Exploring Tutors’ Roles in Supporting Student Mental Health: Expectations and Perceptions in Higher Education

Abstract

Student mental health is a challenging topic in higher education, with institutions aiming to develop policies and practices to support students’ personal and professional development. This study examined students’ expectations of their tutors regarding mental health support. An adaptation of the Mentor-Q questionnaire, focusing on tutor role perceptions, was used to measure student expectations. Confirmatory factor analysis showed a good fit of the adapted instrument. The questionnaire was completed by 869 students at a Dutch university of applied sciences, and the results were discussed with eight students. The findings revealed four role expectations of tutors: awareness raisers, connectors, referrers, and guardians. Latent class analysis revealed six clusters, with connector and referrer roles as well as awareness raiser and guardian roles being almost equally important within the clusters. The main difference between the clusters was the height of expectations. Satisfaction with the tutor was significantly related to cluster membership, such that higher expectations correlated with lower satisfaction. Additionally, students who discussed their mental health with their tutors were more represented in clusters with relatively low expectations. Discussions with the students highlighted the nuanced landscape of expectations and the importance of contextual factors and metacommunications. Implications for policy and practical development are also discussed.

1. Introduction

Addressing student mental health is a current challenge in higher education [1]. A study conducted in 19 colleges across eight countries showed that roughly one-third of first-year students who participated in a self-report survey screened positive for at least one common DSM-IV anxiety, mood, or substance disorder (35.3% lifetime, 31.4% 12 months) [2]. In a recent revision of a psychiatric epidemiological cohort study conducted on the general Dutch population, findings revealed that 40% of young adults aged 18–24 years had experienced a mental disorder within the past year; almost 14% reported having had a mood disorder, over 20% an anxiety disorder, and over 17% a substance use disorder [3]. In addition, mental health disorders were more prevalent amongst students than among their working peers. Furthermore, only a small proportion of students actively seek help. It is estimated that only a quarter to approximately one-third of students with mental health problems receive treatment [4].

The learning environment refers to ‘the social, psychological, and pedagogical contexts in which learning occurs and affects student achievement and attitudes’ [5]. This environment can affect the (mental) well-being of students (e.g., Deunk and Korpershoek [6]). Educators are prominent figures in the learning environment who, through their frequent interactions with students, play a role in shaping important facets of students’ educational experience [7]. Although teachers may have the potential to support young people’s mental health as educators, how they should or could be involved is not well defined [8,9]. Aspects of the teacher-student relationship, such as the quality of interactions, are related to students’ well-being gains during their studies, and the quality of this relationship may mediate the link between mental health issues and study progress [10,11] According to students, important characteristics and behaviours for teachers to be of value in well-being and mental health support are being friendly, personable, and supportive, having an empathetic attitude, being competent, having good communication skills, and being available and approachable [11,12,13]. Although these studies provide insight into teacher attributes that are important for student well-being, they do not focus on students’ expectations regarding the role of teachers in supporting their (mental) well-being.

In today’s higher education, the idea of engaging students more actively to improve higher education has gained attention [14,15]. This also accounts for the support for student mental health in general and the teacher’s role in student mental health in particular, given its importance for students as well as its implicit role so far in education. Therefore, in the present exploratory study, the focus is on studying student expectations of the teacher’s role in relation to students’ mental health. Objectives of our study were:

1.

To identify distinct role expectations and role expectation profiles among students at an applied science university using latent class analysis.

2.

To investigate the relationships between students’ role expectation profiles and several background characteristics to further refine the insight into the profiles, and more specifically, to investigate the relationships between role expectation profiles and the level of satisfaction students experience concerning how teachers fulfil their roles.

In this study, a specific focus was on teachers with tutoring roles, as earlier research has shown that, according to students, these teachers are important actors within the university of applied sciences concerning their mental health [16]. Students have a personal tutor during their studies. This person is often a member of the teaching team of their study program. Tutoring provides both academic and personal support. More precisely, at the university of applied sciences where the study takes place, the tutor serves as the primary point of contact for students, providing guidance throughout their studies and coaching students regarding their academic progress, well-being, and any potential issues or challenges they may encounter. Given the focus on tutors, this term will be used to refer to this specific group of teachers throughout the remainder of this paper.

