Variables affecting the school adaptation of secondary-school students who do not seek help: attachment, coping style, positivity, and prospects

ABSTRACT

School refusal is a severe social issue in many countries. For helping students who experience stress, multiple social supports including school counsellors and teachers have been provided. However, these supports may not be useful for ‘non-help-seekers’ – those who do not seek help even when they have serious problems. Given the scarce research on non-help-seekers’ cognition, it is essential to identify internal factors that affect non-help-seekers’ adaptation in school. The current study examined the associations between non-help-seekers’ school adaptation and attachment, stress-coping style, and three psychological variables: trust, positivity, and prospects. Two cross-sectional studies (N = 960, 49.5% boys; N = 658, 51.8% boys, respectively) revealed the importance of attachment, a possibility of interventions to psychological factors, and the limited effect of stress coping strategies when a student’s problem is severe. Several interventions including retrieval-induced forgetting and modelling were suggested as support for non-help-seekers before their problems become dire.

KEYWORDS:

• Help-seeking
• non-help-seekers
• adolescent
• school adaptation
• stress coping

Introduction

School refusal is a psychosocial problem defined by students’ absence from school due to various difficulties including academic performance, peer relationships, and other emotional distress (Heyne & Sauter, 2013). While truancy reflects non-attendance owing to antisocial behaviours or a lack of interest, school refusal reflects school non-attendance owing to emotional distress (Fremont, 2003). It is not considered a clinical diagnosis; rather, it is a symptom (Dabkowska, Dabkowska, Araszkiewicz, & Wilkosc, 2011). In many countries, both Eastern and Western, school refusal is a social issue that is stressful for students, families, and school personnel (e.g. Fremont, 2003; Inglés, Gonzálvez, García-Fernández, Vicent, & Martínez-Monteagudo, 2015; Park et al., 2015). In Japan, the number of students who refuse to attend school has continued to increase since the 1990s, and, in this time, approximately 120,000 secondary-school students did not regularly attend school without reasons such as illness or economic circumstances (The Ministry of Education, Culture, Sports, Science, and Technology (MEXT), 2018). This number means that, in every class, there are one or two students who do not attend school. Ample research has revealed the short- and long-term risks of school refusal such as poor academic achievement (e.g. Alexander, Entwisle, & Kabbani, 2001; Lamdin, 1996), lack of social skills and social isolation (e.g. Kearney, 2008; Place, Hulsmeier, Davis, & Taylor, 2002), and psychiatric diagnoses such as depressive disorder (e.g. Prabhuswamy, 2018; Prabhuswamy, Srinath, Girimaji, & Seshadri, 2007); therefore, prevention of school refusal is essential for Japan and other countries to reduce various risks that children may face.

The MEXT has attempted to prevent school refusal with various policies. For example, school counsellors are employed, teachers are trained concerning students’ mental health, and students are encouraged to seek help from others (MEXT, 2016). Moreover, improvement and expansion of free consultation service including through the Internet is also supported by national administration (Ministry of Education, Culture, Sports, Science, and Technology, 2016).

As these examples show, current school-refusal-prevention measures focus on enhancing environments where students can feel free to share their problems; although, the use of these services relies on students’ help-seeking behaviour. However, Ishikuma and Onose (1997) reported that 38% of Japanese secondary-school students are ‘non-help-seekers’ – those who do not seek help even if they have problems that they cannot solve by themselves. This implies that current social supports may not be useful for a proportion of struggling secondary-school students. A service gap between social supports and non-help-seekers is an important issue in other countries as well. For example, the World Mental Health Survey data from 21 countries described that 61% of the respondents who were suicidal had not sought treatment of any type in the previous year (Bruffaerts et al., 1999). Youth non-help-seekers exist regardless of ethnicity and culture, and they generally have a higher risk of suicidal behaviour (Goldston et al., 2008). Therefore, reconsideration of the policies and practices designed to prevent school-refusal is necessary, especially for non-help-seekers who have low school adaptation.

Reconsidering support for non-help-seekers

How can we support youth non-help-seekers before they develop severe problems? To answer this question, there are two matters to consider. The first is that the effect of help-seeking behaviour is unclear. Although the current measures provided by the MEXT are supported by the premise that help-seeking behaviours can help students earn more support and enhance their mental health, this premise has not been substantially supported by empirical evidence. In fact, most studies have revealed that help-seeking and school mental health are uncorrelated in contrast to some researchers insist that students who seek help become mentally healthier (e.g. Heerde & Hemphill, 2018; Kato, 2015; Rickwood, 1995). Therefore, the current efforts to build an environment that promotes help-seeking, in which students can seek consultation, may not effectively enhance non-help-seeking students’ mental health.

