Role Perceptions of Teachers Concerning Student Mental Health in Higher Education

Abstract

Student mental health is a relevant aspect of the current academic context due to concerns regarding escalating mental health issues and because of an increasingly holistic perspective on education, learning, and personal development. This research investigates the perceptions of teachers in higher education regarding their roles in promoting student mental health, explicitly focusing on teachers who hold a tutor position. After following a strict procedure, the Mentor-Q questionnaire was designed to measure respondents’ perceptions of their role. This questionnaire was administered online to all Dutch University of Applied Sciences teachers for the study; 180 teachers finished it. Results were discussed and validated using a member check (n = 10). Exploratory factor analysis identified four role perceptions: connector, awareness raiser, referrer, and guardian. While all teachers endorsed these roles, the three clusters that appeared in a latent class analysis varied in their agreement levels. No background variables were significantly related to the role perception profiles. It seems that supporting teachers in recognizing and maintaining boundaries in their role in student mental health is crucial. This study highlights the complexity of the teacher’s role in supporting the student’s mental health and has implications for aligning their responsibilities with higher education goals.

1. Introduction

The mental health of higher-education students is a subject of significant interest. The increasing number of students dealing with mental health issues [1,2], the link between mental well-being, academic achievement, and the drop-out rate [3], along with the concern that most students with mental health problems delay seeking help [4,5], has encouraged research on these adverse trends.

Furthermore, the growing interest in students’ mental health is linked to an increasingly holistic perspective on education, learning, and personal development. This perception intertwines (mental) health and cognitive development due to the emphasis on viewing students as ‘whole’ individuals enacting multiple social roles apart from being students [6]. While recognizing this interconnectedness has gained prominence, especially in light of the COVID-19 pandemic, it is not a novel concept. Decades ago, Tinto [7] already emphasized the significance of the cognitive and academic dimensions of education (academic integration) and the interconnectedness of students with individuals in the educational setting (social integration) for effective learning [8]. Besides education’s qualification and socialization functions, Biesta [9] also highlights a third function called ‘subjectification’: identity formation. This indicates that students’ academic development is intertwined with their personal growth, and consequently, educational institutions should address both aspects in their learning environments to support their students.

The importance of explicitly acknowledging the role of educational institutions in addressing student mental health has been exemplified in several studies. In addition to the availability of support services (e.g., Cohen et al. [10]), there is often an implied role for teachers [11,12,13]. Teachers are not supposed to act as mental health professionals. However, they are acknowledged for their role in supporting student well-being and mental health [14,15]. They can do so by creating a safe and conducive learning environment [16,17]. Their accessibility further underscores the importance of teachers; they are individuals in the educational context closely connected with students and, therefore, should play a frontline role in identifying and supporting students who require assistance [12,18,19,20,21].

Despite their role in supporting and promoting student mental health, only a few studies have explored the experiences and reflections of teaching staff in higher education. Existing literature often focuses on the obstacles and challenges teachers face to fulfill their roles in student mental health. For instance, research indicates that while teachers often acknowledge their roles in student mental health, they struggle with questions regarding the extent of this role, how to integrate it with other responsibilities, their self-efficacy, and the support they receive [13,18,19,20,22,23]. Furthermore, the literature also suggests an inverse relationship between the well-being of teachers and students (e.g., Jayman et al. [24] and Kiltz et al. [25]). As teachers take on a greater role in supporting the mental health of their students, their own well-being may decline. Despite being in a in a good position to provide support, teachers often lack the necessary skills and knowledge due to stigma, limited mental health education, heavy workload, their own mental health needs, and unclear job expectations. These challenges have been highlighted in recent studies by Payne [20] and Semchuk [14].

The findings emphasize the significance of how teachers perceive their roles. Role perception refers to the various attitudes, viewpoints, understandings, and expectations that individuals or groups of people have about their status and position in an organization. Unlike role definition, role perception is subjective and can vary from person to person. These perceptions serve as the foundation for understanding teachers’ roles and provide the motivation for their behavior, according to Jiao et al. [26]. Another rationale supporting the importance of teacher role perceptions is that integrating a focus on student mental health within the teacher role may necessitate adopting new approaches. The literature on the professional learning and development of teachers has shown that successful implementation of changes relies on the beliefs, readiness, willingness, and ability of teachers who have to execute them [27,28].

Another notable issue was that most studies on this topic are qualitative and have small sample sizes [19,20,22,23]. Roles for teachers emerging from these studies are those of confidant/listener [19,22], signposting/referring [19,20,22,23], first responder/recognizer of signals [22,23], and bystander [22].