2. Methods

2.1. Research Design and Procedure of Data Collection

A quantitative survey was conducted at a large university of applied sciences in the Netherlands. The university offers approximately 150 study programmes at the associate degree, bachelor’s, and master’s levels, organised into six clusters of study programmes (faculties). More than 24,000 students were enrolled in the university at the time of this study.

Data were collected in October and November 2022. All students except first-year students (N = 17,296) were invited to participate in the online survey; first-year students were excluded because they had been invited to participate in another study conducted during the same period. Students were informed about the study by an e-mail sent to all students in Dutch as well as in English. A reminder was sent after two weeks; the total period of data collection was three weeks. In the invitation, a link directed students to the survey, distributed via Qualtrics. The respondents were briefed on the survey’s objectives, assuring them that their data would be analysed anonymously and ethically. In addition, they were informed that the research plan had received approval from the university’s internal ethics committee. Before they could start the questionnaire, informed consent was obtained from all the students.

2.2. Participants

A total of 869 students completed the questionnaires. An overview of the respondents’ sample characteristics is presented in

Table 1. Sample characteristics and comparison with student population characteristics at the university.

Characteristics		Respondent Sample (n = 869) %	Student Population (N = 17,296) % 1	Chi-Square Result
Gender				χ2 (1, n = 847) = 96.39, p ≤ 0.001
	Male	28	45
	Female	70	55
	Undisclosed	2	-
Age				χ2 (1, n = 869) = 56.80, p ≤ 0.001
	<20 years old	14	10
	20–22 years old	42	39
	23–25 years old	19	25
	26–28 years old	7	11
	29–31 years old	3	4
	≥32 years old	15	11
Study year				(χ2 (1, n = 869) = 68.38, p ≤ 0.001)
	Second year	35	28
	Third year	22	27
	Fourth year	21	22
	>Four year	12	23
Faculty				χ2 (1, n = 862) = 140.45, p ≤ 0.001
	Business	22	22
	Technology	11	17
	Hospitality	11	19
	Education	25	22
	Health and well-being	25	13
	Law	7	6
Level of education				(χ2 (1, n = 869) = 8.29, p = 0.016)
	Associate degree	4	6
	Bachelor	94	91
	Master	2	3
Type of study program				(χ2 (1, n = 869) = 1.34, p = 0.25)
Full time		82	84
Part time		18	16
Nationality				(χ2 (1, n = 869) = 1.50, p = 0.22)
	The Netherlands	86	88
	International (EU)	12	9
	International (outside EU)	2	3
Mental health discussed with tutor 2
	yes	54	-
	no	43	-
	undisclosed	3	-
Satisfaction with tutor (in mental health support) 3,4
	<5	28
	6–7	34
	8–10	38

¹ As first-year students were excluded from the study, they are not included in the numbers of the total student population. ² No numbers of total population available. ³ n = 731, as this question was only for students who had discussed their mental health with their tutor. ⁴ Respondents could answer on a scale of 1–10 (1 = very dissatisfied, 10 = perfectly satisfied).

along with a comparison with the overall student population at the university of applied sciences.

The sample was representative of the total student population for the type of study program (full-time or part-time) and nationality (national or international student), but not for the other characteristics for which the numbers of the total student population were available. Male students were underrepresented in the sample. For age, study year, faculty, and level of education, several categories deviated from the distribution in the overall sample.

2.3. Instrument

2.3.1. Student Expectations in Mental Health Support Questionnaire (Mentor-MAP)

We used the factor structure of the Mental Health Evaluation by Tutors Offering Support and Resources Questionnaire, Mentor-Q, a validated questionnaire to assess the role perceptions of tutors on their role in the mental health issues of students [8]. The Mentor-Q questionnaire consists of statements measuring role perceptions of tutors concerning student mental health with a 4-point Likert scale (1 = does not fit in my perception at all, 2 = does not fit in my perception, 3 = fits in my perception, and 4 = completely fits in my perception). All 34 statements from this measurement instrument were reformulated into statements that measured students’ role expectations of their tutors in student mental health (Mental Health Education Needs Tutors Offering Resources Mapping Student Perceptions, Mentor-MAP). The conversion process was conducted with eight students who provided feedback on the Dutch version of the instrument. Students were asked to respond to the statements using a 4-point Likert scale (1 = does not fit at all with what I expect, 2 = does not fit with what I expect, 3 = fits with what I expect, 4 = fits perfectly with what I expect). A reformulated statement is, for example, ‘I expect that the initiative to talk about mental health lies with me as a student (and not with a study coach)’ in the Mentor-Map was formulated as ‘I believe that the initiative to discuss students’ mental health lies with the students themselves’ in the Mentor-Q questionnaire. The questionnaire furthermore consisted of questions about background characteristics (gender, age, year of study, faculty of study, level of study program: associate degree, master’s, or bachelor’s degree, type of study program: full- or part-time, whether or not mental health has been a topic of conversation with the tutor, and nationality (national or international student). Students were also asked about their satisfaction with the role fulfilment of their tutor concerning their mental health on a scale of 1–10. After the tutor questionnaire (available only in Dutch) was adjusted, it was translated into English for international students.