The second matter is that adolescent non-help-seekers are not uniform. There are roughly three types of non-help-seekers: those who do not need help because their problem is not severe (Wilsono & Deane, 2001), those who cannot seek help for various reasons such as perceived lack of support (Watanabe et al., 2012) and negative consultation expectations (Nagai & Arai, 2007), and those who can maintain their own mental stability without asking for help (Kuhl, Jarkon-Horlick, & Morrissey, 1997). It is important to understand that there are several types of non-help-seekers, and that the most effective support method will vary depending on type. The existence of adaptive non-help-seekers – those who can maintain their own mental stability without asking for help – suggests that there are other factors that influence the mental health of non-help-seekers. If it can be clarified there are protective internal factors that adaptive non-help-seekers possess, new interventions that support these factors can be considered. Thus, studies that compare internal factors between high- and low-adaptive non-help-seekers are valuable.

Possible factors that affect adaptation among non-help-seekers

This research began by exploring factors that affect differences between adaptive and non-adaptive non-help-seekers. One possible factor is attachment. Attachment theory (Bowlby, 1969) is a fundamental theory that explains a variety of psychological and developmental outcomes such as stress sensitivity (e.g. Howard & Medway, 2004), emotion regulation patterns (e.g. Thompson, 2008), emotional intelligence (Nanu, 2015), and well-being (e.g. Armsden & Greenberg, 1987). Attachment is measured by two variables: ‘avoidance of closeness’ and ‘anxiety’ (Collins & Read, 1990). Attachment is crucial because of its strong association with development; however, it is difficult to intervene in adolescents’ attachment style because it is considered a stable characteristic that is basically built during early childhood (Ainsworth, 1989; Waters, Weinfield, & Hamilton, 2000).

Since it would be advantageous for interventions to identify potential changeable variables, this research also considered three psychological variables that are related to subjective well-being – interpersonal trust (e.g. Catalino, Algoe, & Fredrickson, 2014), prospects (Zimbardo & Boyd, 2010), and positivity (Lauriola & Iani, 2016) – as well as stress coping style. Interpersonal trust was selected as a possible protective factor because it is related to attachment, and the association with multiple psychosocial outcomes including well-being has been supported (e.g. Catalino et al., 2014; Tokuda, Yanai, & Inoguchi, 2008). Individuals who have low trust in others are less likely to seek support when in need than are individuals who have high trust in others (Tokuda et al., 2008); thus, developing the capacity to trust others is essential for a successful social adjustment (Suedfeld et al., 2005). The current research posits that non-help-seekers who have high interpersonal trust will be mentally healthier than will those who have low interpersonal trust, and they may seek help if they really need it.

Prospects were also considered as a possible protective factor based on Socioemotional Selectivity Theory (SST; Carstensen, Isaacowitz, & Charles, 1999) and Time Perspective Theory (Zimbardo & Boyd, 2010). According to SST, people who perceive that they have a future become more active in acquiring knowledge and experience and tend to have social goals. Additionally, future time perspective in Zimbardo’s time perspective theory is related to health behaviours (Henson, Carey, Carey, & Maisto, 2006) and well-being (MacLeod, 2017). Both SST and the time perspective theory insist that human behaviours differ depending on how confident the individual is that a better future will be obtained. Therefore, this study defines prospects as a level of expectation for the individual’s potential in the future; those who are in a difficult situation in which they cannot rely on others for help may maintain their mental health with future goals if their prospects are high.

Positivity was selected as a potential factor because it is a key component of well-being (Lauriola & Iani, 2016); specifically, it has been defined as a general cognitive orientation that associates with dispositional optimism, self-esteem, and life satisfaction (e.g. Caprara et al., 2012). Non-help-seekers with high positivity are assumed to maintain mental health by thinking positively even in difficult situations.

Lastly, stress coping style was considered as a behavioural factor to compare with the psychological factors above. Because help-seeking is theoretically a stress coping strategy, the effect of other coping strategies should also be considered (Carver & Connor-Smith, 2010). In other words, adaptive- and non-adaptive non-help-seekers could be using different coping strategies. The COPE (Carver, Scheier, & Weintraub, 1989), the most frequently used scale in research on stress coping (Kato, 2015), divides various coping strategies into three aspects: problem-focused strategies, emotion-focused strategies, and less useful strategies. In general, problem-focused strategies are regarded as adaptive (Carver & Connor-Smith, 2010). Active coping from problem-focused strategies was selected as a strategy that shows the strongest positive correlation with mental health, while behavioural disengagement from less useful strategies was selected because of its strong negative correlation with mental health (Carver et al., 1989; Kato, 2015). Following previous studies, it was hypothesized that adaptive non-help-seekers would use more active coping and less behavioural disengagement than non-adaptive non-help-seekers.

