The protective role of self-compassion on test anxiety among adolescents

ABSTRACT

A growing body of work has begun to identify the benefits of self-compassion in promoting positive mental health and wellbeing for adolescents. Test anxiety is a form of anxiety elicited in response to examinations, which can negatively impact academic achievement and wellbeing. The current pilot study explored the effect of self-compassion on test anxiety among adolescents, while controlling for the effects of gender and general anxiety. Participants were adolescents (n = 47), aged 16 to 17 years old, enrolled in a post-primary school in Northern Ireland and studying for qualifications after compulsory education. Hierarchical regression analysis found that higher levels of self-compassion were related to reduced test anxiety. However, female adolescents and those with higher levels of general anxiety reported greater test anxiety. The findings highlight the potential value of self-compassion as a protective factor against test anxiety for young people in schools. The implications for practitioners supporting young people with mental health and wellbeing difficulties are considered.

KEYWORDS:

• Self-compassion
• test anxiety
• general anxiety
• protective factor
• adolescent

High levels of anxiety are common among young people (Cummings et al., 2014). Youth may be particularly vulnerable to developing anxiety difficulties during adolescence, as a result of marked physiological and psychological changes, as well as exposure to novel social and environmental stressors (Costello et al., 2003). It is estimated that approximately one in three adolescents, aged 13 to 18 years old, may struggle with an anxiety disorder (Merikangas et al., 2010). It has been suggested that these high prevalence rates of anxiety reflect a growing issue within public health (Johnson & Greenberg, 2013).

School and examination-based pressures are frequently identified as a particular source of anxiety for young people (Putwain et al., 2014). For example, worrying about upcoming examinations has been linked to sleep disturbances, unavoidable guilt when not studying, and feeling overwhelmed when attempting to revise (Chamberlain et al., 2011). Concerningly, examination pressures may also be an antecedent to death by suicide (Rodway et al., 2016). It is therefore important for anxiety around examinations to be taken seriously (Nicholson & Putwain, 2019). Test anxiety, also known as examination anxiety or assessment anxiety, is a form of anxiety elicited in response to a perception of threat and fear of failure in examinations (Putwain & Daly, 2014). Test anxiety can adversely impact both academic achievement and wellbeing. As such, there is a need to identify factors that protect against the onset test anxiety for young people in schools. The current pilot study explored the effect of self-compassion on test anxiety among adolescents.

Adolescents and test anxiety

Test anxiety presents with cognitive (i.e. catastrophic and worry-related thoughts about failure), affective (i.e. panic and fear), and physiological components (i.e. elevated heart rate and nausea; Segool et al., 2014). Test anxiety exists across ages and educational contexts, including primary, post-primary, and higher education (Von der Embse et al., 2018). Test anxiety typically escalates with age, as youth negotiate a growing number of school-based pressures to achieve academic success throughout childhood and adolescence and display a heightened sensitivity to evaluative situations (Blakemore & Mills, 2014; Klinger et al., 2015). These young people are also exposed to more advanced learning tasks and accumulate a greater number of past failures (Zeidner, 2007). Similarly, adolescents take compulsory examinations in post-primary school of which the results are perceived to have much more importance (Von der Embse & Hasson, 2012).

Female adolescents typically report being more test-anxious than males (Putwain & Daly, 2014). Test anxiety may also overlap with other anxiety difficulties, such as clinical anxiety, social anxiety, and specific phobias (Herzer et al., 2014; LeBeau et al., 2010; Putwain & von der Embse, 2021). This overlap across anxiety domains may be explained by the integrative network approach (Borsboom, 2017). This approach argues that anxiety difficulties are not the inherent cause of symptoms, but rather are the subsequent presentations that result from a pattern of interaction between symptoms (Heeren & McNally, 2016). That is, anxiety symptoms exist on an interconnected network. Stronger connections between two symptoms increases the likelihood that one will activate the other. For example, the fear of meeting new people associated with social anxiety has the potential to generalise across other social contexts, including avoiding sitting an examination (Heeren & McNally, 2018). In particular, many general anxiety symptoms, including intrusive and worrying thoughts about failure, may overlap with other anxiety difficulties (Heeren et al., 2018). As such, activation of general anxiety symptoms could spread across the network and influence the emergence of test anxiety among youth. Previous research has also identified general anxiety as the ‘hidden generator’ that renders individuals more vulnerable to anxiety difficulties (Gidron, 2013).