This exploratory study aims to investigate whether teacher roles can be identified and verified within a large group of teachers. This kind of study has not been conducted before. Additionally, we aim to explore possible profiles of role perceptions, which can provide valuable guidance for the professional development of teachers. The research will address the following questions:

1. How do higher-education teachers perceive their role concerning students’ mental health, and to what extent are their background characteristics related to these perceptions?

2. Can teachers be categorized into meaningful role-perception profiles, and are their background characteristics related to these profiles?

This study focuses on teaching staff serving as personal tutors to students. Previous research has indicated that tutors play a significant role in students’ well-being compared to other teachers [29]. In higher education, students are often assigned a personal tutor who is a member of the teaching team for their study program. Approaches to personal private tutoring may include academic and personal support. Nearly all teachers served in a tutoring capacity at the institution where this research was conducted, which is often an integral part of their teaching responsibilities. Thus, it is an allocated role and not a chosen one. Teachers may have received some basic training in tutoring as part of mandatory didactical training for teaching staff without formal teaching qualifications. Furthermore, the university also offers optional training in tutoring. While the university has established general guidelines for teaching and tutoring, these do not specifically address the role of tutors in supporting student well-being and mental health.

2. Methods

2.1. Research Design

A survey study was conducted at a Dutch University of Applied Sciences offering 150 study programs at different levels. These programs are organized into six clusters of study programs (faculties). The university has an enrollment of over 24,000 students and a teaching staff of approximately 1570 employees. The teaching staff comprises individuals facilitating student learning, such as workplace supervisors, instructors, teachers, and researchers working with students.

2.2. Methods and Procedure of Data Collection

An email invitation was extended to all teaching staff (N = 1570) to participate in the study. The invitation specified that the target consisted of tutors in the current or past year, with those not fitting this criterion being encouraged to disregard the invitation. The exact number of teaching staff involved in tutoring is unavailable because tutoring is considered a role rather than a job. However, the university’s educational science department estimated this figure to be approximately 80% (N = 1256). The invitation also contained information about the study objectives and included practical details such as the estimated time investment (10 min) and how sensitive information was handled. Participants were directed to the questionnaire via a link designed and distributed in Qualtrics. 180 teachers with tutoring roles completed the online questionnaire, resulting in a response rate of nearly 12%.

Participants

An overview of the respondent sample’s characteristics is provided in

Table 1. Percentages for background and enrolment characteristics for the respondent sample and teaching staff population at the university.

Characteristics	Respondent Sample (n = 180) %	Teaching Staff Population (n = 1570) (%) ᵃ
Type
Tutoring role only	76	- ᵇ
Teaching and tutoring role	24	- ᵇ
Gender
Male	39	45 *
Female	61	55
Age group
25–34	8	18 **
35–44	27	26
45–54	34	26 **
55>	31	30
Academy c
Academy of Business	22	21
Academy of Technology	24	21
Academy of Hospitality, Leisure and Tourism	11	14
Academy of Education	15	25 ***
Academy of Social Studies and Healthcare	24	14 ***
Academy of Public Management	4	5
Years of working experience in higher education
≤4 years	22	- ᵇ
5–9 years	20	- ᵇ
10–14 years	19	- ᵇ
15–19 years	14	- ᵇ
20–24 years	15	- ᵇ
25> years	10	- ᵇ
Years of working experience in tutoring
≤4 years	29	- ᵇ
5–9 years	22	- ᵇ
10–14 years	20	- ᵇ
15–19 years	14	- ᵇ
20–24 years	8	- ᵇ
25> years	7	- ᵇ

ᵃ since no numbers are available for only those teachers that fulfill a tutoring role, numbers of all teaching staff are used. ᵇ no numbers available ᶜ for academy, n = 2 participants with a teaching position worked in a support-service department and therefore are excluded in analyses concerning academy. * Χ² (1, N = 180) = 2.25, p = 0.13. ** Χ² (3, N = 180) = 15.05, p = 0.002. Age group 25–34 underrepresented in sample, whereas age group 45–54 overrepresented in sample. *** Χ² (5, N = 178) = 21.00, p < 0.01. Academy of Education is underrepresented in sample; Academy of Social Studies and Healthcare is overrepresented.

, along with a comparison to the overall teaching staff population.

Male teachers, teachers between 25–34 years old, and teachers from the Academy of Education were somewhat underrepresented in the sample. Teachers between 45 and 54 years old and those from the Academy of Social Studies and Healthcare were somewhat overrepresented in the sample.

2.3. The Instrument

We created a new instrument for this study as there was no existing validated tool to measure the role perception of tutors in student mental health. The instrument consists of statements that focus on how tutors perceive their role concerning students’ mental health.

