Helpful and unhelpful factors in school-based counselling and pastoral care as usual: analysis of qualitative data from the experience of service questionnaire

ABSTRACT

The purpose of the present study was to explore and compare helpful and unhelpful factors reported by young people who received either schoolbased humanistic counselling and the school’s usual pastoral care (SBHC + PCAU) or pastoral care alone (PCAU) as part of a randomised controlled trial (RCT) of school-based counselling. Participants (N = 305) were 13–16 years of age and had answered one or more of the three free-text questions in the Experience of Service Questionnaire (ESQ). A thematic analysis was conducted to identify helpful and unhelpful factors, and a quantitative comparative analysis was subsequently used to assess differences between the SBHC+PCAU and PCAU groups. The identified helpful and unhelpful factors were consistent with previous research on school-based counselling, which indicated that having an opportunity to talk and be listened to by a supportive adult, who can offer appropriate guidance, is helpful for young people. Significant differences were found between the SBHC+PCAU and PCAU groups regarding three helpful themes (having a positive experience, having an opportunity to talk, and being offered guidance) and one unhelpful theme (the adult wasn’t directive enough). The identified helpful and unhelpful factors were not particularly unique to school-based counselling; rather, they represented a more general statement of what young people believed to be helpful and unhelpful in their interactions with adults.

KEYWORDS:

• School-based counselling
• pastoral care
• humanistic counselling
• secondary school

Psychological difficulties among children and young people are well-documented and have increased in prevalence (Green et al. 2005; Gutman et al. 2015; Marcheselli et al. 2018). Many mental health disorders first occur in adolescence (Blakemore 2019). Current estimates suggest that as many as one in seven 11- to 16-year-olds in England have a mental health condition and that one in 16 meet the criteria for more than one condition (Marcheselli et al. 2018). Whilst many mental health difficulties begin in adolescence, they frequently continue into adulthood (Blakemore 2019). This can result in enduring financial and social consequences for individuals (Chen et al. 2006) as well as significant societal costs (Snell et al. 2013).

School-based counselling (SBC) is one of the most widespread mental health interventions available to young people (Cooper 2013). In the United Kingdom, SBC typically follows a humanistic approach (Harris 2013), which is based on the work of Carl Rogers (Rogers 1957, 1961). By providing a non-judgemental, empathic, and supportive relationship, humanistic counselling aims to help young people identify their own solutions to their problems (Hill, Roth, and Cooper 2014).

Quantitative data from four pilot randomised controlled trials (RCTs) of manualised school-based humanistic counselling (SBHC) indicated that SBHC was associated with positive outcomes, particularly significant and large reductions in psychological distress (Chen et al. 2006; Cooper 2013; Cooper et al. 2010; McArthur, Cooper, and Berdondini 2013; Pearce et al. 2017). More recently, the first fully powered RCT of SBHC (the ETHOS trial, Cooper 2021) compared the effectiveness of counselling plus the schools’ usual pastoral care (the pre-existing provision of support for emotional health and well-being, SBHC+PCAU), with pastoral care alone (PCAU). Pastoral care is ‘the term used in education in the UK to describe the structures, practices and approaches to support the welfare, well-being and development of children and young people’ (Calvert 2009, 267).

Pastoral care is universally offered at U.K. secondary schools and offers a benchmark upon which the quality of each school is assessed (OFSTED 2015). Pastoral care is aimed at ensuring the physical and emotional well-being of students; although it is universally offered across all schools, the offer itself is not standardised and can therefore vary considerably from school to school. Pastoral care often includes mentoring, anger management, and school inclusion interventions, it can be delivered to groups, classes, or the whole school, but a significant proportion of pastoral care typically involves one-to-one support for individual students provided by teaching staff (Kendal, Keeley, and Callery 2014). The ETHOS trial found that, compared to pastoral care alone, SBHC+PCAU led to significant and small to moderate reductions in psychological distress at the end of therapy and at three months posttreatment (Cooper 2021). Whilst RCTs are important for establishing the effectiveness of SBC, further research on perceived helpful and unhelpful factors is required to improve the design, delivery, and effectiveness of interventions.