2.3.2. Member Check

Incorporating different perspectives in the interpretation of results is considered important as it enriches sense-making [17]. Therefore, member checking was conducted by involving research participants to validate the trustworthiness of the results [18]. After completing the questionnaire, the respondents were asked to participate actively in the next stage of the research. In total, 259 students indicated that they were willing to participate, and these students were approached to participate in a focus group session. The researchers set two dates and times, with one session in Dutch as the language of instruction and the other in English. Twelve students signed up to participate in one of the two focus group sessions; four of them dropped out without providing a reason. Thus, eight students participated in the member check. Findings on role expectations and role expectation profiles were presented. The participants were asked about their recognition and possible explanations.

2.4. Data Analysis

To determine whether the factor structure of the Mentor-Q [8] also held for the student questionnaire used here, confirmatory factor analysis (CFA) was conducted. For the confirmatory factor analysis, a diagonally weighted least-squares estimator was used because of the ordinal nature of the 34 indicators. Despite a significant χ²-value (χ² (521) = 1503.626, p < 0.001), which is not uncommon in large samples, other fit indices insensitive to sample size showed excellent fit (Comparative Fit Index = 0.953; Tucker–Lewis index = 0.949; root mean square error of approximation = 0.047; standardised root mean square residual = 0.063), according to widely used cut-offs [19]. A full overview of the CFA is provided in Supplementary Table S1.

Latent cluster analysis (LCA) was used to identify clusters of students with different role-expectation profiles. The four scales of the Mentor-MAP questionnaire were used as indicator variables for LCA. One would expect clusters derived from an adequate model to show large differences in the indicator variables. Therefore, Cohen’s d-values were calculated to determine the standardised mean differences between clusters on the indicator variables, with the expectation that cluster differences would be large.

To answer the second research question, multivariate multinomial logistic regression (MMLR) was used to determine the impact of ‘satisfaction with role fulfilment of tutor’ as a predictor of cluster membership, while correcting for study year, gender, and type of education.

The CFA of the Mentor-MAP was conducted using the lavaan package for r [20]. The LCA was conducted using Latent Gold (version 5.0) [21], and subsequent analyses were performed using SPSS version 27.

The member check was audio-recorded and transcribed by researcher RD. Transcriptions were summarised and coded using Atlas.ti23. First, coding was performed by one researcher (RD) on a line-by-line basis, using an open coding approach to generate initial concepts. These codes were short words or phrases that were more general than the coded text segment itself; however, those that remain close to the original text [22]. These codes were then discussed with one other researcher (JM). Thereafter, related codes were assembled into meaningful categories, based on the outcomes of the quantitative part of the study as the main themes. Finally, a clarity assessment of the categories and themes was conducted by three other researchers (JM, NB, and MP).

3. Results

3.1. Role Expectations

As the CFA confirmed, the factor structure, role perceptions, and labels found in an earlier tutor study [8] can be used to describe student expectations. The role expectations of the students were as follows:

Awareness raiser

Within this role expectation, the tutor is important in supporting the mental health of students. Raising awareness, recognition, and signalling are seen as part of this role. It fits within this role expectation that the tutor takes the initiative in discussing mental health with the student and the teaching team (of course, only with the student’s permission).

Connector

In this role expectation, the tutor is a key person at the university for the mental health support of students. Easy accessibility, availability, interest, and willingness to lend an empathetic ear to students are considered important. A tutor invests in the relationship with the students to ensure that they are comfortable turning to the tutor, if necessary. Initiative for this lies with the student.