Test anxiety has been shown to adversely impact academic achievement (Von der Embse et al., 2018). That is, young people with greater test anxiety are more likely to perform poorer in examinations (Putwain & Daly, 2013). The effects of test anxiety on examination performance has been found to influence young people’s motivation to engage with school, which can have long-term implications for employment and access to further education (Putwain & Aveyard, 2016). Furthermore, test anxiety can contribute to the development of mental health and wellbeing difficulties (Akinsola & Nwajei, 2013; Herzer et al., 2014). For example, youth with higher test anxiety reported reduced life satisfaction and less positive emotional states (Steinmayr et al., 2016). The negative effects of test anxiety on wellbeing have also been shown to adversely influence adolescents’ life trajectories, such as exacerbating the risk of developing mental health difficulties in later life (Esbjørn et al., 2012). Yet, the education system itself can play a significant role in cultivating competitive and pressurised learning environments that prioritise performance and academic success, and in turn, induce high levels of anxiety among young people. Adolescents struggling with test anxiety may also develop a sense of oppression and powerlessness within school settings as a result of both their level of vulnerability and existing inequalities in systemic power differentials, which may further exacerbate mental health and wellbeing needs. As such, there is a need to identify factors that may buffer against onset of test anxiety to promote thriving and growth among adolescents in schools. In particular, self-compassion has been highlighted as an effective approach for coping with difficulties during adolescence (Bluth & Neff, 2018).

The role of self-compassion

Self-compassion involves relating to the self during challenging times with a sense of care and warmth (Neff, 2012). Self-compassion contains three interacting components, which reflect affective, perceptual, and cognitive responses to coping with distress. These components include self-kindness, mindfulness, and a sense of common humanity (Neff, 2003).

A growing body of work has begun to identify benefits of self-compassion for adolescents. That is, self-compassion has been shown to mitigate against negative mental health and psychosocial outcomes (Marsh et al., 2018), such as anxiety (Muris et al., 2016), depression (Castilho et al., 2017), stress (Bluth & Blanton, 2015), aggression (Barry et al., 2015), alcohol misuse (Tanaka et al., 2011), self-harm (Xavier et al., 2016), and suicidal ideations (Zeller et al., 2015).

Self-compassion has also been reported to promote positive youth outcomes related to improved wellbeing (Bluth et al., 2017), including increased self-esteem (Barry et al., 2015), life satisfaction (Bluth & Blanton, 2015), resiliency (Bluth et al., 2018), distress tolerance (Bluth et al., 2017), social connectedness (Neff & McGehee, 2010), and a perception of belonging to a community (Akin & Akin, 2015). In light of these benefits, self-compassion may act as a protective factor for the challenges that young people encounter during adolescence. Furthermore, fostering greater compassion, including self-compassion, has been highlighted as the potential antidote for supporting young people struggling with mental health and wellbeing needs in school settings (Al-Ghabban, 2018; Welford & Langmead, 2015).

To date, no studies have directly explored the relationship between self-compassion and test anxiety among adolescents. However, young people suffering with test anxiety often experience heightened self-criticism (Cunha & Paiva, 2012). As such, the potential to establish a more positive approach for relating to the self when dealing with school demands may have important implications for adolescent wellbeing. Adolescence can provide an opening for the development of more constructive relationships with the self to deal with difficulties, failures, and challenges in everyday life due to the physical, psychological, and social changes that take place at this time (Roeser & Pinela, 2014). The current pilot study therefore aims to explore the effects of self-compassion on test anxiety among adolescents while controlling for the influence gender and levels of general anxiety.

Method

Participants

The post-primary education sector in Northern Ireland (NI) can be conceptualised as comprising of grammar and non-grammar schools. Grammar schools engage in a practice of academic selection. That is, transitioning primary school children are required to take entrance examinations and attain a prerequisite grade to secure admission. These grammar schools tend to place an increased emphasis on academic achievement (Ingram, 2009) and perform better in state examinations in comparison to non-grammar schools (Northern Ireland Department of Education, 2019).