Qualitative studies suggest that young people believe that SBC is helpful primarily because it provides them with an opportunity to talk and be listened to (Cooper 2009). In a systematic qualitative meta-synthesis of findings from nine studies that asked clients about helpful and unhelpful factors in SBC, Griffiths (2013) identified 26 categories related to the opportunity to openly talk and be listened to, which was the most frequently mentioned helpful factor. Other helpful factors included getting things off one’s chest, problem solving, developing self-awareness, receiving guidance from a counsellor, the counsellor’s personal qualities, confidentiality, the counsellor’s independence from the school, feeling understood, being accepted, and the nondirective nature of counselling. Although there is less data available on the unhelpful aspects of counselling, Griffiths found that these tended to relate to practical issues (e.g. missing lessons), discomfort resulting from the counsellor’s lack of directive input (e.g. asking questions, offering advice, suggesting activities, and providing direction), and negative responses from the counsellor (e.g. seeming tense). However, a significant limitation of this research is that it did not compare helpful and unhelpful factors for SBC with other forms of emotional support offered at school. For instance, young people may find talking and being listened to be helpful, but these modalities may also be found in other approaches offered at school. Thus, these experiences may not be unique to SBC; rather, they may more generally indicate what young people believe to be helpful and unhelpful in their interactions with adults. In addition, previous studies involved small sample sizes and qualitative methods. Whilst such studies can be valuable, the analysis of larger datasets with multiple coders is likely to provide more robust and generalisable data. To address this, we conducted a qualitative multi-coder analysis of young people’s experiences of helpful and unhelpful factors in SBHC+PCAU and PCAU, then a quantitative analysis to examine differences in helpful and unhelpful factors identified by each group.

Methods

Design

The current study was a secondary analysis of data collected in 2018 during the ETHOS trial (Cooper 2021) a U.K.-based RCT of person-centred counselling at 18 secondary schools in Greater London. In it, young people who experienced emotional symptoms received either SBHC+PCAU or PCAU alone. See Stafford et al. 2018, for the full trial protocol.

Ethical approval for the trial was obtained according to procedures approved by the University Ethics Committee of the University of Roehampton, Reference PSYC 16/227, 31 August 2016.

Participants

Participants (N = 305) consisted of young people (ages 13–16) with moderate to severe levels of emotional symptoms (as indicated by a score of ≥5 on the Emotional Symptoms subscale of the Strengths and Difficulties Questionnaire [SDQ], range: 0–10; Goodman 2003). As shown in Table 1, 236 participants (77%) identified as female, 65 as male (21%), and four as ‘other’ (1%). In terms of ethnicity, 45% of participants identified as Black, Asian, or mixed, and 55% identified as White or European. Moreover, 16% of participants reported a disability. To be eligible for the tiral, participants had agreed to receive counselling, were not receiving any other emotional support interventions, had good school attendance (85% or higher), and an estimated English reading age of at least 13 years. Participants were randomly allocated to either (a) the counselling + pastoral care group (SBHC+PCAU) group and received up to 10 individual sessions per week of face-to-face counselling, as well as access to the school’s pastoral care support as required or (b) the pastoral care alone group (PCAU), in which they had access to the school’s usual pastoral care support alone (these participants were offered counselling at the end of the trial).

Table 1. Baseline participant characteristics.

	All participants		SBHC+PCAU		PCAU
	(N = 305)	%	(n = 152)	%	(n = 153)	%
Sex
Female	236	77	116	76.3	120	78
Male	65	21	33	21.7	32	21
Other	4	1	3	2.0	1	1
Age (years)
… 13	133	44	69	45.4	64	42
… 14	109	36	56	36.8	53	35
… 15	60	20	26	17.1	34	22
… 16	3	1	1	0.7	2	1
School year
Year 8	54	18	28	18.4	26	17
Year 9	139	46	71	46.7	68	44
Year 10	93	30	46	30.3	47	31
Year 11	19	6	7	4.6	12	8
Ethnicity
White or European	168	55	85	55.9	83	54
Black, Asian, or mixed, or other	136	45	67	44.1	69	45
Missing	1	0	0		1	1
Disability
No	256	84	129	84.9	127	83
Yes	49	16	23	15.1	26	17

All participants were asked to complete the Experience of Service Questionnaire (ESQ; Attride-Stirling 2003) at 12 weeks. Only participants who answered one or more of the free-text questions in the ESQ were included in the current study (SBHC+PCAU, n = 152; PCAU, n = 153). Mean response rates for the free-text questions were 91% and 94.4% in the SBHC+PCAU and PCAU groups, respectively, which represented 92.7% of participants in the ETHOS trial.

Materials

The ESQ is a 15-item self-reported client experience measure and is designed to be completed by 12- to 18-year-olds. It begins as follows: ‘Please think about the appointments you have had at this service or clinic.’ On the first page, respondents are asked to rate the service they received using 12 quantitative response items (e.g. ‘I feel that the people who saw me listened to me’ and ‘Overall, the help I have received here is good’). The second page contains three free-text questions with five lines for participants to fill in: ‘What was really good about your care?’, ‘Was there anything you didn’t like or anything that needs improving?’ and ‘Is there anything else you want to tell us about the service you received?’. Responses to these open-ended questions provided the qualitative data used in the present study.

Procedure

Recruitment for the trial was conducted via the schools’ pastoral care staff. If young people expressed interest in the study, consent was sought from their parent or caregiver, and a member of the research team conducted an assessment and gathered baseline measures. If a young person was found to be ineligible for participation (based on the criteria outlined above) the pastoral care team was notified to allow other options for support to be sought.

Young people were assigned (1:1) to either (a) school-based humanistic counselling with access to the usual pastoral care support (SBHC+PCAU group) or (b) the usual pastoral care support alone (PCAU group). Allocation was masked and centrally conducted via remote access to a secure randomisation procedure.

