Academic Stress in The Final Years of School: A Systematic Literature Review

Child Psychiatry & Human Development
https://doi.org/10.1007/s10578-020-00981-y
ARTICLE
Academic Stress in the Final Years of School: A Systematic Literature

Review
Viviana M. Wuthrich1 · Tess Jagiello1 · Vanessa Azzi1
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
Heightened academic stress in the final years of schooling is a common concern, yet little is known about how stress changes
over time and what individual, school and family factors are associated with distress. We conducted a systematic review to
examine the nature of distress in students in their final two years of secondary school. Sixty studies were eligible for inclu-
sion. The main findings indicated severity of distress differed across the 17 countries sampled and measures used. There was
some consistencies suggesting about 1 in 6 students experienced excessive distress. Female gender and anxiety proneness
were consistently associated with increased distress, and freedom from negative cognitions with reduced distress. There was
some evidence that individual characteristics (perfectionism, avoidance, coping, self-efficacy, resilience), lifestyle (sleep,
homework), school, family and peer connectedness were associated with distress. Overall at-risk students can be predicted
by theoretical models of anxiety and distress targeted with psychological interventions.
Keywords Academic stress · Examinations · Senior school · Test anxiety · High stakes testing
Academic stress is a common concern for youth; with tests, excessive and what individual, school-based or family fac-
homework and grades being the biggest stressors reported by tors exacerbate or lessen this distress. This is an important
secondary school students [1, 2]. Academic stress in the final issue for educators who are responsible for the wellbeing of
years of high school has received particular attention and has students, but also because research indicates that heightened
been found to be associated with very high levels of distress distress can impede academic performance [10]. It is also
in large samples in Australia [e.g. 3], the Netherlands [4], the unclear whether the distress is limited to the examinations
United Kingdom [e.g. 5], and the United States of America per se, or is associated with other factors such as increases in
[6]. In these countries and others (e.g. broader Europe, East workload, increased expectations for independence in learn-
and South Asia, Canada) the final year or two years of school ing, or personal, family or other school based factors. This is
involves a series of examinations and the performance on particularly unclear as most research has examined distress
these examinations forms the basis for an educational certifi- using measures of test anxiety. Although test anxiety refers
cate, pre-university program, or university entrance scores. to a fear of completing tests or exams [11]; common meas-
Due to the large contribution of examination performance ures of test anxiety have been shown to capture anxiety more
on the overall mark (for example over 50% in Australia), broadly with high scores overlapping with anxiety disorders
these examinations are often referred to as high stakes tests, (such as generalised anxiety disorder, social phobia), trait
and these examinations seem to be particularly relevant to anxiety and general anxiety proneness [for reviews see: 12,
increased levels of stress reported by students [7–9]. 13]. It is therefore not clear whether the high levels of test
Although there are many reports of heightened levels of anxiety found in senior school samples [e.g. 5] relates to
distress in students in the final years of secondary school, examinations specifically, or perhaps reflects general high
little is known about the nature of this distress, whether it is levels of distress associated with the academic pressures of
the final years of schooling more broadly, or in fact, pre-
* Viviana M. Wuthrich dominantly capture students with likely anxiety disorders.
Viviana.Wuthrich@mq.edu.au Developmental models of psychopathology suggest
biological by environment interactions play an impor-
1
Department of Psychology, Centre for Emotional Health, tant role in distress during the adolescent period [14, 15].
Macquarie University, Sydney, Australia
13
Vol.:(0123456789)
Environmental stressors associated with senior school and if they primarily examined scale psychometrics, or were
high stakes examinations may interact for biologically vul- focused solely on special populations unrepresentative of the
nerable students to exacerbate or trigger underlying stress general population (e.g. deaf students, students with cystic
vulnerability. These factors might be school based (e.g. fibrosis) and were not specifically related to understanding
increased learning requirements, pressure to perform), home the distress associated the final years of school. Both qualita-
based (e.g. pressure to perform, diet, sleep routines), peer tive and quantitative studies were included.
group based (e.g. social evaluation concerns, social conta-
gion of stress) or other. Developmental models of child anxi-
Data Extraction and Analysis
ety [16, 17] highlight the role of individuals’ cognitive and
behavioural responses to these stressors, as well as parental
All articles retrieved from the database searches were
responses, in exacerbating or reducing the perceived threat
uploaded into citation management software, EndNote.
of these environmental stressors. With adolescent distress
Duplicates were removed automatically. The inclusion and
showing clear trajectories for adult mental disorders [15], it
exclusion of articles based on their title, then abstract, then
is important to better understand the nature of the distress
full-text was overseen by two authors (VA, TJ) with dis-
experienced by senior school students.
crepancies discussed with VW. In addition, references from
Given the importance of understanding the factors associ-
retrieved papers were checked for relevant studies. A total
ated with heightened distress in senior students, the aim of
of 60 articles were eligible for the current systematic litera-
this review was to examine the literature in order to under-
ture review. Data pertaining to the sample (participants, age,
stand the severity of the distress experienced by students in
gender, type of school, country), study method (qualitative,
the final years of secondary school, how distress changes
quantitative, measures used, timing of measurement), and
over time, and to understand the factors that contribute to
outcomes (ranges and means on relevant distress measures,
or protect students from excessive distress. To the best of
relationships between key variables) were extracted. Quan-
our knowledge this is the first review to examine this issue.
titative data on mean distress scores over time on the same
outcomes measures were used where possible to calculate
overall mean distress for the total sample. Other findings
Method were pooled to form a narrative review of the findings. Fig-
ure 1 outlines the search and selection process based on the
Search Strategy and Selection of Studies
PRISMA guidelines [18]. Table 1 summarises the studies
included in the review.
A systematic literature search was carried out using the Psy-
The quality of the studies were assessed using the Critical
cinfo (American Psychological Association) 1806 to April
Appraisal Skills Programme (CASP) Cohort Study checklist
2018 and Education Resources Information Centre (ERIC;
[19] and Qualitative Study checklist [20]. The CASP check-
Institute of Education Sciences) 1966 to April 2018 data-
lists do not provide a scoring system so instead each study
bases as they were considered to be the most relevant for
was rated against each of five study quality criteria as meet-
the topic. The reference lists of relevant articles were also
ing/not meeting the criteria or rated as unclear. For Cohort
used to source additional articles relevant to the review. Key-
studies the quality criteria were: (1) examined a focused
words selected for the search terms were related to hypoth-
issue, (2) appropriate sample, (3) unlikely measurement
esised predictors and outcomes of stress in final year high
bias, (4) appropriate design/confounds considered, and (5)
school students. The final search included terms related to
appropriate analysis/interpretation of results. Similarly, for
high school (high school, secondary school, senior school,
Qualitative Studies the quality criteria were: (1) clear aims,
HSC, high school certificate), distress (anxiety, anxious, test
(2) appropriate sample, (3) appropriate methodology, (4)
anxiety, fear, stress, distress, coping, burnout, resilien$), and
confounds considered, (5) appropriate analysis/interpretation
academic-related stress (exam$, examination$, test$, aca-
of results. Quality was assessed by two raters (VW, TJ) and
demic pressure, academic hardiness, academic buoyan$, fear
disagreements were solved through discussion. See Table 2
appeal, perfectionis$).
for the quality ratings.
Eligible articles were: published in English in peer-
reviewed journals, reported distress or emotional wellbeing
variables as related to academic stress in high school stu-
dents in their final two years of schooling (e.g. Grade 10/11, Results
11/12, 12/13 depending on the country). For studies that
included a broader school grade range they were included if Sixty articles were found that met eligibility criteria
they reported subgroup analyses related to senior students (see Table 1). They reported on qualitative and quantita-
in the final two years of schooling. Articles were excluded tive studies that investigated the distress experienced by
13
Fig. 1 Flow diagram
Arcles idenfied Arcles idenfied
through PsycInfo through ERIC
(n = 2926) (n = 1522)
Arcles combined Duplicates removed

(n = 4448) (n = 512)
Arcles screened by Arcles excluded by

tle tle
(n = 3936) (n = 2921)
Arcles screened by Arcles excluded by

abstract abstract
(n = 1015) (n = 775)
Full-text arcles Full-text arcles excluded

assessed for eligibility (n = 183)
(n =240) No results for students in
the final two years of high
school (n = 133)
Not specifically related to

Addional arcles Total arcles
understanding academic
idenfied through included in the
distress (n = 44)
reference lists review
(n = 3) (N = 60)
Primarily examined scale
psychometrics (n = 5)
Could not source full-text

(n = 1)
students in the final two years of schooling as well as the Quality Assessment
factors that influenced their distress.
The quality of the 60 included studies varied considerably
(see Table 2). Only 26 studies were rated highly on all cat-
Study Characteristics egories. Most studies reported a clear study aim or described
a focused issue of study. Most studies also had adequate
Studies were conducted across 17 countries: Australia samples for the study aims. Measurement bias was not a con-
(n = 12), United Kingdom (n = 15), Turkey (n = 11), United cern in the qualitative studies, with the risk of measurement
States of America (n = 7), Canada (n = 3), the Nether- bias mixed for the cohort studies, with common concerns
lands (n = 1), New Zealand (n = 1), Germany (n = 1), Italy related to the reliability and validity of measures used. Con-
(n = 1), Brazil (n = 1), Nigeria (n = 1), Iran (n = 1), Portu- founds were considered in only some studies, in particular
gal (n = 1), Malaysia (n = 1), Ireland (n = 1), India (n = 1) confounds related to differences in timing of measurement
and one study that compared cohorts in USA and Korea. in relation to the exams, differences in schools or students
Methodologies varied across studies: five studies used a sampled were often ignored. Finally, the use of appropriate
longitudinal design, 55 were cohort studies, and five used analysis and interpretation of results was consistently better
qualitative analyses. in the qualitative studies, with very mixed quality ratings
13

Table 1 Articles included in the review

Authors Sample Sample size Sample age Measures used Summary of findings for senior
students
13
Akca [23] Turkish students preparing for 319 preparing for HSPT Range 12–14 years (Preparing Learned Helplessness Scale for Females had significantly higher
the High School Placement 389 preparing for UEE for HSPT) Children test anxiety than males
Test (HSPT) and University Range 16–19 years (Preparing Test Anxiety Scale Test anxiety and learned
Entrance Exam (UEE) for UEE) helplessness was significantly
higher for students preparing
for HSPT than UEE. Suggests
older students are better able
to cope
No relationship between test
anxiety and learned helpless-
ness
Akcoltekin [95] Turkish high school students in 270 Unreported Time Management Inventory Poorer time management was
grade 12 (final year) Research Anxiety Scale associated with higher research
anxiety
Astill et al. [96] Amsterdam high school stu- 24 M = 17.63 Actigraphy Sleep time, efficiency, and
dents (final year) (SD = 0.10) Daily questions on stress, subjective quality was signifi-
fatigue, and subjective sleep cantly reduced during stressful
quality exam periods, relative to exam-
Daily number of exams and free periods
alcoholic and caffeinated
drinks
Aysan et al. [66] Turkish high school students 59 juniors M = 15.8 (juniors) Perceived Health Scale Students with higher test anxiety
(juniors and seniors) prepar- 54 seniors M = 17.4 (seniors) Test Anxiety Inventory perceived their health to be
ing for exams Coping Strategies Scale poorer
Students who engaged in self
-blame, wishful thinking, or
avoidance coping strategies
tended to have higher test
anxiety
Byrne [53] Australian high school students 224 Range 12 – 18 years The Fear Survey Schedule for No gender differences in self-
(grades 7, 9, 12) (M = 15.05, Children esteem in year 12
SD = 4.63) The Rosenberg Self-Esteem From year 7 to 12, anxiety
Scale significantly increased for girls
State-Trait Anxiety Inven- and decreased for boys
tory—Trait Scale In year 12, boys and girls used
Home life (parental employ- different coping strategies:
ment, number and gender of Girls were significantly more
sibling, family structure) anxious than boys, boys’
coping strategies appeared to
reduce anxiety and fear
Table 1 (continued)
students
Chin et al. [69] New Zealander high school 188 Range 16–19 years Affect and Arousal Scale Negative affect contributed to
students (grades 12, 13) (M = 16.80, Reactions to Tests higher test anxiety. Higher test
SD = 0.72) Examination Performance anxiety was associated with
poorer exam performance
Compared to boys, girls had
higher exam grades and higher
levels of test anxiety and
emotionality
Christensen [65] American high school students 94 Range 16 – 19 years (M and Debilitating Anxiety Scale – Students in accelerated classes
(senior year) SD unreported) Revised had lower test anxiety than
students in “low stream” (low
achieving) classes
Compared to boys, girls had
higher test anxiety
Chukwuorji and Nwonyi [54] Nigerian secondary school 281 Range 16–19 years State Self-Esteem Scale Students whose parents had
students (candidates for high- (M = 17.05, Test Anxiety Inventory higher-status jobs were more
stakes examinations) SD = 1.87) Socio-demographic informa- test anxious
tion Students with lower self-esteem
had higher test-anxiety
No gender differences in test
anxiety
Çirak [73] Turkish high school students 23 Range 17–18 years Semi Structured Interviews Students frequently reported
(final year) (M and SD unreported) emotional and physiological
anxiety about exams
Test anxiety was related to
perceived pressure from family
expectations and the idea of
absolute success or failure
Cunha and Paiva [27] Portuguese high school stu- 449 Range 15–21 years Socio-demographic informa- Lower test anxiety was
dents (grades 10, 11, 12) (M = 16.28, tion associated with higher self-
SD = 1.17) Test Anxiety Inventory reassurance, acceptance, and
Forms of Self-Criticising and mindfulness
Self-Reassuring Scale Higher test anxiety was associ-
Social Anxiety and Avoidance ated with high self-criticism,
Scale for Adolescents high social anxiety, low self-
Children’s Acceptance and reassurance, and low accept-
Mindfulness Measure ance and mindfulness skills
Girls had higher test anxiety
than boys
Test anxiety was significantly
higher for students with mid-
SES levels, relative to high-
SES students
13