Present study

This study examined the potential protective role of several factors (i.e. attachment, interpersonal trust, prospects, positivity, and coping style) on the mental health of adolescent non-help-seekers to suggest a new support method. Two hypotheses were built in line with previous literature. First, this research posited that stable attachment with low avoidance and anxiety would protect the mental health of non-help-seekers. Second, this study posited that non-help-seekers who have high interpersonal trust, prospects, positivity, and an active coping style will display better school adaptation as compared to their counterparts.

The associations between attachment and school adaptation among non-help-seekers were investigated in Study 1, and the associations between other potential factors and adaptation were investigated in Study 2. In addition to examining these hypotheses, this research considered whether the effects of the variables differ depending on the severity of students’ problems.

Methods

Design and participants

Two cross-sectional studies were conducted. Study 1 participants (N = 960; 49.5% men (one did not answer)) were secondary-school students who answered an anonymous self-report questionnaire in October 2017. The participants were from two public secondary schools in different prefectures in Japan: school A is in a rural area of northern Japan (n = 450), and school B is in the Tokyo metropolitan area (n = 510).

Study 2 participants (N = 658; 51.8% men (one did not answer)) were secondary-school students who completed a questionnaire in September or October 2018. The participants were from two public secondary schools in the same prefecture: school A was the same school as Study 1 (n = 451), and school C (n = 207) was in an urban area of the city. All participating schools were selected because of their generality, accessibility, and the variety of students’ socioeconomic status and academic ability. The surveys were conducted by class after classroom teachers explained that participation was not compulsory. There was no incentive for individual participants. The effective response rate was 94.1% in Study 1 and 94.5% in Study 2.

Measures

In Study 1, the questionnaire measured participants’ sense of school adaptation, help-seeking behaviour, attachment, perceived social support, and cognitive assessment of their problems such as severity and controllability. Study 2 measured several stress-coping strategies, trust of others, positivity, and prospects, in addition to the items collected in Study 1. Attachment was not measured. First, participants answered the items measuring the sense of school adaptation. Then they were asked whether they experienced trouble in the past month. Only those who reported some trouble answered the following questions, which included questions regarding help-seeking behaviours.

Sense of school adaptation

The School Adaptation Scale by Furuichi and Tamaki (1994) is a widely used 10-item scale in Japan for measuring one’s degree of enjoyment with school life. To reduce participants’ burden, the current research used five items that showed high factor loading in the preliminary surveys: (a) I look forward to going to school, (b) every day passes quickly because school is fun, (c) I want to go to school even if I feel a little bad, (d) there are many enjoyments at school, and (e) I like this school. This shortened scale showed good reliability in the current study (α = .90–.91). Participants answered about the past month using a 4-point scale: 1 (never true for me) to 4 (always true for me).

Help-seeking behaviour

Four items each measuring emotional- and instrumental-support seeking were derived from the Japanese version of the COPE (Otsuka, 2008). According to the COPE, emotional-support seeking is a strategy to acquire moral support, sympathy, or understanding, while instrumental-support seeking is a strategy to acquire advice, assistance, or information (Carver et al., 1989). This research regarded these two average points as help-seeking scores (α = .87). Participants who had experienced any problematic experiences in the past month answered how often they used these coping strategies with a 4-point scale: 1 (never did so) to 4 (always did so).

Attachment

The Experiences in Close Relationships Inventory (ECR; Brennan, Clark, & Shaver, 1998) is an attachment scale that is used worldwide in diverse countries and regions. This study used the Japanese version of the ECR-GO (Nakao & Kato, 2004), which is a measure of attachment towards general others across two subscales: anxiety and avoidance of closeness. This study used a four-point Likert-type response scale, with higher scores representing more of the construct. The anxiety subscale contained questions such as, ‘I worry that I might become alone’ (α = .93). The avoidance subscale contained questions such as, ‘I do not like to build close relationships with others’ (α = .76).

Perceived social support

After the instruction sentence – ‘If you have a problem, who do you think you can talk to about it? – participants reported the number of consultation partners from a list of multiple choices including parents, siblings, teachers, friends, and acquaintances on the Internet. These options were developed by referring to relevant studies such as Ishikuma (1999); Honda, Arai, and Ishikuma (2008); and Honda (2013). The validity of the selection items was confirmed by several secondary-school teachers.