Referrer

Within this role expectation, the tutor mainly has the role of referring students to facilities of the university of applied sciences. The referrer’s role is characterised by the importance attached to clear frameworks and expectations from their faculty. In this role, it is important to discuss expectations back and forth with the tutor. Unambiguity in how tutors fulfil their role is considered important.

Guardian

In this role expectation, there are few restrictions on tutors’ time and attention regarding students’ mental health. It is appropriate for tutors to maintain close contact with students, for example, by always being available to them and reacting to them with regard to their mental health outside working hours. Aligning with student expectations seems to be important in this role expectation.

In the member check, the students confirmed the validity of the role expectations. In general, they indicated that the four role descriptions covered the full range of tutor roles, although one student mentioned the importance of the tutor in preventing mental health issues such as stress through clear and timely communication about deadlines. Furthermore, the students shared their interpretations of role expectations and elaborated on their meaning and importance.

Regarding the role of the awareness raiser, it was stated by a number of the students that this role reflects the interest of the tutor in a student. A good tutor–student relationship is crucial for this role expectation; otherwise, this role could be considered undesirable or invasive. Remarks regarding the guardian role were that students made a distinction between ‘what would be nice’ and ‘what I expect’. For most students, it would be nice if tutors acted as guardians, but they would not necessarily expect this role. Other students indicated that this role is beyond the scope of the tutor role. For some students, this role expectation was invasive or seen as a sign that professional help was more appropriate. In any case, this role expectation was considered for specific moments or situations, whereas the awareness raiser and connector roles were considered more structural approaches. Specific remarks concerning the referrer’s role included the centrality of the tutor in the support chain of the institution. However, for some students, this role expectation was associated with disinterest and lack of motivation of the tutor. Regarding the role of connector, it was mentioned that this was the role expectation they recognised in the approaches of most tutors.

3.2. Role Expectation Profiles

An LCA model grouped students into an unobserved (latent) cluster [23] that was predictive of students’ responses to the role expectations of their tutors concerning their mental health. Ten models were run to determine the best solution. The first model contained a single cluster, and each successive model included an additional cluster. The log-likelihood (LL), Bayesian information criterion (BIC), entropy r², and classification errors were used as model fit indices. The selection of the best-fitting model was primarily based on parsimony (i.e., the model with the least number of clusters was preferred). The main information criterion was BIC, followed by entropy r² and bivariate residuals. The five- and six-cluster models had almost equally high BIC scores. Although entropy r² was higher and the classification error slightly lower than those of the six-cluster solution, the six-cluster model was considered the best solution because, for the five-cluster solution, the rule of thumb that bivariate residuals should not exceed four was not met [23]. The entropy, r², and classification errors of the six-cluster solution were satisfactory; see

Table 2. Fit indices of potential latent cluster models.

Model	LL	BIC	Npar	Entropy r2	Class. Err.
One-cluster	−9154.878	18,898.513	87	1.000	0.000
Two-cluster	−8736.471	18,095.537	92	0.786	0.060
Three-cluster	−8591.372	17,839.177	97	0.799	0.076
Four-cluster	−8544.702	17,779.673	102	0.724	0.143
Five-cluster	−8517.969	17,760.044	107	0.728	0.143
Six-cluster	−8502.614	17,763.170	112	0.700	0.179
Seven-cluster	−8489.270	17,770.320	117	0.661	0.237
Eight-cluster	−8478.034	17,781.684	122	0.684	0.220
Nine-cluster	−8467.897	17,795.246	127	0.684	0.236
Ten-cluster	−8466.985	17,827.259	132	0.686	0.236

Note. N_par = number of parameters, class. Err. = Classification Error. The selected models are in bold and italics.

. The clusters showed an overall pattern from low to high expectations (Figure 1,

Table 3. Means (and standard deviations) of latent cluster analysis indicator variables across clusters.

	Cluster 1	Cluster 2	Cluster 3	Cluster 4	Cluster 5	Cluster 6
	n = 9	n = 74	n = 55	n = 328	n = 279	n = 124
Awareness raiser	1.96 (0.46)	2.25 (0.30)	2.57 (0.30)	2.77 (0.23)	3.12 (0.24)	3.51 (0.25)
Connector	1.63 (0.43)	2.57 (0.19)	3.11 (0.26)	3.05 (0.21)	3.46 (0.20)	3.77 (0.16)
Referrer	1.63 (0.50)	2.77 (0.29)	3.58 (0.20)	2.99 (0.20)	3.39 (0.25)	3.73 (0.19)
Guardian	1.87 (0.41)	2.32 (0.49)	2.32 (0.40)	2.72 (0.40)	2.87 (0.39)	3.28 (0.41)

). In other words, there were no clusters with a specific mix of low values of certain indicator variables and high values of others. Rather, there was a cluster with low scores on all roles, a cluster with high scores on all roles, while the other clusters showed moderate scores for all four indicator variables (role expectations). The exception was cluster three, which showed a slightly different pattern, with the highest expectations for the referrer role and lower scores for the others.