students
13
Daly et al. [115] English sixth-form college 39 M = 17.31 Raw mark on a mock A-level Students with higher levels of
students (in their first year of (SD = .047) examination test anxiety performed worse
their A-level qualification) Results in the GCSE in their GCSE, on average
Test anxiety questionnaire No relationship found between
Heart rate data heart rate and test anxiety
Einstein et al. [29] Australian high school students 772 Range 15–24 years Multidimensional Perfection- Depression, anxiety, and stress
(grade 12) (M = 17.6, ism Scale symptoms increased signifi-
SD = 0.6) Depression Anxiety Stress cantly as exams came closer,
Scales with 20% of the sample falling
Self-report of hours spent into the severe range
studying, motivation, country Students of immigrant parents
of birth reported higher socially-
prescribed perfectionism,
which was positively related to
depression and anxiety
Erzen and Odaci [56] Turkish students in their final 180 Range 16–19 years (M = 17.3, Test Anxiety Inventory A secure attachment style was
year of high school SD = 0.53) Relationships Scales Question- related to lower test anxiety
naire A fearful or dismissing attach-
Self-Efficacy Scale ment style was related to
Personal Information Form higher test anxiety
Students with low self-efficacy
had higher test anxiety
Female students had higher test
anxiety than males
Flett et al. [85] Canadian secondary school stu- 73 Range 16–18 years (M = 16.8, Child-Adolescent Perfection- Depression was associated with
dents (grades 11 and 12) SD unreported) ism Scale higher self-oriented perfection-
Adolescent and Adult Type A ism, more Type A behaviour,
Behavior Scale and lower self-efficacy
Perceived Self-Efficacy Scale
Children’s Depression Inven-
tory
Psychosomatic Symptom Scale
Guner-Kucukkaya and Isik [24] Turkish high school students in 728 Unreported Demographic information form Students in year 11 had signifi-
grades 9, 10, and 11 Brief Symptom Inventory, cantly higher GSI scores than
Global Severity Index (GSI) students in grades 9 and 10
students
Hodge et al. [21] Australian high school students 445 Unreported Questionnaire on: Biographical Almost half of all the students
(grades 11 and 12) and personal characteris- reported “at risk” levels of
tics, school characteristics, distress
academic self-concept and Year 12 students about to
expectation of HSC outcome, undertake major examinations
coping practices, and percep- reported the highest levels of
tions of support distress. Females had higher
State-Trait Anxiety Inventory- levels of distress than males,
Trait and students from a non-

General Health Questionnaire English speaking background
(GHQ) had higher levels of stress than
Adolescent Coping Scale-Short students from English speaking
Form backgrounds
Self-Description Question- Students who had higher trait
naire III anxiety, lower self-confidence,
lower socio-economic status,
lower academic self-concepts,
or less effective coping strate-
gies were more likely to report
higher levels of distress
Trait anxiety was the strong-
est predictor of examination
distress
Karatas et al. [57] Turkish high school students 194 M = 17.5 Test Anxiety Inventory Females had significantly higher
(final year) Scores on the university test anxiety than males
entrance exam Females with higher test anxi-
Grade Point Average (GPA) ety performed worse on their
university entrance exam
Kouzma and Kennedy [94] Australian senior high school 369 Range 16–18 years (M = 16.6, 1-week homework diary Compared to boys, girls spent
students SD = 0.60) Self-report stress scale significantly more time doing
Profile of Mood States homework, and reported
significantly higher levels of
stress and mood disturbance
Hours spent on homework was
significantly positively associ-
ated with stress and mood
disturbance
13

students
13
Kouzma and Kennedy [74] Australian final year students 423 Range 16–18 years (M = 17.3, Academic Stress Questionnaire Main sources of stress included
(year 12) SD = 0.6) examinations and results, too
much to do, worry over future,
making choices about career,
studying for examinations,
amount to learn, need to do
well imposed by others, and
self-imposed need to do well
Lay et al. [30] Canadian students in final year 63 Unreported Endler Multidimensional Anxi- State anxiety was positively
(Year 13) ety Scales (state and trait related to threat appraisals and
forms) harm appraisals
Procrastination scale Emotion-focused coping was
Threat, challenge, harm, and related to higher state anxiety
gain appraisals Prior to and during a stressful
Ways of Coping Scale exam period, students reported
the highest levels of threat
appraisals
State anxiety was highest one
day before their first final
exam. State anxiety was lowest
five days after their last exam
Lee and Larson [48] Korean and USA high school 56 in Korea, M = 17.8 (Korea) Children’s Depression Inven- Korean students had signifi-
seniors preparing for univer- 62 in USA M = 17.0 (USA) tory cantly higher levels of clinical
sity entrance exam (year 12) Experience Sampling Method depression than USA students
self-report diary (activities, Korean students spent sig-
affect) nificantly more time (2x) on
schoolwork. USA students
spent more time in leisure
activities
Korean students reported more
negative emotional states
during daily activities and
schoolwork, which was related
to depression
students
Leonard et al. [6] USA senior private school 128 M = 16.37 Socio-demographic and back- Greatest sources of stress were
students (year 11) (SD = 0.62) ground information grades, homework, and prepar-
mtvU/Associated Press survey ing for college
Help with stress (developed by Grades were a significantly
authors) greater source of stress for
Self-Report Coping Scale females than males
Academic Motivation Scale 49% of students reported feeling
Parental Expectations subscale a great deal of stress on a daily
of the Multidimensional basis

Perfectionism Scale Females reported significantly
Quick Depression Assessment higher stress than males
of PHQ-9 Females more likely to use
Communities that care youth problem-focused coping and
survey emotion-focused internal
avoidance coping
Males more likely to report
emotion-focused external
avoidance coping
Greatest source of support
identified as friends, followed
by family members, followed
by school staff
26% of students had clinical
levels of depression
38% of students reported drink-
ing alcohol to get drunk in the
last week
Substance use was significantly
associated with high levels of
stress
13

students
13
Lin and Yusoff [43] Malaysian students in final year 382 Aged 19 (M, SD and range Sociodemographic information 47.6% of students reported
of high school (upper 6) unreported) GHQ-12 psychological distress
Secondary School Stressor (2.5 × higher than WHO
Questionnaire expected)
Brief Coping Orientation of The top ten stressors perceived
Problem Experienced by students were related to
academia (e.g. “Afraid of not
getting university placement”)
Positive reinterpretation was a
protective factor (negatively
correlated with the develop-
ment of distress)
Risk factors included behav-
ioural disengagement, self-
blame, denial, venting, self-
distraction (significant positive
relationship with distress)
Locker and Cropley [28] English high school students 508 (Reported separately for gen- Positive and Negative Affect Immediately prior high-stakes
(grades 9 and 11) der and school-type) Schedule examinations, females were
M = 13.9–16.8 Children’s Depression Inven- significantly more anxious (in
(SD = .428—.725) tory terms of worry and physi-
Revised Children’s Manifest ological anxiety) than males.
Anxiety Scale Females also showed higher
Rosenberg Self-Esteem Scale levels of negative affect and
distress at this time
Students at all girls schools had
higher anxiety and negative
affect prior to exams, relative
to students attending mixed
and all boys schools
Lushington et al. [59] Australian students in senior 398 (Reported separately for gen- Questionnaire (demographics, No main effects for culture or
high school (years 11 and 12) der and ethnicity) sleep habits, extracurricular gender on sleep quality
activity, academic expecta- Females had significantly higher
tions, stress, mood, coping, depressive mood and study/
diet, pubertal development) interpersonal stress than males
6-item Depressive Mood Scale Among Caucasian Australians,
4-item Brief Resilient Coping higher study stress was related
Scale to higher daytime sleepiness
4-item Study and Interpersonal On school nights, Asian Austral-
Stress subscale from the ians slept less (went to bed
Student-life Stress Inventory later and woke up earlier) than
Caucasian Australians
students
Manley and Rosemier [64] American school students 2,086 (Grade 12 n = 351, Unreported Test Anxiety Scale for Children Girls had significantly higher
(grades 7 to 12) Grade 11 n = 291) General Anxiety Scale for test anxiety than boys at every
Children grade level
No significant differences in test
anxiety between grades 11
and 12
McCann and Meen [72] Canadian high school students 222 Unreported State-Trait Anxiety Inventory Higher anxiety was associated
(grades 11 and 12) (STAI) with lower intellectual ability
Intellectual ability (Raven’s Higher anxiety was associated