Follow-up tests were conducted at six, 12, and 24 weeks by testers who were unaware of which group the participant had been allocated to. The ESQ was administered as part of the 12-week battery of tests. Participants were informed by the tester that they should rate the ESQ in relation to their counsellor if they had received counselling and ‘any pastoral care that they have had over the past three months, including contact with their pastoral care teacher’ if they had not.

Interventions

SBHC is a manualised form of humanistic therapy founded on evidence-based humanistic competencies that are appropriate for counselling young people aged 11–18 (Hill, Roth, and Cooper 2014). The premise of SBHC is that young people can address their own problems when given the opportunity to explore them with a supportive, empathic, and trustworthy counsellor. SBHC involves a range of practices, such as active listening and empathic reflections. In the trial, sessions were held face-to-face on a one-to-one basis and lasted 45–60 minutes. They occurred weekly over a period of up to 10 school weeks, and participants were permitted to cease counselling at any point. In total, 19 counsellors (who were qualified to at least the diploma level) and 18 schools participated in the trial. Of the schools, 14 out of 18 had one counsellor throughout the trial, and four schools had two nonconcurrent counsellors.

The counsellors were trained to follow an SBHC manual, which was developed specifically for the trial. Adherence to the manual was evaluated by two independent assessors using a young person’s adapted version of the Person Centred and Experiential Psychotherapy Rating Scale (PCEPS-YP; Freire, Elliott, and Westwell 2014). All counsellors exceeded the predefined adherence cut-off point. Throughout the trial, the counsellors received approximately one hour of individual clinical supervision every two weeks.

To maintain ecological validity, we did not standardise pastoral care services, which means that they may have considerably varied between schools and in what each young person was offered. Typically, pastoral care involves time with school staff; however, the amount and frequency considerably varied, from one-off meetings of five minutes or less to ongoing help provided over an entire day or more (e.g. with a learning support mentor). Pastoral care could also involve a referral to community-based specialists (e.g. social workers or police liaison officers).

Data analysis

Responses to the three free-text questions in the ESQ were read several times by the first author, then organised into positive and negative comments through an interactive process. The latter was based on the principles of thematic analysis, a qualitative method for identifying, analysing, and reporting patterns of meaning in data (Braun and Clarke 2006; Clarke and Braun 2018). As part of this process, domains, subdomains, and themes were continuously reviewed, refined, added, and collapsed until a preliminary analytical framework was created. The first and second authors then reviewed this framework and used it to jointly code a random sample of data from 25 participants. This led to some re-organisation, merging, and separation of subdomains and themes and the establishment of an analytical framework with two overarching domains: perceived helpful factors (i.e. conducive to positive change) and perceived unhelpful factors (i.e. detrimental to positive change). Within these two domains, subdomains indicated whether the data were related to clients (i.e. characteristics of clients), adults (i.e. characteristics or actions of counsellors or primary pastoral care figures), or context (i.e. characteristics of the school setting). Within each subdomain, there were a range of themes.

Subsequently, the first author created a codebook to define each component of the framework (i.e. domains, subdomains, and themes). A second round of coding was then conducted on a sample of 50 textual units by the third, fourth, and fifth authors and an independent coder (a PhD student with experience in counselling research) using the codebook. The third and definitive coding round was conducted by the first author and the independent coder, who re-reviewed the themes together and completed their final coding through an iterative process of discussion, clarification, and coding. Neither of these researchers were involved in the original RCT. The aim of this process was to establish an inter-coder agreement of 75% or higher, but the authors were able to establish 100% agreement on the final themes through a robust, iterative process.

To compare the factors that participants in both groups found helpful, we conducted chi-squared goodness of fit tests. We used both an uncorrected alpha and a Bonferroni corrected alpha, dividing .05 by the number of tests conducted for helpful and unhelpful factors (13 and four, respectively). This yielded alpha values of .00 and .01, respectively. This method was adopted rather than a binary regression logistic analysis because a 2*3*4*3*2 analysis, looking at interactions, would have far too little power to adequately identify predictors.

Results

Thematic analysis and coding

Tables 2 and 3 present the main helpful and unhelpful factors identified in the thematic analysis. Main themes were defined as themes attributed to at least 5% of all participants. In the following sub-sections, the themes are presented according to subdomain in descending order of frequency. Quotes are presented verbatim.

Table 2. Helpful factors: themes and frequencies across conditions.