Participants were 47 adolescents, aged 16 to 17 years (72% female, 28% male), recruited from a grammar school in NI (56% opt-in participation rate). The participating school boasted a high level of academic success in examinations and served students from more affluent socio-economic backgrounds (e.g. fewer than 20% eligible for free school meals). All adolescents were enrolled in the first year of a two-year Advanced Level (A-Level) qualification, which is optional after compulsory education and often required to gain admission into university.

Procedure

The pilot study adopted a cross-sectional design. Ethical approval for the study was obtained from Queen’s University Belfast. The school principal provided written permission for the research to be carried out within their school. Packs containing information letters and consent forms were then sent home with all adolescents in the eligible year group. Informed written consent was received from the young person and their parents before taking part in the study. Adolescents then completed a paper-based response booklet in school, which contained a series of questionnaires that measured test anxiety, general anxiety, and self-compassion. The booklet took around 10 minutes to complete.

Measures

Test anxiety

Test anxiety was assessed using the Test Anxiety Inventory (TAI; Spielberger, 1980). The TAI consisted of 20 items that assessed the cognitive, affective, and physiological aspects of test anxiety, which can be combined for a total test anxiety score. Adolescents responded to each item using a 4-point Likert scale ranging from 0 = almost never to 3 = almost always. All items were summed to calculate a total test anxiety score (min = 0, max = 60). Higher scores indicated higher levels of test anxiety. The internal consistency of the scale was excellent (Cronbach’s α = .93).

General anxiety

General anxiety was measured using the trait component of the State-Trait Anxiety Inventory (STAI-T; Spielberger et al., 1983). The STAI-T comprised of 20 items that measured individual tendencies to experience anxiety in daily life. Adolescents responded to these items on a 4-point Likert scale ranging from 0 = almost never to 3 = almost always. All items were summed to compute a total score for general anxiety (min = 0, max = 60). Higher scores reflected greater levels of general anxiety. The scale demonstrated excellent internal consistency (Cronbach’s α = .93).

Self-compassion

Self-compassion was assessed using the Self-Compassion Scale – Short Form (SCS-SF; Raes et al., 2011). The SCS-SF contained 12 items that measured aspects of self-compassion, such as self-kindness, mindfulness, and common humanity. However, due to the brief nature of the SCS-SF, it is recommended that only a total score for self-compassion be calculated rather than individual subscales (Raes et al., 2011). Responses to each item were reported using a 5-point Likert scale from 0 = almost never to 4 = almost always. All items were summed to provide a total score for self-compassion (min = 0, max = 48). Higher scores represented greater self-compassion. The internal consistency of the scale was good (Cronbach’s α = .81).

Results

A hierarchical regression was conducted using IBM SPSS Statistics (Version 26) to examine the extent that self-compassion predicted test anxiety while controlling for the effects of gender and level of general anxiety. Assumptions for hierarchical regression were met, including inclusion of ratio data, linearity between self-compassion and test anxiety, multivariate normality, homoscedasticity, and non-biased degrees of multicollinearity (mean VIF = 1.66). Consistent with the findings of a previous meta-analysis that showed large effects of self-compassion in improving mental health outcomes for adolescents, including anxiety (Marsh et al., 2018), a power analysis found that the sample was adequate to explore effects within the moderate to large range at 80% statistical power and α = .05. The means, standard deviations, and bivariate correlations for all study variables are displayed in Table 1.

Table 1. Means, standard deviations, and bivariate correlations for all study variables (N = 47).

Variables	M	SD	1	2	3
1 Gender	72.34% female, 27.66% male		-
2 Test anxiety	27.99	11.84	.43**	-
3 General anxiety	30.39	11.59	.12	.65***	-
4 Self-compassion	20.43	7.64	.01	– .61***	– .78***

*Significant at p < .05.