Progressive Matrices and with lower achievement for
Canadian Lorge-Thorndike students with lower intellectual
intelligence test) ability
Academic achievement (final Higher anxiety was associated
English grades) with greater achievement for
students with higher intellec-
tual ability
McGraw et al. [3] Australian high school students 941 Range 16 – 19 years Demographics More than 12% were severely or
(grade 12) (M = 17.4, Psychological Sense of School extremely depressed
SD = 0.6) Membership More than 20% were severely or
Family Connectedness Self- extremely anxious
Report More than 10% were severely or
UCLA-R Loneliness Scale extremely stressed
Depression Anxiety Stress Lower connectedness to school,
Scales-21 family and peers was associ-
ated with higher levels of
stress, depression, anxiety, and
negative affect
Higher connectedness to school,
family and peers was a protec-
tive factor against stress, anxi-
ety, and depression
Girls reported stronger peer con-
nectedness than boys, but were
significantly more stressed and
anxious
Moulds [25] Australian high school students 1137 Unreported Main daily hassle in current Students in grade 11 were sig-
(grades 7, 9 and 11) term nificantly more anxious than
Student Stress Inventory those in grades 7 and 9
Actual Classroom Environment
Scale
Rosenberg Self-Esteem Scale
Demographic information
13

students
13
Peluso et al. [31] Brazilian senior high students 222 Median age = 17 years Positive and Negative Affect Negative affect significantly
preparing for the university (M and SD unreported) Schedule – Expanded form increased as the high-stakes
admission exam (PANAS-X) exam drew nearer
Time spent studying Mood correlated with time spent
studying: Negative affect was
higher for students who spent
more time studying
Putwain [8] English secondary school stu- 34 Range 14–16 years (M and SD Semi-structured interviews Examination anxiety was report-
dents (grades 10 and 11) unreported) edly influenced by an anxious
disposition and perceived
assessment importance
Academic stress was reportedly
influenced by deadlines, work-
load, motivation to achieve,
and fear of failure
Examination anxiety reportedly
resulted in difficulty remem-
bering information during
exams
Putwain [52] English high school students 1348 Unreported Test Anxiety Inventory Females had significantly higher
(grades 10 and 11) Student Profile Questionnaire test anxiety than males
Ethnicity influenced test anxiety;
Students from White ethnic
backgrounds had lower levels
of test anxiety than students
from Black, Asian, and other
ethnic backgrounds
Students for whom English was
their second language had
higher test anxiety
Putwain [49] English secondary school 615 Range 14–16 years State Test Anxiety Scale Higher test anxiety, worry,
students (grades 10, 11) (M and SD unreported) Examination Performance and emotionality were all
Examination Stakes (high, mid, associated with poorer exam
low) performance
Students reported lowest levels
of test anxiety and attained
highest grades for mid-stakes
exams
students
Putwain [50] English high school students 557 Range 15–16 years Test Anxiety Inventory Females had significantly higher
(grade 11) (M and SD unreported) Office for National Statics test anxiety than males
categorical scale of socio- Higher test anxiety was associ-
economic background ated with poorer exam perfor-
Grade on GCSE examinations mance
Students from a higher socio-
economic background had
lower test anxiety and higher
exam grades than students

from lower socioeconomic
backgrounds
Putwain [75] English secondary school 34 Range 14–16 years Semi-structured Interviews Gender differences in self-
students (grades 10, 11) (M and SD unreported) efficacy were evident across
different subjects
Test anxiety was reportedly
caused by fear of failure, high
career aspirations, pressure
from teachers, and low self-
efficacy
Preparation, self-confidence,
and an expectation of success
appeared to be protective fac-
tors against test anxiety
Putwain and Aveyard [82] English students in their final 270 M = 15.01 Revised Test Anxiety Scale At low levels of worry, students
year (Year 11) (SD = 0.89) Academic Buoyancy Scale with higher perceived control
Motivation and Engagement performed better on the exam,
Scale relative to students with low
Study Management and Aca- perceived control
demic Results Test At high levels of worry, level of
Examination grades (GCSE) perceived control did not influ-
ence exam performance
Putwain and Daly [5] English secondary school stu- 2435 Range 14–16 years Revised Test Anxiety Ques- 16.4% of all students reported
dents (grades 9, 10, 11) (M and SD unreported) tionnaire high levels of test anxiety
Friedben Test Anxiety Scale Girls were more likely than
boys to be highly test anxious,
worried, tense, and concerned
about being negatively judged
by others
13

students
13
Putwain and Symes [89] English secondary school 273 M = 15.5, Teacher’s Use of Fear Appeals Students who perceived teach-
students (grade 11) (SD = 0.45) Questionnaire ers’ fear appeals as threatening
Revised Test Anxiety Ques- performed worse in the high-
tionnaire stakes exam
Achievement Goals Question- Students with a mastery-
naire approach achievement goal
Exam performance performed better on the exam
that students with a perfor-
mance-avoidance achievement
goal
Putwain and Symes [103] English secondary school 132 Grade 10 Teacher’s Use of Fear Appeals Students who perceived teach-
students (grades 10, 11) M = 14.7 (SD = 0.5) Questionnaire ers’ fear appeals as threatening
Grade 11 Revised Test Anxiety Scale showed higher test anxiety
M = 15.5 (SD = 0.5) Achievement Goals Question-
naire
Putwain et al. [79] English high school students 244 M = 15.6 Revised Test Anxiety Scale Higher test anxiety was related
(grade 11) (SD = .49) Children’s Negative Cognitive to academic cognitive distor-
Error Questionnaire tions
Grade on GCSE examinations Students with cognitive distor-
tions also obtained a lower
grade in GCSE examinations
Putwain et al. [87] English high school students 298 M = 14.9 Revised Test Anxiety Ques- Students with greater academic
(grades 10 and 11) (SD = .70) tionnaire buoyancy reported lower test
Coping with Pre-exam Anxiety anxiety
and Uncertainty Highly test anxious students
Academic Buoyancy Scale tended to use more social
support and avoidance coping
strategies
Putwain et al. [10] English secondary school 705 M = 15.03 Revised Test Anxiety Scale Girls had higher test anxiety and
students (grade 11) (SD = 0.58) Academic Buoyancy Scale academic buoyancy than boys
Scores on GSCE examinations No gender differences in exam
performance
Academic buoyancy was a
protective factor against test
anxiety and enabled better
exam performance
students
Putwain et al. [81] English secondary school 325 M = 15.3 Revised Test Anxiety Scale Higher test anxiety was associ-
students (grade 11) (SD = 0.61) Academic Buoyancy Scale ated with poorer exam perfor-
Coping with Pre-exam Anxiety mance
and Uncertainty Scale Students with lower academic
GCSE Scores buoyancy were more likely to
be test anxious and then per-
form poorly on their exams
Rahafar et al. [60] Iranian high school students 158 M = 17.5 Grade Point Average Students with higher levels of
(final year) (SD = 0.51) Reduced morningness-evening- conscientiousness had lower

ness questionnaire test anxiety and higher GPAs,
NEO five-factor inventory— and were more likely to be
Conscientiousness subscale “morning-type” people
Test Anxiety Inventory Females had higher test anxiety
than males
Riekie et al. [102] Australian students in year 11 618 Range 16–17 years Interviews If students felt supported by
(M and SD unreported) Index of Community Socio- teachers and their peers and
Educational Advantage valued at the school, they had a
What is happening in this greater sense of resilience
school questionnaire School connectedness had a
Wellbeing, resilience, and direct relationship on student
moral identity questionnaire wellbeing
WHO-Five Wellbeing Index School connectedness had the
Resilience scale (adapted) most significant impact on
Moral Identity Scale student resilience
Ringeisen and Buchwald [61] German high school students in 82 Range 17–20 years (M = 18.06, German test anxiety inventory Test anxiety was strongly related
their final year SD = 0.81) Negative/positive state emotion to state threat emotions before
T1 (3 weeks before final the exam but not after
exam), T2 (three weeks after
exam)
Robinson et al. [9] Australian high school students 195 Range 16 – 19 years Adolescent Resilience Ques- 24% boys and 36% girls reported
(grade 12) (M = 17.74, tionnaire severe or very severe symp-
SD = 0.44) Subjective Stress toms of depression, anxiety,
Depression Anxiety Stress and stress (more than twice
Scales-21 that of the normative sample)
Measure of Sleep Girls had greater anxiety than
males
Many students had poor sleep
Resilience and freedom from
negative cognitions were pro-
tective factors
13

students
13
Sarason [71] American high school students 460 Unreported Test Anxiety Scale (TAS) Higher test anxiety was associ-
(grades 11 and 12) Need for Achievement Scale ated with lower academic per-
School and College Ability formance, and this relationship
Test was stronger for females
Sari et al. [62] Turkish high school students 724 Range 16–19 years Sociodemographic question- Females had higher test anxiety
(preparing for university (M and SD unreported) naire than males
entrance exams) Rosenberg Self-Esteem Scale Students with higher self-esteem
Revised Test Anxiety Scale reported lower levels of test
anxiety
Schmidt et al. [76] Italian students in their final 610 Range 18–23 years (M = 18.7, Appraisals of high school exit Three weeks before exams the
year of high school SD = 0.90) exam (18 items developed by strongest emotions were anxi-
authors) ety, fear, hope
Intensity of emotions (18 items Feeling anxious/fearful was
developed by authors) positively correlated with
Emotion regulation strategies frustrated/powerless
(18 items adapted from exist- Feeling anxious/fearful or frus-
ing questionnaires) trated/powerless was associ-
ated with evaluating the final
exam as important and ability
to cope as low
Students who reported higher
anxiety/fear reported a greater
tendency to focus on the exam
and to use drugs
Students who reported more
positive emotions used more
reappraisal and problem-
focused coping strategies
Segool et al. [70] USA students in year 11 1248 Unreported Survey questions (developed by Females reported higher test
(required to take a standard- authors) about self-efficacy, anxiety than males
ised high school exam) special education status, aca- Students with lower self-efficacy
demic achievement, minority reported higher test anxiety
status, socio-economic status, Students who perceived higher
gender, school climate, career teacher and peer anxiety
goals, and test importance reported higher test anxiety
Friedben test anxiety scale Students with lower academic
achievement had higher test
anxiety
students
Smith and Sinclair [22] Australian high school students 130 Unreported Depression Anxiety Stress > 40% of Year 12 students and
(grades 11, 12) Scales-21 25% of Year 11 students had
Patterns of Adaptive Learning symptoms of depression,
Scale anxiety and/or stress above the
normal range
Girls had greater anxiety, depres-
sion, and stress symptoms than
males
Boys with a performance-avoid-

ance goal orientation were
more anxious, depressed, and
stressed
High self-efficacy was protec-
tive against stress, anxiety, and
depression in girls
Smith et al. [32] Australian high school students 63 Unreported Patterns of Adaptive Learning As final exams drew nearer,
(grade 12) Survey students became significantly
Depression Anxiety Stress more anxious and depressed,
Scales used more self-handicapping
and performance-avoidance
strategies, and their self-effi-
cacy decreased
Girls had higher anxiety and
stress than boys. Boys used
more performance-avoidance
and self-handicapping strate-
gies than girls
Students with a performance-
approach goal combined with
limited use of self-handicap-
ping strategies had more stress
Smyth and Banks [77] Irish students in the Junior Cer- 12 case-study schools Unreported Interviews (k = 47 lower Students preparing for final
tificate exam year and Leav- secondary, k = 53 upper Leaving Certificate reported
ing Certificate exam year secondary) “particularly high levels of
Questionnaire survey (n = 897 pressure and stress”
lower secondary, n = 748 Students preparing for the Junior
upper secondary) and Leaving Certificate both
reported they “wanted teachers
who were patient and willing
to explain things clearly”
13

students
13
Sud and Sujata [80] Indian high school students 200 Unreported Self-Handicapping Question- Higher test anxiety was associ-
(grade 11) naire ated with greater worry, emo-
Test Anxiety Inventory-Hindi tionality, self-handicapping,
Study Habit Inventory and with poorer study habits
Academic performance The cognitive component of test
(grades) anxiety (worry) was negatively
related to academic perfor-
mance
Ünal-Karagüven [55] Turkish senior high school 336 Range 17–20 years (M = 18) Demographics questionnaire Students with higher perceived
students Academic Motivation Scale academic achievement had
Test Anxiety Inventory lower test anxiety
Communal Mastery Scale Test anxiety was significantly
negatively correlated with
communal mastery (believing
that one can achieve things
with the help of others), hav-
ing siblings, and perceived
academic achievement
von der Embse and Witmer American high school students 1134 Unreported Friedben Test Anxiety Scale Higher test anxiety was associ-
[51] (grade 11) Demographics ated with poorer exam perfor-
Exam performance mance
Wilkinson-Lee et al. [26] American school students 4198 Unreported Emotional distress (Modi- Students in grades 11 and 12
(grades 6 to 12) fied Depression Scale CDC were significantly more dis-
Compendium) tressed than students in earlier
Family obligations (CSAP grades (6, 7, 8 and 9)
Core Measures) Students in higher grades had
School connectedness (CDC lower school connectedness
Classroom Climate Scale)
Academic achievement
(grades)
Yeni Palabiyik [33] Turkish students in their final 62 Unreported Maslach Burnout Inventory- For 12th grade students,
three years (Years 10, 11, student survey emotional exhaustion was sig-
and 12) Quick Placement Test nificantly higher than overall
burnout
Non-significant increase in burn-
out levels from grade 10 to 12
Male 12th grade students had the
highest level of burnout
in the cohort studies. In consolidating the findings across
significantly greater in females