Domain Subdomain Theme	All participants		SBHC+PCAU		PCAU
Helpful factors	n	%	n	%	n	%
Adult actions
Listening	149	48.7	81	53.6	68	43.9
Offering guidance*	48	15.7	33	21.9	15	9.7
Taking the client seriously	25	8.2	14	9.3	11	7.1
Adult qualities
Kind, caring, and supportive	108	35.3	50	33.1	58	37.4
Understanding	48	15.7	23	15.2	25	16.1
Accepting, non-judgemental, and independent	47	15.4	29	19.2	18	11.6
Warm and friendly	15	4.9	6	4.0	9	5.8
Client processes
Opportunity to talk**	84	27.5	52	34.4	32	20.6
Address problems	47	15.4	29	19.2	18	11.6
Voicing thoughts and feelings	27	8.8	15	9.9	12	7.7
Client outcomes
Positive experience**	42	13.7	31	20.5	11	7.1
Helpful experience	34	11.1	21	13.9	13	8.4
Feel calmer or better	15	4.9	6	4.0	9	5.8

*p < .05 **p < .00. Bonferroni corrected alpha (.05/number of tests).

Table 3. Unhelpful factors: themes and frequencies across conditions.

Domain Subdomain Theme	All participants		SBHC+PCAU		PCAU
Unhelpful factors	n	%	n	%	n	%
Adult actions
Not directive enough**	21	6.9	17	11.3	4	2.6
Client processes
I found it hard to talk	14	4.6	7	4.6	7	4.5
Client outcome
Did not help	20	6.5	10	6.6	10	6.5
School/trial context
The timing of sessions was inconvenient	14	4.6	9	6.0	5	3.2

*p < .05 **p < .00. Bonferroni corrected alpha (.05/number of tests).

Helpful factors

Adult actions

‘Adult actions’ refer to actions taken by adults during the sessions. They include listening, taking participants’ remarks seriously, and offering guidance. The most frequently cited helpful adult action was listening, which was mentioned by 149 (48.7%) participants in both groups. Open body language, full attention, recall of earlier conversations, and confirmation that the adult understood the young person’s remarks were all reported as evidence that the adult was listening.

For instance, one participant stated, ‘She remembered everything that I told her so that made me realise that I was actually being heard.’ Another participant stated, ‘that they [the adult] really listen,’ whilst a third participant wrote that ‘the people who tocked [sic] to [them] know how to listen.’

For 48 participants (15.7%), receiving guidance from the adult (e.g. strategies for navigating academic difficulties, relaxation techniques to ease anxiety, or recommendations for useful apps and websites) was reported as helpful. For example, one participant wrote, ‘I have also [been] given ways to stop myself from self-harming.’ Another participant responded that the adult ‘gave [them] ways to deal with [their] problems.’ A third participant stated that it was helpful when the adult ‘gave [them] ideas of how to distract [themselves].’

The experience of having an adult take their opinions, ideas, thoughts, problems, and experiences seriously was cited as helpful by 25 participants (8.2%). Taking the client seriously was reported to foster a sense of emotional safety. For instance, one participant wrote, ‘I was taken seriously no one would laugh at me and I felt safe and comfortable.’ A second participant stated, ‘My thoughts and emotions were taken seriously also I could be myself and not feel uncomfortable.’ A third participant reported that the adult ‘didn’t complain or tell me I’m being delusional took my problems seriously.’

Adult qualities

‘Adult qualities’ referred to the adult’s interpersonal skills or characteristics. Across conditions, the adult’s genuineness, efforts to help, calmness, and demonstration that they cared were perceived by 108 participants (35.3%) as evidence of the counsellor or pastoral care staff being kind, caring, and supportive. For instance, one participant described the adult as ‘really kind and well-hearted,’ whilst a second participant reported ‘the part where [the adult] supported me and looked after me’ as helpful. A third participant wrote, ‘the conversations are very calm also they are kind and very helpful.’ A fourth participant made reference to the adult ‘trying to make it better,’ whilst another participant said that the adult ‘genuinely [cares] for me.’

The adult’s understanding of the young person was reported as helpful by 48 participants (15.7%). Understanding was associated with asking appropriate questions, self-disclosing commonalities, and noticing nonverbal communication cues. For instance, one participant wrote, ‘they were able to understand what I was trying to say and asked questions about it.’ Another participant said that the adult ‘managed to understand things even when [they] couldn’t fully get it out to them.’

Moreover, an accepting, non-judgemental, and independent stance was identified as helpful by 47 participants (15.4%), who reported that it enabled them to speak freely, honestly, in confidence, and without fear of negative consequences. For example, one participant said that the adult ‘didn’t make [them] feel different or alone.’ A second participant wrote that the adult ‘respecting me and not judging me was helpful.’ In addition, a third participant stated, ‘I felt like I could say anything that was worrying me.’ Another commented, ‘I feel like what I say is kept secret and is not shared with others.’

For 15 participants (4.9%), the quality of being warm and friendly, which manifested through traits such as humour, relatability, and being easy to talk to, was reportedly helpful. For example, a participant noted that the adult was ‘very welcoming so I was comfortable sharing.’ Another indicated that the adult was ‘relatively easy to chat and joke with.’

Client processes

‘Client processes’ referred to factors that related to the participants. Having an opportunity to talk, which involves having a dedicated time and space to speak to an adult, was the most frequently reported helpful client process (n = 85, 27.5%). For example, one participant wrote, ‘I loved having someone to speak to every week.’ Another participant said that ‘it allowed [them] to thourgtoughly [sic] talk.’