Cohen’s d-values of inter-class comparisons (see Supplementary Table S2) of the four indicator variables show almost consistent large (Cohen’s d > 0.8) or very large (Cohen’s d > 1.3 [24]) differences between clusters. A trivial difference was found for the guardian role between cluster two (cluster with relatively low expectations) and cluster three (cluster with referrer as the most expected role). A small effect was found for the connector role between clusters three and four (cluster with medium high expectations) and for the guardian role between clusters four and five (cluster with relatively high expectations). For the awareness raiser, a medium effect was found between clusters one (small cluster with low expectations) and two and between clusters three and four. Furthermore, a medium effect was found for the referrer role between clusters three and six (cluster with the highest expectations).

Within clusters, students showed significant differences in the indicator variables. Students consistently expected tutors to have more connector and referrer roles than awareness raiser and guardian. Connector and referrer roles were roughly equally expected of tutors among students, as were awareness raiser and guardian (see Supplementary Table S3 for within-cluster comparisons).

In the member check, the students recognised the relative importance of all role expectations. According to them, tutors should be able to adapt to specific situations in different ways and sense the best role to assume in a specific situation. Thus, the roles were viewed as falling in a continuum. In line with the outcomes of the latent class analysis, the role expectation of connector was considered the main role of tutors, together with the role expectation of referrer. This was linked to the impression that these role descriptions represented the prevailing vision of the tutor’s role in student mental health in higher education.

The MMLR was conducted on a subcategory of students owing to missing cells. Students consistent with the very low expectations profile were excluded (n = 9). Furthermore, we excluded students who did not specify their gender (n = 33) and had an associate’s or master’s degree course (n = 35 and n = 19, respectively). Students who wished not to disclose whether mental health had ever been discussed (n = 28) and those who indicated that they had a different type of coach (n = 33) were also excluded. The four-year and more than four-year categories of the study-year variables were collapsed into a single category. This left a total sample size of 623 students for the MMLR. The variance inflation factors for the analysis were less than 1.112, indicating no signs of multicollinearity.

Student satisfaction with the role fulfilment of their tutors in mental health was predictive of cluster membership. Overall, the satisfaction variable added to the model fit of the MMLR model compared to an intercept-only model (i.e., a model without any predictors; χ² (4) = 10.78, p = 0.029).

In the MMLR model, cluster six (the cluster with the highest expectations of tutors) was the reference category, and increased satisfaction with a coach was significantly predictive of higher odds of belonging to clusters two (OR = 1.33, 95% CI [1.10, 1.60], p = 0.003), four (OR = 1.15, 95% CI [1.03, 1.29], p = 0.017), and five (OR = 1.14, 95% CI [1.01, 1.27], p = 0.028) than to cluster six. Satisfaction was not predictive of the odds of belonging to cluster three rather than six (OR = 1.15, 95% CI [0.97, 1.37], p = 0.110). Additionally, whether mental health had been discussed with a tutor was predictive of class membership (χ² (4) = 9.76, p = 0.045), with significant lower odds of belonging to clusters 2 and 4 if mental health was discussed with the tutor. The remaining variables did not predict class membership (study year: χ² (12) = 9.20, p = 0.686; gender: χ² (4) = 0.12, p = 0.988; type of education: χ² (4) = 3.14, p = 0.534; type of coach: χ² (4) = 2.43, p = 0.658; faculty: χ² (20) = 25.44, p = 0.185). See

Table 4. Outcomes of multivariate multinomial logistic regression with latent cluster membership as the outcome.