**Significant at p < .01.

***Significant at p < .001.

The predictor variables of gender and general anxiety were entered in Step 1 (Table 2). The model was significant (F(2, 44) = 26.44, p < .001) and explained 52.5% of the variance in adolescents’ test anxiety scores. Both gender (coded 0 = male, 1 = female; β = .36, p < .01) and general anxiety (β = .61, p < .001) were significant positive predictors of test anxiety. That is, being female and experiencing higher levels of general anxiety were related to greater test anxiety. Self-compassion was then entered in Step 2 as a new variable to explore the change in variance while accounting for the effects of gender and general anxiety. The model was also significant (F(3, 43) = 21.06, p < .001), with self-compassion accounting for a further 4.9% of the variance in test anxiety scores (F change (1, 43) = 5.22, p < .05). Gender (β = .39, p < .001) and general anxiety (β = .32, p < .05) continued to be significant and positively associated with test anxiety. However, self-compassion was a significant negative predictor of test anxiety (β = – .36, p < .05). In other words, higher levels of self-compassion were related to reduced test anxiety among adolescents, even after controlling for gender and general anxiety.

Table 2. Hierarchical regression analysis to examine the effect of self-compassion in predicting test anxiety from gender and general anxiety (step 1) and self-compassion (step 2).

		β
Step 1 (R^2 = .55)
	Gender	.36**
	General anxiety	.61***
Step 2 (R^2 = .60)
	Gender	.39***
	General Anxiety	.32*
	Self-compassion	– .36*

Step 2 R² Change = .05.

*Significant at p < .05.

**Significant at p < .01.

***Significant at p < .001.

Discussion

The pilot study explored whether self-compassion was a predictor of test anxiety, having controlled for the effects of gender and level of general anxiety. The findings suggest that self-compassion may influence test anxiety among adolescents. That is, adolescents with greater self-compassion tended to report lower levels of test anxiety. The finding builds on previous research on the role of self-compassion in promoting positive mental health and wellbeing for young people (e.g. Marsh et al., 2018) and extends to test anxiety.

The protective effect of self-compassion on test anxiety may be explained by considering the components that make up self-compassion: self-kindness, mindfulness, and a common humanity (Neff, 2003).

Self-kindness responds to failure with acceptance and understanding, rather than criticism. Self-kindness could have fostered a more supportive approach to dealing with difficulties and reduced fear of failure in examinations. As such, adolescents with greater self-compassion may have been more likely to perceive failure as part of the learning process, which lessened test anxiety. Self-kindness could also have promoted a sense of care and warmth that soothed feelings of anxiety.

Mindfulness focuses attention on the present moment, which may have limited rumination and potential catastrophising about the prospect of failure in the future. Mindfulness could also have generated a more balanced, calm, and non-judgemental perspective for adolescents to understand that experiences of failure are separate from appraisals of self-worth and academic competency. Instead, adolescents with higher self-compassion could simply have viewed failure as an opportunity for further learning.

Common humanity creates a shared sense of connection with others. As such, more compassionate adolescents may have been aware that other young people also struggle with test anxiety, which in turn, normalised difficulties and reduced feelings of isolation. A perception of common humanity can also challenge any viewpoint that adolescents are alone in dealing with problems that are exclusive to them. Furthermore, adolescents may have accepted that encountering failure is an inevitable part of being human.

However, it is also possible that adolescents with higher levels of self-compassion may have been more likely to engage in positive coping strategies that focused on self-improvement and mitigated against test anxiety. That is, research with adult samples indicates that greater self-compassion may be associated with adopting mastery rather than performance goals (Williams et al., 2008), reduced academic avoidance (Neff et al., 2005), and greater preparation for examinations (Breines & Chen, 2012).

In line with previous research, female adolescents reported higher levels of test anxiety compared to males (Putwain & Daly, 2014). The reasons for these effects remain unclear. On the one hand, female adolescents may have adopted a more proactive and focused approach to examination preparation that underscored the importance of upcoming assessments and the consequences of failure (Stöber, 2004). While on the other hand, the finding could reflect gender-based differences relating to the disclosure of school-related anxieties (Rose et al., 2012). That is, females may have been more willing to share their experiences of test anxiety than males.

Adolescents with higher levels of general anxiety were also more likely to experience test anxiety. The finding is in keeping with the integrative network approach that anxiety symptoms exist along an interconnected network (Heeren & McNally, 2016). General anxiety presents with a pattern of symptoms that are central to many anxiety difficulties (Heeren et al., 2018). Shared symptoms may have established strong connections between general anxiety and test anxiety. For example, intrusive and worrying thoughts about failure may also overlap with test anxiety symptoms around the fear of failure in upcoming examinations. As such, activation of general anxiety symptoms could spread across the network and trigger those associated with test anxiety. That is, heightened general anxiety may render adolescents at a greater risk of test anxiety.