Some participants (n = 26, 15.4%) reported that addressing problems, which related to discussing their worries and identifying solutions, was helpful. For example, one participant said, ‘I got to speak about the things that are bothering me.’ Another participant reported that ‘[their] worries and stress were listened to.’ A third participant wrote, ‘[the adult] helped me come up with different solutions to the problems I was facing.’

For 27 participants (8.8%), the ability to voice thoughts and feelings was reported as beneficial. For example, one participant wrote, ‘I could open up about my feelings.’ Another participant commented, ‘I was able to talk about my feeling.’ A third stated that ‘being able to explain all of [their] feelings and emotions to [the adult]’ was helpful.

Client outcomes

‘Client outcomes’ referred to the outcomes that the participants attributed to the support they received (SBHC+PCAU or PCAU). Because of the helpful aspects of emotional support, participants reported experiencing a range of positive outcomes. Most commonly, they reported having a positive experience (n = 24, 13.7%), which they would ‘recommend,’ ‘enjoyed,’ and would ‘do again.’ For example, one participant said, ‘it was good and I am happy I went.’ A second participant wrote, ‘it’s been a fantastic opportunity.’ Another participant stated, ‘I would recommend.’

Thirty-four participants (11.1%) reported that they had a helpful experience. For instance, one participant wrote, ‘I think that the way they helped me was brilliant.’ Another participant said, ‘I found it extremely helpful in multiple ways.’ Another stated that ‘talking really helps.’

For 15 participants (4.9%), receiving support helped them feel calmer or better. They recounted being able to relieve their stress, anger, and worries and improve their emotional wellbeing as a result. For instance, one participant wrote, ‘I always leave feeling better.’ A second participant said that ‘it helped [them] calm down,’ whilst a third indicated that it ‘allowed [them] to ease [their] mind.’

Unhelpful factors

Adult actions

For 21 participants (6.9%), the adult not being directive enough was reported as unhelpful. This typically referred to a lack of questions, response, or suggestions on the adult’s part. For instance, one participant wrote, ‘I am not normally given methods on how to deal with my problems.’ Another participant stated, ‘I didn’t like it when say how you feel and the thing you say is just repeated back to you. E.g. if I say I’m annoyed and the response I get is “hm your annoyed” rather than trying to help etc.’ A third participant said that they would have ‘preferred to have been asked more questions,’ whilst a fourth reported, ‘I didn’t receive any techniques or stratigies [sic] to help me. I often came away feeling worse as my problems were out there with nothing to help.’

Client processes

Some participants reported that the support that they received was unhelpful because they found it hard to talk (n = 14, 4.6%). For instance, one participant stated, ‘I just didn’t really like saying or stating out all of my personal problems because I keep myself private.’ A second participant said that they ‘felt uncomfortable with being able to talk about anything.’ Another participant reported a lack of ‘confidence with people and in [themselves] and talking to people.’

Client outcomes

Some participants (n = 20, 6.5%) reported that the support that they received did not help them. For instance, one participant wrote that, ‘although [the adult] tried to help nothing really happened or changed.’ Another said that ‘it didn’t really help [them] with my problems.’

School and trial context

‘School and Trial Context’ referred to factors associated with context in which the support took place. Some participants (n = 14, 4.6%) found the timing of the sessions inconvenient. For instance, one participant stated, ‘I’m having to leave class to talk to [the adult].’ A second participant said, ‘I didn’t like that I was taken out of one of my favourite lessons.’

Comparison of SBHC+PCAU and PCAU Groups

Regarding helpful factors, significant differences were found in the number of participants in the two groups who were associated with the following codes: having a positive experience, having an opportunity to talk, and being offered guidance. Regarding unhelpful factors, a significant difference was found in the number of participants in the two groups who were associated with the following code: the adult wasn’t directive enough. At the Bonferroni adjusted levels (α ≤ .00 for the helpful domain and α ≤ .01 for the unhelpful domain), a significant difference was found in the number of participants who were coded as reporting having had a positive experience and having an opportunity to talk (helpful domain) and those who were coded as reporting that the adult wasn’t directive enough (unhelpful domain).

Helpful aspects

Significantly more participants in the counselling condition reported having a positive experience than in the pastoral care condition X² (1), = 9.5, p = .00, N _SBHC+PCAU = 31 (20.5%), N _PCAU = 11 (7.1%).

Significantly more participants in the counselling condition reported that having an opportunity to talk was helpful than in the pastoral care condition, X² (1, N = 84) = 4.8, p = .03, N _SBHC+PCAU = 52 (34.4%), N _PCAU = 32 (20.6%).

Significantly more participants in the counselling condition reported that being offered guidance was helpful than in the pastoral care condition, X² (1, N = 48) = 6.8, p = 0.01, N _SBHC+PCAU = 33 (21.9%), N _PCAU = 15 (9.7%). However, this was not significant after the Bonferroni correction.