Cognitive assessment of trouble

The transactional theory of stress and coping (Lazarus & Folkman, 1984) suggests cognitive appraisals of problems influence one’s response to troublesome events. According to Lazarus and Folkman (1984), cognitive assessment can be divided into a primary appraisal and a secondary appraisal. The primary appraisal, severity, means how severely trouble affects one’s daily life; and the secondary appraisal, controllability, means how easily it can be solved.

The Cognitive Appraisal Questionnaire (Okayasu, 1992) is a widely used assessment in Japan that includes 20 items. Following previous research that examined secondary-school students (e.g. Miura, Shimada, & Sakano, 1995), this study used eight items that are appropriate for youths, including (a) it threatens my school life and (b) I know how to solve that problem (α = .83 for severely of problems, and α = .91 for controllability). Participants who experienced any trouble in the past month evaluated it on a 4-point scale: 1 (never felt so) to 4 (always felt so).

Trust, prospects, and positivity

The three psychological variables examined in Study 2 were all measured by widely used measures. First, interpersonal trust was measured by the Sense of Basic Trust Scale (Tani, 1998). The scale was developed based on Rasmussen’s (1964) Ego Identity Scale, and it includes items such as ‘I can expect help from people around me when I am in trouble’ and ‘I think people are credible in general’ (α = .77).

Second, the items that measure prospects were derived from the Japanese version of the Future Time Perspective Scale (Ikeuchi & Osada, 2013), which consists of two dimensions: open-ended time perspective and limited time perspective. This study used items from the open-ended time perspective, which focuses on opportunities expected for the future including ‘many opportunities await me in the future’ and ‘my future is filled with possibilities’ (α = .80).

Lastly, the items that measure positivity were derived from the Stress Mindset Measure (youth version; Park et al., 2018), which consists of two dimensions: a ‘stress-is-enhancing’ mindset and a ‘stress-is-debilitating’ mindset. This study used items from the former, which measures positive thinking towards distress, such as ‘experiencing stress facilitates my learning and growth’ and ‘the effect of stress is positive and should be utilized’ (α = .69). Although this internal reliability score seems to be lower than others, it was regarded as fair since the original scale contains only three items in the subscale (Park et al., 2018), and the number of scale items affects the estimation of the adequacy of internal consistency (Ponterotto & Ruckdeschel, 2007).

Active and passive coping strategies

Active and passive coping strategies were measured using the Japanese version of the COPE (Otsuka, 2008) – the same as help-seeking behaviours. Since the COPE contains many coping strategies, this research employed two representative strategies: active coping and behavioural disengagement. Active coping was selected as an active strategy because of its strong positive correlation with mental health, while behavioural disengagement – a passive strategy – was selected because of its strong negative correlation with mental health (Carver et al., 1989; Kato, 2015). Both coping strategies showed sufficient reliability in the current study (α = .85–.86). Participants who experienced any trouble in the past month answered how often they used these coping strategies on a 4-point scale: 1 (never did so) to 4 (always did so).

Data analyses

The data from students who experienced trouble in the last month and whose support-seeking scores were below average were used for analysis as the sample of students who tend not to seek help. They were then divided into high- or low-school-adaptation groups. Independent t-tests were used to test for group differences in mean scores on attachment and perceived social support in Study 1. The same methods were used for the three psychological variables and active- and passive-coping strategies in Study 2. Lastly, multiple regression analyses were performed to examine the simple main effect of each variable and the interaction effects with problem severity on school adaptation.

All analyses were conducted using the ‘pequod’ package in R version 3.4.3. The accepted level of statistical significance was p < .01.

Results

The data from 288 (50.7% men) and 228 (45.6% men) participants from Studies 1 and 2 were analysed, respectively. Descriptive statistics and correlations among all variables are presented in Table 1. In Study 1, the mean scores for school adaptation were 3.29 (SD = .13) in the high-adaptation group (n = 165, 49.7% men) and 2.13 (SD = .22) in the low-adaptation group (n = 123, 52.0% men). In Study 2, the mean scores for school adaptation were 3.31 (SD = .18) in the high-adaptation group (n = 130, 45.4% men), and 2.07 (SD = 0.24) in the low-adaptation group (n = 98, 45.9% men). There were no sex differences between groups in both studies.