Summary of findings for senior studies, more emphasis is placed on the findings from the
Students with lower academic

GPA’s had higher levels of
Depressive symptoms were
higher quality studies.
depressive symptoms
Outcomes
than males
Distress Increases Over the Senior School Period

students
Cohort Studies Comparisons between student cohorts

Personal information question-
across different grades generally found that the final year

Beck Depression Inventory
of high school was associated with more distress than the

penultimate year [21, 22], and the penultimate year was
naire (demographics)
associated with more distress than the earlier years [23–26]

suggesting that distress is likely to increase through the sen-
Measures used
ior years. Although the majority of studies found increased

distress in later school years, not all did [e.g. 27, 28], and
(BDI)
this may relate to differences in the timing of assessments

or perceived stakes of the assessments. For example Locker
and Cropley [28] assessed UK students in year 11 and year
Range 17–21 years (M = 18.4,
9 twice, both at the same time prior to major exams and

found no differences in distress. Therefore students might
find all examination periods equally distressing (regard-
less of school year). However, it is important to note likely
cohort effects between groups studied, as well as differences
Sample age
SD = 2.3)
in the timing and type of distress measurement particularly

in terms of the proximity and types of upcoming examina-
tions. Longitudinal studies with repeated measurement in
the same sample are needed to understand changes in dis-
tress over time.
Longitudinal Studies Five studies sampled distress in the

Sample size
same students throughout the final year, with most find-

ing increases in distress as major examinations approached
984
[29–33], and decreased distress after the examination period

[30]. Einstein, Lovibond and Gaston [29] measured distress
university entrance examina-
Turkish students preparing for
using the Depression, Anxiety and Stress Scales [DASS;

34] in Australian year 12 students on two occasions prior
to upcoming major examinations; 10 weeks prior and again
10 days prior to the same exam period. They found mean
student distress increased significantly from the first testing
to the second testing from: moderate to severe anxiety, from
Sample
tions
severe to extremely severe stress, and remained in the severe

range (but significantly increased) from first to second test-
ing for depression. Similarly, Smith, Sinclair and Chapman
[32] administered the DASS to 63 Australian students in
February (term 1 year 12) in and again in August (term 3)
just before the major trial examinations. Students’ group
Yildirim et al. [47]
means for DASS measured stress, anxiety and depression

increased significantly overall between the two testing occa-
sions from mild to moderate depression, from mild to mod-
Authors
erate anxiety; however, remained at a mild level for stress

across both occasions.
13
Table 2 Quality ratings of articles included in the review

Quantitative studies Focused issue Adequate sample Unlikely bias Appropriate design Adequate analysis/interpretation
Akca [23] Yes Yes Yes No Unclear

Akcoltekin [95] Yes Unclear Unclear Unclear Unclear
Astill et al. [96] Yes Yes Unclear Yes Yes
Aysan et al. [66] Yes Yes Yes Yes Yes
Byrne [53] Yes Yes Yes Yes Yes
Chin et al. [69] Yes Yes Yes Yes Yes
Christensen [65] Unclear Unclear Unclear Unclear Unclear
Chukwuorji andNwonyi [54] Yes Yes Yes Yes Yes
Cunha and Paiva [27] No Yes Yes Unclear Unclear
Daly et al. [115] Unclear No Yes No No
Einstein et al. [29] Yes Yes Yes Yes Yes
Erzen andOdaci [56] Yes Yes Yes Yes Yes
Flett et al. [85] Yes Yes Yes Yes Yes
Guner-Kucukkaya and Isik [24] Yes Yes No No No
Hodge et al. [21] Yes Yes Yes Yes Yes
Karatas et al. [57] Yes Yes Yes No No
Kouzma and Kennedy [74] No Yes No No No
Kouzma and Kennedy [94] Yes Yes Unclear No No
Lay et al. [30] No Yes Yes Yes Unclear
Lee and Larson [48] Yes Yes Yes Yes Yes
Leonard et al. [6]a Yes Yes Unclear Yes Yes
Lin and Yusoff [43] Yes Yes Yes Yes Yes
Lushington et al. [59] Yes Yes Unclear Unclear Yes
Manley and Rosemier [64] Unclear Yes Yes Yes Unclear
McCann and Meen [72] Yes Unclear Yes No No
McGraw et al. [3] Yes Yes Yes Yes Yes
Moulds [25] Yes Yes Unclear Yes Yes
Peluso et al. [31] Yes Yes Yes Yes Yes
Putwain [49] Yes Yes Yes No No
Putwain [50] Yes Yes Yes Unclear Yes
Putwain [52] Yes Yes Yes Yes Yes
Putwain and Aveyard [82] Yes Yes Yes Yes Yes
Putwain and Daly [5] Yes Yes Yes Yes Yes
Putwain and Symes [89] Yes Yes Yes Yes Yes
Putwain and Symes [103] Yes Yes Yes Yes Yes
Putwain et al. [10] Yes Yes Yes Yes Yes
Rahafar et al. [60] No Yes Yes No No
Riekie et al. [102] Yes Yes Unclear Unclear No
Ringeisen and Buchwald [61] Yes Yes Unclear Unclear Unclear
Robinson et al. [9] Yes Yes Unclear Yes Yes
Sarason [71] Yes Unclear Unclear No No
Sari et al. [62] Yes Yes Yes Yes Yes
Schmidt et al. [76] Unclear Unclear No Unclear No
Segool et al. [70] No Yes No Unclear No
Smith and Sinclair [22] Yes Yes Yes Yes Yes
Smith et al. [32] Yes Yes Yes Yes Yes
Smyth and Banks [77]a Unclear Yes Unclear Unclear Unclear
13
Quantitative studies Focused issue Adequate sample Unlikely bias Appropriate design Adequate analysis/interpretation
Sud and Sujata [80] Yes Unclear Yes Yes Yes