Unhelpful aspects

Significantly more participants in the counselling condition reported that the adult was not directive enough than in the pastoral care condition, X² (1, N = 21) = 8.0, p = 0.00, N _SBHC+PCAU = 17 (11.3%), N _PCAU = 4 (2.6%).

Discussion

The study aimed to identify and compare young people’s perspectives of helpful and unhelpful factors in a school-based intervention (either counselling or pastoral care). We found that being listened to was the most frequently reported helpful factor, which is consistent with previous research (Cooper 2004; Dunne, Thompson, and Leitch 2000; Lynass, Pykhtina, and Cooper 2012; McArthur, Cooper, and Berdondini 2013; Sherry 1999; Verasammy and Cooper 2021). Additionally, we found that having an opportunity to talk, addressing problems, and voicing thoughts and feelings were all identified as helpful factors in both groups. These themes correspond with ‘having an opportunity to talk and be listened to,’ a meta-category identified by Griffiths (2013). Had these three talking-related themes been combined with the listening theme into a single meta-category in line with Griffiths (2013), it would have been the most frequently reported helpful factor by far.

Some participants reported that they did not find their intervention to be helpful, and a small number reported that talking was difficult for them. These results both align with the literature on SBC (Cooper 2004, 2006; Dunne, Thompson, and Leitch 2000; Hough and Freire 2007; McArthur, Cooper, and Berdondini 2016; Sherry 1999).

Given that young people experience talking and being listened to as helpful, it is possible that struggling to talk in counselling contributes to poor experiences. This is logical from a theoretical standpoint, as the notion that the exploration of emotions within the therapeutic relationship eases psychological distress is at the core of humanistic therapy (Pascual-Leone, Paivio, and Harrington 2016). Furthermore, McArthur, Cooper, and Berdondini (2016) study on change processes in SBHC found that the ability to express emotions was fundamental in several of the change pathways that they had identified.

Being offered guidance by the adult was reported as helpful by participants in both groups. This finding aligns with research on counselling with young people (Bondi et al. 2006; Cooper 2004, 2006, 2009; Dunne, Thompson, and Leitch 2000; Freire and Cooper 2008; Hough and Freire 2007; Karver et al. 2006; McArthur, Cooper, and Berdondini 2016; Verasammy and Cooper 2021). Similarly, some participants reported that the adult not being directive enough was an unhelpful factor, which is also consistent with the literature (Bondi et al. 2006; Cooper 2004, 2006, 2009; Dunne, Thompson, and Leitch 2000; Freire and Cooper 2008; Hough and Freire 2007; Karver et al. 2006; Lynass, Pykhtina, and Cooper 2012; McArthur, Cooper, and Berdondini 2016; Verasammy and Cooper 2021).

According to McArthur, Cooper, and Berdondini (2016), the provision of guidance (e.g. advice, suggestions of activities, and asking questions) in counselling has been shown to relieve feelings of awkwardness amongst young people who must sit with an adult and helps to put clients at ease. This may explain why guidance or directive input was frequently cited as helpful by participants who received counselling. McArthur, Cooper, and Berdondini (2016) proposed that feeling at ease enables young people to talk about their emotions, which also potentially facilitates the development of the therapeutic alliance and enables young people to derive the most benefit from their counselling. It may also be that the advice offered was good – and helped the client to improve their life.

Another finding that was consistent with previous research is that several of the adult’s personal qualities were reported as helpful (Freire and Cooper 2008; Hough and Freire 2007; Lynass, Pykhtina, and Cooper 2012; McArthur, Cooper, and Berdondini 2013, 2016; Sherry 1999). The results from this study contribute to the evidence base which suggests that qualities such as being kind, caring, and supportive (Cooper 2004, 2006; Dunne, Thompson, and Leitch 2000; Freire and Cooper 2008; Hough and Freire 2007; Lynass, Pykhtina, and Cooper 2012; McArthur, Cooper, and Berdondini 2016; Sherry 1999); understanding (Cooper 2004; Dunne, Thompson, and Leitch 2000; Freire and M. Cooper 2008; Loynd, Cooper, and Hough 2005; McArthur, Cooper, and Berdondini 2016; Sherry 1999); accepting (Cooper 2006; Hough and Freire 2007; Loynd, Cooper, and Hough 2005); non-judgemental (Hough and Freire 2007; Sherry 1999); independent (Cooper 2006; Dunne, Thompson, and Leitch 2000; McArthur, Cooper, and Berdondini 2016; Sherry 1999); and warm and friendly (Cooper 2004; Hough and Freire 2007; McArthur, Cooper, and Berdondini 2016) are all perceived as helpful by young people who receive support at school.