Table 1. Descriptive statistics and correlations among all variables (upper study 1, lower study 2)

		1	2	3	4	5	6	7	8	9	10	11	12	13
1	Sense of school adaptation		.10*	.18***	−.15***	−.39***	.23***	−.13***	.31***	-	-	-	-	-
2	Emotional help-seeking	.28***		.77***	.13**	−.33***	.27***	.17***	.14***	-	-	-	-	-
3	Instrumental help-seeking	.35***	.77***		.00	−.33***	.30***	.03	.24***	-	-	-	-	-
4	Anxious (attachment)	-	-	-		.14***	−.08*	.37***	−.17***	-	-	-	-	-
5	Avoidance (attachment)	-	-	-	-		−.28***	.13***	−.27***	-	-	-	-	-
6	Perceived social support	.24***	.26***	.28***	-	-		.03	.20***	-	-	-	-	-
7	Severity of problem	−.18***	.11*	.00	-	-	−.06		−.15***	-	-	-	-	-
8	Controllability of problem	.26***	.15**	.25***	-	-	.21***	−.28***		-	-	-	-	-
9	Interpersonal trust	.54***	.38***	.36***	-	-	.22***	−.25***	.30***		-	-	-	-
10	Prospects	.46***	.28***	.29***	-	-	.22***	−.21***	.34***	.54***		-	-	-
11	Positivity	.50***	.40***	.43***	-	-	.27***	−.12*	.32***	.56***	.52***		-	-
12	Active coping	.34***	.39***	.50***	-	-	.27***	−.20***	.48***	.42***	.41***	.45***		-
13	Behavioural disengagement	−.16**	.15**	.07	-	-	−.07	.33***	−.23***	−.17***	−.19***	−.09	−.21***
Study 1
	M	2.93	2.52	2.50	1.71	2.42	2.13	2.75	2.80	-	-	-	-	-
	SD	.66	.90	.86	.60	.54	1.34	.64	.72	-	-	-	-	-
Study 2
	M	3.06	2.52	2.49	-	-	2.40	1.93	2.89	2.88	3.12	2.81	2.85	1.92
	SD	.74	.87	.82	-	-	1.33	.84	.90	.77	.84	.77	.79	.87

M = mean; SD = standard deviation. *p < .05, **p < .01, ***p < .001.

Study 1

T-tests revealed that avoidance of closeness from the attachment scale and perceived social support were significantly different between the two groups (Table 2). Low-adaptation respondents displayed significantly higher avoidance than did high-adaptation respondents. Regarding social-support accessibility, high-adaptation participants averaged significantly more consultation partners than did low-adaptation participants. Frequently, consultation partners were reported as parents or friends, not school counsellors. No difference was observed concerning the anxiety dimension of attachment.

Table 2. T-test results comparing high- and low-adaptation on attachment and the number of consultants

	Low-adaptive non-help-seekers		High-adaptive non-help-seekers
	M	SD	M	SD	t	df	p
Sense of school adaptation	2.13	.22	3.29	.13	22.71	223.34	< .001
Anxious (attachment)	1.88	.60	1.82	.62	.86	249.04	.391
Avoidance (attachment)	2.83	.50	2.49	.50	5.62	247.96	< .001
Perceived social support	1.48	1.27	2.02	1.69	4.00	277.66	< .001
Severity of problem	1.66	.61	1.83	.62	2.59	256.25	.010
Controllability of problem	2.43	.74	2.84	.72	4.72	249.28	< .001

M = mean; SD = standard deviation, df = degrees of freedom.

Study 2

Difference between high- and low-adapted non-help-seekers on trust, prospects, positivity, and coping strategies

A significant difference between the groups was found regarding all three psychological variables. High-adaptation respondents had significantly higher scores for trust, prospects, and positivity. Although the gaps were smaller than for psychological variables above, stress coping strategies (i.e. active coping and behavioural disengagement) also showed significant group differences. Regarding social-support accessibility, a significant difference was found as in Study 1, and high-adaptation participants had significantly more consultation partners than did low-adaptation participants (Table 3).

Table 3. T-test results comparing high- and low-adaptation on trust, prospects, positivity, and coping strategies

	Low-adaptive non-help-seekers		High-adaptive non-help-seekers
	M	SD	M	SD	t	df	p
Sense of school adaptation	2.07	.24	3.31	.18	19.95	190.84	< .001
Interpersonal trust	2.14	.63	2.84	.72	7.87	221.54	< .001
Prospects	2.38	.93	3.13	.77	6.49	184.95	< .001
Positivity	2.12	.65	2.77	.71	7.19	218.31	< .001
Active coping	2.39	.76	2.70	.81	2.97	215.75	.003
Behavioural disengagement	2.07	.88	1.68	.79	3.43	195.66	.001
Perceived social support	1.80	1.19	2.42	1.27	3.77	215.72	< .001
Severity of problem	2.00	.84	1.72	.80	2.54	185.47	.012
Controllability of problem	2.55	.93	2.91	.86	2.95	184.04	.004

M = mean; SD = standard deviation; df = degrees of freedom.