Strengths and limitations

The pilot study highlights the potential role of self-compassion as a protective factor against test anxiety for adolescents. However, the findings should also be considered in the light of several limitations.

First, the study had a small sample size. Within the context of a hierarchical regression model, sample size can impact the validity of the findings and possibility of detecting significant effects. That is, other relations between constructs in the current model may exist, including potential moderating effects, however, these were unable to be detected on account of insufficient statistical power. The small sample size also limits the overall generalisability of findings. This generalisability may be further constrained as the study recruited adolescents from a more affluent grammar school in NI with a history of academic success. There is therefore a need for future research to replicate the findings with a larger and more diverse sample, including those from a range of educational contexts and socio-economic backgrounds.

Second, the study examined the effect of self-compassion on test anxiety using data collected at a single timepoint. As a result, it is difficult to make inferences around causation and the direction of effects. While the findings are consistent with theoretical and empirical evidence on the benefits of self-compassion (Marsh et al., 2018) and overlap between test anxiety and other anxiety domains (Putwain & von der Embse, 2021), other directions could also be possible. For example, adolescents with lower levels of test anxiety may be less threatened by examinations, and in turn, find it easier to treat oneself with care and warmth. Similarly, fear of failure in examinations may extrapolate to other worries about negative outcomes in the future associated with more general anxiety. Future research should unravel the direction of these effects using longitudinal data.

Third, the study used the SCS-SF to assess self-compassion. Given the brief nature of the measure, it is recommended that scores should only be calculated for self-compassion rather than the comprising components of self-kindness, mindfulness, and common humanity (Raes et al., 2011). As a result, the study is unable to identify the psychological processes underlying the benefits of self-compassion for test anxiety. Future research should explore the individual impact of self-kindness, mindfulness, and common humanity on test anxiety among adolescents. It would also be beneficial for research to examine the potential influence of factors, such as achievement goals and coping strategies, to further understand the relationship between self-compassion and test anxiety. In addition, qualitative follow-up may offer a more nuanced understanding of the protective role of self-compassion for young people struggling with test anxiety. This approach may be emancipatory in nature by eliciting the voice of a vulnerable population faced with a sense of oppression and powerlessness within school settings.

Implications for practice

The findings of the pilot study have important implications for practitioners supporting young people’s mental health and wellbeing in schools. While preliminary, the findings contribute to our knowledge on the benefits of self-compassion for youth. In particular, they highlight that self-compassion may be a protective factor against test anxiety. However, the findings also indicate that females and those with higher levels of general anxiety can be more vulnerable to test anxiety.

Self-compassion is modifiable and can be cultivated and developed with practice and training (Neff & Germer, 2013). As such, those that hold power within schools have a responsibility to facilitate the development of greater self-compassion and reduce test anxiety for young people. For example, through the delivery of interventions designed to promote self-compassion (O’Driscoll & McAleese, 2021), which have the potential to improve overall academic achievement and have lasting benefits for employment, education, and mental health as young people transition into adulthood. However, schools may be inhibitors to compassion, often as a result of the inhumane performative orientated education systems that also contribute to the development of test anxiety. As such, there is a need for educators to advocate and effect meaningful systemic change to foster positive learning environments that enable all young people to fulfil their potential. It is imperative that these institutions do not focus on academics at the expense of mental health. The development and implementation of a more compassionate approach across multiple levels of the school ecology may provide a novel lens to identify and respond to emotional and wellbeing needs with a sense of warmth and care. Schools can become compassionate communities. That is, rather than pacifying and placating both young people and staff into acceptance of pressurised and competitive school environments, compassion can provide empowerment to rectify systemic power imbalances and mitigate against growing mental health challenges by facilitating a greater sensitivity to suffering in the self and others and foster a motivation to alleviate or prevent it.

Disclosure statement

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

Additional information

Funding

Dean O’Driscoll is supported by a Northern Ireland Department of Education Educational Psychology training award.