The statement list was developed, using rigorous scale-development criteria [30,31]. Statements were grounded in interviews with students regarding their expectations from teachers and tutors concerning their well-being and mental health (n = 27), the existing literature and research (qualitative) on the teacher–student relationship and the role of teachers in student mental health [16,18,19,22,32,33,34,35,36,37], and exploratory interviews with tutors (n = 4).

The first step of item generation resulted in a preliminary item pool of 386 statements. We removed redundant statements and those that did not align with the purpose of the study (by researcher RD). This process led to the exclusion of the statements that referred to behavior or self-efficacy instead of role perception, resulting in 82 statements. We discussed this set with three other researchers (NB, MP, and JM) and checked for the formulation, clarity, and appropriateness for the study purpose. Further adjustments led to removing eight unclear items, leaving behind 74 statements.

To enhance the content validity and clarity of the study, two focus groups were conducted, each consisting of four teachers with tutoring roles. Although efforts were made for participant heterogeneity, practical considerations, such as availability, led to relatively homogeneous groups consisting of female participants from the faculty of Social Studies and Healthcare, with one participant from the faculty of Business. Results from participant discussions led to further removal of the statements. They were asked to score the statements on relevance, clarity, and conciseness, and we also verified if they missed topics or statements. The appropriateness of statements was measured using a content validity index (CVI) to determine its accuracy [38]. Subsequently, participants were asked to rate the statements for relevance on a scale from one to four (1 = not relevant at all, 4 = very relevant). Following the CVI procedure, six of the eight participants had to judge a statement as relevant (score 3 or 4) for a statement to be considered valid. Statements lacking content validity, scoring one or two by two or more tutors, were removed. The feedback from the focus groups resulted in the elimination of 16 items, leaving 58 items remaining. In the final stage of the instrument development, a second iteration on content validity involved a panel of ten experts. These experts, well-versed in research methodology, educational sciences, (student) mental health, and tutoring, evaluated statements using the CVI procedure. Participants were also asked to provide feedback on clarity, conciseness, and potential omissions. Following CVI criteria, statements were deemed content valid if rated as relevant (score 3 or 4) by seven or more participants. Thirteen items were removed. The procedure also yielded additional information on the clarity of the instructions, which was used to improve the instrument further.

The final version of the instrument (the Mental Health Evaluation by Teachers Offering Support and Resources Questionnaire [Mentor-Q]) contained instructions for participants, in which the definition of mental health was explained. In our study, we use the definition of mental health as described by the World Health Organization [39]: “A state of well-being where one can realize their abilities, cope with normal life stresses, work productively, and make contributions to their community.” Furthermore, an explanation of how to score the statements was provided. The questionnaire consisted of 45 statements which were presented in themes that had emerged during the process of the induction of statements: (1) relation between mental health and study, (2) initiative, (3) involvement, (4) content of role perception, and (5) role perception about others. Participants were asked to rate how much the statements represented their perception of their role in student mental health. Statements could be ranked with a score ranging between 1 and 4 (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).

The questionnaire also contained an open text field for teachers to add comments on the statements and their scorings. Additionally, background information was gathered through questions on gender, age, academy affiliation, and years of working experience in higher education and tutoring, presented in the final section of the instrument.

2.4. Psychometric Qualities of the Mentor-Q

To establish the psychometric qualities of the Mentor-Q, we first examined if the items were suitable for factor analysis. This was the case, with the Kaiser–Meyer–Olkin measure of sampling adequacy being 0.85, above the commonly recommended value of 0.6, and a significant Bartlett’s test of sphericity (χ² (990) = 3135.40, p < 0.001).

We employed exploratory factor analysis to discern if the statements revealed interpretable underlying factors, specifically, distinct role perceptions. Solutions were examined using varimax rotation. The four-factor solution, which explained 45% of the variance, was preferred due to its optimal theoretical interpretability. Eleven items were eliminated because they did not have a factor loading of 0.4 or above. Cross loadings of 0.4 or higher on two factors were checked, and statements were allocated to the factor they fitted best content-wise. All six statements with such cross-loadings were allocated to the factor they loaded highest to. An overview of factors and factor loadings of statements, mean scores, and standard deviations of the four-factor solution are presented in

Table 2. Factor loadings and communalities based on a principal components analysis with varimax rotation for 34 items of the Mental Health Evaluation by Teachers Offering Support and Resources Questionnaire, Mentor-Q (N = 180) (Statements have been translated into English for the purpose of this article. The questionnaire was developed and validated in Dutch.).