Difference of influence of examined variables by the severity of trouble

Multiple regression analyses were conducted to examine the relationship between participants’ sense of school adaptation and the independent variables. In these analyses, the interaction effect of the variables and severity of problems towards school adaptation was also examined. Table 4 summarizes the results. Trust, prospects, and positivity scores were all positively and significantly correlated with school adaptation; however, neither active coping nor behavioural disengagement showed a significant main effect on adaptation. No interaction effect of the variables and severity of problems was confirmed.

Table 4. Results of multiple regression analyses

		t	β	p	F	df	p	adj. R^2
Interpersonal trust					12.98	7.00	< .001	.28
	Grade	1.35	−.08	.179
	Sex	.31	.02	.753
	Trust	7.16	.46	< .001
	Perceived social support	1.89	.11	.059
	Severity of problem	.02	.00	.984
	Controllability of problem	1.75	.11	.082
	Trust* Severity	.12	.01	.901
Prospects					10.32	7.00	< .001	.23
	Grade	1.94	−.12	.054
	Sex	.13	.01	.898
	Prospects	5.91	.41	< .001
	Perceived social support	1.91	.12	.057
	Severity of problem	.88	−.06	.378
	Controllability of problem	1.28	.09	.203
	Prospects* Severity	1.14	−.07	.255
Positivity					14.31	7.00	< .001	.30
	Grade	1.71	−.10	.088
	Sex	.40	.02	.691
	Positivity	7.65	.46	< .001
	Perceived social support	1.88	.11	.061
	Severity of problem	1.71	−.10	.088
	Controllability of problem	1.54	.10	.125
	Positivity* Severity of problem	.84	−.05	.402
AC					5.22	7.00	< .001	.12
	Grade	1.45	−.10	.148
	Sex	.46	.03	.644
	AC	1.29	.10	.198
	Perceived social support	2.41	.16	.017
	Severity of problem	1.77	−.12	.078
	Controllability of problem	2.32	.18	.022
	AC* Severity of problem	1.20	−.08	.230
BD					5.30	7.00	< .001	.12
	Grade	1.40	−.09	.164
	Sex	.96	.06	.340
	BD	1.89	−.14	.060
	Perceived social support	2.62	.17	.010
	Severity of problem	1.03	−.07	.305
	Controllability of problem	3.11	.22	.002
	BD* Severity	.46	−.03	.644

AC = active coping; BD = behavioural disengagement; df = degrees of freedom; adj. = adjusted. *p < .05, **p < .01, ***p < .001.

Additionally, simple slopes tests were preliminarily performed to follow-up the interaction terms even though there was no significance regarding the interaction effect. Some indications about the difference of influence of the examined variables by problem severity were found. Trust, prospects, and positivity were highly associated with school adaptation regardless of problem severity. However, coping strategies were differently associated with adaptation depending on problem severity. Specifically, for low-severity problems, high levels of active coping were associated with increased school adaptation. In contrast, for high-severity problems, there was no association between active coping and adaptation. Moreover, high levels of behavioural disengagement were not associated with school adaptation for low-severity problems, while behavioural disengagement was negatively associated with school adaptation for high-severity problems. The associations between each variable and school adaptation as moderated by problem severity are shown in Figures 1 and 2.

Figure 1. Association between psychological variables and sense of school adaptation as moderated by problem severity

Note. (a) interpersonal trust, (b) prospects, (c) positivity. ***p < .001.

Figure 2. Association between coping strategies and sense of school adaptation as moderated by problem severity

Note. (a) active coping, (b) behavioural disengagement. *p < .05.

Discussion

Internal factors that differ between adaptive and non-adaptive non-help-seekers

Study 1 identified differences in attachment between high-adaptive and low-adaptive non-help-seekers. As a result, among the two axes defining attachment, only avoidance showed a significant difference, while there was no significant difference in anxiety. However, the sample displayed low anxiety overall; therefore, it may not have been possible to clarify the natural differences that should have been seen. School refusal often co-occurs with separation anxiety – a similar concept to attachment anxiety (King, Ollendick, & Tonge, 1995); in detail, separation anxiety disorder is reported to occur in up to 80% of children with school refusal (Masi, Mucci, & Millepiedi, 2001). Because this survey was conducted among students who were attending school, there is the possibility that a survey of clinical groups that had refused to attend school may result in significant differences in anxiety. In any case, as with previous studies, a part of the influence of attachment was also confirmed in the adaptation of non-help-seekers (e.g. Armsden & Greenberg, 1987).