A quantitative analysis of the qualitative data, which compared the two interventions, identified significant differences between the two groups with regard to two (15%) helpful themes (having an opportunity to talk and having a positive experience) and one (25%) unhelpful theme (the adult not being directive enough). This suggests that the helpful and unhelpful factors reported by participants were not particularly distinctive to SBC but were a more general statement of what young people may find helpful and unhelpful in their interactions with adults more generally. Although these findings are perhaps surprising, they are consistent with research that shows that interventions led by professionals (e.g. counsellors) and paraprofessionals (e.g. ‘a mental health care worker, paid or voluntary, who is not formally qualified in the psychological treatment of mental distress’; Cooper 2008, 92) do not have strikingly different outcomes (Atkins and Christensen 2001; Berman and Norton 1985; Montgomery et al. 2010; Tandon et al. 2021). The helpfulness of the pastoral care intervention may have also been influenced by teachers’ understanding of young people at their school (Hester et al. 2004; Leflot et al. 2010; Shelemy, Harvey, and Waite 2020) and their pre-existing relationships with them, which could have facilitated the development of a therapeutic relationship. Likewise, the adult’s positive personal characteristics (e.g. being kind, caring, and supportive), which were frequently cited as helpful by young people, are not unique to counselling. These are all attributes that are also likely to facilitate positive teaching relationships, which could explain the commonalities found between the two groups. This an important finding since, without specific training, research suggests that teachers are not confident in their capacity to manage and support young people who experience emotional distress in the classroom (Cohall et al. 2007; Roeser and Midgley 1997). Moreover, teachers have articulated feelings of helplessness, which stem from their fears of being ill-equipped to support their students’ mental health (Shelemy, Harvey, and Waite 2019). The results from the current study suggest that whilst the combination of SBHC and PCAU was rated more positively than PCAU alone, young people do report the support provided by pastoral care staff as helpful. Therefore, school staff should feel some confidence in their ability to respond to and support young people with emotional symptoms or who are in distress. However, this does not suggest that there is no place for professional counselling at schools; the evidence clearly supports the effectiveness of SBC on a range of issues (e.g. Verasammy and Cooper 2021). However, it also highlights the value and significance of the pastoral care model that already exists in schools and lends support to a pluralistic approach to intervention (Cooper and Dryden 2016; McLeod 2017), in which there are many ways to help clients.

Although there were many similarities between the two groups, we also found notable differences. For instance, significantly more participants in the counselling group (SBHC+PCAU) reported having a positive experience. This was consistent with the quantitative data obtained from the ESQ in the ETHOS trial (Cooper 2021) which indicated higher satisfaction levels in the counselling group (SBHC+PCAU than in the pastoral care group (PCAU). Moreover, significantly more young people in the counselling group than in the pastoral care group reported that having an opportunity to talk was helpful. A study of school counselling that involved adolescent boys (Dunne, Thompson, and Leitch 2000) suggested that the counsellor’s listening skills could facilitate talking. In addition, a study of perceptions of pastoral care (Kendal, Keeley, and Callery 2014) reported that young people were reluctant to speak to teachers and weighed the risks of sharing personal information with them. Moreover, Howieson and Semple (2000) found that pastoral care staff did not fully appreciate young people’s desire for greater privacy in their interactions with them; as a result, they were less likely to seek pastoral support. By contrast, Verasammy and Cooper (2021) found that client safety, which was associated with therapy contracting and explicitly enforcing confidentiality, was a helpful factor in SBC for bullying. Thus, explicit confidentiality and the trained counsellors’ listening skills in the SBHC+PACU group may have encouraged participants to openly talk about their challenges, which could explain the observed differences. Given the pivotal role that talking plays in processes of change, this may also partly explain why more participants in the SBHC+PACU group than in the PCAU group reported having a positive experience. Another benefit of the SBHC+PACU condition was that it gave participants access to more than one type of support, and in doing so, it may have empowered them to shape their own individual package of school-based support by allowing them to choose the service/person most appropriate to their specific problem.

Regarding unhelpful factors, significantly more participants in the counselling group reported that the adult was not directive enough. In both groups, being offered guidance by the adult was reported as a helpful factor. Research on humanistic counselling with young people consistently raises the counsellor’s level of directive input; this is a complex issue, not least because the definition of what is considered directive is broad and open to misinterpretation. Cooper (2021) distinguished between therapeutic non-directivity(being nondirective in sessions) and meta-therapeutic non-directivity (in which the counsellor trusts the client’s own process of self-determination and accordingly responds to their requests for advice or suggestions rather than, for example, reflecting on them). Research has clearly demonstrated that young people value both directive and nondirective input by counsellors and frequently suggested that they want more directive input, not less (Cooper 2004; Hough and Freire 2007; Lynass, Pykhtina, and Cooper 2012; McArthur, Cooper, and Berdondini 2016; Verasammy and Cooper 2021). By contrast, pastoral care is free from the complexities of therapeutic directivity versus meta-therapeutic directivity that are intrinsic to humanistic counselling. As a result, pastoral care staff may have been more willing and able to offer directive input than counsellors during the study. These findings are a reminder that school-based humanistic counsellors should remain collaborative and responsive to their clients and prioritise contact and engagement.