		Cluster 2 n = 74		Cluster 3 n = 55		Cluster 4 n = 328		Cluster 5 n = 279
		OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
Satisfaction with coach		1.33 **	[1.10, 1.60]	1.15	[0.97, 1.37]	1.15 *	[1.03, 1.29]	1.14 *	[1.01, 1.27]
Gender
	Male	1.08	[0.45, 2.60]	0.96	[0.38, 2.44]	1.02	[0.55, 1.91]	0.97	[0.52, 1.81]
	Female	Ref.	-	Ref.	-	Ref.	-	Ref.	-
Study year
	Second year	0.86	[0.21, 3.53]	0.27 *	[0.07, 0.98]	0.98	[0.38, 2.53]	0.69	[0.26, 1.79]
	Third year	0.95	[0.23, 3.94]	0.59	[0.18, 2.01]	0.87	[0.34, 2.27]	0.85	[0.33, 2.21]
	Fourth year	0.90	[0.21, 3.89]	0.43	[0.12, 1.55]	0.74	[0.28, 1.99]	0.61	[0.23, 1.65]
	>Four year	Ref.	-	Ref.	-	Ref.	-	Ref.	-
Type of education
	Full time	0.81	[0.30, 2.17]	1.02	[0.37, 2.84]	1.34	[0.63, 2.86]	1.54	[0.70, 3.41]
	Part time	Ref.	-	Ref.	-	Ref.	-	Ref.	-
Academy
	Business	0.76	[0.14, 4.31]	0.75	[0.13, 4.22]	0.49	[0.14, 1.66]	0.89	[0.25, 3.15]
	Technology	0.63	[0.07, 5.35]	0.41	[0.04, 4.02]	0.82	[0.19, 3.60]	0.86	[0.18, 4.04]
	Hospitality	0.31	[0.05, 2.06]	0.15	[0.02, 1.25]	0.19 *	[0.05, 0.68]	0.55	[0.15, 1.98]
	Education	0.64	[0.11, 3.67]	0.69	[0.12, 3.93]	0.63	[0.18, 2.13]	0.68	[0.19, 2.47]
	Health and well-being	1.02	[0.17, 6.10]	1.50	[0.26, 8.56]	0.97	[0.27, 3.46]	1.11	[0.30, 4.18]
	Law	Ref.	-	Ref.	-	Ref.	-	Ref.	-
Type of coach
	Just coach	0.59	[0.26, 1.34]	1.02	[0.47, 2.24]	0.97	[0.56, 1.68]	0.85	[0.49, 1.47]
	Coach and teacher	Ref.	-	Ref.	-	Ref.	-	Ref.	-
Mental health was discussed
	Yes	0.40 *	[0.18, 0.90]	0.55	[0.24, 1.26]	0.50 *	[0.28, 0.90]	0.78	[0.43, 1.42]
	No	Ref.	-	Ref.	-	Ref.	-	Ref.	-

* p < 0.05; ** p < 0.01.

for outcomes of multivariate multinomial logistic regression. Supplementary Table S4 provides an overview of the cluster membership distributions.

Students in the member check reflected on their role expectations, origin of expectations, and relationship between expectations and satisfaction. They highlighted the importance of the context for expectations, such as faculty (contrary to the outcomes of the MMLR), the scale and culture of their studies, and their support needs. For a proportion of the students, relatively high overall expectations were linked to the importance of mental health. Explanations for their expectations included earlier expectations such as secondary or vocational education, their own needs, and information provided by the tutor or study program. However, expectation management could be improved by providing better information and explicating the tutor’s role in student mental health. Although study year and age were not significant predictors of cluster membership, students in the member check linked these variables to the importance of the individual role expectations of awareness raisers and guardians. For younger students, these roles were considered more appropriate due to their developmental stage, as well as the fact that first-year students might need more intensive support because they are getting used to studying and finding their way in the University of Applied Sciences. Furthermore, students empathised with tutors by stating that roles should fit the tutor’s preferences to some extent and that a balance between prescribing and space for personal interpretation is important.

4. Discussion

This study aimed to explore students’ expectations of tutors regarding mental health support. Therefore, we explored the role expectations and distinct role expectation profiles of students, as well as the relationship between students’ role expectation profiles and background characteristics, specifically the level of satisfaction with tutors’ role fulfilment.

Overall, the findings emphasise the importance of tutors in student mental health from the student’s perspective, confirming the outcomes of earlier studies (e.g., Douwes et al. [16]) and findings on the importance of tutors in higher education in general (e.g., Grey and Osborne [25]; Yale [26,27]). Furthermore, the findings provide more depth to existing knowledge by offering insights into how students expect tutors to play a role in their mental health. For this study, role descriptions found in an earlier study on tutor role perceptions [8] were used, which showed a good fit with the student data. This suggests that the role perceptions of tutors and the role expectations of students are comparable in content. Role expectations were as follows: awareness raiser, connector, referrer, and guardian.