Statements			Factors
	M	SD	1	2	3	4
It is my role to preventively discuss with students what is important for good mental health	2.91	0.77	0.76
It is my role to help students become aware of their mental health	2.87	0.73	0.71
It is my role to recognize and notice students’ diminished mental health	3.07	0.68	0.70
It is my role to address and discuss individual students’ mental health situations within my team (provided the student has given approval)	2.73	0.80	0.60
I have a role in promoting students’ mental health due to its impact on their studies	3.14	0.68	0.59
I view students’ mental health as a topic for the entire teaching team and not just for individual study coaches	3.01	0.76	0.55
It is my role to discuss within my team how we handle students’ mental health in general	2.93	0.74	0.55
If I suspect students have diminished mental health, it is my role to initiate a conversation	3.06	0.66	0.54
I believe that attention to students’ mental health primarily belongs to other actors/services within the university rather than study coaches	2.27	0.88	−0.51
As a study coach, I play a significant role in supporting students’ mental health	2.85	0.71	0.49			0.41
I believe that the initiative to discuss students’ mental health lies with the students themselves	2.43	0.64	−0.45
My role is focused on academic progress, not on students’ mental health	2.10	0.81	−0.41
It is my role to invest in bonding with students so they feel comfortable discussing their mental health with me	3.48	0.56		0.71
Attention to students’ mental health, for me, involves being interested in students and their lives	3.39	0.55		0.69
It is my role to show understanding regarding students’ mental health	3.51	0.51		0.68
It is my role to be easily accessible and approachable for students regarding their mental health	3.60	0.57	0.40	0.57
It is my role to keep an eye on students whom I know have diminished mental health	3.28	0.59		0.55
I am the primary point of contact within the university for students regarding their mental health	2.91	0.76		0.53
It is my role to give attention to students’ mental health in interactions with them	3.45	0.62		0.51
To effectively support students’ mental health, it is my role to invest in creating an informal bond	2.76	0.78		0.44
It is my role to provide a listening ear regarding students’ mental health	3.36	0.57		0.43
It is my role to discuss with students what they can expect from me in terms of support for their mental health	3.28	0.60	0.42		0.58
If students need to temporarily pause their education due to diminished mental health, it is my role to help them resume their studies after the break	3.21	0.71			0.54
It is important that I and my colleagues within the program align our roles in supporting students’ mental health as closely as possible	2.82	0.67			0.54
In my role, it is important that I know when students should seek support for their mental health from other actors/services (e.g., the school counsellors’ office or the student success center) within the university	3.45	0.68			0.52
It is my role to know what students expect from me regarding their mental health	3.08	0.70	0.42		0.50
I am the bridge between students and the university it is mental health services	3.16	0.68		0.46	0.48
When I know something is affecting a student’s mental health, I don’t need to know the details; my role is primarily to refer the student appropriately	3.34	0.62			0.44
For fulfilling my role in students’ mental health, I look at what is expected of me from the study program	2.74	0.72			0.40
It aligns with my role to respond to students who reach out about their mental health during my own free time	2.54	0.86				0.77
It is my role to communicate to students that I am always available if they need assistance with their mental health	2.74	0.89				0.75
If students need to temporarily pause their education due to diminished mental health, it is my role to stay in contact with them	2.47	0.80	0.41			0.59
If students want to discuss their mental health with me, it is my role to make time for them as long as it is during working hours	2.99	0.75				−0.54
In fulfilling my role in students’ mental health, I find it important to meet students’ expectations	2.57	0.70				0.43

. The internal consistency of each factor was examined utilizing Cronbach’s alpha. The first factor emerging from the factor analysis contained 12 statements, $(\overline{x}$ = 2.90, sd = 0.46, $\alpha =0.86$). The second factor consisted of nine statements ($\overline{x}$ = 3.30, sd = 0.39, $\alpha =0.82$). The third factor emerging from the analysis contained eight statements ($\overline{x}$ = 3.13, sd = 0.37, $\alpha =0.68$), and the fourth factor had five statements ($\overline{x}$ = 2.47, sd = 0.53, $\alpha =0.68$). No substantial increases in alpha for any of the scales could have been achieved by eliminating more statements.

2.5. Analysis

First, sample background characteristics were explored by using descriptive measures, and their representation of the total teaching staff population was assessed through chi-square analysis. With independent sample t-tests and analysis of variance (ANOVA), whether role perceptions differed across (categories of) background characteristics of the participants was studied. For the second research question, a latent class analysis (LCA) was performed to categorize participants into meaningful clusters based on role perceptions. Multivariate Multinomial Logistic Regression (MMLR) was used to analyze whether cluster membership correlated to participants’ background characteristics, for example, whether participants with certain background characteristics were over- or under-represented in specific clusters. SPSS version 27 was used for the first research question, while LCA was conducted using Latent Gold (version 5.0), and subsequent MMLR analyses were carried out with SPSS version 27.