The current study contributes to the growing knowledge base about helpful and unhelpful factors in SBC and provides new insights on young people’s perceptions of helpful and unhelpful factors in pastoral care. The finding that there are helpful factors in teacher-led PCAU suggests that school staff may benefit from mental health training to further improve the positive impact it has. The study also highlights the need to collaborate with both young people and teachers to design and deliver effective school-based interventions (Kendal, Keeley, and Callery 2014), and to ensure there are a range of options available.

A major limitation of this study is the lack of standardisation in the pastoral care condition. Although participants were asked to rate any pastoral care that they had received during the trial, may not have accurately recalled this. Another issue is that some participants may have emphasised a specific pastoral care teacher, whilst others may have rated pastoral care more generally. In addition, pastoral care is a highly heterogeneous service, therefore, there was a lack of standardisation in the support that participants would have commented on in their ESQ. Likewise, participants in the counselling group could access pastoral care in addition to counselling; if they used both services, it is possible that they did not accurately recall the type of support provided by each adult. Additionally, participants were not asked to review the coded data; thus, their comments could have been misinterpreted. Our sample was predominantly female, in the younger age ranges of adolescence, and excluded those at risk of serious harm to self or others. This means our findings may not be generalisable to males, older adolescents, and those with more serious mental health issues. We also did not collect data from the young people who chose not to participate in the study – to understand more about what young people find helpful and unhelpful about counselling, we need to understand what prevents some young people from accessing therapy in the first place. Further research with young people eligible to participate in counselling studies but who do not take up the offer of therapy is needed.

A further limitation was that we were unable to run a binary regression logistic analysis to explore the factors that predicted coding within each of our helpful and unhelpful themes: conditions, demographic characteristics (gender, age, ethnicity, disability) and the interactions between conditions and demographic characteristics, because as a 2*3*4*3*2 analysis, which also looked at interactions, we were concerned that we would have far too little power to adequately identify predictors. Future research would benefit from regression modelling of the all the observable characteristics of the participants.

More research is needed to understand how pastoral care is currently delivered in school to facilitate accurate comparisons with other mental health interventions, and to explore how both pastoral care and counselling can be improved.

We can conclude, however, that clients reported many benefits from access to both counselling and pastoral care. Out of these benefits, more were reported by those receiving counselling in addition to pastoral care. This suggests there are unique characteristics associated with receiving professional counselling that complement the pastoral care on offer in schools.

Acknowledgments

The first author is the recipient of a PhD studentship from the British Association for Counselling and Psychotherapy (BACP) and this study forms part of the wider PhD project. We thank the participants for sharing their experience and for taking the time to be part of this research.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request

Additional information

Notes on contributors

Helen Raynham

Helen Raynham is a Chartered Counselling Psychologist and a Psychology PhD candidate. She spent several years as a school-based counsellor and currently works in forensic mental health for the National Health Service. Helen is interested in school-based humanistic counselling and exploring therapeutic change mechanisms, particularly in relation to neurodivergence (principally Attention Deficit Hyperactivity Disorder).

Mick Cooper

Mick Cooper is Professor of Counselling Psychology at the University of Roehampton and a chartered counselling psychologist. He is an internationally recognised figure in the areas of humanistic, existential, relational and pluralistic therapy. Mick’s main areas of research interest have been the therapeutic relationship (including relational depth, shared decision making, and working with goals) and school-based humanistic counselling. He has published both qualitative and quantitative empirical studies, as well as reviews.

Jacqueline Hayes

Jacqueline Hayes Independent practitioner-researcher The Open Door Natural Health and Well-being Centre, Lewes, East Sussex. Dr Hayes is a relational psychotherapist and research psychologist. She specialises in qualitative methodologies and her primary research interests include child and adolescent therapy processes and relational and pragmatic aspects of hallucinations.

John Rae

John Rae is Reader in Psychology, University of Roehampton, UK. He studies talk and body movement in social interaction in informal settings and in service-related settings, in particular, psychotherapy. He is interested in interaction involving neurotypical participants and also in human diversity (e.g., people with aphasia or with an autism spectrum disorder). In addition to research on social interaction, he carries out research in these areas using a range of methods.

Peter Pearce

Peter Pearce is Faculty Head of Applied Social and Organisational Science at Metanoia Institute. He has a long background as a psychotherapist within NHS multi-disciplinary teams in disability and mental health, as a school-based counsellor, and as an academic and trainer. Peter is active nationally and internationally in the development of counselling and psychotherapy, with work focused in particular upon increasing access, free at the point of delivery. He was instrumental in the development of PCE-CfD, an evidence-based, NICE approved person-centred therapy, available in the UK National Health Service, is a member of the NHS Talking Therapies (IAPT) National Expert Advisory Group and chaired the QAA Benchmark Statement review group for these subject areas in 2022. He has published both qualitative and quantitative empirical studies