The outcomes of the latent class analysis showed five large student clusters and one small cluster. Four large clusters of role expectations showed an overall pattern of running from low to high, with no specific mix of low values for certain role expectations and high values for others. One cluster showed a slightly different pattern, with the highest expectations for the referrer role and relatively less expectations for the other three roles, indicating that tutors were not necessarily expected to play a role in the support themselves.

MMLR analysis revealed that satisfaction with the role fulfilment of tutors and faculty members was predictive of cluster membership. Interestingly, students who reported higher levels of satisfaction tended to have lower expectations from their tutors. This correlation between satisfaction levels and tutors’ expectations is not unexpected and is consistent with findings from previous studies [28,29]. However, the underlying reasons for this association in the context of students’ mental health remain unclear. One potential explanation is that students may have unrealistically high expectations, leading to frequent disappointment. Alternatively, some students may not hold particularly high expectations of their tutors in the realm of mental health support, making it easier for them to meet these expectations. Research indicates that many students avoid discussing their mental health issues because of fear of stigma (e.g., [30,31,32]), which could lower expectations in this area.

Alternatively, it is possible that levels of satisfaction influenced students’ responses to the survey questions; dissatisfaction might have prompted students to assign higher scores to statements as an expression of discontent. Furthermore, Martin’s [32] study involving 54 students suggests that, of those who do open up about their school-related problems, the majority find this a positive experience. This may imply that clear expectations may help lower barriers to disclosing mental-health-related topics to tutors. In line with this is the finding in the present study that the variable ‘mental health was discussed’ was a significant predictor of belonging to clusters with low or relatively low expectations, a cluster in which satisfaction was relatively high.

Furthermore, in the MMLR analysis, whether mental health was discussed was predictive of membership in the two clusters (2 and 4) with relatively low role expectations. However, the direction of the relationship remains unclear. It is possible that these students had lower expectations because they had discussed their mental health and that this experience somehow lowered their expectations. It could also be that because of the not particularly high expectations that students hold of their tutors, students discuss their mental health with their tutors more easily.

At first glance, the non-significance of the other background variables in the MMLR suggests a relatively uniform perspective among students regarding their expectations, considering the valence of role expectations. However, a closer examination of the results during member checks revealed a diverse array of viewpoints, expectations, and experiences. This finding underscores the significance of the role expectations identified in the present study. Furthermore, it was apparent that roles were described in broad terms and carried varying meanings and significances across different contexts and students. This highlights the importance of considering personal and educational contexts.

The focus on student expectations of the tutor role in this study furthermore fits within the student-centredness that has increasingly become the standard in higher education [33]. Student centredness refers to pedagogical concepts whereby students and their learning are placed at the heart of the educational process to foster deeper learning processes and outcomes [33]. In such student-centred learning environments, students are more often perceived as important stakeholders in the process of enhancing educational quality and can thus be seen as co-creators. Co-creation entails a close collaboration between learners and tutors to improve teaching and learning by welcoming learners’ perspectives and actively involving them in the educational (design) process [34]. Measuring student expectations, as in the present study, is representative of student-centredness and co-creation, particularly when the aim of measuring student perspectives is to incorporate their views in the development of policies. It is not simply the case, however, that higher educational institutions should always meet and satisfy student expectations. Fulfilling student expectations is not per se the equivalent of providing high-quality support [35]. Students’ expectations do not necessarily align with the educational goals that deliver the best opportunities for long-term growth and development. For instance, the role of guardian in our study is expected to a certain extent and might be helpful for students in the short term. However, a tutor who takes on the role of guardian may not help students in the long-term development of self-regulation and self-efficacy. Therefore, to align with expectations, perspectives on the purpose of university education and the role of educational institutions in students’ mental health are essential.