2.6. Member Check

The quantitative nature of our study stems from the limitation in obtaining detailed insights into the context of participants’ responses. The teaching staff perspective is central to our study, so their involvement is essential in interpreting the results. For this, all willing participants were sent an email invitation to participate in a group session where results would be discussed. Three male and seven female participants responded and participated in this session. The faculties they represented were Business (n = 1), Technology (n = 1), Social Studies and Healthcare (n = 4), Hospitality, Leisure & Tourism (n = 1), and Education (n = 3). In the session, findings were presented and discussed in four steps: recognition, explanation, sufficiency, and improvement of results [40]. This approach is used in practice-based research to interpret results through dialogue. Findings were collected in written form and used as input throughout the interpretation process.

2.7. Ethical Considerations

The ethical committee of the applicable university approved the research plan. Prior to the process, participants were provided with information about the study’s purpose, its content, and the handling of their data. Informed consent was digitally obtained from participants. Responses were gathered anonymously. Only general personal information about their gender, age group, academy, and years of working experience was requested. For participants involved in instrument development and the member check, informed consent was obtained to use their input in the study pseudonymized. During the final preparations of this manuscript, the first author checked the correctness of occasional sentences throughout the document by means of the query: ‘check spelling’ or ‘check grammar’ in ChatGPT version 3.5. The author reviewed the suggestions provided and edited wording if needed.

3. Results

3.1. Research Question 1: Identification of Role Perceptions and Relation with Background Variables

The content of each factor represented an interpretable role perception. Role descriptions are provided below. Names for the role perceptions were developed collaboratively with the participants of the member check.

3.1.1. Role Perception 1: Awareness Raiser

This factor relates to the perception of the teacher’s role in actively supporting students’ mental health. Teachers who hold this view consider themselves essential contributors to the well-being of their students. They engage in activities such as promoting mental-health awareness, identifying signs of mental-health issues, and initiating conversations with students about their mental well-being. They also believe that discussing mental health and addressing individual mental-health concerns of students within their team is a crucial aspect of their role.

3.1.2. Role Perception 2: Connector

Teachers who view themselves as connector prioritize their relationship with students by being accessible, being approachable, and showing genuine interest in them. They create a safe and supportive environment where students feel secure enough to approach them. However, these teachers leave the initiation of conversations about mental health to the students themselves. This approach is unique because it emphasizes the importance of fostering a positive relationship with students, while also giving them the agency to start conversations about their mental health when they feel comfortable doing so.

3.1.3. Role Perception 3: Referrer

A referrer is a teacher who sees themselves as a referral source for students towards appropriate mental-health services within the university setting. Typical for this perception is also the need for clear guidelines and expectations concerning the teacher’s role. Clear expectations from the study program and discussing mutual expectations with students are important. Furthermore, it is considered crucial for all teachers within a team to define and enact their roles in student mental health in comparable ways.

3.1.4. Role Perception 4: Guardian

A guardian’s role is characterized by limitless availability of time and attention for student mental health. Close contact with students, including availability beyond regular working hours, is deemed fitting. Meeting student expectations holds significance within this role perception.

Analysis of mean scores on the role perceptions revealed no significant differences across background variables for the role perceptions of awareness raiser, connector, and referrer. Concerning the guardian role, an independent sample t-test revealed a significant difference in the scores for men ($\overline{x}$ = 2.62, SD = 0.53) and women ($\overline{x}$ = 2.36, SD = 0.51); t(178) = 3.28, p = 0.01. An ANOVA demonstrated a statistically significant difference between age groups (F(3176) = 3.34, p = 0.02). A Bonferroni post hoc test revealed that the mean score on the guardian role perception was statistically significantly lower ($\overline{x}$ = 2.33) for the 35–44 age group compared to the age group of 55> $\overline{x}$ = 2.62), p = 0.03. Furthermore, there was a statistically significant difference between groups (F(5,172) = 3.40, p = 0.01) for academies. A Bonferroni post hoc test revealed that the mean score on the guardian role perception was statistically significantly lower for the Social Studies and Healthcare academies ($\overline{x}$ = 2.32) than the Business academies ($\overline{x}$ = 2.71), p = 0.02. Calculation of Cohen’s d showed medium (gender: 0.50; age: 0.62) to large (academy: 0.88) effect sizes [41]. An overview of all results of t-tests and ANOVAs for background variables is presented in Supplementary Table S1.

During the member check, the findings presented above were discussed. Overall, teachers acknowledged and recognized the role perceptions, confirming that all possible roles were appropriately captured in the results. The guardian role perception was interpreted an extension of the connector and awareness-raiser roles for exceptional situations. The significant differences based on gender, age group, and deployment academy for the guardian role perception were also discussed. According to teachers of the member check, a higher score for men in the guardian perception role could be attributed to the idea that men tend to get less emotionally involved in situations concerning student mental health. Therefore, they are less strict in maintaining boundaries regarding closeness with students and the demarcation between work and private life compared to their female counterparts.