Ünal-Karagüven [55] Unclear Yes Yes Unclear Unclear
von der Embse and Witmer [51] Yes Yes Yes Yes Yes
Wilkinson-Lee et al. [26] Yes Yes Unclear Yes Yes
Yeni Palabiyik [33] Yes Yes Yes Yes Yes
Yildirim et al. [47] Yes Yes Unclear Yes Unclear
Qualitative studies Clear aims Adequate sample Appropriate method Unlikely confounds Adequate analysis/interpretation
Çirak [73] Yes Yes Yes Unclear Yes
Leonard et al. [6]a Yes Yes Yes Yes Yes
Putwain [8] Unclear Yes Yes Unclear Yes
Putwain [75] Yes Yes Yes Yes Yes
Smyth and Banks [77]a Unclear Yes Yes Unclear Yes
a
These studies had both quantitative and qualitative results. Yes = met criterion, No = didn’t meet criterion, Unclear = unclear if criterion was met
Similar findings were found by Peluso et al. [31] in 154 normative data developed in adult samples, and despite its
Brazilian students using the Positive and Negative Affect wide spread use with adolescents, there are no adolescent
Scale [PANAS; 35] administered three times during the norms available. Further, although the three factor structure
last year of high school. They found negative affect scores of the DASS (depression, anxiety, stress) has been generally
increased over time, with no significant change in positive supported in adolescent samples with minor modifications
affect scores. Lay et al. [30] also found that mean state anxi- [40, 41], it has not consistently been supported [42], and
ety scores significantly increased from seven days prior, to so caution is needed in interpreting the results. It is also
one day prior to exams, and then decreased five days after difficult to determine if the distress levels reported across
the exam period in final year students in Canada. In contrast, studies related to distress prior to an examination or general
Locker and Cropley [28] found no significant increases in distress across the senior school period. Studies differed in
distress in a sample of students in four UK high schools the timing of distress measurements to major examinations
from six to eight week prior to exams, to one week prior to and timing was often was not reported.
exams on the PANAS [35], Children’s Depression Inven- Due to the range of measures used, comparisons between
tory [CDI; 36], Revised Manifest Anxiety Scale [RCMAS; studies are difficult; however, the TAI was used in nine stud-
37], or Rosenberg Self-Esteem Scale [38]. Although it is ies and the DASS in five studies enabling some compari-
important to note that there was 21% attrition between the sons across studies to be made. Across the nine studies that
two testing periods and this may have led to missing data at used the TAI (most from the UK), distress was consistently
the higher distress levels. In general studies have shown that reported to be high in final year students [e.g. 5]; however,
stress is heightened in the senior years compared to lower different authors used different definitions for “high” test
years and increases in the lead up to the major exam period. anxiety. In the largest study, Putwain and Daly [5] admin-
istered the TAI to 2435 students from 11 UK secondary
Severity of Distress schools in the final two years of schooling (years 10 and 11;
with a small sample in year 9 who were being accelerated)
Across a wide range of measures and countries, most stud- in the lead up to the final exam period. High test anxiety was
ies reported that students had high levels of distress with defined as scoring in the top 1/3rd of the score range, that
stronger similarities across samples when the same measure is students who reported experiencing anxiety somewhere
was used in the same country (see Table 1). Determining from “often” to “almost always”. Using this definition, they
how “high” the distress was across the samples is difficult found 16.4% reported high test anxiety, with females report-
to ascertain as studies differed in the cut-offs used to define ing significantly higher test anxiety than males (22.5% vs
high distress, with many applying cut-offs from adult nor- 10.3%). It is not clear whether the 16% of students reporting
mative samples. For example interpretation of scores on the high levels of test anxiety is significantly more than in other
TAI were based on normative data from college students years. Similar findings were reported by the other studies
collected 40 years ago, and recent research suggests these using the TAI.
cut-offs are out of date for college students [39]. Simi- Five studies (all Australian) reported on student levels
larly, interpretation of scores on the DASS was based on of depression, anxiety and stress using the DASS [34] in
13
either its full (42 item) or shortened (21 item) form [3, 9, North-East USA scored above the clinical cut off of 10 based
22, 29, 32]. In adult samples, scores can be interpreted as on adult norms [6], and one study used the Beck Depres-
normal, mild, moderate, severe or extremely severe, with the sion Inventory Revised [BDI-R: 46] and reported 45% of
severe and extremely severe categories suggestive of clini- a Turkish sample were in the clinical range for depression
cal levels of distress [34], although it is not known whether (based on adult norms using a cut off of > 16) two months
these cut-offs are applicable to adolescent samples. Rob- prior to their major examinations [47]. Only three studies
inson, Alexander, and Gradisar [9] sampled 195 final year used a scale developed for children, the Children’s Depres-
(year 12) students in South Australia one month prior to sion Inventory [36], and the one study that reported scores
the exams and found that male and female students (respec- in relation to clinical cut offs found that 36% of final year
tively) reported levels in the likely clinical range (severe or Korean students were above the cut-off for clinical depres-
extremely severe on the DASS-21) for: depression (15%, sion, compared to 16% of American students [48]. Across
22%), anxiety (11.7%, 29.9%) and stress (18.3%, 22.3%). these other measures the proportion of students experienc-
This is similar to the findings by McGraw, Moore, Fuller, ing very high distress ranged from 16 to 57% although the
and Bates [3] who surveyed 941 final year students (year majority used normative cut-offs based on adult samples and
12) in term 3 in Victoria Australia and found the propor- so may not accurately reflect clinical distress.
tion of students who reported distress in the likely clinical
range (severe or extremely severe on DASS-21) for: depres-
sion was 12.1%, anxiety (20.9%), and stress (11.3%). Smith Factors Associated with Increased Distress
and Sinclair [22] also measured DASS scores in term 3 for
year 11 and year 12 students in New South Wales, Aus- Across the studies there was evidence that demographic,
tralia, and found the proportion of students reporting severe individual, family and school factors were associated with
to extremely severe levels in the two cohorts (Year 11, Year increased distress in students in the final two years of school.
12) being: 12.5% vs 24% for stress, 11% vs 19% for anxiety
and 13% vs 24% for depression with clear increased distress Demographics Demographic factors such as gender (dis-
in the final year students. cussed below), low socioeconomic status and studying in a
Averaging across the five studies, final year (year 12) stu- second language were found to be associated with increased
dents’ mean level of distress likely to be in the clinical range student distress in the final years of school although when
(severe or extremely severe) was for stress 17.77% (means reported the effect sizes for these effects were very small
ranged from 11.39 to 21.07%), for anxiety 21.03% (means and difficult to differential as a range of demographic vari-
ranged from 17.90 to 24.30%), and for depression 18.12% ables were often lumped together. Four studies [21, 27, 49,
(means ranged from 12.10 to 22.42%). Averaging the mean 50] found a significant difference in distress associated with
score across the scales, 19% of students reported distress in lower socioeconomic status in Australia, United Kingdom
the severe to extremely severe range (as per adult norms). and Portugal, but not Melaka [43], with only two studies
This is similar to the average of 16% with reported high test reporting the unique variance of this effect as very small
anxiety across 11 UK schools in the Putwain and Daly [5] (R-squared change = 0.06–0.006 [21, 50]. Five studies (in
study reported above. So about 1 in 6 final year students is Australian, UK and USA samples) found increased distress
likely to have very high levels of distress which might be of related to coming from a non-English speaking background
clinical concern. [21, 29, 49–51], but not all [52], again with very small
The other studies examined distress severity using a range effects (R-square change = 0.002) [e.g. 21]. Three studies
of other self-report symptom scales and generally reported examined the impact of parental education and found no
high proportions of distress in the student samples (in most effect [9, 43, 53]; however, a protective effect was found
cases as indicated against adult norms). Two studies [21, for higher status of parental occupation in the UK [52] and
43] used the Global Health Questionnaire [GHQ: 44] in Nigeria [54], but not in Melaka [43]. Other demographic
its full and short (12 item) form. Hodge, McCormick and effects were examined in only one or two studies making
Elliott [21] applied a conservative clinical cut-off (≥ 8) on conclusions difficult. Birth order was not associated with
the GHQ 30 item version to Year 11 and Year 12 students distress [9]; however, having a sibling was associated with
in New South Wales Australia and found 42.4% Year 11 significantly reduced distress [55]. No effect was found for
and 56.7% Year 12 students reported distress in the clinical parental marital status [43] whilst one found an effect of
range (based on adult norms). Similarly, Lin and Yusoff [43] parent divorce on males anxiety but not females [53], and
found 48% of senior students in Melaka scored in the clinical increased anxiety in females with household stress but not
range on the GHQ 12 item (using a cut off of ≥ 4). One study males [53]. One study compared final year students living
examined depression using the Patient Health Question- in Korea to students living in America and found that the
naire-9 [PHQ-9: 45] and found 26% of final year students in Korean students had significantly higher depression [48].
13
Individual Differences The relationship between distress South Australia [9], in which 12 internal and external fac-
and individual factors were investigated. The strongest evi- tors associated with resilience were examined. In females,
dence related to female gender, anxiety proneness and free- freedom from negative cognitions was the largest predictor
dom from negative thoughts with emerging evidence for: of resilience to depression, anxiety and stress, with family
perfectionism, academic buoyancy, student motivation and connectedness and confidence also being important. For
coping, hours of study and sleep. males, freedom from negative cognitions, confidence, fam-
ily, peer and school connectedness were significant predic-
Gender Females reported significantly higher distress in the tors of resilience.
final years of high school than males across a wide range of Negative cognitions in test anxiety have also been shown
distress measures and samples [3, 9, 13, 21, 22, 27, 28, 30, to be strongly associated with distress. Test anxiety has been
52, 53, 56–65]. Overall, in the senior school samples females conceptualised as having a cognitive component (e.g. worry
consistently had higher scores for anxiety symptoms than about failure) and an affective-physiological component (e.g.
males [3, 22, 28, 53, 64, 65]. Similarly, females also had somatic symptoms of anxiety) [78]. The cognitive compo-
higher test anxiety scores across a range of senior school nent of test anxiety has been specifically found to be asso-
samples such as in the UK [5, 13, 50, 52], in Chinese stu- ciated with test performance in a range of studies [10, 50,
dents [58], American students [64, 65], Greek students [61], 51, 79, 80], with higher scores being predictive of poorer
Iranian students [60], Portuguese students [27], Turkish stu- test performance on major exams in senior students from
dents [56, 57, 62, 63, 66], but in one study male and female the UK, New Zealand, USA, India, and Turkey [10, 50, 57,
Nigerian students did not differ significantly on test anxiety 69, 80, 81]. In a large study, 705 UK students in the final
[54]. Females had higher depression scores than males in year of school (Year 11) completed measures of test anxiety
some samples [22], but not in others [3, 28]. Females also and academic buoyancy (see below) four months prior and
had higher scores for DASS stress [3, 22]. These findings again two months prior to the final exams [10]. The worry
could relate to the higher emotionality of females and higher component of test anxiety at baseline was associated with
prevalence of anxiety (and depressive) disorders in adoles- lower exam performance four months later. Similar findings
cent female samples [67, 68], or gender differences in will- in a number of UK studies have shown negative cognitions
ingness to report emotional distress. mediate the relationship between test anxiety and exami-
nation performance [79, 82]. Therefore, heighted negative
Anxiety Proneness It was consistently found that greater cognitions and worry about tests appears to be particularly
distress in the final years of school was associated with or associated with student distress as well as reduced examina-
predicted by higher trait anxiety suggesting anxious prone tion performance.
students are more likely to experience distress. For example,
higher state trait anxiety was found to predict 58% of vari- Perfectionism Theoretically, perfectionism is conceptu-
ance in distress (measured on the GHQ) in senior students alised as having dimensions incorporating cognitions and
over and above demographics, school type, coping and gen- behaviours associated with a need to do things perfectly [83,
der [21]. Similarly, negative affect was shown to predict 84]. Hewitt and Flett [84] described perfectionism as being
higher test anxiety in senior school students in New Zea- driven by personal desires for high standards for the self
land [69]. Studies have also reported that high test anxiety, (self-oriented), for others (other-oriented), or a perception
which is a measure of trait anxiety [11] and hence anxiety that others hold unrealistic standards for one’s behaviour
proneness is associated with increased distress, lower self- (socially-prescribed). Perfectionism can be adaptive (high
efficacy and self-esteem in final year students [54, 56, 70], standards without excessive negative self-evaluation) or
and lower examination performance [10, 71, 72]. Therefore, maladaptive (high standards coupled with excessive nega-
anxiety proneness is likely to be a significant predictor of tive self-evaluation) and maladaptive perfectionism has
senior school distress. been associated with distress and psychopathology [83].
A small number of studies have examined the impact of
Negative Cognitions Qualitative and quantitative studies perfectionism on student distress in the final years of high
across a number of different countries (Australia, Canada, school, and have generally found that higher levels of per-
Ireland, Italy, Turkey, Melaka, United Kingdom, United fectionism were associated with increased distress and test
States of America) found heightened distress in the final anxiety [63, 85], mirroring findings in university students
years of school was strongly associated with negative cogni- [86]. In an Australian sample, Einstein et al. [29] found
tions, fear of failure, fear of examinations, homework con- socially-prescribed perfectionism (perceived perfectionistic
cerns and fear of not getting into university [6, 8, 30, 43, standards expected by others e.g. parents) was associated
73–77]. The significance of negative cognitions was par- with increased distress, but self-oriented and other-oriented
ticularly highlighted in a study of 195 year 12 students in perfectionism were not. Therefore, there is some evidence
13
that perfectionistic standards which result in negative evalu- self-efficacy was associated with lower test anxiety scores
ation of self or perceived negative evaluation by others, is in American [70], Australian [9], and Turkish senior school
associated with increased distress in students. As discussed students [55], as well as higher self-esteem in Nigerian [54]
above, it is likely that the negative cognitions that drive and Turkish senior students [56] and lower depression levels
maladaptive perfectionism are an important component in in Canadian senior students [85]. However, test competence
student distress, and more research is needed to examine was not a significant predictor of unique variance in exami-
this further. nation performance when tested in a model also containing
test worry, academic buoyancy and perceived control, sug-
Coping Students who reported using more avoidant or non- gesting that it is test related worry that is most relevant [82].
adaptive coping styles (wishful thinking, self-blame, sup-
pression) had higher distress and test anxiety scores [21, 30, Academic Buoyancy Academic buoyancy is defined as the
43, 66, 76, 87], with adaptive coping (problem solving, task capacity to withstand routine types of academic setbacks,
orientation, reappraisal) associated with lower distress [21, challenges, and pressures experienced by students during
76, 81, 87]. Further, Italian final year students who reported their education such as dealing with competing deadlines,
decreased beliefs about their ability to cope with upcoming poor grades and examination pressure [90]. It has been
examinations had the highest distress [76]. These findings shown to be comprised of higher self-efficacy, higher plan-
mirror other research showing that maladaptive coping, par- ning ability, greater beliefs about control, persistence and
ticularly avoidance, is associated with increased distress and low anxiety [91], and therefore includes many of the ele-
psychopathology generally [88]. ments discussed above. Given that, it is perhaps not surpris-
ing that academic buoyancy has been found in a number
Motivation Four studies looked at the impact of student of studies to be associated with lower test anxiety and dis-
motivation to approach or avoid performance tasks on distress, greater enjoyment of school, class participation, self-
tress severity in the final years of school. A performance- esteem, and better examination performance [10, 81, 87, 90,
approach goal refers to the motivation to outperform peers, 92, 93]. Although academic buoyancy was found to explain
whereas a performance-avoidance goal refers to motivation unique variance (5–10%) in test anxiety in senior school
to avoid demonstrating lack of ability. Smith and Sinclair students over and above coping [87], when the role of aca-
[22] looked at the impact of performance-approach goal ori- demic buoyancy was considered in conjunction with other
entation, performance-avoidance goal orientation, as well predictors of examination performance, only worry about
as self-efficacy and self-handicapping coping strategies on tests (strongest predictor) and perceived control (smaller
levels of depression, anxiety or stress in year 11 and 12 stu- predictor) were significant predictors of unique vari-
dents. They found higher performance-avoidance goal ori- ance, with academic buoyancy and test competence being
entation was associated with higher depression and anxiety non-significant [82]. Therefore, academic buoyancy is an