4.1. Implications

The role expectations found in this study may be helpful in managing expectations between tutors and students regarding this part of the tutor’s role. Furthermore, findings on role expectations may be helpful in the development of educational policies for a learning environment in higher education that is optimal for student mental health and, more specifically, for the tutor’s role in student mental health. Incorporating student voices into the matter of expectations of tutor roles in student mental health is essential, but it must be clear how their opinions are weighted, considering that other (f)actors play a part in explicating the tutor’s role. In line with this, the findings also imply that in discussions of the tutor role, it is important for policymakers and management to acknowledge the difficulties tutors may encounter in their work and contact with students and to offer support for this by offering professionalisation, clear expectations and guidelines, and keeping an eye on the well-being and mental health of tutors.

Second, the findings imply the importance of a good personal tutor–student relationship for tutors to fulfil their roles in student mental health. Here, it is important to acknowledge that the quality of a tutor–student relationship in general, especially in a student-centred learning environment, requires students’ efforts as well. As student expectations seem to be person- and context-specific, insight into the alignment of expectations of students with perceptions of their tutors is of interest for both practice and future research. Further research should also focus on the relationship between clarity in expectations and the support from their tutors as perceived by students regarding their mental health.

Third, in discussions of tutor roles and student expectations, it is important that tutors feel supported by policies and management. In a review of factors impacting teacher educators, supporting student well-being has been shown to be related to decreased educator well-being alongside a perceived lack of institutional support [36]. Although aimed at a specific subgroup of educators, the results of this review are of interest more broadly. Clear guidelines, sufficient time, and opportunities for professionalisation could be helpful for tutors, as student mental health is obviously important, as is the well-being of educators and, in the context of the present study, of tutors.

4.2. Strengths and Limitations

This study provides new insights into tutors’ role expectations concerning students’ mental health in higher education. To the best of our knowledge, this is the first in-depth study of role expectations using a relatively large sample. However, it must be kept in mind that although this study was conducted at only one university of applied sciences, the response rate was relatively low (5% of N = 17,296). Future research should test whether our findings hold for other samples.

The second strength is that the measurement instrument used was an adaptation of a measurement instrument for tutors that was developed by following a rigorous procedure, and that the adaptation was conducted by involving students. However, the questionnaire also has limitations. For instance, satisfaction with role fulfilment was not asked for every statement in the questionnaire but as a single question (which could be scored between 1–10). Therefore, in-depth insights into the elements that foster or hinder these concepts are limited.

Third, we consider it a strength that the study focused on role expectations, as they may serve as a starting point for expectation management. However, some students might have ‘used’ the questionnaire to express their dissatisfaction with their tutor, i.e., used the questionnaire as a student evaluation. Although satisfaction is linked to expectations, this might have biased our results. Furthermore, the students involved in the study had experience with their tutors, which might have generated hindsight bias when they were asked about their expectations. Hindsight bias implies that people generally do not recall the past correctly but rather allow their experience to colour what they claim to have believed initially; thus, students’ memories of their expectations may be biased after the fact [29].

The final strength is that the results show that role expectations are well recognised in a large group of students and that the additional member check has provided more in-depth insight into the importance of the context. The limitation of role expectations may be that they are a simplified representation of reality and that boundaries between roles may not always be as clear or rigid as role descriptions may assume. Despite this limitation, role expectations may serve to manage expectations.

5. Conclusions

In conclusion, this study explored student expectations of their tutors regarding mental health support in higher education by examining role expectations, distinct profiles, and the relationship between these expectations and background characteristics. These findings underscore the importance of tutors in students’ mental health and expand existing knowledge by shedding light on how students expect tutors to play a role in their mental well-being. The study unveiled a nuanced landscape of student expectations and revealed diverse viewpoints and experiences among students, highlighting the significance of context in understanding role expectations and metacommunication regarding the tutor’s role in student mental health support. The correlation between satisfaction with tutor role fulfilment and expectations among students suggests a complex interplay between satisfaction levels, expectations, and student–tutor dynamics. This finding aligns with previous research but underscores the need for further exploration within the specific context of student mental health support. The focus of this study on student-centredness reflects the evolving landscape of higher education, in which students are increasingly recognised as important stakeholders in the educational process, promoting deeper learning outcomes through co-creation and collaboration. However, it is important to note that meeting student expectations does not always equate to providing high-quality support. Understanding and aligning expectations requires consideration of diverse perspectives on the purpose of university education and the role of educational institutions in supporting student well-being. This broad and holistic approach is essential for fostering an environment in which students feel empowered to express their concerns and receive support that aligns with their expectations while also being consistent with the role and possibilities of tutors.