Study 2 examined the differences between three psychological variables: interpersonal trust, prospects, and positivity. As a result, all three variables showed significantly higher values for highly adapted non-help-seekers than those who were less adapted. In addition, the number of consultants, considered as perceived social support, was significantly higher for non-help-seekers who were highly (vs. not) adapted in both Studies 1 and 2.

These results indicate a possibility to improve individuals’ psychological sense of adaptation by performing interventions on psychological variables such as trust, prospects, and positivity instead of encouraging help-seeking. Regarding the number of consultants, it was suggested that mental stability might be obtained by perceiving that there are people who can rely on them when necessary, even without taking actual consulting actions. This research also revealed a difference in stress coping strategies. Although the difference was small compared to the above-mentioned psychological variables and the number of counsellors, the adaptive non-help-seekers used the active coping strategy and less frequently used the behavioural disengagement than did the non-adaptive participants.

Difference in the appropriate approach depending on problem severity

The severity of individuals’ distress is important because the degree of need for assistance varies with the reactive behaviours (Lazarus & Folkman, 1984). Study 2 examined whether the effect of each variable on school adaptation would differ depending on problem severity. Multiple regression analyses revealed that the interaction effect on the school adaptation of each variable and the severity of distress was non-significant. However, a simple regression analysis of the association between coping strategy and adaptation revealed differences between the high and the low severity group. Specifically, active coping and adaptation were positively correlated when problem severity was low, while no correlation was found when it was high. Furthermore, although the group with low problem severity did not show a decrease in adaptation even if behavioural disengagement was high, the group with high problem severity showed significantly declined adaptation when they used the behavioural disengagement strategy. In other words, non-help-seekers who have serious distress are less likely to be positively affected by active coping behaviour, but more likely to be negatively affected by behavioural disengagement. This means that interventions that encourage active coping may be effective only for students with low problem severity. In contrast, the three psychological variables – trust, prospects, and positivity – were hardly affected by problem severity. Therefore, designing interventions for these variables may be effective, regardless of severity.

Suggestions to policymakers and schoolteachers

The primary goal of this study was to identify internal factors that protect the mental stability of adolescent non-help-seekers, and to provide suggestions for improving current school-refusal-prevention measures that are biased towards promoting help-seeking. Based on the findings, four intervention ideas are suggested for schoolteachers and policymakers.

First, it is important for teachers to tell students that they are welcome to consult whenever there is a problem, and to respond sincerely if a student asks for help. This research found that students can be adaptive when they recognize there are multiple people that they can rely on when in need, even if they do not seek help. This research also confirmed the protective effect of interpersonal trust. Since there is no simple intervention to improve interpersonal trust, teachers need to project that they are interested in students and willing to consistently help them. At the administration level, policymakers may better consider smaller class sizes to foster students’ trust towards teachers. This proposal is supported by some evidence that a better teacher-student relationship enhances students’ participation in school and academic involvement (e.g. Roorda, Koomen, Spilt, & Oort, 2011; Yang, Bear, & May, 2018), and that the effect of higher teacher-student interactions is greatest in classrooms with fewer students (e.g. Allen et al., 2013; Blatchford, Bassett, & Brown, 2011).

The second suggestion is to provide stress coping strategies for students. This study showed that adaptive non-help-seekers used more active coping strategies and less behavioural disengagement strategy than did non-adaptive students. This indicates that enhancing students’ stress coping skills would be helpful to maintain adaptation. For example, classroom interventions, brochures with coping tips, and daily knowledge delivery about coping from teachers to students should be considered. However, it is also important to encourage students to seek help if their problems are too severe to solve alone, as this study suggested that problem severity has varied effects on students’ adaptation.

The third is an intervention that enhances students’ positivity. Many previous studies found that intrusive rumination of a stressful event is significantly related to negative thoughts and even posttraumatic stress disorders (e.g. Ehring & Ehlers, 2014; Zhang, Xu, Yuan, & An, 2018). Because no one can positively change others’ mindsets, saying ‘think positively’ or ‘stop thinking about it’ is not appropriate. One possible idea is utilizing retrieval-induced forgetting (RIF, Marsh, Edginton, Conway, & Loveday, 2018). RIF is a phenomenon in which people forget a memory trace when a rival memory trace is retrieved (Anderson, Bjork, & Bjork, 1994). It has been put forward as a potential mechanism by which the positivity bias in memory occurs (Storm & Jobe, 2012). For example, repeatedly remembering a positive experience, such as receiving support from others, can make a related negative memory fade. At school sites, teachers may incorporate RIF with students with negative thoughts.