in year 11 males, and anxiety and stress in year 12 males, important concept in understanding senior school distress;
and depression in year 12 females. Similarly, Putwain and however, it is unclear what the relationship is of academic
Symes [89] also found in 273 final year UK students that buoyancy to the other main predictors of senior school dis-
higher test anxiety was associated with more performance- tress such as the absence of negative cognitions and anxiety
avoidance and mastery-avoidance (avoidance of increasing proneness, and further studies are needed to understand the
competence). In a sample of Indian students in their penul- specific and unique contribution of academic buoyancy on
timate year of high school, those who reported higher test student distress.
anxiety also indicated poorer study habits and using self-
handicapping coping strategies [80]. These findings are Time Spent Studying Four studies examined the relationship
similar to the findings above related to greater distress being between the number of hours spent studying and student
associated with using avoidance as a coping strategy, as well distress and found that more hours studying was associ-
as negative beliefs about their ability and the consequence ated with greater distress and negative mood [31, 48, 59,
of poor performance. 94]. However, the causal relationship between distress and
time spent studying is not clear. In an interesting cultural
Academic Self‑efficacy In qualitative interviews students comparison, Lushington et al. [59] found the relationship
reported that stress in the senior years related specifically between greater hours of study and greater distress was only
to how confident and competent they felt about specific sub- evident in Caucasian-Australian senior students and not in
ject matter [8], with higher confidence in their competence Asian-Australian senior students, and that hours of study
for exam and assessment tasks relating to lower distress. only related to stress levels and not depressed mood. Lee and
The relationship between higher academic confidence and Larson [48] found that increased time spent on homework
lower distress was also found in quantitative studies; Higher was only related to increased depression in those students
13
(Korean and American) who experienced increased nega- boys’ schools [25, 28], and one also found single sex girls’
tive affect during homework. This suggests that students’ schools had significantly greater negative affect scores [28].
attitudes or negative cognitions towards or during homework This school type effect is likely to be accounted for by the
might be a mediating factor between hours of study and dis- findings that females scored higher on anxiety and depres-
tress. Lee and Larson [48] also found that more time spent in sion measures, rather than the school type specifically. One
active leisure was associated with reduced depression, sug- Turkish study reported school based differences in test based
gesting that it might not be the amount the number of hours anxiety [56]; however, how the school types differed was not
spent studying specifically, but the portion of hours studying clearly described and so it is difficult to draw conclusions.
compared to hours spent doing pleasant (mood enhancing) In general it appears that school based differences are minor,
activities. Interestingly, a study with Turkish senior stu- with the main differences related to single sex girls’ schools
dents found that poorer time management when completing having higher distress, which likely reflects the strong effect
school assignments was associated with increased anxiety of female gender on distress, although further research is
[95]. More research is needed to understand the relationship needed to understand if female gender coupled with an all-
between study time and distress, and to examine individual female school environment increases the distress.
differences in study patterns, emotional wellbeing and aca-
demic performance. Family, Peer and School Connectedness Student distress
varied by how strongly connected students felt with their
Sleep Three studies examined the relationship between schools (sense of belonging to their school community),
sleep and stress in final year students and found links their peers (satisfaction with peer relationships) and with
between increased stress and poorer sleep. In a small sam- their family (satisfaction with care and support from family,
ple (n = 24) in the Netherlands, stress experienced in the family cohesion). All five studies found that a more posi-
final examination period was found to be associated with tive relationship with school was associated with reduced
significantly reduced total sleep time (17.5 ± 8.2 min), distress in final year students [3, 9, 26, 63, 102]. Strong
sleep efficiency and increased wake bouts [96]. Similarly, peer connectedness was also shown to be associated with
in a large (n = 195) Australian sample final year students reduced distress. For example, in a large study of 941 Year
one month prior to their final exams reported high rates of 12 Australian students in their final year of school, McGraw
inadequate sleep, with more distressed students reporting et al. [3] found after controlling for gender that connec-
greater rates of daytime napping [9]. Also Lushington et al. tion with school, peers and family were associated with
[59] found that 20% of students (n = 398) had missed class reduced depression (accounting for 42% of the variance),
in the previous month because they had overslept, and that with the strongest predictor being connection with peers.
increased daytime sleepiness was associated with increased Male gender, peer connectedness and family connectedness
stress and depressed mood. Across the studies, it is not clear were all significant protective factors for anxiety and stress
whether sleep is a predictor of distress or a consequence of (explaining 21–22% of the variance respectively), with peer
distress. Both are likely true as there is clear evidence that connectedness again being the strongest protective factor.
inadequate sleep is associated with increased anxiety and Peer connectedness was also related to increased resilience
depressed mood [97–99], as well as anxiety and depres- in an Australian sample [102]. In Turkish senior students,
sion being associated with poorer quality sleep and fatigue perceived support from peers (and teachers) was associated
[100, 101]. More research is needed to understand the link with less test anxiety [63]. Robinson et al. [9] also found
between distress and poor sleep in senior students, what role that peer, school and family connectedness were predictive
excessive study might play in the reduced hours of sleep, of resilience to distress in males in the final year of school,
and whether it is chronic or acute sleep deprivation that is while only family connectedness (and not peer or school
most significant. connectedness) was protective of distress in female students.
In an American sample, students in higher grades reported
Family, Peer and School Factors The evidence related to the lower school connectedness and higher anxiety than younger
impact of school type, family, peer and school connected- students [26]. Therefore, a strong sense of connectedness
ness, perceived pressure, and fear appeals are examined in with the school, family and peers is important for reduced
relation to distress in senior students. distress in senior students.
Type of School In general no significant differences were Fear Appeals and Perceived Pressure Fear appeals refer to
found in distress levels across different school types or attempts by teachers (or parents) to motivate students by
boarding compared to day students [6, 62]. Two studies highlighting the consequences of failing or doing poorly
found significantly higher mean anxiety scores in single [103]. The impact of fear appeals on students’ levels of
sex girls’ schools compared to coeducational and single six distress and on subsequent examination performance has
13
been mixed and differs depending on how students’ inter- that distress increased as examination periods approached,
pret the fear appeals. For example, in a sample of 132 UK however, it is unclear whether the severity of distress expe-
students in their final two years of schooling, fear appeals rienced by these students is dissimilar to younger students
by teachers to year 10 students (penultimate year of high approaching examinations that might be considered to be
school) were found to be associated with increased worry associated with lower academic stakes [4]. More research
and tension related to their major mathematics exam later in using longitudinal designs and carefully timed assessment of
Year 11, but this was only true for students who perceived distress in examination and non-examination periods across
the fear appeals to be threatening [103]. For other students, grade levels is needed.
perceived threat from fear appeals in year 10 was associated Given consistent reports that many students reported very
with increased fear of failure but also an increased motiva- high levels of distress, the question of whether distress expe-
tion to improve their competence (mastery-approach), hence rienced by students in their final years is too much is a topical
the fear appeals helped to increase their commitment to one. This question is difficult to answer given the differences
study. Similarly, in a sample of 273 final year UK students, in measurement tools used and use of age appropriate nor-
Putwain and Symes [89] found that fear appeals by teachers mative cut-offs were generally absent. Examining the data
that were perceived as threatened associated with a greater from the two countries with the most research, using the TAI
performance-avoidance approach, and resulted in poorer in the UK an average of 16% of students might be considered
examination performance. Although research recently to have distress that was excessive, and using the DASS in
found that when fear appeals were used more frequently, the the Australian data an average of 19% might be considered
tendency to appraise them as threatening increased [104], to have excessive distress. This equates to approximately
with related research in university students suggesting that one in six students. This rate of distress is very similar to
efficacy appeals (instead of fear appeals) resulted in reduced the findings from two large Australian national surveys. In
distress and increased examination performance [105]. the Youth Mental Health Report [2] 21.2% of adolescents
More research is needed to understand the links between aged 15–19 years reported high levels of distress, and in the
fear appeals, threat perception, and examination perfor- Australian National Survey of the Mental Health of Chil-
mance. The emerging research that efficacy appeals might dren and Adolescents [68] 19% of adolescents reported very
produce better emotional outcome and test performance is high or high levels of distress. It is not clear if the distress
particularly important. captures in these national surveys predominately captures
Similarly, qualitative studies found students with per- school based stress, or if the distress captured in the studies
ceived pressure from parents, teachers and peers reported in this review capture distress that is different from normal.
higher distress [6, 8, 73, 75]. Çırak’s [73] study in Turk- It is unclear what proportion of the one in six students with
ish senior students identified that the pressure from parents high levels of distress identified in this review had distress
related to a desire by students to not disappoint their parents that was in the clinical range, and whether that was higher
who had often made significant sacrifices to give them an than expected. The recent Australian National Survey of the
educational opportunity. As reported earlier, Einstein et al. Mental Health of Children and Adolescents [68] found that
[29] found that higher socially-prescribed perfectionism, 14.4% of 12–17 year olds met criteria for a mental disorder
which captured perceived pressure to achieve unrealistic in the previous 12 months, and 7% met criteria for an anxiety
goals determined by significant others, was a significant disorder. Therefore it is likely that at least this proportion
predictor of distress (depression and anxiety) in final year of students in the final years of school are likely to be clini-
students prior to a set of major exams. Therefore the findings cally distressed. Although this distress might be transient,
suggest that perceived pressure by teachers to do academi- it is clear that some students are very distressed and would
cally well is predictive of heightened distress in the final benefit from additional assistance to manage their distress.
years of school and mediated the impact on student distress Establishing methods to identifying these students as soon
and examination performance. as possible is an important direction for the future.
In addition, to the proximity of examinations, a number
of other factors were found to be associated with increased
Discussion distress in high school students in the final years. The most
consistent effects related to individual differences such as
This review aimed to understand the nature of, severity of, female gender, anxiety proneness and freedom from negative
and correlates of distress in secondary students undergo- cognitions. The higher prevalence of distress in females fits
ing the final years of schooling. The global interest in this with gender based differences in emotionality and incidence
topic was demonstrated by studies represented student sam- rates of internalising mental disorders [67, 68]. Findings
ples from 16 different countries, with the majority report- that greater anxiety proneness was associated with increased
ing high levels of distress in students. There was evidence distress fits with longitudinal research that shows early child
13
anxiety predicts later adolescent anxiety [106, 107]. It also poor sleep and excessive time spent on homework, with
fits theoretically with the Diathesis-Stress models [108] that academic self-efficacy, academic buoyancy and resilience
assert that the combination of a diathesis or predispositional being protective factors. Although more research is needed
vulnerability coupled with a sufficient stressor results in psy- to understand the unique contribution of these additional
chological disorder or other poor psychological outcomes. predictors over and above gender, anxiety proneness and
As such the heightened stress associated with major exams freedom from negative cognitions, and how these factors
or increased academic demands when coupled with a predis- interact with each other.
posing vulnerability towards distress, results in heightened Finally, there was also emerging evidence for the role
distress that may be of clinical severity. Therefore, anxiety of perceived pressure from parents and teachers (through
proneness is likely to be a significant predictor of senior fear appeals) on student distress, although fear appeals were
school distress, and early identification of anxiety prone- associated with increased distress and poorer examination
ness in early school years could be used to predict students performance only in some students. It is likely that anx-
likely to experience excessive distress prior to the start of ious prone students are likely to interpret fear appeals or
the senior years and thus be targeted for interventions prior pressure from parents and teachers to perform as threaten-
to the senior school years. ing and subsequently experience greater distress, although
The finding that freedom from negative thinking was a more research is needed to examine this relationship. Further
significant protective factor aligns with the findings that research might also look at the underlying factors associ-
the cognitive elements of test anxiety and worry were the ated with the use of fear appeals by teachers and parents.
strongest predictors of student distress. The presence of It is likely that teacher/parental concerns related to subop-
negative cognitions are theorised to play a prominent role timal student performance drives this effect. As academic
in the development and maintenance of child and adolescent performance metrics are being used more frequently in the
anxiety disorders [16, 17]. For example, anxious students United Kingdom, Australia and elsewhere in accountability
are likely to have the tendency to misinterpret ambiguous practices to judge school and teacher effectiveness [113],
information (e.g. average marks), overinflate the likelihood there is also emerging evidence for increasingly high levels
and consequences of perceived negative events (e.g. “I will of stress experienced by teachers who feel pressured to help
fail”, and “It will be a disaster if I fail”), and have poorer their students get excellent results [114], this might result in
coping strategies for dealing with this heightened distress teacher behaviours (such as the use of fear appeals that exac-
(e.g. avoidance, procrastination), exacerbating and main- erbate student distress and reduce academic performance).
taining distress. Furthermore, a number of studies demon- Future research might track teacher stress and use of fear
strated that students with greater worry about tests actually appeals and other teacher behaviours longitudinally to see
had poorer examination performance. This highlights the how it relates to student’s distress in large studies. Strategies
need to intervene with students who show high distress or to reduce teacher distress might also prove a future target for
report high levels of worry about tests early so that they are interventions to relieve student distress.
not disadvantaged educationally. Given the important role of Limitations of the study need to be considered. Firstly, the
negative cognitions in test anxiety, distress and subsequent quality of the studies varied considerably with the majority
examination performance, more attention needs to be paid to being of lower quality. Secondly, studies varied in the meas-
interventions that target negative cognitions particularly in ures used to measure distress as well as individual correlates
the lead up to high stakes tests. There is some evidence that of distress. Thirdly few studies had control or comparison
cognitive behavioural interventions are effective in reducing groups and so it is difficult to determine the difference in
distress in senior years when delivered as prevention pro- distress and relationships with correlates as they compare to
grams [109] or as targeted programs for those experiencing other age groups. Finally, the studies were generally cohort
acute academic distress when delivered in school settings studies with only one measurement occasion such that most
[110, 111], and perhaps even in single session universal studies report on correlates of distress (and often grouping
interventions [112]. These cognitive behavioural interven- multiple factors together), such that individual contributions
tions focus on teaching students how to manage negative of factors or predictors of distress cannot be conclusively
cognitions, feelings of stress, and how to reduce avoid- determined.
ant behaviours, which are the key factors identified in this
review as predictors of distress.
Given evidence for the impact of peer, family and school Summary
connectedness on student distress, future interventions might
include strategies to bolster peer, school and family relation- This systematic review examined the nature of and factors