Differences in scores for age groups were explained by noting that individuals in 35–44 age groups typically experience hectic family lives, which could be a ‘natural buffer’ impacting their availability. The significantly higher score for teachers from the Business academy compared to the teachers from the Social Studies and Healthcare academy was surprising for participants of the member check because the latter group would score highest on this role perception. However, after rethinking this assumption, they could explain the difference because most teachers at this academy have a background in helping and supporting people and, therefore, avoid involving themselves in student mental-health matters. The reason for the higher score at the Business academy could be, according to participants, that networking is essential in most studies offered at this academy, which leads to more informal connections and blurred boundaries between teachers and students.

3.2. Research Question 2: Teacher Role-Perception Profiles

Latent Cluster Analysis (LCA) was employed to determine clusters of teachers with varying role perceptions concerning the support of student mental health. The mean scores on the items belonging to the four role perceptions were used as indicator variables for LCA. These scores are ordinal variables. A total of 10 models was run to determine the best solution. The first model contained a single cluster, and every successive model included an additional cluster.

Table 3. Fit Indices of Potential Latent Cluster Models.

Model	LL	BIC	AIC	Npar.	Entropy r2	Class. Err.
One-cluster	−1874.61	4097.13	3883.21	67	1.0000	0.0000
Two-cluster	−1809.62	3993.14	3763.24	72	0.7838	0.0546
Three-cluster	−1793.25	3986.35	3740.49	77	0.7380	0.1097
Four-cluster	−1780.57	3986.95	3725.13	82	0.7902	0.0905
Five-cluster	−1772.98	3997.73	3719.95	87	0.7356	0.1402
Six-cluster	−1763.02	4003.79	3710.04	92	0.7686	0.1299
Seven-cluster	−1755.34	4014.39	3704.67	97	0.7911	0.1504
Eight-cluster	−1750.5	4030.68	3705.00	102	0.7949	0.1406
Nine-cluster	−1745.81	4047.26	3705.15	107	0.7954	0.1427
Ten-cluster	−1743.43	4068.47	3883.21	112	0.8096	0.1709

Note. Selected model in bold and italics.

shows evaluative information for all models tested.

The main information criterion was BIC, following entropy r² and bivariate residuals. The three-cluster model and four-cluster model have an almost equally high BIC score. Although entropy r² was higher for the four-cluster model and the classification error slightly lower compared to the three-cluster solution, the three-cluster model was deemed to fit best because of lower bivariate residuals. For the four-cluster model, the rule of thumb that these scores should not be over 4 [42] was not met (see Supplementary Table S2).

Figure 1 shows the clusters and scores on the role perceptions for the three-cluster model.

No precise profiles were visible in these clusters, as they did not show a specific mix of low values of certain role perceptions with high values of other role perceptions. Instead, clusters differed in the extent to which the roles match teacher perceptions. This is also pointed out by the moderately high entropy score of 0.74. Cluster 1 covered the most significant proportion of teachers (n = 104). This cluster contained teachers who endorsed the role perceptions of the connector, awareness raiser, and referrer, whereas the guardian role fitted less within their perception. Compared to teachers in other clusters, they endorsed the roles of connector and awareness raiser more than teachers in cluster 2 (n = 41) and less than teachers in cluster 3 (n = 35). Cluster 2 covered teachers who scored highest on the referrer role perception compared to other role perceptions. Cluster 3 displayed a similar pattern to cluster 2 (with connector scoring the highest; next, the awareness raiser and referrer; and, ultimately, the guardian role), but scores were higher across all role perceptions.

Table 4. Mean scores and standard deviations of role perceptions within clusters.

	Mean (n = 180)	Cluster 1 (n = 104)	Cluster 2 (n = 41)	Cluster 3 (n = 35)
Awareness raiser	2.90 (0.46)	2.89 (0.30)	2.40 (0.35)	3.48 (0.21)
Connector	3.30 (0.39)	3.35 (0.24)	2.78 (0.18)	3.79 (0.17)
Referrer	3.13 (0.37)	3.10 (0.11)	2.97 (0.36)	3.43 (0.33)
Guardian	2.47 (0.53)	2.37 (0.49)	2.28 (0.42)	2.97 (0.47)

provides mean scores on role perceptions within clusters.

Calculation of Cohen’s d revealed substantial to large differences between clusters, except for the referrer and guardian role perceptions between clusters 1 and 2, where the mean scores were minor. Cohen’s d of differences between clusters is presented in

Table 5. Cohen’s d of between-cluster comparisons of role perceptions.

