PTS. However, this research shows that a transnational PTS in China should be aware of the students' emerging adulthood, as a stage of psychological development towards independence and strategic exploration. Therefore, a tertiary PTS should appeal to the students' developing autonomy and their preference for strategically choosing between existing sources of informal and formal support. Based on the understanding that tertiary-level Chinese students are emerging adults, further research should explore a conceptual framework that would encourage them to engage with their internationalized institution's personal tutorial system of their own volition.