ships. There was also emerging evidence for other significant associated with student distress in the final two years of
predictors of distress including maladaptive perfectionism, secondary school. The findings indicated academic stress
13
during this period is a clear concern across a large number 12. King NJ, Ollendick TH, Prins PJ (2000) Test-anxious children
of countries with a significant number of students expe- and adolescents: psychopathology, cognition, and psychophysi-
ological reactivity. Behav Chang 17(3):134–142
riencing very high levels of distress. Research indicated 13. Putwain D (2008) Deconstructing test anxiety. Emot Behav Dif-
distress was associated with individual differences in anxi- ficult 13(2):141–155
ety proneness, gender and freedom from negative cogni- 14. Gross C, Hen R (2004) The developmental origins of anxiety.
tions, as well as connectedness to family, peer and school. Nat Rev Neurosci 5(7):545–552
15. Weems CF (2008) Developmental trajectories of childhood anxi-
Other factors also appear to be associated with increased ety: Identifying continuity and change in anxious emotion. Dev
distress, although more research is needed to understand Rev 28(4):488–502
their unique contributions. More methodologically rigor- 16. Rapee RM (2001) The development of generalized anxiety. In:
ous research tracking distress over time is needed to better Vasey MW, Dadds MR (eds) The developmental psychopathol-
ogy of anxiety. Oxford University Press, New York
understand chronicity and severity of distress and moder- 17. Spence SH, Rapee RM (2016) The etiology of social anxiety
ating and mediating factors. Such information will assist disorder: an evidence-based model. Behav Res Ther 86:50–67
endeavours to apply interventions preventatively or in a 18. Moher D, Liberati A, Tetzlaff J, Altman DG (2009) Preferred
targeted fashion for at-risk students. reporting items for systematic reviews and meta-analyses: the
PRISMA statement. Ann Intern Med 151(4):264–269
19. Critical Skills Appraisal Programme (2018) CASP Cohort Study
Checklist [online]: https://casp-uk.net/casp-tools-checklists/
Compliance with Ethical Standards 20. Critical Skills Appraisal Programme (2018) Qualitative Study
Checklist [online]: https://casp-uk.net/casp-tools-checklists/
Conflict of interest The authors declare that they have no conflict of 21. Hodge GM, McCormick J, Elliott R (1997) Examination-induced
interest. distress in a public examination at the completion of secondary
schooling. Brit J Educ Psychol 67(2):185–197
22. Smith L, Sinclair KE (2000) Transforming the HSC: affective
implications. Change 3(2):67–79
23. Akca F (2011) The relationship between test anxiety and learned
References helplessness. Soc Behav Pers 39(1):101–111
24. Guner-Kucukkaya P, Isik I (2010) Predictors of psychiatric symp-
tom scores in a sample of Turkish high school students. Nurse
1. Huan VS, See YL, Ang RP, Har CW (2008) The impact of Health Sci 12(4):429–436
adolescent concerns on their academic stress. Educ Rev 25. Moulds JD (2003) Stress manifestation in high school students:
60(2):169–178 an Australian sample. Psychol Schools 40(4):391–402
2. Ivancic L, Perrens B, Fildes J, Perry Y, Christensen H (2014) 26. Wilkinson-Lee AM, Zhang Q, Nuno VL, Wilhelm MS (2011)
Youth mental health report. Mission Australia and Black Dog Adolescent emotional distress: the role of family obligations and
Institute, AU school connectedness. J Youth Adolesc 40(2):221–230
3. McGraw K, Moore S, Fuller A, Bates G (2008) Family, peer 27. Cunha M, Paiva MJ (2012) Text anxiety in adolescents: the role
and school connectedness in final year secondary school stu- of self-criticism and acceptance and mindfulness skills. Span J
dents. Aust Psychol 43(1):27–37 Psychol 15(2):533–543
4. Dewald JF, Meijer AM, Oort FJ, Kerkhof GA, Bogels SM 28. Locker J, Cropley M (2004) Anxiety, depression and self-esteem
(2014) Adolescents’ sleep in low-stress and high-stress (exam) in secondary school children. Sch Psychol Int 25(3):333–345
times: a prospective quasi-experiment. Behav Sleep Med 29. Einstein DA, Lovibond PF, Gaston JE (2000) Relationship
12(6):493–506 between perfectionism and emotional symptoms in an adolescent
5. Putwain D, Daly A (2014) Test anxiety prevalence and gender sample. Aust J Psychol 52(2):89–93
differences in a sample of English secondary school students. 30. Lay CH, Edwards JM, Parker JD, Endler NS (1989) An assess-
Educ Stud 40(5):554–570 ment of appraisal, anxiety, coping, and procrastination during an
6. Leonard NR, Gwadz MV, Ritchie A, Linick JL, Cleland CM, examination period. Eur J Person 3(3):195–208
Elliott L, Grethel M (2015) A multi-method exploratory study 31. Peluso MA, Savalli C, Curi M, Gorenstein C, Andrade LH (2010)
of stress, coping, and substance use among high school youth Mood changes in the course of preparation for the Brazilian uni-
in private schools. Front Psychol 6:1028–1044 versity admission exam—a longitudinal study. Rev Bras Psiquitr
7. Chamberlain S, Daly AL, Spalding V (2011) The fear fac- 32(1):30–36
tor: students’ experiences of test anxiety when taking A-level 32. Smith L, Sinclair KE, Chapman ES (2002) Students’ goals, self-
examinations. Pastor Care Educ 29(3):193–205 efficacy, self-handicapping, and negative affective responses: an
8. Putwain D (2009) Assessment and examination stress in Key Australian senior school student study. Contemp Educ Psychol
Stage 4. Br Educ Res J 35(3):391–411 27(3):471–485
9. Robinson JA, Alexander DJ, Gradisar MS (2009) Preparing for 33. Yeni Palabiyik P (2014) A study of Turkish high school students’
Year 12 examinations: predictors of psychological distress and burnout and proficiency levels in relation to their sex. Novitas-
sleep. Aust J Psychol 61(2):59–68 ROYAL 8(2):169–177
10. Putwain D, Daly A, Chamberlain S, Sadreddini S (2015) 34. Lovibond S, Lovibond P (1995) Manual for the depression anxi-
Academically buoyant students are less anxious about and ety stress scales. Psychology Foundation of Australia, Sydney
perform better in high-stakes examinations. Brit J Educ Stud 35. Watson D, Clark LA, Tellegen A (1988) Development and
85(3):247–263 validation of brief measures of positive and negative affect: the
11. Spielberger CD (1980) Preliminary professional manual for the PANAS scales. J Pers Soc Psychol 54(6):1063–1070
Test Anxiety Inventory. Consulting Psychologists Press, Palo 36. Kovacs M (1985) The children’s depression inventory (CDI).
Alto Psychopharmacol Bull 21:995–998
13
37. Reynolds CR, Richmond BO (1978) What I think and feel: a Asian-Australian and Caucasian-Australian adolescents. Int J
revised measure of children’s manifest anxiety. J Abnorm Child Ment Health 44(1–2):139–157
Psychol 6(2):271–280 60. Rahafar A, Maghsudloo M, Farhangnia S, Vollmer C, Randler
38. Rosenberg M (1965) Society and the adolescent self-image. C (2016) The role of chronotype, gender, test anxiety, and
Princeton University Press, New Jersey conscientiousness in academic achievement of high school
39. Szafranski DD, Barrera TL, Norton PJ (2012) Test anxiety inven- students. Chronobiol Int 33(1):1–9
tory: 30 years later. Anxiety Stress Coping 25(6):667–677 61. Ringeisen T, Buchwald P (2010) Test anxiety and positive and
40. Szabó M (2010) The short version of the Depression Anxiety negative emotional states during an examination. Special Issue
Stress Scales (DASS-21): factor structure in a young adolescent Test Anxiety 14(4):431–447
sample. J Adolesc 33(1):1–8 62. Sari SA, Bilek G, Celik E (2018) Test anxiety and self-esteem
41. Willemsen J, Markey S, Declercq F, Vanhuele S (2011) Negative in senior high school students: a cross-sectional study. Nord J
emotionality in a large community sample of adolescents: the Psychiatry 72(2):84–88
factor structure and measurement invariance of the short version 63. Yildirim I, Ergene T, Munir K (2007) Academic achievement,
of the depression anxiety stress scales (DASS-21). Stress Health perfectionism and social support as predictors of test anxiety.
27(3):120–128 Hacet U J Educ 34:287–296
42. Patrick J, Dyck M, Bramston P (2010) Depression Anxiety Stress 64. Manley MJ, Rosemier RA (1972) Developmental trends in
Scale: is it valid for children and adolescents? J Clin Psychol general and test anxiety among junior and senior high school
66(9):996–1007 students. J Genet Psychol 120(2):219–226
43. Lin HJ, Yusoff MS (2013) Psychological distress, sources of 65. Christensen H (1979) Test anxiety and academic achievement
stress and coping strategy in high school students. Int Med J in high school students. Percept Mot Ski 49(2):648
20(6):672–676 66. Aysan F, Thompson D, Hamarat E (2001) Test anxiety, cop-
44. Goldberg DP (1972) The detection of psychiatric illness by ques- ing strategies, and perceived health in a group of high school
tionnaire. London University Press, London students: a Turkish sample. J Genet Psychol 162(4):402–411
45. Spitzer RL, Kroenke K, Williams JB, Group P (1999) Valida- 67. Merikangas KR, He J, Burstein M, Swanson SA, Avenevoli S,
tion and utility of a self-report version of PRIME-MD: the PHQ Cui L et al (2010) Lifetime prevalence of mental disorders in
primary care study. J Am Med Assoc 282(18):1737–1744 US adolescents: results from the National Comorbidity Sur-
46. Beck AT, Rush AJ, Shaw BF, Emery G (1979) Cognitive therapy vey Replication-Adolescent Supplement (NCS-A). J Am Acad
of depression. Guilford, New York Child Adolesc Psychiatry 49(10):980–989
47. Yildirim I, Ergene T, Munir K (2007) High rates of depres- 68. Lawrence D, Johnson S, Hafekost J, Boterhoven de Haan K,
sive symptoms among senior high school students preparing Sawyer M, Ainley J et al (2015) The mental health of children
for National University Entrance Examination in Turkey. IJSD and adolescents: Report on the second Australian child and
4(2):35–44 adolescent survey of mental health and wellbeing. Australian
48. Lee M, Larson R (2000) The Korean "Examination Hell": long Government Department of Health, Canberra
hours of studying, distress, and depression. J Youth Adolesc 69. Chin E, Williams MW, Taylor JE, Harvey ST (2017) The influ-
29(2):249–271 ence of negative affect on test anxiety and academic perfor-
49. Putwain D (2008) Do examinations stakes moderate the test mance: an examination of the tripartite model of emotions.
anxiety–examination performance relationship? Educ Psychol Learn Individ Differ 54:1–8
28(2):109–118 70. Segool NK, von der Embse NP, Mata AD, Gallant J (2014)
50. Putwain D (2008) Test anxiety and GCSE performance: the effect Cognitive behavioral model of test anxiety in a high-stakes
of gender and socio-economic background. EPIP 24(4):319–334 context: an exploratory study. School Ment Health 6(1):50–61
51. von der Embse NP, Witmer SE (2014) High-stakes accountabil- 71. Sarason IG (1963) Test anxiety and intellectual performance.
ity: student anxiety and large-scale testing. J Appl Sch Psychol J Abnorm Soc Psychol 66(1):73–75
30(2):132–156 72. McCann SJ, Meen KS (1984) Anxiety, ability, and academic
52. Putwain D (2007) Test anxiety in UK schoolchildren: prevalence achievement. J Soc Psychol 124(2):257–258
and demographic patterns. Brit J Educ Psychol 77:579–593 73. Çırak Y (2016) University entrance exams from the perspective
53. Byrne B (2000) Relationships between anxiety, fear, self-esteem, of senior high school students. J Educ Train Stud 4(9):177–185
and coping strategies in adolescence. J Adolesc 35(137):201–215 74. Kouzma NM, Kennedy GA (2004) Self-reported sources
54. Chukwuorji JC, Nwonyi SK (2015) Test anxiety: contributions of of stress in senior high school students. Psychol Rep
gender, age, parent’s occupation and self-esteem among second- 94(1):314–316
ary school students in Nigeria. J Psychol Afr 25(1):60–64 75. Putwain D (2011) How is examination stress experienced
55. Ünal-Karagüven MH (2015) Demographic factors and communal by secondary students preparing for their General Certifi-
mastery as predictors of academic motivation and test anxiety. J cate of Secondary Education examinations and how can it be
Educ Train Stud 3(3):1–12 explained? Int J Qual Stud Educ 24(6):717–731
56. Erzen E, Odacı H (2016) The effect of the attachment styles 76. Schmidt S, Tinti C, Levine LJ, Testa S (2010) Appraisals, emo-
and self-efficacy of adolescents preparing for university tions and emotion regulation: an integrative approach. Motiv
entrance tests in Turkey on predicting test anxiety. Educ Psychol Emot 34(1):63–72
36(10):1728–1741 77. Smyth E, Banks J (2012) High stakes testing and student per-
57. Karatas H, Alci B, Aydin H (2013) Correlation among high spectives on teaching and learning in the Republic of Ireland.
school senior students’ test anxiety, academic performance Educ Assess Eval Acc 24(4):283–306
and points of University Entrance Exam. Educ Res Rev 78. Liebert RM, Morris LW (1967) Cognitive and emotional com-
8(13):919–926 ponents of test anxiety: a distinction and some initial data.
58. Liu YY (2012) Students’ perceptions of school climate and trait Psychol Rep 20(3):975–978
test anxiety. Psychol Rep 111(3):761–764 79. Putwain D, Connors L, Symes W (2010) Do cognitive distor-
59. Lushington K, Wilson A, Biggs S, Dollman J, Martin J, Ken- tions mediate the test anxiety–examination performance rela-
nedy D (2015) Culture, extracurricular activity, sleep hab- tionship? Educ Psychol 30(1):11–26
its, and mental health: a comparison of senior high school
13
80. Sud A, Sujata (2006) Academic performance in relation to self- 100. Ramsawh HJ, Stein MB, Belik SL, Jacobi F, Sareen J (2009)
handicapping, test anxiety and study habits of high school chil- Relationship of anxiety disorders, sleep quality, and func-
dren. Psychol Stud 51(4):304–309 tional impairment in a community sample. J Psychiatr Res
81. Putwain D, Daly A, Chamberlain S, Sadreddini S (2016) 43(10):926–933
‘Sink or swim’: buoyancy and coping in the cognitive test 101. Tsuno N, Besset A, Ritchie K (2005) Sleep and depression. J Clin
anxiety—academic performance relationship. Educ Psychol Psychiatry 66(10):1254–1269
36(10):1807–1825 102. Riekie H, Aldridge JM, Afari E (2017) The role of the school
82. Putwain D, Aveyard B (2018) Is perceived control a critical fac- climate in high school students’ mental health and identity for-
tor in understanding the negative relationship between cogni- mation: a South Australian study. Br Educ Res J 43(1):95–123
tive test anxiety and examination performance? Sch Psychol Q 103. Putwain D, Symes W (2011) Teachers’ use of fear appeals in the
33(1):65–74 mathematics classroom: worrying or motivating students? Br J
83. Frost RO, Marten P, Lahart C, Rosenblate R (1990) The dimen- Educ Psychol 81(3):456–474
sions of perfectionism. Cogn Ther Res 14(5):449–468 104. Symes W, Putwain D, Remedios R (2015) The enabling and
84. Hewitt PL, Flett GL (1991) Perfectionism in the self and social protective role of academic buoyancy in the appraisal of fear
contexts: conceptualization, assessment, and association with appeals used prior to high stakes examinations. Sch Psychol Int
psychopathology. J Pers Soc Psychol 60(3):456–470 36(6):605–619
85. Flett GL, Panico T, Hewitt PL (2011) Perfectionism, type A 105. von der Embse NP, Schultz BK, Draughn JD (2015) Readying
behavior, and self-efficacy in depression and health symptoms students to test: the influence of fear and efficacy appeals on
among adolescents. Curr Psychol 30(2):105–116 anxiety and test performance. Sch Psychol Int 36(6):620–637
86. Eum K, Rice KG (2011) Test anxiety, perfectionism, goal ori- 106. Ollendick TH, King NJ (1994) Diagnosis, assessment, and treat-
entation, and academic performance. Anxiety Stress Coping ment of internalizing problems in children: the role of longitudi-
24(2):167–178 nal data. J Consult Clin Psychol 62(5):918–927
87. Putwain D, Connors L, Symes W, Douglas-Osborn E (2012) Is 107. Rapee RM (2014) Preschool environment and tempera-
academic buoyancy anything more than adaptive coping? Anxi- ment as predictors of social and nonsocial anxiety disorders
ety Stress Coping 25(3):349–358 in middle adolescence. J Am Acad Child Adolesc Psychiatry
88. Compas BE, Jaser SS, Bettis AH, Watson KH, Gruhn MA, Dun- 53(3):320–328
bar JP et al (2017) Coping, emotion regulation, and psychopa- 108. Zuckerman M (1999) Diathesis-stress models. In: Zuckerman M
thology in childhood and adolescence: a meta-analysis and nar- (ed) Vulnerability to psychopathology: a biosocial model. Ameri-
rative review. Psychol Bull 143(9):939–991 can Psychological Association, Washington, US
89. Putwain D, Symes W (2011) Perceived fear appeals and exami- 109. Werner-Seidler A, Perry Y, Calear AL, Newby JM, Christensen
nation performance: facilitating or debilitating outcomes? Learn H (2017) School-based depression and anxiety prevention pro-
Individ Differ 21(2):227–232 grams for young people: a systematic review and meta-analysis.
90. Martin A, Marsh H (2009) Academic resilience and academic Clin Psychol Rev 51:30–47
buoyancy: Multidimensional and hierarchical conceptual fram- 110. Lowe C, Wuthrich VM, Hudson JL (2020) Randomised con-
ing of causes, correlates and cognate constructs. Oxf Rev Educ trolled trial of Study Without Stress: a Cognitive Behavioural
35(3):353–370 Therapy group program to reduce academic stress in students in
91. Martin A, Marsh H (2006) Academic resilience and its psycho- the final year of high school. In preparation
logical and educational correlates: a construct validity approach. 111. Wuthrich VM, Lowe C (2015) Study without stress program.
Psychol Schools 43(3):267–281 Department of Psychology, Macquarie University, Sydney, AU,
92. Putwain D, Daly A (2013) Do clusters of test anxiety and aca- Centre for Emotional Health
demic buoyancy differentially predict academic performance? 112. Varlow M, Wuthrich V, Murrihy R, Remond L, Tuquiri R, Van
Learn Individ Differ 27:157–162 Kessel J et al (2009) Stress literacy in Australian adolescents.
93. Martin A, Colmar S, Davey L, Marsh H (2010) Longitudinal Youth Stud Aust 28(4):29–34
modelling of academic buoyancy and motivation: do the ’5Cs’ 113. Saeki E, Pendergast L, Segool NK, von der Embse NP (2015)
hold up over time? Br J Educ Psychol 80:473–496 Potential psychosocial and instructional consequences of the
94. Kouzma NM, Kennedy GA (2002) Homework, stress and common core state standards: implications for research and
mood disturbance in senior high school students. Psychol Rep practice. Contemp Sch Psychol 19(2):89–97
91(1):193–198 114. von der Embse NP, Pendergast LL, Segool N, Saeki E, Ryan
95. Akcoltekin A (2015) High school students’ time manage- S (2016) The influence of test-based accountability policies on
ment skills in relation to research anxiety. Educ Res Rev school climate and teacher stress across four states. Teach Teach
10(16):2241–2249 Educ 59:492–502
96. Astill RG, Verhoeven D, Vijzelaar RL, Van Someren EJ (2013) 115. Daly AL, Chamberlain S, Spalding V (2011) Test anxiety, heart
Chronic stress undermines the compensatory sleep efficiency rate and performance in A-level French speaking mock exams:
increase in response to sleep restriction in adolescents. J Sleep an exploratory study. Educ Res 53(3):321–330
Res 22(4):373–379
97. Baum KT, Desai A, Field J, Miller LE, Rausch J, Beebe DW Publisher’s Note Springer Nature remains neutral with regard to
(2014) Sleep restriction worsens mood and emotion regulation jurisdictional claims in published maps and institutional affiliations.
in adolescents. J Child Psychol Psychiatry 55(2):180–190
98. Sarchiapone M, Mandelli L, Carli V, Losue M, Wasserman C,
Hadlaczky G et al (2014) Hours of sleep in adolescents and its
association with anxiety, emotional concerns, and suicidal idea-
tion. Sleep Med 15(2):248–254
99. Short MA, Gradisar M, Lack LC, Wright HR, Dohnt H (2013)
The sleep patterns and well-being of Australian adolescents. J
Adolesc 36(1):103–110
13