	1 vs. 2	1 vs. 3	2 vs. 3
Awareness raiser	1.55 ᵈ	2.08 ᵈ	3.66 ᵈ
Connector	2.56 ᵈ	2.02 ᵈ	5.79 ᵈ
Referrer	0.36 ᵃ	1.00 ᶜ	1.33 ᶜ
Guardian	0.19 ᵃ	1.24 ᶜ	1.57 ᵈ

Note. Effects sizes: ^a Trivial effect; ^b Small effect; ^c Medium effect; ^d Large effect. (Sullivan & Feinn, 2012).

.

The MMLR revealed that none of the background variables were related to cluster membership (gender: χ² (2) = 3.07, p = 0.22; age group: χ²(6) = 5.62, p = 0.47; academy: χ²(10) = 12.06, p = 0.28; years working experience in higher education: χ²(10) = 15.07, p = 0.11; years working experience tutoring: χ²(10) = 16.63, p = 0.08.

Participants in the member check reflected on outcomes, noting that roles cannot be viewed in isolation. Rather they reflect a continuum on which teachers position themselves based on factors such as a student’s specific circumstances, available time, and personal considerations like teacher well-being and motives for supporting students. Some participants mentioned that individuals might adopt a guardian role for likability reasons. Additionally, some participants suggested that the clusters represent a hierarchy of roles, with the connector role forming the ‘base layer’, a role perception that all teachers have or should have. Additional roles can be added depending on the specific situation.

4. Discussion

There is a need for a better understanding of the teacher’s role in supporting student mental health in higher education. This study aimed to investigate teacher perceptions of student mental health and identify possible role-perception profiles. The study focused on a specific group of educators who also have tutoring responsibilities. In addressing the first research question, the study identified four role perceptions—awareness raiser, connector, referrer, and guardian—among a relatively large sample of teachers. In all four perceptions, teachers are engaged in student mental health but differ in how this engagement is perceived. The extent to which teachers perceive these role perceptions for themselves remained consistent across background variables, except for some differences in the guardian role concerning age and academy. Roles showed similarities with some earlier conducted studies in secondary and higher education [14,19,20,23] and contrasted with findings in a proportion of studies that some teachers do not see a role for themselves in support of student mental health [22]. Furthermore, elements characterizing a good teacher–student relationship (TSR) are reflected in the role perceptions. Studies on the importance and impact of this relationship have often been conducted in primary and secondary education. The limited literature available on the nature of this relationship in higher education indicates two essential dimensions in the quality of a teacher–student relationship: (1) an affective dimension, in which elements such as honesty, trust, respect, and care are highlighted, and (2) a support dimension, which describes the support that must be provided through TSR for students’ success at university (e.g., setting clear expectations, answering emails promptly) [43,44]. In the role perceptions found in our study, these dimensions seem to be recognizable, with variation in how these dimensions are fulfilled within the role perceptions.

Addressing the second research question, which aimed to classify teachers into meaningful role-perception profiles and explore the relationship of background characteristics to these profiles, LCA revealed three distinct clusters. The clusters exhibited a valence of role perceptions, indicating that the main difference between clusters lay in the extent to which teachers identified with the role perceptions. No background variable was significantly related to cluster membership. Our finding that role-perception clusters are unrelated to the field of employment differs from findings in other studies, which indicate that different cultures around mental health may develop within different subject areas. There are studies indicating that teachers working in behavioral sciences or disciplines have, for instance, higher mental-health literacy and therefore are more likely to support students with mental-health issues than the staff of other faculties due to their higher mental-health literacy [45,46]. Previous research has also revealed that teachers involved in the healthcare-education sector face difficulties in identifying and maintaining boundaries due to competing academic and professional boundaries [47]. These boundary issues do not appear explicitly in our findings. Contrarily, in the member check, Social Work and Healthcare academy teachers mentioned that their professional backgrounds made them more aware of their boundaries in supporting student mental health. This finding may be impacted by the focus on the theoretical nature of the concept of role perception; identifying and maintaining boundaries in practice may be less clear.

The valence in role perceptions implies that teachers can adopt multiple role perceptions, which may complicate taking on a specific role and maintaining boundaries. There are likely situations in which teachers do not struggle with their role but are confident to take on a specific role. Complexity arises in situations where boundaries between roles and that of the total set of role perceptions are in play. Those can be seen as situations in which an appropriate balance between caring, closeness, and availability on one hand and distance and objectivity on the other has to be found [48]. The outcome of this balancing act has implications for the support provided. A possibly complicating factor in maintaining boundaries is the student-centric focus in higher education, which implies a more personal teacher–student relationship [32,49] with a flattened power structure [50]. The existing literature describes TSR as a balancing act, requiring teachers to be mindful of not becoming overly close and supportive with students [32,43]. Furthermore, tension may arise because TSR in higher education reflects an adult–adult relationship [43], but also a relationship characterized by a difference in power, which implies a balancing act for students between being autonomous and dependent [43,51].