Lastly, the use of modelling to enhance prospects is suggested. There is still no practice nor research on interventions that address adolescents’ prospects. However, considering said interventions is critical having a lack of prospects is a strong risk factor for suicide and violent crimes (e.g. Hall, Platt, & Hall, 1999; Stoddard, Henly, Sieving, & Bolland, 2010). Modelling the experiences of others who have overcome similar painful events should be considered. Specifically, guest speakers could be utilized, or online intervention such as the ‘It Gets Better’ project, which included former President Obama as a contributor (It Gets Better project, n.d.), could be disseminated.

This suggestion was derived in line with SST, however, there is another important theory to be considered, the Mindset Theory by Dweck (2006). According to mindset theory, there are two types of mindset: a growth mindset and a fixed mindset (Dweck, 2006). A growth mindset is a belief that abilities can be developed; whereas a fixed mindset is a belief that abilities are fixed (Dweck, 2006). There is rich evidence that people who have a growth mindset achieve more because they tend to exert a higher level of effort with a belief that efforts enhance their abilities. In contrast, people who have a fixed mindset do not believe that efforts will enhance their abilities, and therefore exert less effort (e.g. Schleider & Weisz, 2017; Yeager et al., 2019). Taking a perspective of mindset theory, if a student’s mindset is fixed, there may not be motivated to learn even if they have a high prospect. Therefore, one thing to keep in mind when utilizing modelling is ensuring that interventions encourage a growth mindset. In line with mindset theory, a person who overcame distress with effort or strategies may be more appropriate as a model than a person who resolved similar adversity without effort or who used innate resources such as parents’ economic power. Although a positive correlation between a future time perspective and a growth mindset has been found (Kwon, 2018), related studies are scarce. Further investigation regarding the association between prospects and mindset theory is needed.

Thus, although there are various possible ways of intervention, the important thing is that researchers and stakeholders seek to understand the diversity of adolescent non-help-seekers and their environment, and that interventions are not biased towards one method like simply encouraging help-seeking.

Limitations and future discussions

This study had three noteworthy limitations. The first limitation concerns sample selection. Although participants were from public schools located in different areas, the sample cannot be regarded as random. Other surveys with a larger sample or clinical sample are needed to enhance the generalizability of the results. Second, a cross-sectional design is a limited investigation method that cannot be used to examine long-term adaptation differences among non-help-seekers. For a deeper understanding of the effects of protective factors, longitudinal or retrospective studies with adult samples are required. Third, this research only included a limited number of factors. Since there are numerous potential internal and external factors, qualitative studies using interview surveys may help to identify other factors and to clarify any possible relationships.

Conclusions

This study contributes to the understanding of protective factors that differ between high- and low-adapted adolescent non-help-seekers and suggested improved support methods for them. These findings confirm the importance of attachment for school adaptation and indicate that interventions into the psychological variables of trust, prospects, and positivity can enhance school adaptation. The current study also revealed that interventions that include stress coping strategies may only be effective for students whose problems are not severe. Finally, ensuring there are enough consultation partners available can improve non-help-seekers’ adaptation, even if they rarely seek actual help. These findings offer new insights into more effective methods for supporting adolescent non-help-seekers before their problems become very serious.

Disclosure statement

No potential conflict of interest was reported by the author.

Data availability statement

The data that support the findings of this study are available from the corresponding author, K. A., upon reasonable request. https://www.researchgate.net/profile/Kyoko_Amai

Additional information

Funding

This work was supported by the Japan Society for the Promotion of Science under Grant number 18J20397.

Notes on contributors

Kyoko Amai

Kyoko Amai is a PhD student in school psychology at the University of Tokyo. She is currently conducting her dissertation project that considers support methods for adolescent non-support-seekers. Her research interests centre primarily on issues of access and quality of care for adolescents who do not seek nor accept support from others. In particular, she focuses on identifying internal and external protective factors and lifelong development influences of adolescent non-support-seekers. Her project aims to suggest innovative support methods that approach non-support-seekers’ internal changes besides current social policies. She was awarded a Research Fellowship for Young Scientists (DC1) from the Japan Society for the Promotion of Science. She also works as a project researcher at the Division of Evidence-Based Research in the Centre for Advanced School Education and Evidence-based Research, the University of Tokyo. In the past, she had experiences to work as a school health and nursing teacher at each two public primary and secondary schools, and as an assistant professor in Chiba University, Japan.