Academic Stress in The Final Years of School: A Systematic Literature Review

Uploaded by

Copyright:

Available Formats

Academic Stress in The Final Years of School: A Systematic Literature Review

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Academic Stress in The Final Years of School: A Systematic Literature Review

Uploaded by

Copyright:

Available Formats

What is the purpose of the systematic literature review?

What is the purpose of the systematic literature review?

What factors are associated with increased academic stress according to the review?

What factors are associated with increased academic stress according to the review?

Child Psychiatry & Human Development

Academic Stress in the Final Years of School: A Systematic Literature

© Springer Science+Business Media, LLC, part of Springer Nature 2020

Arcles combined Duplicates removed

Arcles screened by Arcles excluded by

Arcles screened by Arcles excluded by

Full-text arcles Full-text arcles excluded

Not specifically related to

Could not source full-text

Table 1 Articles included in the review

Trait and students from a non-

of the Multidimensional basis

Intellectual ability (Raven’s Higher anxiety was associated

exam grades than students

(final year) (SD = 0.51) Reduced morningness-evening- conscientiousness had lower

Boys with a performance-avoid-

in the cohort studies. In consolidating the findings across

significantly greater in females

Students with lower academic

Distress Increases Over the Senior School Period

Cohort Studies Comparisons between student cohorts

across different grades generally found that the final year

of high school was associated with more distress than the

associated with more distress than the earlier years [23–26]

ior years. Although the majority of studies found increased

this may relate to differences in the timing of assessments

9 twice, both at the same time prior to major exams and

in the timing and type of distress measurement particularly

Longitudinal Studies Five studies sampled distress in the

same students throughout the final year, with most find-

[29–33], and decreased distress after the examination period

using the Depression, Anxiety and Stress Scales [DASS;

severe to extremely severe stress, and remained in the severe

Yildirim et al. [47]

means for DASS measured stress, anxiety and depression

erate anxiety; however, remained at a mild level for stress

Table 2 Quality ratings of articles included in the review

Akca [23] Yes Yes Yes No Unclear

Sud and Sujata [80] Yes Unclear Yes Yes Yes

You might also like

Arcles combined Duplicates removed

Arcles screened by Arcles excluded by

Arcles screened by Arcles excluded by

Full-text arcles Full-text arcles excluded

Table 1 Articles included in the review

Table 2 Quality ratings of articles included in the review