4.1. Implications

The role perceptions identified in our study can serve as valuable input for discussions among policymakers and teachers experiencing uncertainty about their roles in student well-being. Here, it must be noted that policymakers should be informed by the perspectives of teachers to ensure that teachers are not expected to take on specific roles they do not feel comfortable with. The existing literature on the topic emphasizes that teachers should be able to recognize signals as preventive measures, and the focus should be on enhancing mental-literacy skills [13,23]. This would fit within the role of an awareness raiser in our study. From a preventative perspective, this focus is understandable. However, it does not seem to resonate with the valence of role perceptions that teachers see for themselves. In our introduction, we have highlighted the importance of alignment between educational developments and teachers’ perspectives.

The findings imply that, in terms of professionalizing teachers, training programs focusing on recognizing and maintaining boundaries could prove advantageous because LCA shows that all role perceptions are, to some extent, recognized by teachers. For this, insights into the boundaries and the circumstances under which they are experienced are needed. In practice, the different role perceptions and switching between/combining roles might cause role conflicts and ambiguity for teachers and between teachers and students. Uncertain role perceptions make establishing and maintaining boundaries more complex, confusing teachers and students. If teachers do not have a transparent and comparable perception of their role, it is tough for students to know what to expect. Furthermore, clarity in roles, boundaries, and lines of referral ensures that teachers are not taking on pseudo-counselling roles [23].

The third implication urges further research on unraveling the teacher’s role in student mental health; existing theories in the educational sciences could be used as foundational frameworks. Although attention to the mental health of students in higher education seems like a recent phenomenon, focus on the importance of well-being in education and the acknowledgment of personal development as a function of education has been established by well-known theories (e.g., Biesta [9] Tinto [7]) for decades. Students’ mental health can be seen as an element in the attention to well-being and personal development. Utilizing these theories is also beneficial for maintaining the awareness that attention to students’ mental health is not a purpose in itself but should be seen in the light of the broader functions of education—qualification, socialization, and personal development. This also calls for reflection in the higher education sector on the boundaries of the teacher’s role in the mental health of students to determine which roles are desirable and to which extent a student should need support from others for their mental health.

4.2. Strengths and Limitations

This study is one of the first to investigate teacher role perceptions in student mental health in higher education. More specifically, to the best of our knowledge, it is the first to examine the subject by using a quantitative design. The strength of our study is that the members of the target group were involved in developing the questionnaire and interpreting results.

Furthermore, another strength of the study is its developing of a questionnaire that adhered to rigorous guidelines. Replication studies are recommended to confirm the usefulness of the role perceptions in other samples and contexts.

The third strength is that it focused on teachers from all academies of the University of Applied Sciences. Often, research concerning the topic is done in one or two academies, primarily in academies offering social, behavioral, or nursing study programs. Our findings indicate that differences between role perceptions across disciplines might not be as present as is often assumed. Further studies could focus on further investigation of this finding and the measurement invariance of the questionnaire across disciplines.

Although the sample was representative of most background characteristics, it is likely that teachers with a high interest in the topic are overrepresented. This could serve as an excellent initial point for those drawn to the topic. However, it may be an explanation for the relatively high teacher scores on the role perceptions and scores of the clusters.

A limitation of this study is the potential occurrence of socially desirable responses, as most teachers displayed high endorsement across all roles. This could suggest a response bias, where participants might have been inclined to present themselves in a more favorable light. To consider this, we have interpreted the perception profiles from LCA in a manner that allows us to compare profiles against each other instead of with scores on the measurement scale.

Additionally, while the focus on role perceptions is considered a strength, the study does not provide insights into the actual behaviors of teachers or the relationship between their role perceptions and real-world actions. Therefore, studying this relationship is also recommended for further research to enhance our understanding of how teachers’ perceptions translate into their behaviors.

5. Conclusions

This study focused on role perceptions and role-perception profiles in a large sample of teachers concerning student mental health in higher education. It was conducted at a University of Applied Sciences in the Netherlands. Four role perceptions were found: connector, referrer, awareness raiser, and guardian. Three groups of teachers showed support for all role perceptions but they differed in how well the roles matched teacher views. It is important to provide support to teachers in recognizing and maintaining boundaries in their role in supporting student mental health. This helps both teachers and students to manage their expectations effectively. Our research findings can provide valuable insights into understanding the role of teachers in promoting mental health among students in higher education, within the context and objectives of the institution.